global symposium for hazard risk reduction - gfdrr · pdf filewebsite — ......

GlobalSymposium for

Hazard RiskReduction

Lessons Learned from the Applied Research Grants for Disaster Risk Reduction Program

26–28 July 2004

Washington, DC

The World Bank

A P P L I E D R E S E A R C H G R A N T S F O R D I S A S T E R R I S K R E D U C T I O N

WORKING PAPERS

To explore innovative ways to address human vulnerability tohazards(natural and human-made), the ProVention Consortiuminvited a special disaster management audience—the world’s youth.The Consortium launched a program of Applied Grants for DisasterRisk Reduction in 2002. Sixty-five grants were awarded and fifteenprojects have now been selected for presentation at this GlobalSymposium. This outreach program was managed by the WorldBank’s Hazard Management Unit (HMU) in collaboration with theAsian Disaster Preparedness Center (ADPC), Cranfield DisasterManagement Centre (CDMC) and the University ofWisconsin–Disaster Management Center (UW-DMC).

This collection of working papers presents current research ondisaster management issues and practice.

For additional regional information, please contact:

Asian Disaster Preparedness Center (ADPC)P.O. Box 4, Klong Luang, Pathumthani 12120 THAILANDWebsite — http://www.adpc.net/

Cranfield Disaster Management Centre (CDMC)Cranfield University (RMCS), Shrivenham, Swindon SN6 8LA, UK Website — http://www.rmcs.cranfield.ac.uk/ddmsa/dmc/view

University of Wisconsin–Disaster Management Center (UW–DMC)University of Wisconsin-Madison, 432 North Lake Street, Madison,WI 53706, USAWebsite — http://dmc.engr.wisc.edu/

COVER PHOTOS

Aliou Mamadou Dia, Chipo Muvezwa, Gabriela Muñoz Rodriguez, Rose Berdin

COVER DESIGN

Susan Kummer, Artifax

APPLIED RESEARCH GRANTS FOR DISASTER RISK REDUCTION

The World Bank

July 2004

Global Symposium forHazard Risk Reduction

Lessons Learned from the AppliedResearch Grants for Disaster RiskReduction Program

26–28 July 2004

Washington, DC

WORKING PAPERS

v Preface

vi Acknowledgements

1 Symposium Working Papers

1 LESSONS LEARNED 1 — EARTHQUAKES

1 Martha Liliana CarreñoExpert system for post earthquake building damage evaluation and massive risk occupancy – Colombia

15 Hayk KtunyanMain principles of earthquake early warning system creation around criticalfacilities in seismic active zones – Armenia

23 Bilgen SungaySeismic conservation of historical and cultural treasures of a world city: sizingthe need and formulating an action plan for the museums of Istanbul – Turkey

37 LESSONS LEARNED 2 — FLOODS AND EROSION

37 Aliou Mamadou DiaEarth observation and GIS-based flood monitoring in the Senegal River estuaryand valley – Senegal

51 Rose BerdinCoastal erosion vulnerability mapping along the southern coast of La Union,Philippines – Philippines

69 Keshav Prasad PaudelMapping and assessing risk and vulnerability of water-induced disasters in theTinau Watershed, Western Nepal – Nepal

83 LESSONS LEARNED 3 — MULTI-HAZARDS

83 Dash BiswanathIndicators for disaster preparedness – India

95 Chipo MuvezwaStrengthening regional capacity for disaster management – Zimbabwe

109 W. WarsaMultidimensional SNMR modeling for groundwater exploration – Indonesia


119 Vishnu PrabhirAction advocacy to ensure right to livelihood risk reduction and beyond – India

129 Rutere Salome KagendoStrengthening rural community bonds as a means of reducing vulnerability tolandslides – Kenya

139 Archana PradanGIS and remote sensing for flood disaster identification: a case study of the Koshi River Basin – Nepal

iii

65

RE

SE

AR

CH

GR

AN

TSC

OLL

AB

OR

ATI

NG

CE

NTE

RS

Table of Contents

WO

RK

ING

PA

PER

S

Ea

rth

qu

ak

es

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

continued next page


149 Gabriela I. Muñoz Rodríguez: Local and popular folklore and culture onhazard and vulnerability meets geographical information system for the riskreduction preparedness of the people of the Barrio San Antonio of Naiguatá,Estate Vargas – Venezuela

163 Avanish Kumar: Communication strategies for industrial disaster riskreduction – India

173 Nadejda Petrova: Risk reduction through first aid capacity building of youthemergency teams – Bulgaria

177 The 65 Research Grants

193 The Collaborating Centers

iv

v

On behalf of the ProVention Consortium, the World Bank's Hazard Management Unit (HMU) organized a competitive forum to support innovative disaster risk management projects and the promotion ofcompetent young professionals (under 35 years of age) dedicated to reducing disaster risk in developingcountries. The ProVention Consortium is a global coalition of governments, internationall organizations,academic institutions, the private sector, and civil society organizations aimed at reducing disaster impactsin developing countries. The Applied Research Grants for Disaster Risk Reduction program has beenadministered by the HMU in cooperation with three collaborating centers.

The program’s objective is to support innovative risk management projects dedicated to reducing disasterrisk and developed by disaster professionals in developing countries. Young researchers from around theworld were invited to propose creative projects to reduce disaster impacts, and the most promising ideaswere awarded small grants for implementation. The applications were screened by an independent panelof disaster management specialists, who nominated for award those proposals that best fulfilled the grantcriteria in three categories: (1) hazard and risk identification; (2) risk reduction; and, (3) risk sharing /transfer. Sixty�five individuals and/or teams won grants of up to US$5,000.

The ProVention Consortium launched the program in December 2002 and awarded the grants in June2003. The young researchers completed their projects in January 2004. For more information about thegrants program, please visit the website at: http://www.proventionconsortium.org/projects/ appliedres.htm.

Awards were granted to students and young professionals from the following 27 countries: Argentina,Armenia, Bangladesh, Barbados, Bhutan, Bulgaria, China, Colombia, Costa Rica, Georgia, India, Indonesia,Kenya, Mexico, Nepal, Nigeria, Pakistan, Philippines, Senegal, South Africa, Sudan, Tajikistan, Turkey,Uzbekistan, Venezuela, Vietnam, and Zimbabwe.

After a series of independent and peer reviews, fifteen projects were selected as representative of themost innovative and sustainable activities; summaries and lessons learned from these projects comprisethis document. Team leaders from these projects present their findings at the "Global Symposium forHazard Risk Reduction," July 26�28, 2004 at World Bank headquarters in Washington, DC.

The fifteen selected project leaders have prepared Working Papers reporting the results of their researchactivities. Those projects have been arranged as scheduled for presentation at the Symposium. Followingthose papers, there is a complete list of all sixty�five projects with contact information for those involved ineach. We encourage you to contact the project teams if you are interested in more details about theresults and lessons learned.

At the end of this publication, there is brief information about each of the collaborating centers: the AsianDisaster Preparedness Center (ADPC), Cranfield Disaster Management Centre (CDMC) and the Universityof Wisconsin�Disaster Management Center (UW�DMC).

ADPC was responsible for coordination and support for the grantees in two regions, East Asia Pacific andSouth Asia.

CDMC was responsible for coordination and support for the grantees in one region, Sub�Saharan Africa.

UW–DMC was responsible for coordination and support for the grantees in two regions, Europe & CentralAsia and Latin America & Caribbean; they also had global administrative responsibilities for the grantapplication process, the independent review panel for grant awards, maintenance of the program's websitefor project sharing by grantees, the independent review panel for selection of representative projects, theregional webconferences for project peer review, project publications and the Symposium.

Preface

vi

The Working Papers in this publication were prepared as part of the Applied Research Grants forDisaster Risk Reduction program, which was organized by the World Bank's Hazard ManagementUnit under the umbrella of the ProVention Consortium. These papers are the basis for the "GlobalSymposium for Hazard Risk Reduction," a meeting held July 26�28, 2004 at World Bankheadquarters in Washington, DC.

The ProVention Consortium would like to thank all who submitted research proposals and thosesixty�five who were awarded grants, along with their team members and advisors. Next, we thankthose who served as independent reviewers for the proposals and for selection of the mostrepresentative projects for presentation at the Symposium: William Anderson, Claude DeVille, JimGood, Bruno Haghebaert, Ailsa Holloway, Terry Jeggle, Alcira Kreimer, David Peppiatt, Sheila Reed,Juan Pablo Sarmiento, John Telford, Paul Thompson, and Jaime Valdés.

We would also like to thank program leaders and staff at the Asian Disaster Preparedness Center,Cranfield Disaster Management Centre and the University of Wisconsin–Disaster ManagementCenter.

We are also grateful to the following individuals for their assistance and timely support: MargaretArnold, Michelle Addison, David Alexander, Maxx Dilley, Lynn Gross, Ian Davis, Susan Kummer,Manisha Malla Shrestha, Lorraine Ortner�Blake, Alcira Kreimer, Linda Martin�Chitturi, MaryvonnePlessis�Fraissard, Maria Eugenia Quintero, Loy Rego, Tim Randall, Maya Schaerer, Don Schramm,M. Vitor Serra, Eva von Oelreich, and Zoe Trohanis.

The ProVention Consortium would especially like to thank the United Kingdom's Department forInternational Development (DFID) and the Government of the Kingdom of Norway's Royal Ministryof Foreign Affairs for their generous support.

Acknowledgements

1

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

Expert system for post-

earthquake building damage

evaluation and massive risk

occupancy decision-making

TEAM LEADER

Martha Liliana Carreño T.<[email protected]>

PROJECT ADVISORS

Omar D. CardonaAlex H. Barbat

Colombia

SYMPOS IUM NOTES

2

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

Lessons learned or importantelements that will have impacts ondisaster risk reduction – Colombiaand worldwide

The experience in all countries has demonstrated thatwhenever a moderate or strong earthquake occurs,that causes widespread damage in an urban center,serious problems emerge for the processes of riskevaluation of the affected buildings. These problemshave serious legal, economic and securityimplications. A similar situation was repeated in thecase of the building damage evaluation after theearthquake of January 1999 in coffee growing areaof Colombia.

These problems have appeared, partly, due to thelack of experience of the inspectors and theunavoidable need to have involving in the processnon�expert people to make difficult decisions ofmuch responsibility about the security of thebuildings. Therefore, it is common, that errors arebeing made, like demolishing or evacuatingconstructions that probably were not in so seriousconditions, or underestimating damage, placing indanger the life of the occupants.

Due to this situation, within the Program of DamageEvaluation, that was implemented by the AsociaciónColombiana de Ingeniería Sísmica (AIS) and theOficina Municipal de Prevención y Atención deDesastres (OMPAD) of Manizales, Colombia, it wasagreed to develop a computational tool, based onArtificial Intelligence (AI) to avoid the error�makingbefore mentioned, and with the purpose ofcomplement the conventional methods of damageevaluation –inspection forms, guidelines andevaluation procedures– that were elaborated for riskestimation in case of an earthquake in the city.

This Expert System will help to avoid the loss of life, identify the unsafe buildings and prevent itsoccupation. Also, it will contribute to identify the safe places that can be used like shelters for peoplewho have lost their homes or that have had to beevacuated. The system is a support for the decision�making by the non�expert inspectors. Then, it willfacilitate the massive and right evaluation ofdamage, risk, habitability and reparability of buildings after an earthquake.

This system is also being implementing officially by the city of Bogotá and it is hoped that will beadopted by othre cities of Colombia with thetechnical support of AIS. Also, it could be used in

3

other countries, once it become adapted inagreement with their construction classes,contributing to a fast and reliable decision making;problem of risk and disaster that must be faced andthat has deep social and economic implications forthe society when an important earthquake strikes.

Summary on the implementation ofthe project and its results

Using techniques of Computational Intelligence (CI)an Expert System was developed, based on expertcriteria, in order to employ properly subjective andincomplete information about the damage ofaffected buildings after an earthquake, and to carryout linguistic or qualitative qualifications about theirlevel of risk, habitability and reparability in a reliableway. This system was designed using fuzzy sets andfuzzy logic and it was calibrated by means of anartificial neural network (ANN); modern technologiesof AI suitable for the handling of vague andqualitative information.

In order to carry out the proposed tool it wasnecessary to participate in some activities in theframework of the Program of Damage Evaluationthat was made in Colombia by AIS and the OMPAD.These activities have been fundamental for thesuitable understanding of the problem and todevelop a useful product, witch was not had initiallyinto consideration within the mentioned program.These activities to develop the Expert system were:

• To participate in the review of the main existingmethodologies of building damage evaluation.

• To participate in the development of theguidelines and inspection forms for buildingsaffected by earthquakes in Manizales city.

• To formulate the conceptual and mathematicalmodel of neuro�fuzzy expert system based onComputational Intelligence.

• Development, calibration and test of the expertsystem using information from evaluations madeduring the January 1999 Quindio earthquake.

A summarized description of each activity ispresented in the following paragraphs.

Review of the main existingmethodologies of building damageevaluation

The main elements for a methodology of buildingdamage evaluation are: classification of buildingdamage, definition of the possibilities to use the

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

constructions that suffered damage, organization forcollection of data, and the analysis and informationprocessing. This was considered very useful to studythe different methods and to compare the mostrecent formats and criteria of evaluation of thedifferent methodologies with the purpose of identify�ing the main conceptual differences among. Thisactivity was made partially in Barcelona within theacademic activities of the author and during visitsmade to Bogotá, D.C. It meant the careful revision ofthe methodologies before mentioned, with the pur�pose of obtaining fundamental inputs for the devel�opment of the Expert System object of this project.

For the review of the existing methodologies a groupof specialists of AIS was comprised. The author of thisproject was part of the group and contributed withher work related to the topic made during her grad�uate studies. It was carried out a bibliographical compilation on the existing methodologies ofdifferent countries. The main methodologiesevaluated were from Macedonia (old Yugoslavia),USA, Japan, Mexico, Italy, Greece, Turkey, andColombia. This review was made with the purpose ofanalyzing comparatively the different methodologiesconsidering aspects such as: their objectives andscope, criteria for description and location ofconstructions, criteria for damage evaluation andclassification, type of recommendations andemergency measures proposed, among others.

The conclusion, after a rigorous review of allmethodologies, has been that the more completeapproaches are: the ATC�20 from USA, themethodology developed by the Mexican Society ofSeismic Engineering, the method promoted by theMinistry of the Construction of Japan and by theNational Seismic Service of Italy. Also it is importantto acknowledge the previous techniques developedin Colombia.

In general, the methodologies are relatively recent,in almost all countries the forms and guidelines havebeen updated due to the improvement of knowledgeafter each earthquake. The damage evaluationprocedures normally are applied by different stages,which have been classified in fast, detailed andengineering evaluations, being the first two of specialinterest for the object of this work.

The fast evaluation of habitability of constructions is applied usually in the short term to define thepossible occupation and use of the building. After theevaluation, if the buildings are safe, they can beused. In addition, some recommendations are issued

with the purpose to reduce risk for their inhabitants.The detailed evaluations that describe the level ofstructural damage and its classification should bedeveloped due to several reasons, but in general theyare made with the objective to verify the safety levelof the constructions on which there are some doubtsdue to a very fast evaluation or due to the lack ofexperience of the inspectors.

To participate in the development of theguidelines and inspection forms forbuildings affected by earthquakes inManizales city

From the review of the work made in Colombia todate and other places of the world, it was possible toconclude that the existence of a detailed fastevaluation with aims of emergency response hasbeen in order to verify doubtful evaluations or tomake a look by a person with more criterion andexperience to make decisions. Therefore, it wasconcluded that it is more efficient to apply a singleevaluation form and that the inspectors teams shouldbe comprised, at least, by two people to have agreater trustworthiness in the concepts. Also, it waspossible to conclude that is better to avoid making adouble evaluation and that in case of need thesecond professional concept would be easier if itcould be made in the same form in the second visit;hoping that these cases are exceptional and not therule. However, it was ratified that with the aid of anExpert System based on AI, made to support thedecisions of less experienced inspectors, it is possibleto solve this problem.

Considering the Program of Damage Evaluation thatAIS and OMPAD make in Colombia, with the purposeof preparing the institutions of the city and theprivate sector to carry out the estimation of risk inthe buildings affected in case of earthquake, thedesign of the inspection guidelines of constructionsafter an earthquake was made. Also, the damageinspection forms and the warning formats ofhabitability were developed. This work was carriedout in a participative way with the potential users inthe city and with the coordination of a team ofspecialists of AIS in which the author of his projectwas involved.

The design of the field guidelines was taking intoaccount that the persons who will apply them will beprofessionals related to construction sector, like civilengineers, architects or technicians, advancedstudents of the same professional areas and, ingeneral, people that will even arrive from other cities

4

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

in case of an extreme event. The guidelines, theinspection forms and the warnings formats ofhabitability are the output of the research processmade with the purpose of consolidating a basicmethodology useful not only for the city of Manizalesbut also for other cities and the national level. Theobjective of the inspection guidelines is to have amethodology for building damage and safetyevaluation that might be applied after an earthquake.This methodology allows defining the habitability ofconstructions quickly and facilitates the later actionsof rehabilitation and reconstruction.

The scope of the guidelines was limited byagreement of all participants. The classification of thebuilding damage and the habitability are based onthe results of the building global inspection, thedamage in the structural and non�structural elementsand the ground conditions of surroundings. Therefore,the guidelines do not include the procedures toevaluate the definitive rehabilitation of theconstruction. For this, each owner should contract astructural engineer, who makes detailed andcomplete inspection or carries out tests on the qualityof the materials, the state of reinforcement, etc. Onthe other hand, it was not developed procedures toquantify the detailed economic and social impactgenerated by the earthquake, but with the guidelinesis possible to obtain a gross estimation of themagnitude of the disaster, useful for rehabilitationand reconstruction planning.

Formulation of the conceptual andmathematical model of neuro-fuzzyexpert system based on ComputationalIntelligence

In agreement with the previous remarks, the conven�tional damage evaluation consists on identificationand registry of damage occurred in the affectedbuildings, with the purpose of determining in a fastway if they are at risk. In other words, if they are safeor they must be evacuated to protect the life of hisoccupants. This process demands experience, criterionand training; therefore, it is a need that all peopleinvolved in the process must be experts in theevaluation process. Unfortunately, the professionalswho usually fulfill this requirement are very few. Onthe other hand, the evaluation is an urgent andmassive task that should be made by severalprofessionals and the current methodologies are notvery objective if non�expert people are involvedbecause they have the tendency to aggravate or tounderestimate the real level of damage. Therefore, inthe case of the Program of Damage Evaluation for

5

Manizales, promoted by AIS and OMPAD, it wasconcluded about the need to complement thecurrent methodology of damage evaluation with aninnovating tool that help to solve the deficienciesbefore mentioned.

The evaluation of the damage and the level of riskthat presents a construction after an earthquake area complex activity that demands much criterion andknowledge and can have serious implications forsecurity of life. The information that takes part in theevaluation is highly subjective and depends on theconception and the inspector perception in eachcase. Mistakes can be made, such as the demolitionof buildings that probably were not in so seriousconditions, or buildings can be evacuated unneces�sarily. Also, it is possible that non�expert inspectorsignore failures in the construction that jeopardizetheir stability, putting in danger the life of hisoccupants. For that reason it is so important toguarantee that the more experienced people andwith the better knowledge on structural behaviorand pathology make the evaluations.

The levels of damage in one building are defined inmost of the evaluation methodologies by means oflinguistics qualifications such as slight, minor,moderate, average, severe or heavy. These conceptscan have a remarkable variation in their meaningaccording to the knowledge and experience of eachperson. Therefore, a defined frontier between thesevaluations does not exist clearly. What for a personcould be moderate for another can be severe, as well as it can be in the middle of both concepts foranother one. For this reason it is necessary to unifythe sense of these concepts and to make theevaluation more quantitative as possible,determining percentage of affected elements, sizeand type of cracks, etc.

In agreement with the research made by the author,using fuzzy sets it is possible to represent qualitativeor subjective information in a numerical way, such asit is required in damage evaluation after anearthquake. Fuzzy sets do not have defined limitsperfectly; that is to say, the transition between themembership and the non�membership of an elementto a set is gradual. On the other hand, the ANNshave been used to face nonlinear complex problems,simulating the operation of the nervous system ofthe humans. These algorithms are demonstrating tobe very useful for several applications, like patternand image recognition, signal processing,optimization and the automatic control.

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

Taking into account this features a computationaltool was designed that helps to avoid the mistakesusually made by non�expert people before mentionedand considering they use subjective and incompleteinformation. In addition, this tool allows carrying outthe massive and suitable evaluation of damage,habitability and reparability of buildings affected byearthquakes, being based on linguistics or qualitativequalifications, using artificial neural networks andfuzzy logic.

Development, calibration and test of the expert system using the informa-tion of the evaluations made during theJanuary 1999 Quindío earthquake.

The source code of the expert system was developedin Visual BASIC with registry modules made inAccess. The computer program “Earthquake DamageEvaluation” of buildings (EDE) for Manizales runs onWindows 95, 98, 2000 or XP. The computer programis the direct product of this project.

The neural network in which the system is based wascalibrated by means of a Kohonen�type learningalgorithm. The learning process was made usingdamage evaluations achieved after the 1999 Quindioearthquake in Colombia. The training of the networkdid not consider structural systems such as wood andsteel frames because the absence of these structuralsystems in the area affected by the earthquake. Onlyfew composite (frame and wall) reinforced concretebuildings were used because there are a few of themin the area.

In order to carry out correctly this training an exhaus�tive review of the evaluation forms used for theMinistry of Development for the census of affectedbuildings in the area was made. This information wasanalyzed with the purpose of identifying the cases(building class, levels of damage, diagnosis) and theinspectors who made the evaluations. Thus, AIS madea detailed review of the registries to develop thedatabase for the calibration and training of theneural network. It meant also the work of expertsand the administrative and logistical support ofpeople related to AIS to collect and adjust theinformation.

Once the system was calibrated using the database it was necessary to make several tests with hypothe�tical cases. They allowed to the experts verify theperformance of the system. Also, it was made aprocess of qualification and verification with localexperts in Manizales city, with the purpose ofexploring feasible cases and to guide the right use of

the system by people who will have the responsibilityto carry out the evaluations in case of an earthquakein the future, with the coordination of OMPAD.

The Expert System EDE was delivered to the city, likea complementary tool for damage evaluation, forindividual and complex cases, within the Program ofDamage Evaluation made by AIS and OMPAD inManizales. At present, EDE is also implementing inthe city of Bogotá, the capitol city of the country, andit now part of the Program for Post�earthquakeBuilding Damage Evaluation that AIS develops withthe Dirección de Prevención y Atención deEmergencias (DPAE) from Bogotá.

The author thanks to the ProVentium Consortium bythe grant. This aid has been an important support toinclude the doctoral research activities of the author–developed in the Technical University of Catalonia–in the Program of Damage Evaluation that is made inColombia for the city of Manizales and Bogotá. Alsothe author is very grateful with the professors Alex H.Barbat and Omar D. Cardona by their academicdirection and with AIS due to the opportunity to bepart of its program and to include the modelproposed for practical application. Particularly, theauthor acknowledges the guidance of Omar D.Cardona as president of AIS and Ana Campos astechnical coordinator of the program for Manizalesand of the second stage of the program in Bogotá, inexecution at present.

Brief description of the model

Computational model

The presented model uses a fuzzy logic approach,required by the subjective and incomplete characterof the information. Post�earthquake damageevaluations use qualitative and linguistic expressionsthat are appropriately handled by the fuzzy setstheory. In addition, an artificial neural network (ANN)is used to calibrate the system using the judgementof specialists. Artificial neural networks were inspiredby the modelling of neural networks of naturalneurons in a human brain. In this case a networkwith fuzzy signals is used. This enables the use ofcomputational intelligence for the damage evaluationby non�experts.

The artificial neural network (ANN) structure of thismodel consists of three layers. The neurons in theinput layer are grouped in four types, namelystructural elements (SE), non�structural elements(NE), ground conditions (GC), and pre�existentconditions (PC). Each one contributes with informa�

6

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

tion to the neurons in the intermediate layer; theyonly affect the intermediate neuron in the group towhich they belong. The input neurons number orvariables in the model is not constant; this numberdepends on the structural system and on the impor�tance of the groups of variables for the evaluation.The number of neurons of the input layer used forthe structural elements group changes according tothe class of building. Table 1 shows the structuralelements or variables considered according to thestructural system.

A qualification is assigned depending on theobserved damage using five possible damage levelsthat are represented by means of fuzzy sets. Forstructural and non�structural elements, the followinglinguistics damage qualifications are used: none (N),light (L), moderate (M), heavy (H) and severe (S).Figure 1 illustrates the membership functions forthese qualifications.

7

a)

b)

c)

d)

Figure 2 Damage in structural elements. a) Severe damage in a reinforced concrete

joint, b) Moderate damage in a reinforcedconcrete beam, c) Heavy damage in a masonrywall, d) Heavy damage in a bahareque wall

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

0

0.2

0.4

0.6

0.8

1

1.2

0 0.2 0.4 0.6 0.8 1Damage Indices

Mem

ber

ship

(¬

x) None

Light

Moderate

Heavy

Severe

Figure 1. Membership functions for linguistic qualifications

Figures 2 and 3 show some damage levels in differ�ent structural and non�structural elements. The mem�bership functions of fuzzy sets achieve their maxi�mum membership point for the values of the damageindices whose selection will be explained forward.

Structural system Structural elements

RC frames or (with) shear walls

Columns/walls, beams, jointsand floors

Steel or wood framesColumns, beams,connections and floors

UnreinforcedReinforcedConfined masonry

Bearing walls and floors

Bahareque or tapial walls Bearing walls and floors

Table 1. Structural elements according tostructural system

Figure 2.Damage instructuralelements. a) Severedamage in areinforcedconcrete jointb) Moderatedamage in areinforcedconcrete beamc) Heavydamage in amasonry walld) Heavydamage in abahareque wall

a

c

b

d

Damage in the non�structural elements does notaffect the stability of the buildings, but may put atrisk the security of the occupants. Table 2 presentsthe classification of the non�structural elements:elements whose evaluation is compulsory andelements whose evaluation is optional.

Table 2. Non-structural elements

The ground and pre�existent conditions variables arevalued through the qualification of their state at theevaluation moment. The linguistic qualifications usedare: very good (VG), good (G), medium (M), bad (B)and very bad (VB). Ground conditions consist ofvariables that can affect the stability of the building,such as landslides and soil liquefaction; examples ofthese situations can be observed in Figure 4. Pre�existent conditions are related to the quality of theconstruction materials, plane and vertical shapeirregularities of the building, and the structuralconfiguration, illustrated in the Figure 5; theseconditions may increase the seismic vulnerability of a building.

8

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

a

b

a

b

Figure 3.Damage in non-structural elementsa) Severe damage in partitionsb) Severe damage in stairs

Compulsatory evaluationelements

Partition

Elements of façade

Stairs

Optional evaluationelements

Ceiling and lights

Installations

Roof

Elevated tanks

Figure 4 Ground conditionsa) Soil settlement due to liquefactionb) Landslides and ground failure

Figure 5 Pre-existent conditions: a) Bad construction qualityb) Vertical shape irregularity: soft floorc) Plane shape irregularityd) Bad structural configuration: lack of

frame continuity

9

This model is a user�friendly computer programcalled Earthquake Damage Evaluation of Buildings,EDE. It offers aids for the inspectors, such as detaileddescriptions and damage photographs. Recently,another computer program has been developed (ATC2003). Unlike, this other program does not supportthe decision�making and basically its objective is tofacilitate the data collection of the existinginspection form of ATC�20 (ATC 1989).

Description of the ANN

Input layer of the artificial neural network. Thefuzzy sets for each element or variable i (for instancecolumns, walls or beams) in the input layer areobtained from the inspector’s linguistic qualifica�tions of damage Dj at each level j and its extensionwj. The damage extension, or percentage of eachdamage level in each element, varies from 0 to 100and it is normalized

(1)

The accumulated qualification of damage Di foreach variable is obtained as the union of the scaledfuzzy sets, taking into account the damagemembership functions mDj(Dj), and its extensionsor weights assigned by the inspector

(2)

(3)

The union in the theory of the fuzzy sets is repre�sented by the maximum membership or dependency.By means of defuzzification, using the centroid ofarea method (COA), a qualification index Ci isobtained for each variable of each group of neurons

(4)

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

a

b

c

d

Each variable has predefined the basic membershipfunctions for the fuzzy sets corresponding to the five possible levels of damage. The linguisticqualifications change in each case. Figures 6 and 7show some examples of the data input of thecomputer program.

Figure 6. Input of the damagequalifications a) Damage extension at each damage level for a structural elementb) For a non-structural element

Figure 7. Input of the qualifications forground conditions and pre-existentconditions a) Landslides and ground failure b) Structural configuration

Intermediate or hidden layer of ANN. This layerhas four neurons corresponding to each group ofvariables: structural elements, non�structuralelements, ground conditions and pre�existentconditions. Figure 8 shows a general scheme of theevaluation process. In this neural network model,the inputs of the four neurons are the qualificationsCi obtained for each variable of the each group ofneurons and its weight Wi or degree of importanceon the corresponding intermediate neurone. Theseweights were defined with the participation ofexperts in earthquake damage evaluation. Usingthese qualifications and weights for each variable i,a global index could be obtained, for each group k,from the defuzzification of the union or maximummembership of the scaled fuzzy sets. The membershipfunctions mCki(Cki) and their weights Wki showthe notation for the group of structural elements.

(5)

(6)

The groups of variables related to ground and pre�existing conditions are optional, thus they can be orcan be not considered within the evaluation. If thishappens, the habitability and reparability of buildingis assessed only with the structural and non�structural information.

Output layer of ANN. In this layer, the globalindices obtained for structural elements, non�struc�tural elements, ground and pre�existent conditionscorrespond to one final linguistic qualification in each case. The damage level (qualitative) is obtainedaccording to the “proximity” of the value obtainedto a global damage function of reference. Thetraining process of the neural network is achieved inthis layer. The indices that identify each qualitativelevel (center of cluster) are changed in agreementwith the indices calculated in each evaluation andwith a learning rate. Once the final qualifications aremade, it is possible to determine the global buildingdamage, the habitability and reparability of thebuilding using a set of fuzzy rule bases.

Training process of the ANN

The neuronal network is calibrated in the outputlayer when the damage functions are defined inrelation to the damage matrix indices. In order tostart the calibration, a departure point is defined,

10

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

a

b

a

b

( )( )

()

[]

centroidSEiCSEi

1SEC1SESE

CW

,...,CWI

SEi

1SE

µ×

µ×= max

( ) ( )( )

()SEiCSEi

1SEC1SECSE

CW

,...,CWC

SEi

1SE

µ×

µ×=µ max

that means the initial indices ofeach level of damage. Theindices proposed by the ATC(1985), Park, Ang and Wen(1984), the fragility curves usedby HAZUS (FEMA 1999), andthe indices used by Sanchez�Silva and García (2001) havebeen considered. The values ofthese indices correspond to thecenter of the area for everymembership function related toeach damage level. Table 3shows the indices proposed inthis work.

11

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

For structural elements

* According to structural system:

-Reinforced concrete fram-Confined masonry-Reinforced masonry-Unreinforced masonry-Bahareque walls-Tapial walls-RC frame with shear walls-Steel frame-Wood frame

S tr uc. e l em. 1*

S tr uc. e l em. 2*

Structural Elements

IS E

WS 1

WS 2

W Si

cr ispfuz zy

Partitions

Elements of facade

Stairs

IN S

W N S1

W N S2

W NS3

Non-structural Elements

Configuration

Plane ShapeIrregularities

Quality

Pre-existing Conditions

Vertical ShapeIrregularities

cr isp

fuz zyIP C

Ground Conditions

Liquefaction

Landslides

IG Ccr isp

fuz zy

cr isp

fuz zy

0.01

D1 Level I

0.17

D 2Level II

0.33

D 3 Level III

0.55

D 4

Level IV

0.76

D 5

Level V

S tr uc. e l em. i*

Fuzzy rule basis

ISE

INS

IG C

IP

Pre-existingConditions

GroundConditions

Non-structuralDamage

HabitabilityReparability

CP C4 , WP C4

CP C3 , WP C3

CP C2 , WP C2

CP C1 , WP C1

CG C2 , WG C2

CN S3 , WN S3

CN S2 , WN S2

CN S1 , WN S1

CS E i, WS E i

CS E 2, WS E 2

CS E 1, WS E 1

StructuralDamage

CG C1 , WG C1

W G C1

W G C2

W PC1

W PC2

W PC3

W PC4

Figure 8. Structure of the proposed artificial neural network

Table 3. Comparative table for damage indices

Damage Level Park, Ang and Wen Sanchez-Silva Proposed

Very light < 0.10 0.07 0.10

Light 0.10 – 0.25 0.175 0.20

Moderate 0.25 – 0.40 0.325 0.35

Severe 0.40 – 0.80 0.6 0.60

Destruction >0.80 0.8 0.90

Some authors consider that collapse occurs for avalue equal to 0.8 and others propose a collapsethreshold of 0.77. Therefore, a value of 0.76 hasbeen selected to describe the destruction level indexor collapse. In the selection of the damage index, theauthor decided to be conservative, since the indicescorresponding to severe and moderate damage havebeen highly discussed, and doubts exist on whetherthey should be smaller.

The calibration is performed for each damage leveland only the indices corresponding to the groups ofvariables considered in each case are calibrated. Thenetwork learning is made using a Kohonen network

(7)

where Ikj is the value of the index of the variablesgroup k recalculated considering a learning rate a, afunction with exponential decay, and the differencebetween the resulting index of the present evaluationand the previous indices in each damage level j. Thelearning rate is defined by

(8)

where t is the number of times that has been usedthe index that is calibrated. The damage evaluationsmade after the Quindío earthquake in Colombia(1999) were used for training.

Fuzzy rule bases for decision-making

The building habitability and reparability are assessedbased on previous results of damage level of thestructural and non�structural elements, the state ofthe ground and pre�existent conditions. The globallevel of a building damage is estimated with thestructural and non�structural damage results. This hasfive possible qualifications: none, light, moderate,heavy and severe damage. The global building stateis determined taking into account the rule bases onground conditions, and by this way the habitability ofthe building. The linguistic qualification for thebuilding habitability has four possibilities: usable,restricted use, dangerous and prohibited. They meanhabitable immediately, usable after reparation, usableafter structural reinforcement, and not usable at all.The building reparability depends besides on otherfuzzy rule bases: the pre�existent conditions. Thebuilding reparability has four possibilities: not any orminor treatment, reparation, reinforcement, andpossible demolition.

References

ATC, (1985). Earthquake damage evaluation data forCalifornia, ATC�13, Applied Technology Council, RedwoodCity, California.

ATC, (1989). Procedures for postearthquake safety evaluationof buildings, ATC�20, Applied Technology Council, RedwoodCity, California.

ATC, (2003). Users manual: Mobile postearthquake buildingsafety evaluation data acquisition system (Version 1.0), ATC�20i, Applied Technology Council, Redwood City, California.

FEMA, (1999). Earthquake Loss Estimation MethodologyHAZUS, Technical Manual, Vol. I, II and III, First edition 1997,National Institute of Buildings Sciences of Federal EmergencyManagement Agency, Washington.

Sanchez�Silva, M. and L. García, (2001). Earthquake DamageAssessment Based on Fuzzy Logic and Neural Networks, EERIEarthquake Spectra, Vol. 17, N. 1, February, pp. 89�11 2.Oakland, California.

Y.J. Park, A. Ang and Y. Wen, (1984). Seismic Damage Analysisand Damage�Limiting Design of R.C. Buildings, StructuralResearch Series, Report No 516, University of Illinois at Urban�Champaign.

How does your project address the linksbetween poverty and disaster risk reduction in developing countries?

The disasters generate poverty and the povertygenerates disasters. The constructions of the low�income socioeconomic layers are the most vulnerableand therefore they are more affected whenearthquakes happen. In this sense the benefits of thisproject are more for the people that, as result of asuitable damage evaluation, are protected in theirlives, avoiding they occupy buildings at risk, orprotecting their patrimony, avoiding the unnecessarydemolition of their houses when in fact they do notthreaten ruin. In any case, this project contributeswith an innovating tool, from the technical point ofview, that will be useful for all socioeconomic layersof a community affected by an earthquake, both indeveloping as in developed countries.

How will your project be sustainable beyondthese first six months?

The tool developed into this project is now acomponent of technical support within the Programof Building Damage Evaluation that AsociaciónColombiana de Ingeniería Sísmica (AIS) iscoordinating in the cities of Manizales and Bogotá(the latter in process of implementation, phase II).The AIS�400 Committee, related to Vulnerability of

12

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

( ) ( ) ( ) ( )[ ]kjkjkjkj ItIttI1tI −α+=+

( ) ( )t1.0exp1.0t ×−×=α

Buildings and Damage Evaluation will promote theprogram in other cities of the country in the nextyears. Also, together with the Advisory NationalCommission of Seismic and Volcanic Risk and theNational Commission for Earthquake ResistantBuilding Code, AIS will promote a unified nationalmethodology. This methodology of national level willinclude the Expert System, EDE developed in thisproject like a key tool for the preparedness, trainingand response of the engineers in support to theNational System for Disaster Prevention and RiskMitigation of Colombia, in case of an earthquakeanywhere within the country.

How will your project have ownership amonga broad community and institutional base?

Initially the project was developed as part of theProgram of Damage Evaluation developed forManizales city and now for Bogotá, under the direc�tion of the organizations in charge of the preventionand attention of disasters (or disaster risk manage�ment) in each city, with the technical support of AIS.Although the project was developed with anacademic scope, from the beginning it was proposedwithin the frame of the institutional action andwithin an existing program of preparedness for thedamage evaluation in case of an earthquake in a cityinterested in the subject (Manizales). Due to theoutcomes and advances, a second city (the capital)

13

was interested in the Expert System, object of thisproject, for its adoption as official tool to support thebuilding evaluations in case of an earthquake. Atpresent a phase of training is implemented andinvolves hundreds of engineers, several universitiesand numerous public organizations. In the two citiesa remarkable advance has been obtained that, like areplica, surely other cities and the national level willfollow.

What are the components of learning in your project?

The project corresponds to a highly technicaloutcome where advanced Computational Intelligence(CI) techniques have been explored, like the fuzzylogic and artificial neural networks. This tool wasdeveloped to face a real problem of seismic securityof remarkable complexity, such as the decision�making on habitability and reparability of buildingsaffected by earthquakes it is. The developed toolusing these techniques improves the effectiveness ofthe damage evaluation and risk of the constructions,avoiding that non�expert people evaluate in a wrongway the level of security of buildings. Decision�making about the possibilities of building occupationand reparation, by this way, is more reliable andquick, favoring the efficient management to attendthe community in case of a seismic disaster.

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

14

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

15

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

The main principles of

earthquake early warning

system creation around critical

facilities in seismic active zone

TEAM LEADER

Hayk Ktunyan<[email protected]>

TEAM MEMBER

Aroyan Gevorg

PROJECT ADVISORS

Dr. Valery ArzumanyanDr. Zaven Khlghatyan

Armenia

SYMPOS IUM NOTES

“National survey for seismic protection” agency emergency managementadministration under the government of the Republic of Armenia

Davidashen�Massiv 4, 375054Yerevan, Republic of Armenia

16

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

Before the Spitak (Armenia) 1998 destructiveearthquake the territory of Armenia was constructedwith buildings and structures designed for a groundacceleration of 0.1�0.2g corresponding to seismiceffect with an intensity of 7�8 by scale MSK�64,according to the seismic zonation map of that time.

However, the Spitak earthquake with a magnitudeM=7.0 and with intensity influence of 10 in theepicentral zone causing more than 25 thousanddeaths and the destruction of most buildings andconstructions in all Northern regions of Armenia,has shown that the level of seismic resistance ofbuildings and structures in Armenia is considerablylower than the level of seismic hazard. Taking thisfact into account the new seismic zonation map ofthe territory of Armenia was developed in theNational Survey for Seismic Protection (NSSP) RA in1995. This map was designed by a detailed analysisof the statistic of the region’s seismisity includingprehistorical, historical and instrumental periods.According to this map the level of seismic hazard hasbeen found to be I=9 with expected horizontalground acceleration of up to 0.4g in Yerevan (thecapital of Armenia).

Nearly the entire territory of Armenia is located inzones of high seismic risk. In addition, the presenceof Nuclear power plant, large chemical facilities anddams creates the threat of occurring technogenicsources and flood in case of strong earthquake.However, without doubt undoubted the mostdangerous zone is the territory around the Armeniannuclear power plant (ANPP).

The Armenian NPP is located in the southwest ofArmenia, 28 km from the Yerevan city, capital ofArmenia with 1.3 million population, 4 km from theMetsamor village and 16 km north of the Araks River.

The Armenian NPP was originally designed in 1969to withstand seismic loads of 7 bal. After 1977Vrancea (Romania) earthquake the seismic resistanceof structures of ANPP was increased to 8 bal, andsome constructions up to 9 bal. The Armenian NPPhas two units of 408 mwat (gross capacity) each.Both are WWR 440�270 Soviet type reactors, which isa version of the WWER 440�230 model. The units 1and 2 came into commercial operation in December1976 and January 1980, respectively.

17

During the Armenian NPP operation time theearthquake hazard at ANNP site was revised severaltimes:

• original site intensity, (0.1g)

• following 1977 Vrancea (Romania) earthquake re�evaluated site intensity (0.2g)

• following 1988 Spiak (Armenia) earthquake,re�evaluated seismic hazard, using the deterministapproach.

Peak ground accelerations (pga) obtained for twodifferent fractile values (50% and 84%) of those pgaare 0.21g and 0.35g respectively.

The last earthquake hazard assessment at the ANPPsite was performed on the basis of seismotectonicmodel which was scaled by Armenian NPP in 1994[Karakhanian at al, 1994]. According to that modelthe nearest two faults to the NPP site are wellexposed Araks active fault and Nearyerevan burieddeep fault, which activity at least in holozone notrevealed. Another remarkable peculiarity is thepresence of dissipating seismisity zone with M=5.5in the radius of 10 km from ANPP and the source ofthe strongest near�located earthquake.

The new attenuation model ofground motion for Armenia andadjacent area

Attenuation relationships play a vital role inearthquake hazard analysis and seismic design ofstructures. In recent years, a large number of investi�gations were carried out in this field and manyregional relationships between earthquake, magni�tude, source�station distance, local site conditionsand ground motion parameters such as peak accel�eration and spectral acceleration have been derived.

Based on the acceleration time histories recordedbetween June 1990 and September 1998 with thepermanent and temporary digital strong motionnetwork in the Caucasus and adjacent area anempirical PHA attenuation model for Caucasus areawas developed [6]. This attenuation model is the firstmodel in Caucasus region developed on the basis ofaccelerograms registered in Caucasus region.

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

The equation for larger values of peak horizontalacceleration is:

were pha is the peak horizontal acceleration in(cm/sec2), M is the surface�wave magnitude and D is the hypocentral distance in [km], p is 0 for 50percentile values and 1 for 84 percentile.

The empirical attenuation relationship is consideredto be valid for hypocentral distance between 4 kmand 230 km and surface�wave magnitude between 4 and 7.

The predicted peak ground motion accelerationvalues are in good agreement with correspondingmodels from Western�North America. Due to thecomplex structure of the Caucasus the scatter andabsorption of ground motion is somewhat higherthan in European areas.

Using this new attenuation model the expected peakground acceleration was estimated for ANPP site.

The performed calculations showed that expectedpeak ground acceleration at the ANPP site can reach0.329g, which corresponds to accepted now PGA forANPP site �0.35g.

Earthquake Early Warning System

It is internationally accepted that the Soviet designednuclear power plants require seismic upgrading tovarious degrees (Fraas et al., 1997). The InternationalAtomic Energy Agency (IAEA) has been carrying outassessments of seismic upgrading of the Sovietdesigned nuclear power plants over the past decade(Gulpinar et al., 1997). In the case where seismicupgrading by strengthening of the buildings andequipments is an expensive procedure and notfeasible due to economical, political and timingreasons, an active reactor protection system based onan earthquake early warning system like alternativeapproaches have to be taken into consideration.

The idea of alarm system is the following: to detectthe earthquake motion as early as possible near the source in order to prepare against the earth�quake before seismic motion reaches the site,using difference of transmission velocities, electriccommunication (300.000 km/sec) and seismic waves(~3.5 km/sec). Such systems can provide pre�warningtime of up to a few dozens seconds before the arrivalof the destructive ground motion.

The main concept of Early Warning System was firstlyintroduced by J. D. Cooper, M. D. Ou in San FranciscoDaily Evening Bulletin on 3rd November 1868. Thereport explained the concept as follows:

Since the Japanese magnet indicator has proved afailure, we are now obliged to look for some meansof prognosticating this fearful convulsion, and I wishto suggest the following mode by which we maymake electricity the means, perhaps, of savingthousands of lives in case of occurrence of moresevere shocks than we have yet experienced. It iswell known that these shocks are produced by awave–motion of the surface of the earth, the wavesradiating from a center just as they do in water whena stone is thrown in. If this center happens to be farenough from this city, we may be easily notified ofthe coming wave in time for all to escape fromdangerous buildings before it reaches us. The rate ofvelocity, as observed and recorded in Dr. J. B. Trask’swork in Earthquake in California from 1800 to 1864,is 61.5 (six and one fifth) miles per minute, or a littleless per hour (40 miles) than the tidal wave isreported to have traveled across the ocean to thisport from the Sandwich Islands or Japan.

A very simple mechanical contrivance can bearranged at various points from 10 to 100 miles fromSan Francisco, by which a wave of the earth highenough to do damage, will start an electric currentover the wires now radiating from this city, andalmost instantaneously ring an alarm bell, whichshould be hung in a high tower near the center ofthe city. This bell should be very large or peculiarsound, and known to everybody as the earthquakebell. Of course, nothing but the distant undulation ofthe surface of the earth should ring it. This machinerywould be self�acting, and not dependent on thetelegraph operators, who might not always retainpresence of mind enough to telegraph at themoment, or might sound the alarm too often. Assome shocks appear to come from the west, a cablemight be laid to the Farallone Islands, 25 milesdistant, and warning thus given of any danger fromthat direction (Fig. 1).

Of course there might be shocks the central force ofwhich was too near this city to be thus protected butthat is not likely to occur once in a hundred times.

18

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

( ) 21

22 5.4

28.000231.0log44.072.0log

+=

∗+∗−−∗+=

DR

PRRMpha

J. D. COOPER, M.D.

Figure 1. Concept of the Front Alarm by Dr. J. D. Cooper.

At that time, no system could realize this idea. After acentury, based on this concept the several EarthquakeEarly Warning System were constructed in the World.

They are:

Urgent earthquake detection and alarmsystem (UrEDAS) in Japan.

The special feature of this system is the rapid alarmupon detecting the earthquake using informationfrom P�wave data. By monitoring the earthquakemotion of a single observation point in real time, anUrEDAS detects initial P�wave motion, estimatesepicenter azimuth and magnitude, calculatesepicentral distance within about three seconds afterdetecting P�wave. System for different railways hasbeen in operation since 1983 (Nakamura, 1996)

Seismic Alert System (SAS) for Mexico City.

The seismic detector system has 12 digital strongmotion field stations along 300km of the Mexicancoast at 25 km spacing. This system consists of threeunits: seismic detection, telecommunications, andradio warning. The warning time in Mexico City variesbetween 58 and 74 seconds [2].

19

Seismic early warning system for Ignalinanuclear power plant.

This Seismic early warning system was installedrecently in Ignalina Nuclear power Plant (INPP) inLithuania [7]. Six seismic stations were installed in a ring centered at the plant at a distance ofapproximately 30 km, and the seventh station wasplaced in the plant. The stations are uniformlydistributed as shown in Fig. 2. Each consists of threeindependent substations, which are approximately500 m apart. The ground motion at each station ismeasured and recorded continuously by threeaccelerometers and seismometers. The data istransmitted via telemetry to the control center at INPP.

The time for emergency stop for RBMK reactors atIgnalina is only 2.5 seconds. The pre�warning time,provided by the seismic alarm system for theIgnalina NPP is 4 seconds. Therefore, the nuclearreaction can be stopped before the earthquakearrives.

Concept of ANPP Earthquake EarlyWarning System

Concept of ANPP Earthquake Early Warning Systemwas developed on the basis of Yerevan EEWSproject, considered in [1]. According to this project,the system can be based on 15 radiotelemetriccontrol points located by circular arrangementsaround Yerevan city at a distance of approximately30 km. According to this project, the warning timefor Yerevan city is about 3–6.5 seconds. It issuggested that the nets of accelerographs not onlybe used for EWS, but also for current monitoring ofweak earthquake.

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

Figure 2. Layout of earthquake early warning system of Ignalina Nuclear Power Plant (INPP) (YT: accelerometer, ST: seismometer)

This concept can be used for other industrial facili�ties, such as chemical plants, life�line systems etc.

The earthquake detection is based on S�waves. Inorder to provide the maximum pre�warning time theseismic field stations will be installed approximatelyin circular arrangements around ANPP. The distancefrom ANPP to the field stations should be optimal.It cannot be too short, because in this case the timeof the alarm will not be enough to initiate protectiveactions. And it also should not be too long for thefollowing reasons: a lot of seismogeneous zones willbe out of range; second, due to large perimeters ofthe location of the field stations in all directions.

Taking into account the location of seismogeneouszones around ANPP, as well as above mentionedarguments, it is useful to set optimal distance to 25�35 km. From this point of view, the ten fieldstations are required. Due to the availability ofseveral monitoring sites around ANPP, eight of them

will be located at existing NSSP telemetric stations.Another two stations will have to be new installationpoints (Fig. 3).

The existing monitoring system applies relay stationsat several locations. These will also be used for EEWS.At the two new installation points, necessity of relaystations should be investigated. One accelerometerper station will be required. The principle of the alarmshould be based on positive reaction of at least twoneighboring stations in a given temporal window � in order to exclude the possibility of a false alarm incases of equipment failures.

In a setup with ten stations, the distance betweenthe field stations is approximately 15�20 km. Theprobability is extremely low that ground shaking fromsources other than seismic events, e.g. from railwaytraffic, construction works or other man�made events,affect two stations at the same time at a distance15�20 km.

20

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

Figure 3. Map of existing telemetric stations around ANNP and new installation points

ARUCH – Aruch

AKSZ – Araks

ANPPZ – ANPP

ARRZ – Ararat

EREZ – Yerevan

GNIZ – Garni

KAPZ – Kaputan

PARZ � Parakar

VNNZ – Vanand

VRNZ – Vardanashen

21

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

Each field station will include an accelerometers anda digitizer (Fig. 4).

Figure 4. Simple scheme of investigation station

The digital signal will lead to the differentiationsoftware processor. The special software willdifferentiate each event on seismic event and non�seismic event. One factor in this context is thefrequency. Above 10Hz, a non seismic event isindicated, at lower frequencies it may be an earth�quake. For seismic events the frequency peak must bebetween 2 to 10 Hz. The Fast Fourier Transformation(FFT) provides valuable information about frequencycontent of a seismic event. If the FFT shows adominant frequency within the predeterminedbandwidth, a switch is set to FFT: Yes.

Seismic events with frequencies beyond 10Hz arepossible. However, these are minor events, which donot cause any problems for the nuclear power plant.

The second factor is Cumulative Absolute Velocity(CAV). The CAV algorithm is applied to the seismicdata, which is filtered between 2 to 10Hz. If, within1 second, the data exceeds a predefined level, theacceleration value is multiplied by 1 second(resulting in a velocity) and added to the CAV sum. IfCAV exceeds a predetermined value, a switch is setto CAV: Yes.

For this purpose, the Seismic Alarm System Software,SEISDIFF, which was conceptually developed by EWE�GeoSys, should be useful. The time required bySEISDIFF to detect a relevant Seismic event is1second following the arrival of the S�wave at thefield station. Since the event is checked fromdifferent points of view, the reliability of the alarm isimproved and false alarms are eliminated. Thisincreases the reliability of the Seismic Alarm System.

As to threshold, acceleration values above which theinstalled accelerographs at control points mustrespond should be estimated taking into accountthat at ANPP site acceleration should be more than0.1g. By the studying the seismic waves attenuationwith the distance, as well as ground condition atfield points we estimated the threshold value ofinstruments on control points. It is equal to 0.15g.

Finally, about the mean time of the alarm. The simpleestimation including the configuration of EEWS andconsidering velocity of seismic shear waves of about 3.5 km/sec show that the earthquake will bedetected 7�10 seconds before the arrival at the plant site. Differentiation of each event to seismic eventand non�seismic event requires 2 seconds, also 1�2seconds will be lost on receiving the alarm signalsfrom the two neighboring field stations. Thereforethe pre�warning time should be 3�7 seconds.

Application to WWER-type reactors

The WWER Soviet built nuclear reactor is pressurizedwater reactor. The emergency shutdown of theWWER nuclear reactor is characterized by anincreased insertion time of the control rods up to12 seconds. The pre�warning time provided byAEEWS may not be sufficient for the control rods to reach their lower position in the reactor core. Butit will be sufficient to issue control rods insertionsignal and to release them from the positioningmechanism. The control rods may not have reachedthe lower position in the reactor core before thedestructive earthquake arrives at the site, but thedropping down of the rods is at least in progress.

22

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

Conclusion

The concept of Earthquake Early Warning System forANPP is developed. In this system, the earthquakedetection is based on S�waves. The seismic fieldstations will be installed approximately in circulararrangements around ANPP. The pre�warning timeprovided by AEEWS is about 3�7seconds, that maynot be sufficient for the WWER type nuclear reactorcontrol rods to reach their lower position in thereactor core, but it will be sufficient to issue controlrods insertion signal and to release them from thepositioning mechanism, and dropping down of therods will be at least in progress. The principle of thealarm should be based on positive reaction of atleast two neighboring stations in a given temporalwindow� in order to exclude the possibility of a falsealarm in case of instrument failures. More than onealgorithm for differentiation between seismic andnon–seismic events will be used. This concept alsocan be used for other industrial facilities, such aschemical plants, life�line systems and others.

References

Balasanian S., Nazaretian s. et al (1997) The New SeismicZonation Map for the Territory of Armenia. Natural Hazard 15,Kluwer Academic Publishers, Netherlands, pp 231�249

Balassanian S., Martirossyan A., Arsumanyan V., (1998) Projectof Creation of an Earthquake early warning system forArmenia. International IDNDR conference on early warningsystem for the Reduction on natural Disasters (1998; Potsdam,Germany) Early warning system for natural Disaster reduction(2003) (edited by J. Zchau, A. Kuppers)

Espinosa – Aranda G.M. et al (1996), Results of the Mexicocity early warning system, Proc. 11�th world conference onEarthquake Engineering, Acapulto, Mexico

Fraas K. et al. (1997), Seismic re�evaluation and upgrading ofBohunice NPP, Proc. SMIRT 14 Post conference seminar no.16, Vienna

Gurpinar A., Zola M. (1997), Overview and generalcomparison for Paks NPP, Proc. SMIRT 14 Post conferenceseminar no. 16, Vienna

Nakamura Y. (1996), Real�time information systems for hazardmitigation, Proc. 11�th world conference on Earthquakeengineering, Acapulco, Mexico.

Smit P., Arsumanyan V., Gavakhishvili Z., Arefiev S., MayerRosa D., Balassanian S., Chelidze T. (2000). The digitalAccelerograph Network in the Caucasus. Kluwer AcademicPublishers. Earthquake Hazard and Seismic Risk Reduction(edited by S. Balassanian, A. Cisternas, M. Melkoumian).

Wieland M. et al (1997), Seismic alarm system for IgnalinaNuclear Power Plant Proc. SMIRT 14 Post conference seminarno. 16, Vienna, August 1997.

Willy H. K. Lee, Juan Manuel Espinosa�Aranda EarthquakeEarly Warning Systems: Current Status and Perspectives Earlywarning system for natural Disaster reduction (2003) (editedby J. Zchau, A. Kuppers)

23

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

Seismic conservation of

historical and cultural treasures

of a world city: sizing the need

and formulating an action plan

for the museums of Istanbul,

Turkey

TEAM LEADER

Bilgen Sungay <[email protected]>

TEAM MEMBER

Nevra Erturk

PROJECT ADVISOR

Marla Petal

Turkey

SYMPOS IUM NOTES

24

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

Nevra Erturk, Research Assistant, Yildiz Technical University,Faculty of Art and Design, Museum Studies Graduate Program

Bilgen Sungay, Project Development Specialist, BogaziciUniversity, Kandilli Observatory and Earthquake ResearchInstitute, Disaster Preparedness Education Program

NOTE: Authors Nevra Erturk and Bilgen Sungay formed the project imple�mentation team. The team is advised by mentor, Marla Petal, Coordinator,Bogazici University, Kandilli Observatory and Earthquake Research Institute,Disaster Preparedness Education Program. The project team is grateful forthe valuable contribution of Republic of Turkey, Ministry of Culture andTourism; Republic of Turkey, Istanbul Provincial Directorate of Culture andTourism; Dr. Jerry Podany, Conservator, John Paul Getty Museum; Prof. TomurAtagok, Chairperson, Yildiz Technical University, Faculty of Art and Design,Museum Studies Graduate Program; Prof. Dr. Gulay Barbarosoglu, Director,Bogazici University, Kandilli Observatory and Earthquake Research Institute;Prof. Dr. Mustafa Erdik, Chairman, and Assoc. Prof. Eser Durukal, EarthquakeEngineering Department, Bogazici University, Kandilli Observatory andEarthquake Research Institute; Mr. Suha Ulgen, Interactive Media andGeographic Information Systems Inc.; Ms. Suheyla Sezan from DisasterPreparedness Education Program, Bogazici University, Kandilli Observatoryand Earthquake Research Institute; Mr. Kevin Marshall, Preparator, J. PaulGetty Museum; Mr. Omur Tufan, Museum Professional, Topkapi PalaceMuseum; directors and museum staff of all Istanbul Museums andProVention Consortium Disaster Risk Reduction Program.

The exposure of the Marmara and Aegean regions ofTurkey to a major and devastating earthquake in thenear future is a scientific fact. An earthquake will putthe rich and irreplaceable cultural heritage of worldcivilizations which are exhibited and stored inIstanbul Museums, at great peril. The tourism sectorin the Marmara Region is largely dependent on theintegral part of the world cultural heritage andcultural tourism. Protecting the tourism sector of theeconomy involves disaster preparedness education,and business resumption planning. Most importantlyit involves seismic mitigation of the collectionsthemselves. Disaster preparedness is needed for theprotection of museum visitors, staff and the museumcollections.

OBJECTIVES

This project aimed to make the knowledge aboutdisaster preparedness focusing on non�structuralmitigation more widely available in order to savelives and prevent injuries of museum staff andvisitors; to preserve our cultural heritage for futuregenerations; to protect business continuity in thetourism sector and to assist this sector in prioritizingand developing practical non�structural seismicmitigation action plans.

MILESTONES

The project accomplished the following:

1. Compiled examples of hazards and best practices

2. Prepared educational electronic slidepresentation explaining non�structural hazardsand mitigation methods for museum collectionsboth on display and in storage. Two hundred andfourteen slides covered the following topics:

• What Happens During An Earthquake?

• What Is Non�Structural Mitigation?

• Principles for Non�Structural Mitigation

• How Are Objects Damaged?

• Reducing Risk in Exhibits

• Reducing Risk in Storage

• Reducing Risk in Public Facilities, Offices andLibraries in Museums

• Where Can We Start?

• What Is Being Done in Istanbul Museums?(Green et al., 2003; Marshall; Podany,2001a; Podany, 2001b; Podany, 2001c)

25

Topkapi Palace Museum, Photograph: T Mimarlik Dekorasyon.Taahhüt San. Ve Tic. Ltd.

Sadberk Hanim Museum, Photograph: B.U., K.O.E.R.I., DisasterPreparedness Education Program

3. Held a seminar for museum directors and staffwith sixty�one participants from 31 museumsand organizations. The slideshow was shared bythe project team (see note previous page) at theseminar.

4. Developed Non�structural Hazard Survey Forms

• Museum Information Form to collectinformation about management and budget,museum building, museum collection,disaster experience and preparedness ofeach museum.

• Rapid Room Survey Form to quantify non�structural mitigation needs.

• Rapid Room Survey Summary Form

• Object Risk Identification Form to tackleproblem�solving and decision making foreach object. (Marshall; Podany, 2001a;Podany, 2001b; Podany, 2001c).

5. Museum visits: Exhibit areas in fourteenmuseums were visited and surveyed, andstorage areas in six museums were surveyedamong 50 museums in Istanbul to test surveyforms and to quickly identify and quantify risksand develop potential approach. These museumswere selected according to the following criteria:

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

• The institutions they work under: Museums’management; budget, staff and technicalpossibilities differ according to theinstitutions they work under.

• Collection content: Museums that havedifferent type of collection content, havedifferent needs in the field of non�structuralmitigation.

• Museum type: A variety of types of museumssuch as palace museums, monumentalmuseums, house�museums have differentexhibition and storage conditions.

• Number of visitors: The most visitedmuseums within their directorate will bethose whose contribution to the localeconomy seems significant (Atagok, 1999).

6. Data analyzed and report prepared: A full reportwas prepared for the Ministry of Tourism andCulture, and disaster mitigation advocates.Recommendations include rapid implementationof easy non�structural mitigation measures and aseries of short and long�term steps to developmore expertise and research�based solutions inthis area.

PROJECT RESULTS

Identifying non-structural risk in museums

Risks

Non�structural risks were evaluated from theperspective of the potential risks to visitors, staff, andthe collections themselves. The most common risks atthe museums are as follows:

Risk of broken glass• Window and door glass• Showcase glass• Mirror glass• Balustrade and elevator glass.

Risk of free�standing objects and riggings on the floorFree�standing objects and riggings on the floor,which are taller than they are wide (or deep),are at risk of overturning. Shorter or widerobjects may slide. Large objects that have beenfastened only from their bottom or objectsfastened on a free�standing unsecured base orrigging (esp. bust fastened on pedestal) are alsoat risk of overturning These are:

• Showcases• Free�standing objects on the floor or

standing on unsecured base • Humidity controllers and air conditions • Fire extinguishers• Footed storyboards• Furniture (for instance; bookcases, buffets or

tables at palace museums etc.)• Computers for visitors’ usage• Barriers• Folding screens

Risks within showcasesIn addition to fastening the showcasesthemselves, the objects within the showcasesneed to be stabilized and secondary dangersalso need to be mitigated. These risks are:• Objects falling by overturning, sliding, slipping

out from their places or hitting of hangingand swinging objects,

• Objects hitting into each other because ofcrowding

• Pedestals, mannequins or mounts used toexhibit/fasten that are not fixed to theshowcase base may fall or slide,

• Glass or panels in the showcase’s ceilingunder the lighting fixture may break,

• Fluorescent light bulbs may fall,• Glass shelves within the showcase may break.

Risks of hanging objectsObjects hanging on walls or from a case orceiling can swing and hit other objects or slipout of their places and fall. This risk takes threeforms:• Unsecured objects hung on the wall

(eg. with only one nail)• Objects hung from ceiling with open hooks,• Objects which are secure, but can hit other

objects if they swing.

Risks from the ceiling of the buildingPiece or objects, which can fall from the ceilingof the building by breaking off, can damageuncovered objects below. These are as follows:• Plaster relief pieces on ceiling breaking and

falling off (esp.; plaster relieves at palacemuseums),

• Pipes falling from above or pipes cracking andreleasing water,

26

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

• Lighting fixtures falling, sliding or hittingobjects,

• Suspended ceiling pieces falling and hittingobject,

• Roof window breakage. (Marshall; Podany,2001a; Podany, 2001b; Podany, 2001c).

Forms

The project team developed three survey forms thatare aimed to help museum staff in quantifying non�structural risks in the museums.

Method

The method used in the forms and during the on�sitesurveys at the museums begins with categorizing themeasures into 3 categories; easy, medium and hardto apply. These categories also correspond roughly tocost of application. The reason for this is our belief inthe general principle that anything that can be doneeasily and inexpensively should be tackled as a highpriority as there are few barriers to safety, beyond adecision to act.

Easy methods are considered to be low�cost and canbe easily�applied. These methods include:• Using museum wax, monofilament and steel wire

to fasten objects to horizontal or vertical surfaces,to reduce risk of tipping and falling

• Using metal hooks on objects that are hung onwalls, or from ceilings to reduce risk of falling

• Placing sand bags inside the objects in order toreduce the risk of toppling

• Placing rubberized shelf mats under small objectsto reduce risk of sliding

• Using mechanical latches to prevent cupboarddoors from opening

• Placing restraints on open shelving to reduce riskof objects falling.

Medium methods are considered to be ones that costsomewhat more, are more time consuming andrequire more labor power. Some methods may alsobe in this category because they require specialpermission to accomplish. These include:• Fastening objects to surfaces using specially�

produced mounts• Using padding between objects• Preserving objects in storage in boxes or

containers• Covering glass surfaces with security film• Bolting or screwing objects to surfaces in order to

reduce risk of toppling

27

Difficult methods are considered to be those that areexpensive, require special production or may be veryhard to find an appropriate solution due to thedifficulty in fastening to a wall of a historicalbuilding or the aesthetic concerns, etc. Thesemethods include:• Designing & producing or buying a new shelving

system• Producing a base isolation for the particular object• Producing special solutions for objects that have

special conditions

The project team spent time with museumprofessionals to verify the logic of this categorizationsystem. While there is general consensus, thedecisions made during a rapid assessment arenecessarily subjective. The project team found that itwas possible to spend time discussing the dilemmasposed by any one object, and many considerationsbeyond time and money might also influencedecisions about a particular method, or applicationof a method. Some of these are: aesthetics, adjacentobjects, available skills and materials, medium andlong�term plans for the exhibit, the frequency ofexhibit change, the durability of method, and soforth might all.

During the research on non�structural risk identi�fication in museums, the project team’s purposewas to conduct a rapid survey. Rather than to makefinal decisions about mitigation methods, theobjective was to gain an overview of needs. It isrecommended therefore, that each museumundertake this process using it’s own staff bothfor rapid survey and prioritization as well as for

making individual decisions about the stabilizationmethod to be used with any particular object orgroup of objects.

How to use the formsStep 1 – Form 1A:RAPID ROOM SURVEY FORM

The aim in using Form 1A is to quickly and easilyseparate the objects that need to be mitigated intoeasy, medium or difficult methods. The reason forthis is to focus on the easy methods that can berealized quickly and help taking action.

By looking at the room as a whole, and consideringthe range of objects, materials and display methods,a pre�classification and sifting can be done byreferencing the Rapid Room Survey form. Makingthis pre�classification does not require looking at theobjects separately. Instead, museum specialists can

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

easily decide the category of the mitigation methodfor groups of objects in general, and then quicklycount the objects in each group.

One could generally determine the size of the objector material for the mitigation method by using thecolumns “small, medium and large” on same form ifdesired. The logic of this is that larger objects requiremore material to create a fastner, and that thisseparation would further help in estimating the costof mitigation.

Step 2 – Form 1B:RAPID ROOM SURVEY SUMMARY FORM

After surveying and collecting information using Form1A in every room, Form 1B can be used to summarizethis data, consider the results museum�wide anddetermine the approximate cost for the mitigationmethods. This is a tool for museum administrators touse in planning and budgeting.

Step 3 – Form 1C:OBJECT RISK IDENTIFICATION FORM

This form can be used after Form 1A and/or Form 1B,to help in making decisions about how to secureobjects that can not be secured by an easy method.When medium or difficult methods are required itbecomes important to examine each objectindividually. Form 1C can be used in two differentways:

• Only the upper part including the sectionsnamed “Photograph” and “Notes on themitigation method suggested” can be used. Itis important to write the inventory number, thename and the place of the object in order to beable to identify particular objects among manyforms filled. If a museum specialist is able todecide what kind of method is needed tosecure the object and what the priority level isat first glance, then it would be enough to fill

the upper part only. However, the photographof the object with its surrounding and asketch/notes on what kind of method isthought to be appropriate for that object,would help a team in considering the optionsand deciding on details. It would also aidmount�makers in reviewing and planning thework to be done.

• Using the whole form: If museum specialists orteams have trouble in assessing the risk to theobject, or how to approach mitigation, thescoring system at the lower part of the formcan be helpful both in deciding the priority level of the object and in specifying itsvulnerabilities. The questions investigatesubjects like physical condition of the object,the possibility of toppling, and secondarythreats. In the scoring part at the end, thecheckmarks in each column are counted, andthe priority level set based on the highest score.When two scores are very close to each other,the higher risk level should be accepted.

In the end, these decisions will be subjective. It may change according to the people working on the subject, and may vary from situation tosituation.There may be places where an easy method can turn into “hard to apply” because of avariety of challenges. The scoring is offered only tohelp professionals become conscious of the variablesand considerations, and as consistent as possible indecision making.

It is not easy to create a standard under thesecircumstances. Comments on the use of the formsvary from museum to museum. Museums that are big in size and have many objects mention that the forms may be useful, while on the other hand,smaller museums mention that it may not benecessary to use such forms as decision�making and action may more easily be controlled.

28

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

29

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

30

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

31

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

32

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

REDUCING NON-STRUCTURAL RISK IN MUSEUMS

Challenges facing non-structuralmitigation work at museums

Some challenges are faced when undertaking non�structural mitigation work for the protection ofpeople and objects. They differ from museum tomuseum according to their administrative status,physical conditions, earthquake consciousness of themuseum staff and their collection content.

Some of the museum buildings in Istanbul arehistorical, therefore special permission is needed tofasten objects to the walls, ceilings and floor wherethese are assumed to alter or damage the structure.The temperature and humidity control are alsodifficult in historical buildings and extra space cannotbe added. However, extra space is mostly needed asthe collections are very large in most of the museumsand the number is constantly increasing. The increasein the number of objects also cause changes at theexhibition and storage that also increases the cost oftaking measures and the time period for takingmeasures. On the contrary, it is also difficult to obtainapproval for decreasing the number of exhibitedobjects to prevent overcrowding in showcases.Decreasing the number also requires extra space instorage areas which as it is discussed, is not apossible solution for many museums under theexisting circumstances.

Some of Istanbul Museums also face somebureaucratic difficulties. For instance, when themanagement changes, the agenda may also changemodifying the priorities of the work to be done in themuseum. Additionally, it may require time to obtainapprovals for taking specific measures from theadministrators of the institutions they are under.

In most of the Istanbul Museums, the limited budgetis making actions like obtaining materials or assuringstaff to focus on non�structural mitigation workdifficult. The existing staff are overburdened and haveinsufficient time for any additional responsibility inmost of the museums. Therefore, there is a need foradditional staff who will focus working on the non�structural mitigation work. They also should beconscious and qualified professionals. However, thesecriteria limits to find qualified staff as the subject isyet new. It should be mentioned that there is a lackof volunteer system in most of the Istanbul Museums,which could help with duties of the existing staff.

There are also some aesthetic worries in takingmeasures. Some of the museum professionals believethat the measures taken change the value of theobject as it is an addition to it. Additionally, museumstaff is suspicious whether the methods woulddamage the objects and thus there is a need toresearch and design on the field which will taketime. The measures will be subjective as they may bedesigned in several different ways according to thepeople working on it.

From the functional point of view, there are concernslike; necessity of taking measures in storage areasleaving the objects visible so that they are easilyreachable when the researchers need access ormaking custom mounts not specific only to oneobject so that when there is a change in exhibition,it could fit to hold another one. On the other hand,mitigation measure work may make changes ofexhibits and periodic care more time�consuming andit may pose additional risks to objects whilemounting and unmounting. Museum staff would liketo be sure about the specifications and limits ofcertain methods as there is a lack of research dataor engineering knowledge.

There are also psychological and social worriesabout the subject. There are question marks like;whether the people’s morale will be brought down,if earthquakes are always of concern or whether themuseums will get the support they need from thecommunity to undertake these measures. At some ofthe Istanbul Museums, we should emphasize thatthe necessity in strengthening the museum buildingstructurally receives as a competing priority.

It is very important to form the earthquakeconsciousness among different groups in the societyto be able to handle non�structural mitigation workand all of the challenges to be faced along the way.

Suggestions

For the existing situation, although there is mostly aproblem with limited budget, it is possible to startwith cheap and practical non�structural mitigationmethods immediately. This action would further helpto focus on the more complex methods afterwards. Itis observed that there is relatively less difficulty inmitigating storage than exhibits, mostly because themitigation measures do not need to includeaesthetic considerations. Therefore, some exhibitsmight be protected by decreasing the existingnumber of same type of objects on display, andboxing these safely in storage, thus rememberingthat it is important to organize storage areas in a

33

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

way that the objects can be seen and identified easilywhich will provide easy access to the objects duringresearch or display changes. It is also important toorganize storage areas to be able to have easy andquick access in emergency situations. Additionally,specially�designed storage facilities protecting objectsfrom fire and water as well as earthquake areextremely important.

There is a need to employ assistants to the curator,conservationist, restorationist, carpenter, etc. to focusworking on non�structural mitigation againstearthquake.

Mitigation efforts should be taken againstearthquake not only in the exhibition galleries andstorage areas, but also in offices, museum shops,exits, corridors, halls and other public service spaces.Measures should be checked periodically and shouldbe continuous and routine. Additionally, special effortshould be given during exhibition changes andcleaning. The quality of both application and materialused are very important for both efficiency andeffectiveness. Knowing which methods areappropriate for which objects is also important for the efficiency of the application.

For future applications, it is very important to design new museums bearing earthquake risks ofboth structural and non�structural components inmind thus avoiding the cost of doing this as anafterthought. Museums need new and wide storageareas both to store the existing objects in betterconditions and to use for salvage operations after a disaster.

The protection of objects against earthquake needsto become a routine consideration at the beginningof acquisition, exhibition, and storage becoming ascommon as protection from fire and theft. When theexpense of non�structural mitigation work become aline�item in the budget, it will speed up andencourage the necessary works to be done.

Standard Museum Emergency Plans should beadapted for each museum. Emergency planning isneeded to develop safe and efficient procedures forevacuation of people and objects. Museum staff andtour guides should be trained in basic disasterawareness, structural and non�structural awareness,and community emergency response and the planshould be practiced in all museums.

It is important to publish the research and workrealized in this field in order to be able to put the

subject on the agenda of scientific field and of publicopinion. This will also increase the potential foridentifying and receiving financial support for thisundertaking.

CONCLUSION

The need

In conclusion the project team’s experience doingthis research has led a number of points that may beimportant in accomplishing non�structural mitigationboth in Istanbul’s museums and in other places.

It seems important to develop some interdisciplinaryteams that can undertake the extensive initial effortsof non�structural mitigation not for each museum in isolation, but across museums, in a consultativecapacity. There is a clear need for professionals whowill specialize in the subject of non�structuralmitigation planning, problem�solving, training andimplementation, and will accumulate and share thevery specific knowledge needed for the future. Suchgroups working with each museum should, of course,include at least one lead member of the staff withthe interest and authority to coordinate museum staff participation. Small teams within the museum,working with guidance and support from outsideteams dedicated to this task may be an effective wayto approach this comprehensive task. There are anumber of distinct tasks that might be addressed bydifferent multi�disciplinary teams:

• Structural safety investigation of museumbuildings: The need for triaging buildings whichmay be structurally sound, may require modest orextensive retrofitting, or may in fact not be anadvisable place to locate treasures.

• Non�structural mitigation assessment andplanning: Risk identification, prioritization,budgeting, and action�planning.

• Scientific research on methods and materials fornon�structural mitigation: This research shouldinclude both materials testing for safety andpreservation of objects and shake�table testing andengineering calculations to determine: applicabilityand size/weight and configuration limitations ofexisting and new materials and techniques (incollaboration with exhibition designers).

• Implementation of non�structural mitigationmeasures: This includes work in exhibits andstorage, mount�making, boxing, packing and otheractivities.

34

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

• Education: Update and deliver educationalmaterials and programs, especially in relation toresearch on the subject worldwide.

• Public relations, advocacy and promotion:Identification of sponsors and champions who willtake on this mission until it becomes part of theongoing work of all museums in seismic risk areas,and ongoing research.

Regular and repeated training programs for people invarious roles should include:

• Education to raise public awareness of thevulnerability of cultural heritage to naturaldisasters.

• Education within tourism sector in order to raiseinterest and support from industry leaders andworkers,

• Education of museum directors and staff, bothdecision�makers and project implementers tosecure their understanding and commitment,

• Education of students, who study in relateddepartments of universities (museologists,earthquake engineers, interior designers, architects,archeologists),

• Education of trade school students, skilledcraftspeople and restorationists who will work onmount�making, packing for storage, and otherapplications. These individuals should be carefullyselected for their meticulous care, maturity, andtheir awareness of the gravity of this responsibility.

The project team recommends the following short�term projects:

• Preliminary research can be given to students fromrelated departments of universities as homework,projects or theses.

• The technical specifications and limits of existingmethods should be experimentally researched withshake�table testing. New methods should also beinvestigated in this manner.

• Sources for purchasing non�structural mitigationmaterials in local markets should be researchedand chemical content of these materials should betested as necessary at the Directorate of theCentral Laboratory for Restoration andConservation.

• A multi�disciplinary group of people from variousfields (museum studies, earthquake engineering,architecture, chemical engineering, etc.) andvolunteers who want to become specialists in thissubject is to be identified to support and to

participate in local and international trainingprograms.

• Basic non�structural mitigation training availablefrom Bogazici University Continuing EducationDepartment and Kandilli Observatory andEarthquake Research Institute can be madeavailable to all professional museum staff(administrators, curators, restorationists,preparators, conservationists, designers, etc.)

• A mobile training and mount�making unit can beestablished to tour museums and provide on�siteconsultation and training on simple mitigationtechniques and acrylic mount�making.

• One or two museums can be selected asdemonstration sites to implement and showcasecomprehensive non�structural mitigation,emergency planning, and staff training.

• A variety of national and internationalgovernmental, non�governmental and privateresources must be called upon to support thisimportant work.

The project team recommends the following long�term approach to assure ongoing leadership andattention to these tasks:

An institute, which will concentrate inter�disciplinaryinterest and expertise and workspace for research�development, education, consulting, technicalsupport, specialized and temporary storage, andsalvage operations after disaster can be created forlong�term systematic implementation and extensionof these lessons throughout Turkey and the region.

Achievements and the sustainability of the project

The need for a broad�base of ownership has beenaddressed by the development of several importantnew collaborative efforts.

• Several Istanbul Museums have already taken orbegan to take impressive measures on non�structural mitigation against earthquake and arecontinuing to be the leaders and the advocates forthe action. These museums have provided projectteam with information and are continuing toencourage other museums by sharing informationand provide collaborative impetus in applyingnon�structural mitigation measures, emergencyplanning and education.

• Bogazici University, Kandilli Observatory andEarthquake Research Institute and Yildiz TechnicalUniversity, Faculty of Art and Design, Museum

35

WO

RK

ING

PA

PE

RS

Ea

rth

qu

ak

es

Studies Program directors have met and statedtheir commitment to working together on futureprojects. We are going through the process ofdeveloping official agreements with otheruniversities to create wider collaboration.

• Two training programs in non�structural mitigationfor museum collections have been integrated intolessons on Maintenance & Conservation andCollection Management at Yildiz TechnicalUniversity, Faculty of Art and Design, MuseumStudies Graduate Program.

•A slide presentation about non�structural damagemitigation in museums has been prepared in bothEnglish and Turkish, and will be shared overinternet via web�site of Bogazici University, KandilliObservatory and Earthquake ResearchInstitute(KOERI), Disaster Preparedness EducationProgram (AHEP); www.ahep.org.

“Seismic Conservation of Historical and CulturalTreasures of a World City: Sizing the Need andFormulating an Action Plan for the Museums ofIstanbul, Turkey” is a pilot project implemented inthe field of cultural heritage protection in connectionwith museology and earthquake preparedness. Thisproject’s results are believed to provide a basis forthe forthcoming projects and have importantinfluence for museums in Turkey and all developingcountries throughout the world.

References

Atagok, Tomur (Comp.), Yeniden Muzeciligi Dusunmek.Istanbul: Yildiz Teknik Universitesi Yayinlari, 1999.

Erturk, Nevra; Sungay, Bilgen and Marla Petal, SeismicConservation of Historical and Cultural Treasures of a WoldCity: Sizing the need and Formulating an Action Plan for theMuseums of Istanbul, Turkey, Final Report to World BankProVention Consortium, January 2004.

Green, Rebekah; Caliskan, Omer; Sungay, Bilgen and ZaraPaci, Non�Structural Mitigation Handbook. Istanbul: Bo¤aziçiUniversity, Kandilli Observatory and Earthquake ResearchInstitute, 2003. available at: www.ahep.org

Green, Rebekah; Caliskan, Omer; Sungay, Bilgen and ZaraPaci, YOTA Yapisal Olmayan Tehlikelerin Azaltilmasi El Kitabi.Istanbul: Beyaz Gemi Yayinlari, 2003.

Green, Rebekah, Caliskan, Omer, Sungay, Bilgen and ZaraPaci, YOTA Yapisal Olmayan Tehlikelerin AzaltilmasiPowerpoint Sunumu. Istanbul: Bogazici Universitesi, KandilliRasathanesi ve Deprem Arastirma Enstitusu Afete HazirlikEgitim Projesi, 2003.

Marshall, Kevin L., Slide presentation entitled Packing forStorage in Seismically Active Regions. J. Paul Getty Museum.

Podany, Jerry, Muze Koleksiyonlari Icin Afet Hazirliklari.Istanbul: Bogazici Universitesi, Kandilli Rasathanesi ve DepremArastirma Enstitusu, Istanbul Afete Hazirlik Egitim Projesi,2001a.

Podany, Jerry, Muze Koleksiyonlari Icin Sismik GuvenlikCalismasi. Istanbul: Bogazici Universitesi, Kandilli Rasathanesive Deprem Arastirma Enstitusu, Istanbul Afete Hazirlik EgitimProjesi, 2001b.

Podany, Jerry, Slide presentation entitled DisasterPreparedness for Museum Collections. J. Paul Getty Museum,2001c.

T.C. Kultur Bakanligi, Turkiye Müzeleri. Ankara: T.C. KulturBakanligi Anitlar ve Muzeler Genel Mudurlugu, 2002.

Tufan, Omur, Slide presentation entitled Japon Muzeleri’ndeDepreme Karsi Alinan Tedbirler, Topkapi Palace Museum,October 2001.

36

WO

RK

ING

PAP

ER

S

Ea

rthq

ua

ke

s

37

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Senegal

SYMPOS IUM NOTES

Earth observation and

GIS-based flood monitoring

in the Senegal River Estuary

and Valley

TEAM LEADER

Aliou Mamadou Dia <[email protected]>

PROJECT ADVISOR

Dr. Alioune Kane

38

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

The Senegal River basin

The Senegal River basin is located in West Africa and occupies an area of roughly 300 000 km2. Theentire basin, including the upstream catchments isdrained by the 1700 km long Senegal River and itstributaries. The area of the river basin accounts forabout 1.6% of the African continent and lies withinthe territory of four different countries: Guinea, Mali,Mauritania and Senegal (Figure 1).

Hundreds of thousands of people live along theSenegal river which rises in the Fouta Djallonmountains of Guinea and flows north towards theedge of the Sahara desert before swinging west toempty into the Atlantic Ocean. Its water is usedirrigate large areas of rice, sugar cane, maize andtomatoes as well as grazing land for livestock. Thebasin is divided into three distinct regions: the upperbasin which lies in the mountains of Mali andGuinea, the middle valley which forms the 500 kmlong borderline between Senegal and Mauritania,and the delta in the lower valley where the SenegalRiver discharge into the Atlantic Ocean.

The flow rate of the river is determined mainly by therainfall in the upper basin. The high�water seasonlasts from July to October; the low�water season,with a steady decrease in volume, begins inNovember and lasts until May or June. The high�water season peaks at the end of August orbeginning of September and quickly ends duringOctober.

Another important feature of the Senegal River priorto the construction of its two dams was the inter�annual irregularity in its flow volume. For a long time

39

this inter�annual flood irregularity posed a majorproblem for the valley, as it decreased the potentialfor guaranteed agricultural production in this narrowgeographic area. The arable land area that couldeffectively be farmed after the flood could varybetween 15,000 ha and 150,000 ha, depending on the magnitude and duration of the flood.Exceptionally high water levels caused widespreaddevastation in 1890, 1906, 1950, 1994, 1999 and 2003.

The particularly low water level during the dryseason resulted in a deep intrusion of the ocean’ssalted waters into the riverbed. During the 1970s asaltwater wedge penetrated more than 200 kmupstream of Saint�Louis. To address the problemsassociated with the significant inter�annualvariability in rainfall and water flow of the SenegalRiver, three of the four main bordering countries(Mali, Mauritania, and Senegal) entered into a treatyto form the Senegal River Authority, the Organisationpour la Mise en Valeur du Fleuve Sénégal (OMVS),and related organizational structures in 1972.

The tasks of the OMVS were to attain the goal offood self�sufficiency for the Senegal River Basininhabitants; reduce the economic vulnerability of theorganization’s member states to climatic fluctua�tions as well as to external factors; accelerate theeconomic development of member states; conserveecosystem balance in the sub�region, particularly inthe basin; and secure and improve the incomes ofbasin inhabitants. To accomplish these goals theOMVS was charged with constructing and managinga regional infrastructure consisting of two majordams. The first to be completed (in 1986) was theDiama dam, located 27 km upstream from the city of

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Figure 1: The Senegal River basin

St Louis (Senegal). It was built to stop the dry�seasonintrusion of sea water along the river bed which,during the drought years, would penetrate over 100km inland. The second is the storage dam atManantali in Mali (completed in 1990) on the Bafing,the main tributary of the river, which suppliesapproximately 50% of the annual flow. The reservoiris theoretically capable of stocking 11 billion m3 ofthe strongly seasonal rainfall on the Fouta DjalonMountains in Guinea. The water can then begradually released over a longer period than thenatural flood. The two dams should provide enoughwater to achieve the following developmentobjectives: irrigate 375 000 hectares of formerfloodplain, especially for rice production; producehydropower (800 Gwh per year); make the rivernavigable all year round between Saint Louis at theriver mouth and Ambibédi in Mali.

Water level and discharge measurementsin the estuary

Discharge information from Diama dam is availablefrom July 1986 to now. As shown below, releasesfrom the barrage occur each year from approximatelyJuly through to December. At other times of the yearthere is no flow through the barrage.

Over the measurement period the data showsvariations in both mean sea level and tidal range.Subsequent analysis of the data is presented, whichdisplays tidal range both at St Louis (from the

40

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 2: Time Series of Discharge, Diama Barrage (1986-2003)

Figure 3: Water level at Saint-Louis 1986-2002

measured data) and the ocean (from tidalpredictions) and mean water level at St Louis. Thewater level measurements from Diama Barrage arealso shown. Note that as these measurements areonly daily recordings, tidal range cannot be extracted.

When water levels exceed 1.2 m above MSL floodingoccurs in Saint�Louis. According to the measurementsat Saint�Louis, this has occurred nine times since1964. Since the barrage was constructed in 1985,this has only occurred three times. The dischargesfrom the barrage from 1985 onwards are also shownbelow. Note that the peak discharge greater than3000 m3/s recorded in 1987 is considered to beincorrect, and should be approximately 1500 m3/s(Gilif, 2002).

OBJECTIVES OF THIS PROJECT

In 1999, the western part of Africa experiencedhigher precipitation rates, resulting in higher riverdischarge in the Senegal River and its tributaries andthus larger inundations in the river valley and deltathan seen during the last 30 years. Several villagesand irrigation infrastructures were destroyed. Peoplehad to abandon their houses and rice field cropswere lost. Further downstream, Saint�Louis, theformer capital of Senegal, experienced large damagesdue to inundation of areas built up during the dryeryears in the 80s. Floods risk reduction is related topoverty in the sense that floods consequences

(removal of population, housing problem, healthproblem, lost crops, etc.) reinforce poverty; thusreducing flood risk will inevitably cut down the rate of poverty.

Therefore, local and regional decision�makers needmanagement tools and materials for flood monitor�ing. The ultimate objective of this project is todevelop new tools base on satellite images for floodmonitoring and forecast in the Senegal River valleyand estuary that can be used on regional, nationaland local level by relevant authorities to improvewater management and to reduce impacts ofextreme events like floods.

Accurate mapping of the areal extent and duration of the yearly flood of the Senegal River is desirable in order to monitor hazards on people, irrigatedagriculture, buildings and infrastructure and also toassess the potential for the traditional flood recessionagriculture. Furthermore it has been highlighted thenecessity to have precise Land Cover Maps as ameans to assess flood damage but also as a way ofmodelling the hydrological functioning to bettercharacterise the watershed in the Senegal River valley and estuary.

METHODOLOGY

Data and image acquisition

Data acquisition was an important stage of thisproject. Several missions were carried out in theestuary and the lower valley because floodmonitoring in the Senegal River valley and estuaryrequire a large number of data sets. The most criticaldata include atmospheric weather data, soil physicaldata, current and historical land�use data for ageneral description of land cover and land�usemanagement around the town of Saint�Louis.

41

The SPOT 4 (1999, 2001, 2002) and SPOT 5 (2002)images were acquired near the University of Dakar incollaboration with University of Marne la Vallée(France). The SPOT Xi image of October 13, 1998 isrecovered to the GILIF1 project developed in 2002 onthe Senegal river by the DGPRE2, (UNEP) but belongsto former UTIS3 center. The available satellite dataare listed in Table 1.

This kind of approach based on satellite images haspreviously been used by Sandholt and al., (2000) inthe lower valley around the city of Podor. This pilotstudy focused on different Landsat images incombination with data from other sensors likeAVHRR and radar data from ERS.

Pre-processing

Geometric correction

The SPOT data were preprocessed for data extractionand geometric correction. The images we use in thisstudy were not directly georefenced and at thebeginning they have deformations and must bestood up to be put to the better orientation inconformity with the geographic reality, in a plan ofprojection. All images have been geometricallyrectified to UTM Zone 28 North. Because of thedifferences in spatial scale and thus areal coveragefor the images, different areal coverages areavailable in each case, so direct comparisons areonly possible in regions with overlap.

The Ground Control Points (GCP) we have chosen at the beginning with the Japanese InternationalCooperation Agency (JICA) maps (1/50 000), hadnot known conclusive results because the RMS oftenwas being enough raised. The precision of thosemeasures was not satisfied. The choice was beingparticularly focussed on remarkable points, asintersections of roads, trees, angles of fields, etc.

In order to correct these ground control points andto have a better precision in the choice of GCPs,we carried out a mission to complete the points withthe GPS. This solution has permitted to obtain aprecision of 5 m. These points thus made it possibleto rectify the images with a high degree of accuracy.

Data fusion

In this case of images fusion, we have searched tocombine judiciously the spatial information of highresolution panchromatic image of January 16, 2001

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Date Mode Season Observations

13�10�1998 XS 4 Rainy high floods

23�10�1999 XS 4 Rainy high floods

31�10�2001 XS 4 Rainy low floods

16�01�2002 Panchromatic Dry no floods

Table 1 : Satellite data

1 Gestion Intégrée du Littoral et du Bassin Fluvial2 Direction de la Gestion et de la Planification des Ressources en Eau 3 Unité de Traitement de l’Imagerie Satellitaire

(10 m of resolution), with themultispectral SPOTimages of October 23, 1999 and of October 31,2002 (20 m of resolution). This permits to obtain animage richer simultaneously in spatial and spectralinformation (Figure 4). The fusion is most certainly animportant manipulation to realize in the case of sucha study, but we realise that the panchromatic imagebrings noise on the Spot XS image and thisinformation is very difficult to analyse in theclassification process.

The Classification process

The first unsupervised classification process is amethod of reiterated classification and this processdoes not lean on any exterior data. The pixels areregrouped according to the method of minimumdistance. The K�Means unsupervised classificationmethod has given satisfied results. It is based on thecalculation of classes to obtain a result distributedregularly in the data space, then iteratively regroupsthe pixels according to the class the more near usingthe method of minimum distance. From the greatclasses obtained in a coarse way, we refined theclassification with the method of supervisedclassification (Figure 5).

A standard supervised maximum likelihoodclassification was carried out using the ENVI imageprocessing system. A total of five spectrally differentclasses are identified corresponding to spectralvariability in the image. Differences on waterreflectance are caused by turbidity and water depthand presence of varying amounts of greenvegetation. Training and test areas were definedusing field observations and delimitation carried outfrom a visual interpretation of SPOT images.

42

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 4: Fusion process between SPOT image (23/10/1999and 16/01/2001)

Figure 5: The classification process

MAPPING FLOODEXTENSION AROUNDTHE CITY OF ST LOUIS

The relevance of the study hasbeen accentuated by extensiveflooding in 1999, when thegreatest river flow in 30 yearswas reported. Because ofexceptionally high rainfall inFouta�Djalon in the rainyseason of 1999, peak flow atBakel reached 4440 m3/s(IRD—Institut de Recherchepour le Développement, 2001).The result was the destructionof many villages and irrigationinfrastructure. Substantial partsof low lying Saint�Louis by theAtlantic Ocean was struck byflooding. Man�made factorscontributed to the seriousnessof the event. Dikes constructedalong the river channel toprotect populations hindered or delayed up�streamflooding of the wide river valley, exacerbatingproblems downstream (Figure 6).

Cartography of flooded area in October 1999

The cartography of flooded surfaces permitted toisolate the different regions reached by the flood inOctober 1999. It is a multi�temporal approach whichconsists in comparing the level of water surfacesbetween the flooded period and the normal period.The characteristics of the 2 multispectral SPOTimages used are presented in the table below.

We took as reference the 2001 image, because it hasthe same seasonal and temporal characteristics thatthe 1999s image, except only that it was acquired ina normal period with out floods. Figure 7 showssatellite imagery of the lower estuary around the cityof Saint�Louis during the flood period in October1999 (left) and during the normal period in 2001(right).

43

Figure 7a: Satellite Imagery of Senegal RiverLower Estuary, Flood conditions in October1999 (left) and no flood conditions in October2001 (right)

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Figure 6: A view of flooded area in Saint-Louis in October 1999

Date Mode Season Observations

23 10 1999 XS 4 Rainy High flood

31 10 2001 XS 4 Rainy not flood

Table 2. Characteristics of multispectral SPOTimages

One observes on these two extracts, theimportance of the surfaces covered by thefloods in October 1999. The principal bed ofthe river is largely overflowed and all theregions in the north of the town of Saint�Louiswere considerably flooded (Figure. 8).

The floods devastated very important surfaces. With acomparison of these two satellite images (1999 and2001), it appear that the different flooded surfaces havesome difference. Slightly flooded surfaces account for64.86% of flooded surfaces. Strongly flooded surfacesaccount for 21.52% of flooded surfaces and the other isthe flooded and muddy surfaces (Table 3 and Figure 9).

44

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 7b: Extract of spot 4 xs of 23/10/199 (left) and Spot 4 xs of 31/10/2001 (right).

Figure 8: Flooded area in the region of Saint-Louis in October 1999

These 3 layers of vectors were extractedautomatically and exported with the Shape format tobe integrated in the GIS (Figure 10a and 10b).

45

Water surfaces represent an important part of the totalarea considered. The total surface of water is 20.48%of the total surface of the area considered. Turbidwater accounts for 16.93% of the total surface water.

Comparison of flood extension between 1998 and 1999

In October 1998 and 1999, important floods occurredin the area of Saint�Louis. Their respective extents werecompared with satellite images by a multi�temporalapproach. It arises that the floods were of 1999 weremore important in term of surfaces flooded comparedto those of 1998 (Table 4).

Table 4: Statistics of the two SPOT images(1998–1999)

The 1999 floods have affected non flooded surfaces in 1998. The extent of flooded surfaces in 1999represents 70,109 km2 more ever than that of 1998.

THE DIGITAL ELEVATION MODEL

The relief of the area of St Louis is very low. The firstpeaks of altitude 10 to 15 m are located at severalkilometers of the river. In the interior of the city thetopography does not exceed 3m. The maps weredelineated using topographic base maps with contourintervals of 10 feet, except in the city, where thecontour intervals were five feet. The status of thetopographic maps and the lack of high�resolutionelevation data pose great difficulties. Availability anduse of a DEM would expedite planning anddevelopment of land use for precision agriculture,drainage systems, land subdivision, utilities,

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Types of objects Number ofobjects Min Mean Max Summon Standard

deviation

Flooded surfaces(low)

1332 400 150759.16 28595200 200811200 1432383.37

Flooded surfaces(high)

2836 400 23493.37 3207600 66627200 136738.7

Flooded and muddysurfaces

2665 400 15821.84 1881600 42165200 67334.53

Year No. ofobjects Min Mean Max Somme Stand.

Dev.

1998 510 400 41915.29 5104800 21376800 317499.44

1999 1062 400 86145.01 13822800 91486000 618155.77

Table 3: Statistics of the surfaces flooded in October 1999, the surfaces are expressed in meters

0,00 10,00 20,00 30,00 40,00 50,00 60,00 70,00

Slighty flooded Strongly flooded Muddy & flooded

Figure 9 : Flooded surfaces in 1999 (%)

Figure 10b. Flooded surfaces in 1999 super-imosed on image of fusion of Spot 4xs of31/10/2001 and panchromatic of 16/01/2002.Scale 1/100 000

Figure 10a. Extract of the vector layersrelating to the surfaces flooded in 199 in lightblue, surfaces slightly flooded; in dark blue,strongly flooded surfaces; in red, flooded andmuddy surfaces.

commercial and industrial districts, etc., and improvethe quality of soils mapping.

The solution chosen to establish a DTM was toexploit the altimetric data provided in the form ofpoint sides and level lines by JICA5 and SEGECOTmaps. The precision of these altimetric data isunfortunately not easily measurable. Indeed, wedon’t know the reliability of the original data,neither the methods employed to obtain these data,and even less the geoids.

The traditional method of satellite data combination(Spot 4) and the gathering of important informationmade it possible to produce a GeographicalInformation System to monitor floods in the lowerestuary of the Senegal River valley. Remote Sensing,DTM and GIS seem to be powerful tools for combin�ing important information for a better comprehensionof the floods and the characterization of surfacequalities on the estuary.

We developed a Geographic Information Systemincluding approximately many layers of local and/orregional spatial data. We backdated relevant para�meters such that the GIS is both uniformly formattedand historically accurate. With this GIS in place, weare able to assess the flood extension.

By a multi�temporal approach, we established thequalitative and quantitative impact of floods on thevarious geographical objects, a detailed cartographyof the occupation of the ground, the surfaces floodedin 1998 and 1999. The study undertaken to St Louismade possible to consider surfaces flooded in 1999and to understand the width of these floodscompared to those of 1998. The constitution of a toolof decision�making aid makes possible to haveinformation relating to the limits reached by theflood, the surface of flooded surfaces and to detectthe more exposed zones (the most reached) in orderto establish a hierarchical map according to thepercentage of exposure to the risk of the geograph�ical objects touched by the floods (populations), roadinfrastructures and tracks, medical and social infra�structures, perimeters of cultures (agriculture), etc.

LESSONS LEARNED

African countries know unceasingly and almost eachyear dramatic natural disasters. It is the case inparticular of Algeria (with repetitive earthquakes),Mozambique, Zimbabwe, and Senegal (which wasstruck by floods). Many of these disasters especiallyaffected poor populations and had serious effects onthe already fragile and unstable local economies.Their consequences are all the more serious as theseaffected populations do not have the means of beingprotected from such catastrophes and the Statescould not set up true policies or strategies ofprevention or reduction of the natural disasters. Thefew rare strategies are often inappropriate anddedicated to the failure.

Disaster risk reduction and management approach isrelatively new in concept as well as in practice.Although a few countries have adopted risk

46

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 12a. DTM of the estuary area

Figure 12b. DTM of the town of Saint-Louis

management concepts and principles in disastermanagement, most countries, especially developingcountries, remain unfamiliar with this approach. Theprevailing practices, particularly in Africa, are moreinclined towards managing response to disasters(which requires preparedness) than towardsmanaging risks and the underlying conditions thatlead to disasters (which requires, among others, riskassessment, vulnerability reduction, and capacityenhancement).

In the Senegal River valley and estuary for example,poor and socially disadvantaged groups are usuallythe most vulnerable and affected by floods. Thesedisasters, in turn, are a source of transient hardshipand distress and a factor contributing to persistentpoverty. Indeed, at the household level, poverty is thesingle most important factor determining vulner�ability. In Senegal, the prevalence of poverty is veryhigh. In 1994, the first Household Budget/Consump�tion Survey (ESAM�I)4 made it possible to estimatethe proportion of households below the povertythreshold (fixed at 2,400 calories per adult equivalentand per day) at 57.9 percent. On the basis ofextrapolations made from the CWIQ5 (2001) data,the percentage of households below the povertythreshold is about 53.9 percent, i.e. a slight dropcompared with 1994, as a result certainly of theincrease in per capita income over the period 1995�2001. However, these figures are well below thefindings of the EPPS6 (2001) according to which 65percent of the households interviewed (same sampleas the CWIQ) consider themselves poor and 23percent even rate themselves as very poor. Moreover,64 percent of the households declare that povertyhas worsened over the past five years, a perceptionthat is contrary to what is stated above. Thisapparent contradiction undoubtedly results fromdifferent criteria for assessing poverty.

Poverty in Senegal is located for a large part in therural areas and more especially in the rural zones ofthe Center, South and Northeast. This concentrationof poverty in rural areas is also confirmed by theEPPS (2001): in point of fact, in rural areas theincidence of poverty varies between 72 percent and88 percent while in urban zones it ranges between44 percent and 59 percent. In both cases, theincidence of poverty remains high. This situation isespecially worsened by the floods which strike therural populations. For example, the floods that swept

47

through Senegal between 9�11 January 2002 led tothe loss of 28 lives, with over 100,000 otheraffected. A damage assessment revealed that anestimated 105,471 head of livestock had perishedand 13,993 homes here demolished. As much as581 ha of crops were washed away. Approximately1,537 tons of rice was also destroyed.

The 1994, 1999 and 2003 floods in the basin werealso unusual, for both their depth and duration.Unlike the normal floods, which cover large parts ofthe valley for several days or weeks during August toSeptember, the floods in 1999 lasted until mid�October in many areas, killing people and destroyingroads, houses, crops, and other assets. Flood impactshave been severe in the zone because of the highlevels of vulnerability and low levels of resilience ofthe population, the lack of adequate physicalprotection infrastructure and changing floodingpatterns due to environmental change and theimpact of the dam’s structures.

Thus, to attenuate the impact of the natural disasterson the poor populations, it is important to undertakein�depth studies on the relation between disasterand poverty. Our project tries to tackle this questionin the Senegal River basin by using space technologyto map the floods extension and thus to identifyarea and populations which are touched or whichare in danger. The establishment of this projectprovided a beginning of coordinated approach to a flood mitigation strategy for the Senegal estuaryand lower valley. Since 1999, the local scientistcommunity and local authorities try to made goodprogress in addressing the existing flood threats inthe area.

The Government of Senegal, for instance, identifiesfloods in the Senegal River as one of the factorseroding the income of the poor populations viacrisis�related expenditure and reductions in incomeearning capabilities. Furthermore, it recognizes thatpoverty alleviation cannot be achieved simply byincreasing income, but instead requires a range ofother measures, including the strengthening of localcapacity to protect the poor against these floods.

In this case, recent catastrophic floods (2001 and2003) all over the basin have raised new questionsas to traditional approaches in dealing with suchextreme events. The increasing occupation offloodplains around the city of Saint Louis in the

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

4 Household Budget/Consumption Survey5 Core Welfare Indicators Questionnaire6 Household Survey on Perception of Poverty in Senegal

estuary, competing and conflicting developmentaldemands in the lower valley have exacerbated theimpacts of floods on society and the environment.Furthermore, the concerns of human vulnerability andan environment that can be further mismanaged orabused, have focused attention to the need for moreintegrated, anticipatory, and far�reaching waterpolicies and strategies. Understanding and respond�ing to floods requires a comprehensive view ofintervening environmental, social and economicfactors. This calls for joint approaches by all relevantnational agencies, as well as for the development of integrated support strategies by internationalagencies with expertise on the subject, as UNEP(United Nations Environment Programme) or UN/ISDR (International Strategy for Disaster Reduction).

The Senegalese Government should continue to give priority attention to its strategy for DisasterReduction as a common platform for responding tothe challenges presented by the increased incidenceand scale of floods in the lower valley.

The analysis and lessons learned from priorexperiences of floods help to define profiles of riskattached to people, activities and places that shareattributes, in the face of particular potential sourcesof damage. Understanding risk relates to the abilityto define what could happen in the future, given arange of possible alternatives. Assessing risks, basedon vulnerability and hazard analysis, is a requiredstep for the adoption of adequate and successfuldisaster reduction policies and measure in theSenegal River estuary.

The project allows us to carry out several fields worksand to collect many information and data related tothe floods and to carry out a GIS. The investigationsgave an idea on the overall organisation of the studyzone in particular on the occupation of the easilyflooded area around the town of Saint�Louis andsome villages and small towns in the lower valley.The investigations allowed by stepping of testimoniesto define the limit of the extension of the past floodsand to index the level of the various historical risings.In this study the issue of flood hazard mapping hasbeen addressed from the perspective of differentmapping scale in a GIS environment. The flood

hazard map is particularly handy for the planners andadministrators for formulating remedial strategy. Italso makes the process of resource allocation simpleresulting in a smooth and effective implementation ofthe adopted flood management strategy. The aim ofthis regional study is to broadly identify the highhazard area in the area around the city of Saint�Louisand in the lower estuary of the Senegal River valley.Our project eventually leads to identification of thehigher hazard zone.

This project meet a double aim: on the one hand tobetter include/understand the dynamic of the floodsin the estuary, on the other hand to producedocuments for early alarm in direction to theauthorities and to place at their disposal tools ofdecision making. Its purpose was to provide apreliminary approach that can be used as ademonstration of the capabilities, applications andadvantages of satellite images, and as a guide forfuture investigations. Remote Sensing technology hasits special superiority and potentiality for floodmonitoring and assessment, so it has been appliedfor this purpose in this project, especially for thedisaster resulting from floods. A lot of scientific andpractical achievements have been obtained in thestudy area. The information on inundated area andthe variation of river channel was successfullyobtained. In 1998 and 1999, by using the SPOTsatellites images, the floods occur in the lowerestuary of the Senegal River were investigatedseparately. After that, a lot of local expertsrecognized the importance of remote sensing imagedata for disaster risk reduction especially for thetown of Saint�Louis in the lower valley and suggestedto set�up the real time transmission system ofairborne remote sensing for disaster monitoring withdata providers.

This is a pilot study, and has not been subject to therigorous calibration and validation exercise normallyexpected in a more advanced risk reductionassessment. This means that the results and findingsshould be considered as preliminary. Our project hasthus the merit to apply such a step and to havesatisfactory results. The GIS developed constitutes atool of decision making aid for governmental andlocal authorities.

48

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Bibliography

Codata (2002): Scientific Data for decision making towardsustainable development Senegal River Basin: Case Study.Workshop, March 11�15, 2002, Dakar, Senegal.

Diop Isabelle Niang, Dansokho Mamadou, LY Ibrahima.,Niang Seydou (2002): Senegal national report. Phase 1:Integrated problem Analysis. GEF MSP Sub�Saharan AfricaProject (GF/6010�0016): “Development and Protection of theCoastal and Marine Environment in Sub�Saharan Africa”

Gilif (2002: Gestion intégrée du littoral et du bassin fluvial duSénégal, Rapport final, 2002

IRD, ( 2001): Programme d’Optimisation des Reservoires.Rapport préliminaire.

Sandholt and al., (2000): Remote Sensing Techniques forFlood Monitoring in the Senegal River Valley.

Republic of Senegal (2002): Poverty reduction strategy paper.

Vengroff Richard, (2000): Decentralization, Democratizationand Development in Senegal, Department of Political Science,University of Connecticut, Storrs, CT 06269�1024. Paperprepared for delivery at the Yale Colloquium onDecentralization and Development, January 21, 2000.

49

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

50

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

51

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Coastal erosion vulnerability

mapping along the southern

coast of La Union, Philippines

TEAM LEADER

Rose Berdin <[email protected]>

TEAM MEMBERS

Christina T. RemotiguePeter ZamoraYvainne Y. Yacat

PROJECT ADVISOR

Dr. Fernando P. Siringan

Philippines

SYMPOS IUM NOTES

Rose D. Berdin, Cristina T. Remotigue, Peter B. Zamora, and Ma. Yvainne Y. Yacat�Sta. Maria

National Institute of Geological Sciences, University of the Philippines

Diliman, Quezon City 1101, PHILIPPINES

52

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

INTRODUCTION

The coastal zone of the Philippines covers approxi�mately 11,000 km2, which accounts for less than 4%of the country’s total land area but is home toalmost 50 million people or 63% of the country’stotal population. As population and investments ininfrastructures steadily grow along with rapid urbani�zation and expansion, coastal communities becomemore and more at risk or vulnerable to coastalhazards such as tsunamis, storm surges and erosion:the latter being the most widespread and persistentin the past several decades.

Coastal erosion is common in the Philippines despitethe present�day elevated sediment yield of rivers.Erosion could be due to natural factors, for instance,river mouth switching in the bayhead of LingayenGulf (Mateo and Siringan, submitted), mass wastingas a result of rapid sedimentation at the mouth ofthe Bucao River in Zambales (Siringan and Ringor,1996), and return to normal sediment yield of riversyears after a volcanic eruption in Pamatawan Delta,Zambales (Siringan and Ringor, 1996). However,there are just as many coastal segments whereerosion is attributed directly to human activities.Erosion along the shores of Pampanga and Cavite inManila Bay is attributed to dam construction andoffshore sand mining, respectively (Siringan andRingor, 1997; Doruelo, 2000). Thus, for propermitigation, if amenable, the cause or causes oferosion should be established. Including theabovementioned work, only few studies, however,address this problem (e.g., Siringan et al., 1999;Siringan and Jaraula, 2002).

This study sought to develop a coastal vulnerabilityindex for southern La Union. Coastal segments wereclassified according to the degree of vulnerability toerosion based on established shoreline trends, knowncoastal characteristics and available socio�economicinformation. Long�term or decadal trends of shorelinechanges were derived from time�series analysis ofmaps and anecdotal accounts. Probable causes ofshoreline changes and the common response of thegovernment and affected communities are alsopresented. The results of this study hopefully will beused to guide the government and stakeholders tocome up with development plans and disastermanagement strategies, and to refine or formulatepolicies that would consider the consequences ofcoastal erosion.

53

STUDY AREA

Situated in northwestern Luzon is the 60�km north�south trending coastal stretch of southern La Union,from San Fernando in the north to Sto. Tomas in thesouth (Fig. 1). Except in San Fernando, wherein somecoastal segments are made up of rocky headlandsand some are fringed by coral reefs, the entirecoastline of the study area is characterized by sandybeaches along gentle coastal plains. Two large riversystems drain into the area, Bauang and AringayRivers, which have watershed areas of 516 km2 and397 km2, respectively (Siringan and Mateo, 1999).Wind data indicate predominant winds coming fromthe northwest. At wind speed ranging from 2 to 6m/s during normal conditions, waves with significantheight of less than 1 m are generated; at 10 to 20m/s, wave heights reach over 3 m (Woodward Clyde,1999). Wind�driven circulation model by Delas Alas(1981) indicate the predominance of southerlycurrents in the area. Sediments coming from Bauangand Aringay Rivers, coupled with the predominantsoutherly longshore currents, formed the Sto. Tomasspit, an elongated sand body that extends from thedelta flanks of Aringay southwards to approximately12.5 km The study area covers the towns of SanFernando, Bauang, Caba, Aringay, Agoo and Sto.Tomas (Fig. 1), of which 48 coastal barangays(barangay is the smallest political unit in thePhilippines) are included.

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Figure 1. The study area consists of six coastalmunicipalities in southern La Union. Boxesshow the coastline extent of each municipality.

METHODS AND DATABASE

Several steps were undertaken before the coastalvulnerability index (CVI) was developed. Physical andsocio�economic data were acquired from primary andsecondary sources. The physical data included estab�lishing trends of shoreline and bathymetric (or waterdepth) changes. Shoreline trend analysis was comple�mented by interviews of long�time coastal residents.Collectively, these data were used to construct thecoastal vulnerability index for southern La Union.

Shoreline trend analysis

Long�term changes in shoreline position weredetermined by overlaying digitized shoreline from oldmaps and from a recent GPS survey (Table 1). Thepresent shoreline was mapped by taking GPSreadings, with average accuracy of ±8m, everyapproximately 30 to 40 m while walking the coast.To properly position the GPS readings on the map,GPS fix of ground control points, such as major roadintersections and bridges, were also obtained. Thedifferent shorelines were corrected for differences inprojection and were superimposed on vertical aerialphotographs taken in 1990, again using prominentfeatures such as roads and intersections as controlpoints. Although the aerial photographs were notcorrected for distortion, possible distortion wasminimized by cropping the edges and making use ofthe central part of each frame before stitching themtogether using a graphics software.

The maximum amount of shoreline translation wasmeasured perpendicular to the shoreline orientationalong a particular segment. In the succeedingsections, shoreline position based on the 1944 and1956 maps is denoted as 1940s because the infor�mation used in the maps was based on aerial photo�graphy from the 40s. Similarly, the 1970s shoreline isderived from the 1977 topographic map; althoughthe data could have been acquired earlier as thesource of information was not indicated in the map.

Ground�truthing of these changes was performedduring the course of the GPS survey by taking note of coastal features associated with erosion andaccretion. Photographs were also taken fordocumentation.

Bathymetric change

Although the year of the survey was not indicated,the 1977 map represents water depth informationsometime within 1940s to 1970s because it containsinformation not reported in the early 1900s survey.The water depth data from the 1977 map were geo�referenced and digitized to extract the oldbathymetry.

Except off the coast of Caba, which was surveyed inAugust and January, 2002 (Siringan and Jaraula,2002), bathymetric survey for the rest of the studyarea was conducted in April and September, 2003using Garmin GPS sounder with an accuracy of ±15m for position and ±10 cm for water depth. Along azigzag route perpendicular to the coast, at anaverage cruising speed of 10 km/h, water depth and

54

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Year Type of Data Area of coverage Reference

1901 Map San FernandoIn Meimban, 1997. La Union: The

Making of a Province (1850�1921).

1944 1:50,000 topographic map San Fernando to BauangUSC&GS

Map sheet 3160 I (based on 1944aerial photography)

1956 1:50,000 topographic map Bauang to Sto. TomasCorps of Engineer, US Army map sheets

3068 I and 3069 II (based on 1946,1947 and 1959 aerial photography)

1977 1:50,000 topographic map San Fernando to Sto. TomasNAMRIA map sheets 7075 I and 7076

II (source of data not indicated)

1991 1:40,000 vertical aerial photographs San Fernando to Sto. Tomas NAMRIA

2003 GPS survey San Fernando to Sto. Tomas This study

Table 1. Database used in determining long-term trends of shoreline changes.

position were acquired every two seconds. Both oldand new raw data were then gridded using a 50 x50 m square grid. Changes in bathymetry werecalculated by subtracting the gridded output. Ablanking file was generated from the plotted newdata such that only the changes within the areacovered by the recent survey are considered.Correction for tides and waves was not applied.

Social survey

Interviews of longtime residents with good recall ofevents were conducted to validate, augment andcomplement field data and to understand how thelocals perceive the cause of erosion and how theyrespond to it. The survey questionnaire, withemphasis on shoreline and land use changes, wasmodified from the instrument used by Siringan andRodolfo (2002) and was first tested in Wenceslao,Caba (Siringan and Jaraula, 2002). In each munici�pality, barangays showing significant long�termshoreline changes as assessed from spatial data wereselected: six from San Fernando City, four from Sto.Tomas, three each from Caba and Agoo, and twoeach for Bauang and Aringay. From the 20 barangayschosen, 63 respondents aged 30 to 84 years old,mostly between 55 to 60 years old, were inter�viewed. Almost all of the respondents are fishermenand barangay officials.

Coastal Vulnerability Index

The Coastal Vulnerability Index (CVI) for the studyarea was developed following mainly the method ofMcLaughlin et al. (2002). Variables contributingsignificantly to the study area’s vulnerability toerosion were chosen and grouped into three sub�indices: coastal characteristics (wave buffers,substrate, shoreface slope and shoreline evolution),natural (wave exposure) and anthropogenic forcings(mining history), and socio�economic characteristics(population, land use, roads and bridges, cultural andhistorical landmarks, and conservation status). Eachvariable is then ranked on a scale of 1 to 5, with 5being the most vulnerable and 1 the least, accordingto their perceived level of vulnerability. After thescores per sub�index are added, the final CVI scorewas calculated by getting the sum of the partialweight of each sub index: 40% for coastal character�istics, 20% forcing and 40% socio�economic. Calcu�lation for CVI was limited to barangay level, which isthe highest resolution of the available socio�economic data. Four categories were used in classify�ing the vulnerability of the coastal barangays: low,medium, high and extremely high. A color code isthen assigned to each category to derive a coastal

55

vulnerability map. Below is a brief description of thevariables used and how they were ranked.

Coastal characteristics

Coral reefs, sand dunes and mangroves are naturalcoastal features, which serve as buffers against theerosive energy of waves. Thus, the presence of thesefeatures lessens a coastal segment’s vulnerability toerosion. Areas where any of these natural buffers ispresent are given a score of 1 and where not one ofthese features is present, a score of 5.

In the study area, the substrates encountered arebare rock or cliff, mud and sand. Among the three,bare rock or cliffs are the least vulnerable to erosionbecause they are more indurated and therefore,more difficult to entrain. Sandy beaches, on the otherhand, would be the most vulnerable to erosion;although larger in grain size and heavier as opposedto mud, sand is less cohesive and therefore, relativelyeasier to erode. A score of 1 is given to areas withbare rock or rocky beach, mud flats 3 and sandybeaches 5.

Shoreface slope on a regional scale affects thewaves that reach the coastline. Areas with steeperslope allow waves to break closer to the coast andthus, enhancing erosion. Along coastal segmentswith gentler slope, waves approaching the shorebreak at a greater distance from the shore therefore,dissipating much of the wave energy before hittingthe shoreline. Shoreface slope for each coastalbarangay was calculated from the new bathymetricdata. Slopes values were divided into five equalranges: the lowest range was given a rank of 1whereas the highest range a rank 5.

Ranking for the shoreline evolution variable isderived from the established long�term trend ofshoreline changes in the study area (see Results andDiscussion): rapidly translating, rapidly retreating,slowly retreating, highly variable and relatively“stable”. This classification considers the time factor– how the coast behaved in the past severaldecades – which may also give us an idea of how itmight respond to processes and events in the future.Furthermore, it takes into account the geomorph�ological setting (e.g., delta, spit) and relativeproximity to major sediment sources, which may ormay not be an advantage (see Results andDiscussion). Barangays along rapidly translatingcoastal segments, although not necessarily eroding,are given the highest score because of the highpotential for erosion. Consistently rapidly eroding

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

shorelines are ranked 4, and the retreating, but at aslower pace, shorelines ranked 3. Less vulnerable,with score of 2, are highly variable coastal segments,wherein no consistent trend of change occurs. Leastvulnerable to erosion are the relatively “stable”coasts; in the past several decades, the shorelinealong these coastal segments appears to haveremained constant.

Natural and anthropogenic forcings

Only two variables are included in this sub�index:wave exposure and coastal mining history. Inevaluating a coastal segment’s relative exposure towaves, the predominant wind direction, which isnorthwest, was used as proxy. The most vulnerablesegment, therefore, would be one facing northwestwhereas a south�facing segment the least vulnerable.

Magnetite sand mining by Filmag in the 60s to 70salong almost the entire La Union coast negativelyimpacted the coast. This activity physically removedhuge volumes of sand from the beach, loosenedmaterials thereby making it easily transported bywaves and removed a natural beach armor, the heavymagnetite sand. The negative effect of coastal miningtrailed long after the operations stopped; therefore,we assigned a score of 1 for coastal segments notmined and 5 for those mined by Filmag.

Socio-economic characteristics

Information about the socio�economic variables arebased largely on reports obtained from each localitynamely, the comprehensive land use plan of SanFernando City, Aringay, Caba, Agoo and Sto. Tomas;municipal development report of Bauang and coastaldevelopment framework of La Union. These aresupplemented by secondary data obtained from eachmunicipal/city hall.

Instead of classifying the coast according to settle�ment type, population density, or the number ofpersons per hectare, was used. Population densitiesin coastal barangays range between 2 and 362persons per hectare. Only Caba and Aringayregistered lower figures than the average populationdensity of 4 persons per heactare in the entire LaUnion. On the other hand, three barangays (IlocanosNorte, Ilocanos Sur and Baluarte) have densitiesgreater than 100 persons per hectare; hence, a rankof 5 was assigned to areas with at least 100 personsper hectare and a rank of 1 to areas with less than10 persons per hectare. The difference between theupper and lower limits was equally distributed overthe remaining ranks.

The ranking was simpler for the other categories. Inland use, residential area has been omitted since it isalready incorporated in the population ranking. The

56

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

VARIABLE 1 2 3 4 5

Natural buffers (coral reefs,dunes, mangroves)

present none

Substrate Bare rock, cliff MudflatSandy beach/Sand flat

Shoreface slope <0.8° 0.8�1.4° 1.4�2.1° 2.1�2.7° 2.7�3.4°

Shoreline evolutionRelatively“stable”

Highly variable Slowly retreating Rapidly retreating Rapidly translating

VARIABLE 1 2 3 4 5

Wave exposure relative toNW winds

S�facing SW�facing N�facing W�facing NW�facing

Mining history none present

Sub-index 1. Coastal characteristics.

Sub-index 2. Natural and anthropogenic forcings.

rest of the categories were ranked according to theirrelative economic significance in the area. Highestranked are industrial areas, followed by commercial,agricultural, institutional and natural habitats. Thisranking does not belittle the socio�economic impor�tance of natural habitats; the presence of suchhabitats could make the shoreline more stable and therefore, less vulnerable to erosion.

Roads and bridges were categorized into barangay,municipal/city, provincial or national. Because of thedifficulty to distinguish one from the other, barangay,municipal and provincial roads were lumped into asingle category and given a rank of 3. National roadswere assigned a rank of 5 due to its scale and higherconstruction costs, which translates to greaterbudgetary requirements for repairs, if and whenneeded.

The last two variables – landmarks and conservationstatus – are dichotomous and limited to presence orabsence. Historical and cultural landmarks includelighthouses, bell towers, monuments, old churchesand buildings and even railways. A major conduit tothe north from early 1900s until the 1970s, theprovincial government plan to restore the railways. Interms of the last category, many coastal areas in LaUnion are part of the Network of Protected Agricul�tural Areas (PLUC�TWG, 1996), which primarily limitsfurther industrialization of important agriculturallands. Existing natural habitats such as CarlatanLagoon are also under conservation status.

57

RESULTS AND DISCUSSION

Trends and causes of shoreline changesalong the coast of southern La Union

Long-term changes

Time�series analysis of maps and images coupledwith social survey results (Fig. 2) show that erosionfrom the 1940s to present is prevalent along the 60�km long coast of southern La Union. This iscorroborated by changes in bathymetry, whichindicate an overall deepening in the nearshore areas.Spatially, the coastal stretch can be divided into fivetypes based on the magnitude and trend of shorelinechange: rapidly translating, relatively “stable”,rapidly retreating, slowly retreating, and highlyvariable coasts.

Coastal segments along the southern flanks of theBauang and Aringay River deltas experienced largeshoreline translations during the past six decades(Fig. 2). In Parian Oeste and Payocpoc Norte/Oeste inBauang, the shoreline advanced by 650 m from1940s to 1970s, 800m from 1970s to 1991 and600m from 1991 to present. Shifting of river mouthand the elongation of the spit accompanied thisseries of net land gain from the northern end of themouth towards the south. In contrast, the adjacentcoastal stretch of Payocpoc Sur, Pilar and Santiagoretreated by 550 m from 1970s to the present.Residents from Pilar, however, noted as much as1700 m of land loss since 1940s. The land area of

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

VARIABLE 1 2 3 4 5

Population <10 per ha. 10�40 per ha. 40�70 per ha. 70�100 per ha. >100 per ha.

Land useNatural habitat/Open spaces

Institutional/Parks/Health Facilities

Aquaculture/Agriculture

Commercial/ResortsIndustrial (plants and ports)

Roads and bridges NoneProvincial/Municipal/Barangay

National

Cultural/Historicallandmarks

None present

Conservation status None present

Sub-index 3. Socio-economic characteristics.

this barangay was reduced to 3 ha from the original11.5 ha. The coastal segment immediately south ofthe Aringay river mouth, in Alaska, gained 400 m of land from 1940s to 1970s, but lost 500 m whenthe shoreline retreated since then. Locals from this barangay observed 480 to 1300 m of seaencroachment since the 1960s. Farther south, inDulao, Sta. Rita Central and Sta. Rita West, theshoreline consistently prograded and is at present800 m of its 1940s position. Respondents fromDulao reported a consistent trend, whereas Sta.Rita West folks had the opposite observation. Theselarge and rapid changes have been occurring evenprior to the 1940s as these regions are labeled assuch in the 1940s maps.

A marked contrast occurs north of the deltas; alongPagdalagan Sur through Pugo and Santiago Surthrough Samara, the shoreline appears to haveundergone minimal changes (Fig. 2). Based on themaps, slight accretion occurred along these coastalsegments. This is, however, inconsistent with the 200to 500 m land loss observed by Baccuit Sur residents.

Shoreline retreat has been predominant since the1940s along Bagbag through San Carlos in Caba,and Sta. Rita Sur through San Isidro in Agoo (Fig. 2).Maps indicate 150�250 m of net erosion, but thelocals give larger estimates of land loss: 500 to 1000m in Wenceslao and San Carlos, 1000 to 2000 m inBaluarte, and 300 to 535 m in San Manuel. Atpresent, a series of groins and bulkheads line thecoast of Agoo. Construction of these structuresstarted in the 1980s. Similarly, the entire coast of SanFernando has been retreating, but at a slower pace(Fig. 2). Maps dating back to 1900s, indicateshoreline fluctuations throughout the San Fernandocoast in the order of a few tens of meters. This trendis consistent with the anecdotal accounts: amaximum land loss of 70 m since the 1960s.

The western coast of Sto. Tomas is on a southeast�accreting spit. This thin strip of land includes Baybay,Cabaroan and Narvacan, the westernmost shorelineof which can be characterized as highly variable (Fig. 2). Based on the maps, no consistent shorelinetrend exists and variation occurs along 1 to 3 km�long segments of the coast; however, residents fromNarvacan narrated that 300 to 1000 m of retreatoccurred since the 1960s. The most conspicuouschange occurred in the southernmost terminus of the spit, which extended by as much as 1000 m inabout 50 years.

The apparent inconsistencies between the magni�tudes of erosion derived from the time�series analysis

58

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 2. Changes in shoreline position overdecadal time scales. Also shown are thebarangay limits and corresponding populationdata for year 2000.

of maps and the social survey may be explained bythe different perspectives provided by these datasets.Shoreline positions from maps, because they werederived from aerial photographs, provide a snapshotof the area during a specific time only – that is,when the aerial photo was taken. On the other hand,the observations of the locals span a longer period oftime. However, errors may arise in the respondents’estimate of distances, which vary from person toperson. This can be tested by checking how a parti�cular change at a particular time was consistentlyquoted by the respondents in a barangay. None�theless, the general trends derived from the mapsand from anecdotal accounts are consistent.

Changes in bathymetry during the past decadesconfirm the abovementioned results. Along coastalsegments where significant erosion took place,deepening occurred in the immediate vicinity. This isexhibited off Payocpoc Sur through Wenceslao,Samara, Alaska, and Sta. Rita Sur through SanManuel Norte, where as much as 4 m of deepeningoccurred. Likewise, San Fernando Bay generallydeepened, except for its central part whichunderwent shallowing. Maximum deepening of about8 m took place in the narrow passage of the mouthof San Fernando Bay. Conversely, shoaling occurredwhere progradation is considerable. This is seenadjacent to the mouth of Bauang and Aringay River.In Bauang, the offshore area fronting Parian Oesteand Payocpoc Norte/Oeste became shallower by 3 to9 m. The same amount of shoaling took place infront of the Aringay River mouth, whereas offshoreDulao through Sta. Rita West, water depth shallowedby 5 m. Shoaling of approximately 4 m in theoffshore area shows two along�coast elongatepattern originating in front of the Bauang Rivermouth and extending south, 3 km off the coast ofSantiago Sur. A similar pattern is observed south ofAringay River. The area of shallowing close to thecoast, however, extends only in front of Sta. RitaWest whereas the more distal area of shoalingreaches about 1 km offshore Baybay. These spatialvariations in bathymetric change indicate sites wheresediments are being removed and where they arebeing deposited. The trends observed in the studyarea show that in general, sediments that are erodedfrom the coast are delivered offshore. Furthermore,materials coming from the rivers are mostlydeposited to the south, which is consistent with thepredominant southward longshore drift in the area.

59

Response to coastal changes

Individually or as a community, the people of LaUnion responded to the prevalent erosion either bytemporarily evacuating, relocating, building riprapsor sandbags, or if funds are available, by construct�ing seawalls (Fig. 3). Short�term erosion that occursduring storms or typhoons prompts affected commu�nities to temporarily evacuate to a safer place;nevertheless, when conditions return to normal andpartial or full beach recovery occurs, residents goback and restore whatever’s left of their properties.In a lot of cases, however, beach recovery does nothappen and people are forced to relocate, usuallyjust several meters inland. Relocating has becomefrequent for many, especially for those living inpersistently eroding coastal segments. In the case ofPilar in Bauang, residents have transferred five timesalready since 1975. To protect properties andinfrastructures from erosion, private individuals andsome communities resorted to building structureslike ripraps, sandbags and seawalls. The ripraps andsandbags, according to the locals, were not effectivein preventing erosion and maintaining them becamecostly in the long run. Those who have moreresources, such as along Carlatan in San Fernando,construct more robust structures.

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Figure 3. Cheaper alternatives to the moreexpensive coastal protection structures:sandbags in Wenceslao (top) and ripraps inAlaska (bottom). In the long run, however,maintaining these structures also becomescostly.

On the Government’s part, huge funds have beenallocated to construct coastal engineering structuresto counter erosion. In 1980, a 1.5 km�long seawallwas built in Sta. Rita to protect the Agoo�DamortisNational Park. The entire structure was completed in1988; however, its southern end had to be repairedperiodically due to weak foundation. The 1990earthquake further damaged the base of the seawall,which prompted the construction of its second layer.This, and the original construction of the Sta. Ritaseawall over a period of 14 years, from 1980 to1994, cost approximately PhP250 M. Since 1984, aseries of groins and bulkheads have been put up toprotect the Agoo�Sto. Tomas coast. Initially built toprotect the pavement just south of Agoo Playacompound, groin construction was further extendedto San Julian Norte, Agoo in the north and to Sto.Tomas in the south. At present, more than 60 groins,costing around PhP57 M, dot the Agoo to Sto. Tomascoastline. More groins are being proposed forconstruction to arrest the erosion of areas downdriftof these structures.

On the contrary, new coastal lands formed by progra�dation are promptly occupied and even titled, thoughby law it should be public land (Commonwealth Act141, Chapter IX). In Sta. Rita Central, Aringay, shore�line prograded by approximately 350 m from 1940sto 1970s. This newly�accreted land is now inhabitedand the Agoo�Damortis National Park, which includesa basketball court, church, resort, cemented accessroad and the seawall, is built on it (Fig. 4). In the

past 30 years, another 450 m of land was added tothis coastal stretch and is also being slowly occupied(Fig. 4). This prompt occupation of newly�formed landleads to a lot of problems later on: in Caba, forexample, progradation in the late 1930s until 1960slured people to settle in, but were later forced torelocate several times because of the subsequenterosion that took place (Siringan and Jaraula, 2002).This led to loss of properties. Thus, the localgovernment should be prompt in declaring accretedland as public property to avoid a similar scenario.

Causes

Changes in shoreline position along the southerncoast of La Union during the 60�year period can beattributed to natural and anthropogenic factors.Natural causes of erosion include local forcings suchas waves, earthquakes and shifting position of rivermouth. Constant wave action on the coast bringsabout erosion by removing materials from the beachand by wearing away hard materials that it cannotcarry. Hence, coastal segments where wave energy ishigh are more prone to erosion than those in moreprotected settings. The general north�southorientation of La Union’s shoreline renders it opento waves originating from South China Sea. Winddata from observations in Poro Point, San Fernandoindicate predominant winds coming from thenorthwest. Typically, waves with significant height ofless than a meter are generated at wind speedranging from 2 to 6 m/s during normal conditionsbut, at wind speed of 10 to 20 m/s, can reach ashigh as 3 m (Woodward Clyde, 1999). Furthermore,waves during storms, in the form of storm surges, canbe even higher; documented storm surge height inIlocos Sur, north of the study area, ranges from 3to10 m (PAGASA, 2002).

60

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 4. In 1980, a 1.5-km long seawall wasconstructed to protect communities and theAgoo-Damortis National Seashore Park in Sta.Rita (left). At present, however, approximately450 m of land accreted in front of the structure(below) due to the natural southward drift ofsediments coming from Aringay River, 5 kmnorth of the area.

In July 16, 1990, a 7.8�magnitude earthquake rockedcentral Luzon. This devastating earthquake causedseveral areas in La Union, specifically along thecoasts of Aringay, Agoo and Sto. Tomas to subsideinstantaneously due to liquefaction: 1 to 2 m inAlaska (anecdotal account), 0.5 m in San NicolasWest, 1 to 2 m in Cabaroan, 1 to 3 m in Narvacan(Geomatrix, 1995), among others. Widespreadinundation accompanied the lowering of coastallands; the most severely�affected area is Alaska,wherein residents recounted about 1 km of seaencroachment (Fig. 5). After the earthquake, thelocals in Sto. Tomas observed slight shorelineprogradation; but because the land is now lower, itbecame more prone to erosion. Torres et al. (1994)attributes liquefaction in La Union and in theneighboring provinces to the nature of the underlyingmaterial: relatively unconsolidated and water�saturated sands. Aringay sits on a delta, which isbuilt by rapid and continuous deposition of materialsfrom Aringay River. Agoo and Sto. Tomas, are atopvery young sandy deposits formed by accumulation ofsediments transported by the longshore current.

In contrast to the large shoreline retreat thatoccurred in Alaska, the coastal stretch to the south,in Dulao and Sta. Rita, experienced rapid landaccretion after the earthquake (Fig. 4). These areasare at the receiving end of the eroded materials fromAlaska and the sudden influx of sediments from therivers due to earthquake�induced mass wasting in theupstream. Rapid progradation south of the BauangRiver is ascribed to the southward migration of theriver mouth coupled with the south�directedlongshore currents. However, because these lands arerelatively new, they are also likely to experienceliquefaction in the future. Furthermore, because thesesegments are very close to the river, future shifts inriver course may lead to rapid erosion. The northernflanks of the deltas, although seemingly “stable” at

61

present, may also be at risk of rapid erosion asshifting of the rivers northwards is also probable. Amore northerly position of the river mouths in thedistant past, no longer covered by the period withmaps, is indicated by scars of previous streampositions, which can be seen in aerial photographs.

On the global scale, widespread erosion can beascribed to the effects of global warming: decreasedprecipitation, increased storminess and sea�level rise.El Nino Southern Oscillation (ENSO) record for thepast several decades indicates a shift from a LaNina�dominated to El Nino�dominated climate in the1970s. With El Nino occurring more frequently, theamount of precipitation decreases, hence, riverdischarge to the coast also goes down. Peak annualdischarge data from 1946 to present of the twomajor rivers in the study area indicate a decline(Streamflow data 1980�2000, 2001), which may berelated to the observed shift in climate regime.

Storms and typhoons in the Philippines are veryfrequent and cause the greatest damage to proper�ties and the most number of lives lost among all theother natural disasters (ADRC, 2002). Residents ofLa Union attribute about 5 to10 m of erosion totyphoon passage. In the past several decades, thelocals observed that the beach is able to recoveralmost after every typhoon; however, the series ofstrong typhoons that hit the area in the 1990shindered the shoreline from returning to its pre�storm position. Typhoons such as Trining (Inter�national Code Ruth; October 24, 1991), Mameng(Cybil; September 27, 1995) and Feria (Utor; July 3,2001) caused significant destruction to the coastalcommunities of La Union. Should storm frequencycontinue to increase due to climate change, erosionwould accelerate because the affected coastalsegments are not able to recover to pre�stormconditions.

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Figure 5. Present shoreline of Barangay Alaska is littered with tree stumps indicating that treesused to thrive along the coast. During the July 1990 earthquake, a large area of Alaska subsided by1 to 2 m as a result of liquefaction. According to the locals, prior to the earthquake, the shorelinewas approximately 1 km farther offshore.

Tide�gauge records in the Philippines registered a rise in sea level in the past decades (Siringan et al.,2000). Over the next century, global sea�level rise isprojected by the International Panel on ClimateChange (IPCC) to range from 15 to 90 cm, with 48cm as its best estimate (Wigley and Raper, 1992).The values may seem minimal, but the effect of risingsea level on gentle coastal plains, such as in thestudy area, would be significant as it would allow thesea to encroach farther landwards. Therefore, as sealevel continues to rise, erosion will continue tothreaten coastal communities.

On top of these natural causes, certain anthropogenicactivities exacerbate coastal erosion. The mostextensive, and probably the most detrimental, ofthese activities is the mining of magnetite sand byFilmag Inc. For 10 years, from 1964 to 1974, Filmagextracted approximately 2M cubic meters of magne�tite along almost the entire La Union coast, whichstretches for about 100 km The total volume of sandextracted would translate to an average deepening of2 m and shoreline retreat of 10 m per meter ofcoastal segment. This is a conservative estimatebecause not the entire La Union coast was mined;hence, certain sections most likely have deepenedand retreated more than the estimated values.Furthermore, extracting materials from the coastentails loosening of the substrate, thereby renderingthe sediments more susceptible to transport by wavesand currents. More importantly, the heavy magnetitesand serves as beach armor; thus, removing them isanalogous to destroying the beach’s naturalprotection.

Another human activity that contributes to erosion isthe destruction of natural resources, such as coralreefs, sand dunes and native coastal vegetation (e.g.,mangroves, aroma trees and the like), which act asnatural wave buffers. In San Fernando, for example,coral reefs that extend out to the northeast of Poroand fringe Dalumpinas, Lingsat and Carlatan protectthe coastal areas along San Fernando Bay byattenuating waves coming from South China Sea. Atpresent, however, large areas of these reefs havebeen damaged already due to illegal fishing practices(e.g., dynamite fishing). Thus, erosion along thesecoastal segments can be due partly to increasedimpact of waves, which in turn is due to reefdestruction. Furthermore, coral reefs contributeconsiderable amounts of sediments to the beach. Themost prolific sand producers are the foraminifers,minute one�celled organisms that secrete CaCO3tests, which thrive in reef flats with healthy sea grasscommunities. Destruction of the reef flat or coral reef

would deprive the beach of a major sediment source.This is critical for San Fernando, because the majorrivers are far and are located downdrift of the bay.

Sand dunes also offer a natural means of protectingthe coast. They take the brunt of the waves and alsoserve as sand reservoir. Sand dunes in DalumpinasOeste, according to anecdotal accounts, have beenleveled to provide a mining company access to thebeach sometime in the 60s to 70s. With the use ofloaders, sand was also quarried for constructionpurposes. At present, the height of the remainingdunes in the area is only a third of its previous level.Further dune degradation is occurring toaccommodate more houses and resorts in the area.

To protect properties from erosion, ripraps andmakeshift seawalls in Carlatan were built using theexposed rock platform along the coast as primarymaterial. According to a resident, attempts were alsomade to level this platform, which extends severaltens of meters offshore, by blasting the protrudingrocks so that Carlatan can have a sandy beachinstead of a rocky shore. What the community isunaware of is that the rocky platform also acts as awave buffer, similar to coral reefs; instead of wavesbreaking on the coast, incoming waves expend theirenergy offshore when they encounter the platform,hence reducing erosion.

Erosion is also aggravated by artificial structures thatinterfere with sediment distribution. A groin promotesbeach accretion on the updrift segment of the coastbut deprives sediments to the downdrift end, thus,enhancing erosion along that side. The series ofgroins in Aringay and Agoo (Fig. 6), therefore, isdetrimental to the Sto. Tomas coast because theydecrease the amount of sediments that normallyreaches Sto. Tomas. Geomorphological evidenceindicate that Sto. Tomas, a spit, was formed by theaccumulation of sediments transported through thesouth�directed longshore drift. Piers can also serve asgroins if the structure’s design does not allowsediments to pass through. An example of this is the

62

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 6. Series of groins dotting the Agoo-Sto.Tomas coastline.

Bauang Power Plant Company pier in Pilar, whichmay enhance erosion to the south by trappingsediments on the north side of the structure. Thepresence of this structure is critical especially sincethe coastal stretch south of it has been undergoingsevere erosion. Armoring the shoreline with seawallsand revetments, although effective in protectingproperties and fixing the shoreline, cause beachnarrowing and loss in retreating coasts as demon�strated in Oahu, Hawaii (Fletcher et al., 1997). Thesestructures reduce sediment delivery to the beach bypreventing the natural process of “upland” erosion,thus refocusing wave energy on the beach in front ofthe structure. Furthermore, decreased sedimentamount in the updrift beaches means decreasedsupply to the downdrift segments, hence, amplifyingerosion. This appears to be the case along severalcoastal segments in San Fernando wherein beachesin front of armored structures are either very narrowor completely lost. A similar scenario was observed infront of a series of resorts south of San Fernandothrough north of the Bauang River (Fig. 7), whichindicates that these lighter structures may alsoinduce an effect similar to seawalls and revetments.

63

The net accretion in front of the Sta. Rita seawall,which at present sits behind 450 m of newly�accreted land, is a different case. This is because it issituated along the southern flank of Aringay deltawhere sediment supply is abundant and transportdirection is towards the south. Furthermore, basedon the maps (Fig. 1), the seawall and the Agoo�Damortis National Seashore Park are on top ofnewly�accreted land. Land gain, therefore, is not due to the structure.

Activities in the watershed may also contribute tocoastal erosion. Initially, deforestation andurbanization would lead to rapid soil erosion in theuplands, which would translate to large sedimentinput to the coast and thus, allow progradation.However, as surfaces are covered by grass andcemented pavements, upland erosion decreases and sediment input to the coast declines. This thentranslates to shoreline erosion. To date, there are no data on the watershed history of the study area;however a decrease in the river’s sediment yielddue to changes in vegetation cover as documentedby Kummer et al. (1994) in other parts of thePhilippines could have also occurred in the water�sheds of rivers draining into southern La Union.

Vulnerability mapping

Figure 8 shows the vulnerability map derived fromcomputing the coastal vulnerability index (CVI) ofeach barangay. Coastal segments are markedaccording to their degree of vulnerability to erosion:extremely high, high, moderate and low. Alsopresented in the figure are the vulnerability profilesof each barangay represented as pie charts. Thismethod was suggested by previous studies (Gornitz,1993 in McLaughlin et al., 2002) to distinguish therelative strengths of each sub�index in a particularsite. By viewing the index and the profile side byside, one can then have a better understanding ofthe factors that make an area vulnerable to erosion.

Barangays situated within the immediate vicinity ofthe mouths of Bauang and Aringay rivers rankedmost vulnerable with respect to coastal character�istics. These coastal segments are classified as rapidlytranslating and thus, are easily affected by shifting of river mouths and sediment delivery from thewatershed. The adjacent barangays in Agoo, Sta.Rita Central and Sta. Rita West, where a 1.5�km long seawall was built, also ranked highly vulnerable.Lowest scores were obtained for Poro, San Francisco,Canaoay and San Vicente, all situated in theLingayen Gulf�facing southern communities of SanFernando City.

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Figure 7. Resorts and restaurants line thesouthern coast of San Fernando and extendsto north of Bauang. These structures providevery narrow shoreline setback for beachadjustment, thereby causing an effect similarto impermeable structures like seawalls andrevetments. These photographs were taken inSan Vicente; the top photo is the viewlooking north; bottom is looking south.

64

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 8.Vulnerability

map of thesouthern coasts

of La Union.Relative scoresof the coastal

characteristics,coastal forcing

and socio-economic

sub-indices arerepresented by

pie charts.

65

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

In terms of coastal forcing, NW�facing areas wheremining occurred in the past such as DalumpinasOeste in San Fernando, Taberna and Pugo in Bauang,Samara in Alaska, and Sta. Rita, San Julian Norte andSan Julian West ranked highly vulnerable. Whereas,the more protected localities along coast of SanFernando City are the least vulnerable. These includeCarlatan and the barangays that ranked low in thecoastal characteristics sub�index. Ironically, Carlatanhad the highest socio�economic scores, together withSan Nicolas West in Agoo. Carlatan hosts a numberof industries and like San Nicolas West, has a richcultural heritage. Even more ironic is the fact thatthe barangays that were previously classified ashighly vulnerable with respect to the physicalvariables had the lowest scores.

The combined scored for the three sub�indices identi�fied Carlatan and San Nicolas West as extremelyvulnerable to erosion. Carlatan, although low innatural and anthropogenic forcings, has the highestscore in the socio�economic sub�index and alsoweighed heavily in the coastal characteristics sub�index. As for San Nicolas West, high scores wereregistered in all three sub�indices. Classified as highly vulnerable to erosion are Lingsat and IlocanosNorte in San Fernando, barangays north and south of Bauang River (except Pilar), barangays in Agoo(except Sta. Rita Sur, San Nicolas West), andCabaroan in Sto. Tomas—all situated along west�facing coastal segments. These areas ranked relativelyhigh in the coastal characteristics and the forcingssub�indices and moderate in the socio�economicaspects. The rest of the barangays, mostly rural areasand situated along the flanks of Bauang and Aringay,are identified to be moderately vulnerable to coastalerosion. The least vulnerable areas are Poro and San Francisco in San Fernando and Pagdalagan inBauang. Socio�economically, Poro and San Franciscoranked high, but low in coastal characteristics.Pagdalagan is not an industrial and commercial area unlike the two others.

Closer examination of the weighted scores of thesub�indices shows that the coastal characteristics and natural and anthropogenic forcings do not varymuch. Excluding the scores for Poro (2) and Canaoay(1.6), weighted scores for coastal characteristicscluster from 3.2 to 6.4; Canaoay and Poro are atopbedrock whereas the rest are mostly on sandy coastal plain. Weighted scores for natural andanthropogenic forcings range from 0.4 to 2. Whattipped the scale of the vulnerability index are thesocio�economic characteristics wherein weightedscores vary from 5.2 to 12.8. This indicates the

significant contribution of socio�economicinformation in assessing an area’s vulnerability.

LIMITATIONS OF THE STUDY

Direct relationships between rates of erosion, timingand ultimate causes cannot be established with thedata gathered. The time frame provided by maps and aerial photographs is limited; therefore, moreinterviews are needed to obtain a more statisticallyreliable data set. We are also constrained by theavailability of secondary information, specifically thesocio�economic profiles, because of poor recordkeeping of the offices concerned. The vulnerabilityindex could be refined with more data.

CONCLUSIONS ANDRECOMMENDATIONS

Anecdotal accounts are important sources ofinformation on the changes of shoreline position.However, it should be filtered and tempered withother data.

Rates and primary cause/s of erosion vary alongcontiguous coastal segments; thus, mitigationmeasures, if still amenable, may also differaccordingly. The cause/s may also change throughtime. As coastal erosion is a complex problem, itshould be studied seriously for more informed course of actions.

Uncontrolled rapid occupation of accreted coastallands will lead to conflicts with the naturallychanging landscape. The local government should,therefore, assert the public domain nature of thesenew lands to help lessen damages due to subse�quent erosion. Since the problem transcendsmunicipal boundaries, the actions for mitigationshould therefore be agreed upon by the localgovernment of affected municipalities.

There are three main stakeholders in the area, themarginalized group; the business group; and thegovernment. However, the local government must atthe outset protect the interest of the marginalizedgroup, mostly fisherfolks, and look for alternativesolutions without compromising or sacrificing theirmain source of livelihood.

Increase information advocacy to raise the level ofawareness on the problem among stakeholders isvital. A shared understanding and recognition of the risk involved in coastal erosion may lead to theaffected community’s willingness to accept effective

66

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

mitigation approaches and to support anysustainable use practice introduced. A well�informedcommunity readily accepts the idea of relocatingwhen mitigation through engineering structures maynot be effective and relocation is the only bestsolution. Informed managers from all levels of thegovernment can rechannel limited funds to otherprojects that can benefit the community.

Of equal importance is the need for the governmentto refine existing policies and formulate new onesthat would specifically address coastal erosion. Atpresent, there are no coastal laws, ordinances ordecrees that directly tackle this extensive andpersistent problem. Once effective policies areendorsed and approved, the next step is to havestrong, political will to institute reforms.

ACKNOWLEDGEMENT

Our team greatly appreciates the enthusiasticassistance of Lea Soria, Nap Villanueva, KarenRodriguez, Titan Quiña, Nelle Monteclaro, NelBaluda, Eden Baliatan and Grace Asio in gatheringfield data. Special thanks to Zoan Reotita whoprovided the much needed help in organizingmaterials, data and a lot more other things. We alsothank Rose Pang�ot for lending us those hard�to�look for reports and Kelvin Rodolfo for introducingus to the Emery method of beach profiling, whichmade surveying a lot more easier and fun for us. Tothe people of La Union, for sharing their knowledgeabout the coast, Dios ti Angina. To our resourcepersons, Samuel Sanchez, Mary May C. Diaz and theothers who wanted to remain anonymous, we arevery grateful. To Ando, our mentor, thank you somuch for the unwavering guidance and support.

This research benefited so much from results of anongoing work funded by the City Government of SanFernando, La Union and a previous study funded byMs. Lourdes Eristingcol.

67

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

BIBLIOGRAPHY

Asian Disaster Reduction Center (ADRC), 2002. PhilippineDisasters Datasheets.

Comprehensive land use plan of Agoo, La Union.

Comprehensive land use plan of San Fernando City, La Union.

Comprehensive land use plan of Sto. Tomas, La Union.

De las Alas, J.G. 1986. Prediction of the levels and trends ofnon�oil pollutants in the Lingayen Gulf. University ofthe Philippines Science Research Foundation. Diliman,Quezon City.

Doruelo, J.C.S., 2000. The dynamics of sediment transport offthe coasts of Cavite. MSc Thesis. Univeristy of thePhilippines.

Fletcher, C.H., Mullane, R.A. and Richmond, B.M., 1997.Beach loss along armored shorelines on Oahu, HawaiiIslands. Journal of Coastal Research, 13(1): 209�215.

Geomatrix Consultants, Inc., 1995. Geohazard and seismiczonation mapping of the Agoo and Caba urban studyareas, unpublished report.

Kummer, D., Conception, R. and Canizares, B., 1994.Environmental degradation in the uplands of Cebu.The Geographical Review. 266�276

Mateo, Z.R.P. and Siringan, F.P., submitted. Tectonic and sealevel controls of high�frequency Holocene deltaswitching and fluvial avulsion in Lingayen Gulfbayhead, northwestern Philippines.

McLaughlin, S., McKenna, J., and Cooper, J.A.G., 2002. Socio�economic data in vulnerability indices: constraints andopportunities. Journal of Coastal Research, SpecialIssue 36: 487�497.

Meimban, A. O. 1997. La Union: The Making of a Province(1850�1921), 341 pp.

Municipal development report of Bauang, La Union.

Municipal development report of Caba, La Union.

Municipal development report of Aringay, La Union.

PAGASA. 2003. Information on tropical cyclone: province ofLa Union (1948�2002).

PLUC�TWG. 1996. La Union Physical Framework Plan(Cooperative Provincial Land Use Plan (1993�2002).

Rivera, P. C. 1997. Hydrodynamics, sediment transport andlight extinction off Cape Bolinao, Philippines.Disseration submitted in partial fulfillment of therequirements of the Board of Deans of WageningenAgricultural University and the Academic Board of theInternationl Institute for Infrastructural,Hydraulic andEnvironmental Engineering for the Degree ofDOCTOR. 244 pp.

Siringan, F.P., Berdin, R.D., Doruelo, J.S., Jaraula, C.M.B.,Mateo, Z.R.P., Ong, J.B.T. and Yacat, M.Y.Y., 1999,Sediment transport patterns along Macajalar Bay andCagayan de Oro River derived from geomorphic andsedimentologic parameters. Abstracts andProceedings “Tectonics of Southern Philippines:Implications to Natural Hazards, Mineralization,Magmatism and Environmental Change”, Aurelio,M.A. and Pajarillaga, N. (eds), Geological Society ofthe Philippines, Quezon City. pp. 28�33.

Siringan, F.P. and Jaraula, C., 2002. Land configurationchanges along the Aringay�Bauang coastal zone:possible causes and mitigation measures,unpublished report.

Siringan, F.P., Maeda, Y., and Rodolfo, K.S., 2001. Short�termand long�term changes of sea level in the PhilippineIslands. Global Change and Asia Pacific Coasts(Mimura and Yokoki, eds.): Proceedings ofAPN/SURVAS/LOICZ Joint Conference on CoastalImpacts of Climate Change and Adaptation in theAsia�Pacific Region, pp. 143�149. Siringan, F. P., andRingor, C. L., 1997, Predominant nearshore sedimentdispersal patterns in Manila Bay, Science Diliman, 9:29�40

Siringan, F. P. and Ringor, C. L., 1996, Changes in the positionof the zambales shoreline before and after the Mt.Pinatubo eruption: controls of shoreline change,Science Diliman, 7 and 8: 1�13.

Siringan, F. P. and Rodolfo, K. S. 2003. Relative sea�levelchanges and worsening floods in the northwesternPampanga Delta: causes and some possiblemitigation measures. Science Diliman vol. 15, no. 2,pp. 1�12.

Torres, R.C., Paladio, M.L.O., Punongbayan, R.S., Alonso, R.A.,1994. Liquefaction inventory and mapping in thePhilippines: a worst case scenario. National DisasterMitigation in the Philippines DOST�PHIVOLCS, 71�98.

Wigley and Raper, 1992. Implications for climate and sealevel of revised IPCC emission scenarios. Nature, 357,293�300.

68

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

69

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Mapping and assessing risk

and vulnerability of water-

induced disasters in the Tinau

Watershed, Western Nepal

TEAM LEADER

Keshav Prasad Paudel <[email protected]>

PROJECT ADVISORS

Prof. Dr. Vidya Bir Singh KansakarNarendra Raj Khanal

Nepal

SYMPOS IUM NOTES

70

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Introduction

Background

High relief, steep slopes, relatively steep rivergradient, and active geology with intense monsoonrainfall make the young Himalayas as one of themost hazardous areas in the world and it frequentlysuffers from different kinds of water induceddisasters like soil erosion, landslide, debris flow,flood, glacial lake out�burst flood, landslide damout�burst flood etc. Against the backdrop of highpopulation density and expansion of infrastructureresulting in changes in natural slope, rivermorphology and the drainage system, land use/landcover; the frequency in the occurrence of thelandslide and flood hazards has been increasing inrecent years. The loss of lives by flood and landslideand avalanches comprises about 29% of the totalloss from all types of disasters.1 As a result ofincreased hazard, magnitude of lives and property atrisk has also correspondingly increased. There is needto identify the areas that are susceptible to waterinduced disaster and comprehensive risk assessmentfor designing mitigation measures in order to reducethe risk and vulnerability. Such activities provide asound basis for disaster mitigation planning andhence for allocation of financial and other resources.Risk assessment includes an evaluation of thepotential occurrence of hazard events, assessment ofvulnerability and potential damages, and theassessment of stakeholders’ capacity simultaneously.However, there are few studies and research inlandslide and flood hazard mapping with risk andvulnerability assessmentin Nepal. With in thisbackground the presentresearch project wasimplemented formapping and assessingrisk and vulnerability ofwater induced disasterin Tinau sub watershed.

The present researchproject is thecomponent ofProvention Consortiumresearch grant fordisaster reduction. Theproject aims to identifythe frequency,magnitude, and extentof water induceddisasters experienced inthe study area, map

71

hazard and risk, access the socio�economic impact ofwater induced disaster and explore the response andrecovery capacity of local people.

Project Area

Tinau watershed lies in Palpa District of WesternDevelopment Region of Nepal. It extends from27°45’00” to 27°52’51” north latitude and83°25’30” to 83°42’30” east longitude (Map .1).The total area of watershed is 234.2 sq. km.Elevation ranges from 330 m at the confluence ofTinau and Jhumsa Khola to 1893.4 m at GhustungLekh. The study area has highly rugged terrain withsteep slopes and deeply incised valley of TinauKhola, bounded in the south by the Siwalik range.The Mahabharat range occupies the central andnorthern part of the watershed. Being in between ofMain Boundary Thrust (MBT) and Central BoundaryThrust (CBT), this watershed is tectonically veryactive and geomorphologically very unstable.

The sub�watershed covers complete or partialportion of 20 VDCs and one municipality. Theestimated population of sub watershed based onfocus group discussion is 66665. The populationcomposition of watershed is dominance of Magarcaste ethnic group. It occupies 42.6% of totalpopulation followed by Brahamin/Chhetri 31.3%,Occupational or so called “Dalit “ (untouchable)caste 12.9%. Other hill ethnic caste occupies 4.6%and other caste population occupies about 2% ofthe total population.

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Map 1. Study area

The main source of income of the local people isagriculture. About 92% households are engaged inagriculture as a main source of income. However the landholding pattern is not even. About 6%households are landless and 24% households aremarginal having up to 0.25ha land.

Project Implementation

The main focus area of over all projectimplementation is landslide and flood hazardmapping and risk identification of Tinau subwatershed. During the project implementationsperiod following activities were carried out.

Information requirement analysis

Various literatures, documents, knowledge anddiscussion with advisor were used as references forinformation requirement analysis and identification ofdata. Various methods for mapping and assessingcommunity risk and vulnerability of natural disasterwere reviewed and the detail work plan for thisproject was prepared. Checklist for observation,interview guide, schedule and focus group discussionguide were also prepared.

Preliminary field observation

A short field observation was carried out during thelast week of July to obtain basic information ofwatershed. The verification of watershed boundary,identification of organization and institutions workingin the field of flood and landslide disaster, and somesecondary information collection from local levelorganization was done during this field trip. GPSobservation of 50 sample site of landslide and floodlocation and major landmarks, (for RS imageryanalysis) was also taken. A pilot survey of focus

group discussion, household survey and RRA wasconducted in MadanPokhara to check the checklist offocus group discussion and household survey. Thechecklist and interview guide and schedule werefinalized on the basis of pilot survey.

Acquisition of secondary information

Various government and non�governmentorganizations identified during informationrequirement analysis were visited. Both spatial andnon�spatial data were collected from respectivesource as follows:

• Toposheet maps (Scale: 1: 25000) published byTopographical Survey Department, Nepal. 1958, &1993.

• Land system, Land utilization and land capabilitymap published by Land Resource Mapping Project,Nepal, 1986.

• Geological maps prepared by Department of Minesand Geology, Nepal, at the scale of 1: 50000.

• Digital version of IRS Satellite imagery acquired in2001 (obtained from Forest Department).

• Aerial photos (1978, 1996) prepared byTopographical Survey Department, Nepal.

• Hydrological and climatological data of last 20years from Department of Hydrology andMeteorology.

• The recorded information of loss of lives andproperties by Disaster in Nepal (1983 – 2002)from Department of Water Induced DisasterPrevention and Ministry of Home.

• Demographic and other socioeconomic censusdata (2001) from Central Bureau of Statistics.

• Various socio�economic information were alsoobtained from District Profile of Palpa district(2000).

• The information on mitigation measure, rescueoperation for landslide and flood hazardimplemented by Disaster Management Committeesand Nepal Red Cross Society.

Aerial photo interpretation and digitalimage processing

Information on channel course, landslide extent, andflood wash area were identified and preliminarydistribution map of these parameters were preparedthrough aerial photo (1996) interpretation and digitalimage processing of IRS image (2001).

72

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Landless

0.51–1 ha

1–2 ha

>2 ha

0–0.25 ha

0.26–0.5 ha

6%

27%

4%

24%

26%

13%

Figure1. Landholding pattern in Tinau Watershed

Field survey and group discussion

A field survey was carried out in September 2003.The field survey was carried out to map flood pronearea, flood wash area, landslide extent distribution,channel course diversion; verify and update theinformation obtained from maps, aerial photos, andimagery interpretation. Focus group discussions ateach hazard prone area with the key informants wereheld in order to obtain additional information onlandslides and floods and also on the number ofhouseholds and population. A social map onlandslide and flood hazards was also prepared withthe help of the local people. During the groupdiscussions the participants were asked to show anddelineate the areas susceptible to landslide and floodhazards.

Preparation and acquisition of digitaldatabase

The information collected and mapped on toposheetshowing distribution of landslide, flood wash area,flood prone area, channel shifting was digitized andprepared spatial digital database. The map ofgeology, toposheet (1958), Land utilization map(1986), Land utilization/ land capability map (1986)were also digitized. The digital data set of river, road,landuse, settlement and administrative unit of basemap were also prepared. The digital data set ofcontour and spot height was acquired from SurveyDepartment.

The layers of landslide, and flood area, and channelcourse were prepared from aerial photointerpretation (1978, 1996) and digital imageprocessing (IRS image 2001). The parameters layer ofDEM, slope, relief, rainfall interpolated, slope aspect,distance from lineament, distance from stream,drainage density were prepared.

Preparation of hazard and risk maps

The landslide and flood hazard maps of Tinau sub�watershed was prepared using GIS based analysis.Subsequently, risk map of each hazard were alsoprepared (Fig. 2).

At first landslide distribution maps was preparedbased on field survey and image interpretation.Eleven parameters: lithological unit, lineaments, slopegradient, slope shape, slope aspect, relative relief,drainage density, water table and drainage condition,distance from road, landuse, and vegetation densitywere used for landslide hazard mapping. Landslidehazard map was prepared by using the bivariatestatistical method. The landslide hazard map was

73

categorized into three categories according to theprobability of occurrence of landslide as high hazardarea represent high probability and low hazard arearepresent low probability of occurrence.

Similarly, flood hazard mapping was done using theDigital Elevation Model generated from contours andspot heights. In addition to this information obtainedfrom land system maps, and channel morphologywas also used. The flood hazard was classified in tofour categories according to the probability ofoccurrence of flood and channel shifting.

Risk map was prepared by combining thevulnerability and the hazard maps. For the analysisof vulnerability only four sets of parameters werechosen. These included: a) population, b) land use,c) infrastructure and d) irrigation canal. Populationdensity at the VDC level, distance from roads andtrails and distance from cannel, and the land cover(based on use type) were taken in order to evaluatethe degree of potential loss from the landslides.

Household survey

Based on two criteria distance from river bed andthe hazard zone (the high, medium and low) fivesettlements were selected for intensive householdsurvey. The household survey was conducted todifferentiate the severity, risk, and vulnerability ofhouseholds and the response and recovery capacityof household level with varies hazard zonation.Census of all the households was conducted in thosesettlements where the total number of households isless than 30. In settlements where the numbers ofhouseholds are more than 30, at least 30 or 10percent of the total households were selected forinterview.

Vulnerability and Response Recoverycapacity Analysis

The socio�economic and physical impact and vulner�ability to landslide and flood has been analyzed interms of the degree of losses and the socioeconomicand physical capability of the people and infra�structure to respond to and recover from thepotential losses from landslides, debris flow, floods,and riverbank cutting in the watershed based oninformation collected from both group discussionsand household surveys.

Seminars and Public AwarenessProgramme

One day seminar was held in Tansen with therepresentative of key stakeholders and organization

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Weight: Geology

Weight: Lineament

Weight: Slope gradient

Weight: Slope aspect

Weight: Slope shape

Weight: Relative relief

Weight: Landuse

Weight: Water table

Weight: Drainage density

Weight: Distance from road

Weight: Vegetation density

LandslideOccurrence map

Calculate the weight in the attribute table for each parameter:Wi = ln Dens Class/Dens MapWhere,Wi = the weight given to a certain parameter classDens Class = the landslide density within the parameter classDens Map = the landslide density within the entire mapln = Natural logarithm

Attribute tables of weight of each parameter class

Sum of weight maps

Landslidehazard map

Topography

Old channelcourses

Ground watertable

Lithology/Structure

Historical flood dataDistance fromstream /rivulets

Flood hazard mapAttribute tablesof rating score of eachparameter class

Score maps ofeach parameter class

Land use/landcover map

Distance to road/trails

Populationdensity

Distance toirrigation cannel

Parameters layers

Reclassification

Reclassification

Reclassification

Map

com

bin

ation

Map

co

mb

inat

ion

Parameters layers

Rating

Rating

Attribute tables Weight maps

Sum of score maps

Score tables ofeach parameter class

Score maps ofeach parameter class

Sum of score maps

Vulnerability maps

Flood risk map

Landslide risk map

Geology

Lineament

Slope gradient

Slope aspect

Slope shape

Relative relief

Landuse

Water table

Drainage density

Distance from road

Vegetation density

Satellite imageAerial photoToposheetGeology map

Land system map

GPS observation

Sources

Overlay

Geology

74

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Figure 2. Flow chart for landslide and flood risk mapping

75

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

working in the watershed area, former president ofidentified high hazard and risk VDCs, VDCs secretary,and some members of local community developmentorganization were invited to share the results. As per the suggestion of seminar one�day discussionprogramme with the local community and represen�tatives of community development local organiza�tions in high hazard area was organized in Madni�phnat and in Chhap Koldada in order to raiseawareness of local community. The discussion wasconcentrated to identify hazard areas and neededemergency and mitigation measures.

The hazard and risk mapsprepared were provided tothe local community. Mutualrelationship was establishedwith District Committee ofParopakar, the local socialorganization. Maps and majorfinding were provided to thisorganization so thatcommunity�based programmeof public awareness andbioengineering measures isdeveloped to reduce the riskof landslide in Koldada.Paropakar is working in thedisaster awareness andbioengineering measure tolandslide. During theimplementation of projectcloser relationship was alsobeen established with NepalRed Cross Society, PalpaBranch, and Department ofSoil Conservation andWatershed Management,District Office that are activein disaster management inthe study area.

Results

About 202 active landslideswere identified based onaerial photo/satellite imageinterpretation, toposheet(1993) and field observation.Out of 81 landslide scarsidentified in aerial photo of1978, 63 landslide scars arefound stopped. About 103 old landslide scars wereidentified. These landslides

appear to be more dormant and more active thanstabilized type. Most of the landslides are foundcomposed of mud, soil and rock. About 55% oflandslides are found depth in nature (i.e. above 3mdepth) The two largest landslides in terms of thelength, coverage and depth occurred in watershedwith in last 40 years namely landslide of Masyamand Bausidadna of Chhap were found located incultivated land and most destructive in nature. Thelandslide of Chhap appears as dormant, whereas thelandslide of Masyam is active. Whereas The landslidehazard map as generated following above mentioned

Map 3. Flood hazard map.

Map 2. Landslide hazard map.

76

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

methodologies shows that about 17.6% land lie inthe high hazard zone and 36.7% in the moderatehazard zone (map 2).

In terms of flood hazard zone 11.5% of the land liesunder the high and moderately high hazard zone and4.7% of land falls under the moderate hazard zone(map 3). Regarding geographical location Madniphant, the fertile valley of watershed is foundseverely affected by frequent flood, channel shiftingand water logging. The problem of channel shifting,chute cut and neck cut is frequent in this part ofregion. An attempt was made to assess the impact ofhazards on different land use/land cover in order tounderstand how the people have been managing theland resources in the light of the probability ofoccurrence of damaging events of landslides andfloods. About 84.7% of total 27.04 km2 high andmoderately high flood hazard zone lie in cultivatedland, which is about 20.3% of the total cultivatedland of Tinau Watershed. It is also the most fragileland of the whole watershed. In terms of distributionof cultivated land in landslide hazard zone, only7.5% of the total cultivated land is in the high andmoderately high hazard zone, which is about 20% of

the total high and moderately high landslide hazardof the watershed. Whereas about 73.3 of highhazard and moderately hazard area lie in the forestlanduse, which, shares 27.5% of the total area offorest land.

Similarly, an attempt was made to assess thelocation of the house units in areas under varioushazard levels. Nearly 2% of the house/buildings, i.e.134 houses, are located in the high�hazard area and10.8% i.e. 707 houses, in the moderately high area.About 73.74% or 4844 houses are in the lowhazard area. Regarding the location of house interms landslide hazard area, 5.4% houses are inhigh hazard area, 74.4% of houses are located inlow hazard area and rest houses are in moderatehazard area.

Distribution of road length by road type by differenthazard type is shown in Table 1. There are manyactive and dormant landslide scars along theSiddharth Highway. Distribution pattern shows thatmore than one third length of highway of sub water�shed is lie in the high hazard zone. About 0.3 km

Area

in h

a.

190

80

70

60

50

40

30

20

10

0

HighModerateLow

Brush/shrubland

Cultivatedland

Forestland

Others

Figure 3. Landuse by landslide hazard

Brush/shrubland

Cultivatedland

Forestland

Others

Area

in h

a.

HighModerateLow

120

100

80

60

40

20

0

Figure 4. Landuse by flood hazard

High

Moderate

Low

20%

5%

75%

Figure 5. Distribution of house unit in differentflood hazard zone

High 2% Moderately

high

Moderate

Low

13%

11%

74%

Figure 6. Distribution of house unit in differentlandslide hazard zone

77

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

and 7.4 km of the major road and minor trailsrespectively lie in high flood hazard zone. About 4.04km feeder and major road and 24.873 km minortrails are in moderately high hazard area. For land�slide hazard 13.2 km out of 35.29 km highway, 6.3km of feeder road are in high hazard area (Table 1).

About 12.4% of the total area is under the highlandslide risk zone in terms of expected loss/damageto property. Similarly, about 31.3% and 56.3% of thearea is under moderate landslide risk and low land�slide risk zones respectively (Map 4). About 7.8% oftotal area is under the high flood risk zone in termsof expected loss /damage to property. Likewise, about7.4% and 84.9% area is under moderate and lowflood risk (Map 5).

As reported by focus group discussion out of 6716households in the watershed, 2327 households areexposed to hazards of different types. Out of them37.26% of households belong to Magar and otherhill ethnic groups, and they are found more vulner�able to hazard. In terms of economic class about32.4% households are either marginal or small farmhouseholds. The settlement system in watershed istypical that the houses are located in the hill sideand the valley and flood plain area is mainly used forcultivation. There are very few houses located in veryhigh flood hazard zone in Madiphnat. So most of theproperties such as livestock, house/building, fodderand fruit trees etc. are safe from flood hazard in thisregion. However the cultivated land located in floodprone area is only means of livelihood of manypeople residing nearby areas. So frequent floodingand channel shifting adversely affect the peoplewhose land is located there. The canals, roads, cropsand cultivated land in flood prone area are annuallyswept away by river flooding and regular channel

shifting. In the northern part of the Madiphant somesettlements are situated in very high flood risk zone.Mostly affected people form river bank cutting,channel shifting and flood in this area are poorlandless tenant families. Out of 39 households ofMadiphnat surveyed 12 households were of poortenants.

The estimated value of element of risk (excluding liveloss) in the watershed is Rs. 5.4 billion. The highestvalue in susceptible zone is Rs. 1.9 billion inMadanpokhara VDC. The loss of past major waterinduced hazard in watershed is found varies with theintensity of events. During the last 42 years the lossfrom river bank cutting is Rs.19.8 billion with perevent loss of Rs. 800 thousand. Like wise the lossfrom major floods in last 42 years is Rs. 28.2 billionwith the per event loss of Rs.2.6 million. The totalloss from landslide events in past 42 years is 6.5billion with per event loss of Rs. 300 thousands. Theloss from water logging in Madiphnat since last 13years is Rs.39 million with per annual loss Rs. of 3million. About 86 to 98.3% of total exposed value isprivately owned. The past major events of flood,landslide and river shifting, total estimated loss, perevent estimated average loss and the recurrenceinterval of such major hazard in study area is shownin Table 2.

The damage to road and road closure due tolandslide is the common phenomena of thiswatershed. During the past three years the totalvalue of Highway repair after the damage bylandslide is Rs 714400. The maximum loss from roaddamage by landslide is Rs. 420000 during the lastthree years. During this period altogether 39landslide events caused the Highway closure for446.5 hours (Table 3).

Table 1. Length of Road (in km) by different hazard type

Road typeFlood hazard Landslide hazard

High Moderately High Moderate Low High Moderate Low

Highway 0.295 4.043 1.696 29.259 13.191 10.931 11.161

District road 2.013 2.443 4.455

Feeder road 0.32 1.501 26.396 6.3 6.997 14.834

Other road 0.211 6.746 0.865 0.609 5.477

Trail 7.355 24.873 23.101 356.349 52.159 143.017 216.191

78

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

Map 4. Landslide risk mappling in Tinau Watershed

Map 5. Food risk mappling in Tinau Watershed

79

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

VDC Duration Period No. of Event

RecurrenceInterval

Total Loss (Rs.)

Per Eventloss (Rs.)

Flood

Chitrundhara 2038�60 22 4 5.5 2710680 677670

Bandi Pokhara 2038�56 18 2 9 214000 107000

Madan Pokhara 2018�60 42 33 1.27 116644000 3534667

Kusumkhola 2041�60 19 8 2.38 480000 60000

Kaseni 2018�60 42 33 1.27 4290000 130000

Jhadewa 2054�60 42 19 2.2 152000 8000

Telgha 2032�60 28 9 3.1 140460003 15606667

Pokharathok 2018�60 42 33 1.27 1485000 45000

Koldanda 2038�55 17 3 5.6 1253800 417950

Gothadi 2038�52 14 3 4.7 3000000 1000000

Chidipani 2018�60 42 33 1.27 9570000 290000

Humin 2018�60 42 19 2.2 1567500 82500

Devinagar 2018�60 42 6 7 385260 64210

Rupse 2018�60 42 33 1.27 231000 70000

Major Landslide

Tansen 2038�59 21 7 3 3109100 444167.7

Chirtundhara 2036�45 19 9 2.1 1431000 159000

Bandi Pokhara 2038�55 17 5 3.4 1707000 341400

Madan Pokhara 2027�55 28 13 2.2 10936250 841250

Thimure 2032�60 28 42 0.67 4171999 99333.33

Kusumkhola 2042�60 18 16 1.13 3200000 200000

Masyam 2027�55 28 40 0.7 333333 83333.33

Khasauli 2036�57 21 28 0.75 6499080 232110

Koldanda 2038�60 22 68 0.32 28786664 423333.3

Rupse 2042�60 18 12 1.5 3264000 272000

Gothadi 2037�52 15 11 1.36 1540000 140000

Major river bank cutting/channel shifting

Dobhan 2018�60 42 53 0.8 2863590 54030.43

Kaseni 2018�60 42 64 0.7 3040000 47500

Jhadewa 2018�60 42 81 0.5 32400000 400000

Pokharathok 2018�60 42 83 0.5 1660000 20000

Humin 2018�60 42 63 0.6 126000000 200000

Chidipani 2018�60 42 81 0.5 864003 10666.7

Madanpokhara 2018�60 42 93 0.4 23250000 250000

Rupse 2018�60 42 83 0.5 8300000 100000

Table 2. Recurrence interval and estimated loss of major hazards in Tinau watershed

Source: Field Survey,2003

80

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

The strategies adopted by household to minimize therisk of landslide, flood and other geomorphic hazardsinclude evacuation from hazard area to other area,construction of small structure to control rive bankcutting and landslide, retaining wall, and treeplantation. The local people are also adopting cropcalendar and stock of certain grains/crops for futureinsecurity to cope with uncertain disaster in future.Similarly, in Madniphnat area the local communityhave settled their settlement in hillside area, and theflood plain is used for only cultivation purpose. It alsohelped them to safe the lives and loss of other assetsfrom frequent flood hazard.

However the condition of poor people is found tosevere. There always food deficit to the landless,marginal and poor people. The information collectedon focus group discussion shows that about 17.2%households of watershed area lives in food deficiencycondition for more than 6 months and 21.3%households have food deficiency of about 1 to 3months. The household survey of high flood andlandslide hazard area shows that about 53% ofhouseholds have not any such food stockingmechanism due to the food deficiency. Most of thedisaster stricken poor households are found earningother lands as a tenants. However, the underprivi�leged caster and ethnic and women headed house�holds have to pay additional production (certain extraportion) locally called Gunjais to the landholder.

The impact of disaster between gender is also foundunequal in terms of workload, decision makingpower, financial status, and roles and responsibilities.It has major impact on women especially in times ofdisaster who disproportionately assume a greaterresponsibility of family.

The strategies adopted by other GOs and NGOs aremostly post disaster measures and relief distribution.Some structural works have been carried out. It isonly after the disaster events. Most of such effortsare found driven by the traditional relief and disasterpreparedness, centralized top down approach.Contrast to it some local organization, and NGOsand District Soil Conservation office recently imple�menting the community based disaster mitigationprogramme. In this approach local communities areexpected to be involved in decisions from which theywere previously excluded. However, in many casesthe community�based approach also found unable toaddress the social difference across gender, class,castes etc., which also highly differentiate the impact,response and recovery capacity of people towarddisaster. The underprivileged community member like

DateBlockageduration

in hrs

Landslidetype

Cost(Rs.)

17/6/03 22 Rock fall 35000

18/7/03 14 slide 35000

18/7/03 3 slide 50000

17/7/03 3 slide 10000

17/6/03 2 Rock fall 3875

16/6/03 16.5 slide 12400

2/7/03 5 slide 4000

4/7/03 5 slide 4000

1/10/02 5 slide 285000

8/10/02 2 slide 110000

2/10/02 6 slide 3000

25/9/02 3 slide 1670

2/7/02 2.5 slide NA

17/6/02 21 slide 18000

5/6/02 6 slide NA

17/7/01 42 slide NA

17/07/01 33.5 slide NA

19/7/01 45 slide NA

25/6/01 3 slide 4200

22/06/01 2 slide 420000

29/5/01 9 slide 5500

31/07/01 3 slide 4500

1/8/01 5 slide 7500

18/8/01 2 slide 3000

19/7/01 60 slide 80000

22/8/01 6 slide 9000

24/8/01 7 slide 10500

19/7/01 3 slide 4500

22/9/01 6 slide 9000

19/7/01 2 slide 3000

3/8/01 13 slide 7200

30/7/01 slide 7500

30/7/01 2 slide 3000

1/8/01 slide 3000

5/8/01 81 slide 106000

23/8/01 4 slide 10000

29/7/01 2 slide 7000

2/8/01pattiallyblocked

10000

Table 3. Major events of highway closure bylandslide and value of repair after the damage(2000–2003)

women, so called Dalit, and ethnic people are foundexcluded from decision�making and access andcontrol over resources. During the period of focusgroup discussion, seminar and later discussionprogramme local people raised such issues. Accordingto the local community member the poor and under�privileged member of community have beenpurposively excluding from the relief distribution. Inmany cases, the structural construction measures areconstructed suppressing their views on behalf of elitemembers of community.

Lesson Learned

The GIS and remote sensing tools can fruitfully beused for landslide and flood hazard and risk assess�ment. The risks indicated by the combination ofvulnerability maps (based on population, economicvalue of the property, and infrastructure) with hazardmaps could be useful in prioritizing areas for theimplementation of disaster preparedness plans andmitigation measures.

The failure of past relief and post disaster activities indisaster management activities shows the lack ofparticipation of local people in the entire process ofdisaster management activities. There is need for theestablishment of local institutions responsible fordisaster preparedness. Provision should be made tostrengthen such institutions through training andtechnical and financial support. Efforts should also bemade to create awareness among the local people.

Mass poverty and low level of off�farm activities,illiteracy, and poor service facilities have contributedto the low response and recovery capacity to dealwith disasters. Unless the response and recoverycapacity of the local people is improved, the loss anddamage are likely to increase. Therefore, disasterreduction and preparedness strategies should includecomponents such as poverty reduction and empower�ment of women and other disadvantaged groups inthe community and overall development activities insustained way.

Structural measures have been adopted in someareas in order to reduce the risk of landslides, debrisflow, floods, and riverbank cutting. In the absence ofconstruction standards and regular monitoring andmaintenance of already constructed structures, therisks to disasters have further been increasing. It is in this context that equal emphasis should be givento monitoring and maintenance of the structuresalready developed. Similarly, efforts should be made

81

to develop construction standards for buildings andinfrastructures.

Prevailing high conflict among different groups ofcommunity in terms of resource use, decision makingin community, access, share and controls overresources use or entitlement and access approachshould be addressed in any disaster managementactivity.

Early warning system has not yet been developed.Keeping in view the lead�time of flooding betweenhighland and lowland areas, it is essential to developa mechanism of community�based warning system.The magnitude of landslide, debris flow, and floodevents can be reduced if strong conservationmeasures are implemented.

Mechanisms to share the cost and benefit ofconservation measures between highlands andlowlands should be introduced. It is in this contextthat the highland�lowland interaction should betaken into consideration while selecting areas toimplement community based disaster managementprogrammes in the future.

Reference

ADPC, 1991, Disaster Mitigation in Asia and the Pacific,Manila: Asian Disaster Prevention Center.

Carter, Nick W., 1991, Disaster Management : A DisasterManager’s Handbook, Manila: ADB.

Chettri, M.B. and Bhattarai, D. 2001. Mitigation andManagement of Floods in Nepal, Nepal: MOH/HMG

Deoja, B.; Dhital M.; Thapa B.; and Wagner, A. 1991. Eds.Mountain Risk Engineering Handbook. Kathmandu, Nepal:ICIMOD. Vol. 1 and 2, 875 p.

Dhital, M. R., Khanal, N. R., and Thapa, K. B. (1993). The roleof extreme weather events, mass movements and land usechange in increasing natural hazards: A report of thepreliminary field assessment and workshop on causes andrecent damage incurred in south central Nepal (July, 1993),ICIMOD, Kathmandu

DHM(1998). Climatic records of Nepal. Department ofHydrology and Metereology, Kathmandu.

DHM(1998). Hydrological records of Nepal. Department ofHydrology and Meteorology, Kathmandu, Nepal.

Fordham, Maurren and Ketteridge, Anne�Michalle, 1995,Flood Disasters�Dividing the Community, (A Paper Presentedat the EMERGENCY Planning ‘95 Conference 2�6 July 1995),Lancaster.

Ghimire, M (2000) Slope instability assesment, hazard andrisk zonation using Remote sensing and GeographicalInformation System: A case study of Banganga watershed inthe Mahabharata and the Siwaliks of Central Nepal, reportsubmitted to CSSTE�AP, Dehradun, India, Unpublished

WO

RK

ING

PA

PE

RS

Flo

od

s &

Ero

sio

n

Paudel, Keshav Prasad,2001, Landslide and its Consequencesin Nepal (1971�200), Seminar report submitted to CentralDepartment of Geograpy,T.U.

Poudyal Chhetri, Meen B. and Bhattarai, Damodar (2001),Mitigation and Management of Flood in Nepal, Kathmandu:Ministry of Home Affairs/HMG Nepal.

The Integrated Land and Water Information System: ILWIS(1997). User’s Guide, ILWIS Department, InternationalInstitute for Aerospace Survey and Earth Science Enschede,The Netherlands. Pp 419�420.

Tianchi, LI.(1999). Mountain hazard mitigation and riskengineering in the Hindu Kush�Himalayan Region, ICIMOD,Kathmandu

UNDP and ICIMOD, 2001, Community Risk and VulnerabilityAssessment, Kathmandu UNDP, ICIMOD.

UNDP and ICIMOD, 2001, Hazard and Risk Mapping,Kathmandu: UNDP, ICIMOD.

82

WO

RK

ING

PAP

ER

S

Floo

ds &

Ero

sion

1 HMG/N, Ministry of Home Affairs, 2000

83

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Indicators for disaster

preparedness

TEAM LEADER

Dash Biswanath <[email protected]>

PROJECT ADVISOR

Prof. Vinod Kumar Sharma

India

SYMPOS IUM NOTES

84

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

The project aims at building an instrument to assessdisaster preparedness for a set of hazards (earth�quake, drought, cyclone, flood, fire and epidemics) ata community level. Such instrument is expected to behighly useful for funding agencies and policy makersin making objective evaluation and monitoring ofpreparedness programs. It would also provide feed�backs for designing appropriate programs and beuseful as guidelines for implementing agencies. Thespecific objectives of the project are a) to identify thekey parameters involved in community preparednessb) to select a set of appropriate indicators formeasuring such parameters c) to design an indexingmethod for combining these indicators together d) totest�trial the instrument (i.e. the set of selectedindicators together with the indexing technique)adequately in different communities for verifying itsreliability and e) to disseminate assessment resultsamong those communities for their opinion andawareness. The most important feature of this aboveinstrument is its representation of social, economicand cultural condition of people in the selection ofindicators and thereby providing direct link betweenpoverty and risk reduction.

Methodology

The project methodology can be described to beconsisting of five stages. In the first stage, compre�hensive literature review was undertaken to definecommunity preparedness and suggest a conceptualframework that explains community preparedness onthe basis of ten parameters. To measure theseparameters, total of eighty�five indicators wereidentified as possible candidates. In stage�II, DelphiAnalysis was carried out where a panel of twentyexperts were constituted and presented with theseset of indicators to select the most appropriate onesamong them. Based on their opinion, total of thirty�five indicators (i.e. if all the six hazards areapplicable) were finalized. In stage� III, an indexingmethod was developed. This method while combiningvarious quantitative and qualitative indicators alsotried to be simple and make minimum use ofstatistical tools. In stage�IV, the designed instrument(i.e. the set of selected indicators together with theindexing technique) was put into practical use forassessing preparedness in six communities located inthree different states of India. These communities areamong the most disaster prone in the country anddiffer considerably in terms of hazards they face andtheir cultural practices etc. Data collected from suchcommunities were fed into the designed instrumentfor measuring preparedness. In stage V, the findings

85

from such assessment were disseminated among thesame communities for member’s opinion aboutsuch assessment and also to create awarenessamong people about probable hazards and theirpresent state of preparedness.

Stage-I

What is preparedness?

Preparedness as a core concept in disaster studieslacks consensus in its conceptualization.1 It isdefined in different ways ranging from actionoriented steps to education. The various approachesto preparedness can be classified under three broadcategories2 a) on the basis of attributes e.g.existence of warning system, awareness, availabilityof resources etc. b) preparedness as scenarioplanning e.g. drawing possible disaster casescenarios and making plans for it3 and c)preparedness as psycho�social processes whichemphasizes individual decision making processes 4.Preparedness can also be thought of at differentlevels e.g. individual, household, organization,community, country etc.5 For example, Cottrell et al.(2001) have found that within a community even ifindividuals are reasonably well prepared, thecommunity as a whole may not be prepared. Thus itis important to recognize the difference that existsbetween these levels as procedures for assessmentcan vary depending on such levels6. The objective ofpreparedness is often described to be “to enhancethe ability to respond well.”7 In such conceptuali�zation, preparedness activities are mostly orientedtowards good response (e.g. warning dissemination,evacuation plan, stockpiling of essential suppliesetc.) and fails to emphasize post� disaster recoveryaspects. In the context of developing countrieswhere majority of people live under poverty, post�disaster recovery however is as important assurviving the initial impact.8 Thus preparedness forsuch societies must look beyond the response stageand need to include post�disaster recovery as one ofits important objective.9

Defining community preparedness

The way preparedness is conceptualized in this studyis viewing it as a state of readiness (that may be atan individual, household or community level) to facehazards at a given point of time. It is a constantlychanging process. The dynamic nature of the conceptcan be seen from the way preparedness varies evenduring different times of a day e.g. people receivingwarning at night are relatively less prepared for

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

evacuation. Similarly such change can also occur dueto variety of other reasons such as an official’s visit,a game of football or a local festival. Communitypreparedness is thus defined as the state of readi�ness of the community to face probable hazards sothat minimum of losses occur from it and smoothrecovery takes place. This view of preparednessrelates it to the concept of mitigation as; prepared�ness providing measure of effectiveness of variousmitigation efforts. To illustrate further, existence of acyclone shelter (structural mitigation) in a communityif not located properly or awareness campaigns (i.e. non�structural mitigation measure) similarly if not well planned do not contribute to the overallpreparedness of the community. In other words,effectiveness of these mitigational measures deter�mines to a large extent the level of communitypreparedness.

Measuring community preparedness

There have been several previous attempts tomeasure disaster preparedness of a community withindicators though, most of such attempts are limitedto specific hazard only e.g. for earthquake10, wild�fire11, terrorism12 etc. Cottrell et al. (2001) in theirattempt have tried to assess community�prepared�ness for multi�hazards and have identified fourparameters as important; a) population character�istics e.g. number of children, squatter settlement etc.b) building and critical infrastructures such as road,drinking water, communication network, health,

sanitation etc. c) physical environment and d) socialenvironment e.g. ethnic groups.

The framework suggested here for measuringpreparedness tries to build on this work of Cottrell etal. (2001). It views community preparedness to be anumbrella construct made of ten parameters (Fig.1)which are; a) Physical safety i.e. how safe are thecommunity members in view of the physical dangerfrom these hazards? The parameter essentially triesto measure how effective structural mitigationmeasures are e.g. availability of cyclone shelter, itscapacity, resistance of building structures for earth�quakes etc. b) Hazard awareness i.e. awareness levelabout hazards which have reasonable probability ofoccurrence c) Organizational preparedness i.e. howfar the community is organized to face disaster e.g.existence of disaster committee, plans, volunteers etc.d) Infrastructures and Services which tries tomeasure current state of these services and how wellplans and procedures have been developed forrestoring critical services as and when disruptionsoccur e) Recovery ability i.e. ability of the communitymembers to recover from the impact of the hazard f) Physical environment i.e. state of environment toface hazards e.g. condition of sub�surface aquifers,vegetation cover etc. g) Social capital i.e. degree towhich social networking and cooperation existsamong community members h) Psychologicalpreparedness i.e. how safe and prepared do community members feel in view of these hazards?i) Cultural capital i.e. cultural richness such as

86

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Hazard awareness Physical safety

Hazard dependent parameters

Community prepareness

Social capitalInfrastructure and servicesOrganizational preparedness

Household preparedness

Psychological preparedness

Physical environment

Recovery ability

Cultural capital

Figure-1: Community Preparedness (Suggested framework)

existence, recognition and use of traditional copingmechanisms j) Household preparedness i.e.preparedness at a household level. Three of theabove ten parameters (i.e. physical safety, hazardawareness and physical environment) can be seen tobe hazard dependant which, means to measure theseparameters one needs to consider the specifichazards which are applicable for the community andbased on it select indicators. The hazard independentnature of most of the parameters in communitypreparedness emphasize that preparedness programsshould focus as much on the disaster agent as itshould try and reduce existing structuralvulnerabilities.

Stage-II

Delphi analysis

The framework as suggested in Fig.1 views thatmeasuring community preparedness requiresmeasuring the parameters involved in it and thencombine them for getting overall preparedness.Literature search was conducted, resulting in theidentification of eighty�five indicators (complete listof these indicators are given in project report,appendix�B) considered as potential candidates formeasuring the ten parameters. Preliminary reviewe.g. modification and reformulation of theseindicators reduced its total number to fifty�eight. Toselect the most appropriate ones from these fifty�eight, a panel of twenty experts were constituteddrawn from different backgrounds1 and werepresented these indicators in questionnaire form fortheir evaluation. The exercise was completed in twophases over a period of three weeks. (DelphiAnalysis) The expert’s selections were subsequentlytabulated to rank these indicators and determinetheir suitability to measure individual parameters.Subsequent review and analysis provides thefollowing list of thirty�five indicators presented underthe corresponding parameter as the most suitableones for measuring. The numerical figures given atthe end of each indicator represents its relative score(in the Delphi Analysis) i.e. number of expertschoosing the indicator for its suitability.

Physical safety: The following five indicators areselected (one of which is a composite indicator) tomeasure the parameter though exact selection wouldalso depend on the specific hazards applicable forthe community.

• Earthquake: Percentage of houses in thecommunity that are earthquake�resistant (20)

87

• Cyclone: Adequacy of the cyclone shelter i.e.maximum accommodation capacity of the sheltersas a percentage of total population (20)

• Drought: Public works program i.e. readiness ofconcerned agencies to initiate such programswhich includes response time for implementingthe programs (15)

• Flood: Percentage of population with floodprotection that includes non�structural measuressuch as flood proofing etc. (14)

• Health Care Services i.e. for all hazards: Theindicator is a composite one which takes threesub indicators for its value. (20)

– Distance of the hospital from the community – Structural resistance of the hospital building for

probable hazards – Kind of hospital i.e. the category should be

indicative of available facilities

Hazard awareness: Three indicators have beenselected to measure hazard awareness.

• Percentage of population which attendedawareness campaigns in last five years (15)

• School curriculum and adult literacy materialscontaining hazard information (12)

• Awareness about probable hazards (awarenesshere is to be considered as knowing aboutprotection measures and how to reduce thehazard impact) (10)

Organizational preparedness: Six indicators areselected to measure the parameter.

• No. of volunteer per hundred population (17)

• Average number of training days per volunteer(17)

• Existence of disaster committee and how welldifferent social groups e.g. minorities and womenetc. are represented in such committee (15)

• Acceptance of the disaster committee i.e.percentage of population who accept that thecommittee is representing their interests (15)

• Comprehensiveness of the disaster plan. The planis to be measured against criteria such as hazardanalysis, vulnerability analysis, resources inventory,provision for disabled, senior citizen and childrenetc. (15)

• Plan awareness i.e. percentage of populationaware about the plan and its content (14)

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Infrastructure and services: Five indicators arechosen to measure readiness of the community interms of infrastructure and services.

• Emergency plan for hospital i.e. if such plan exists,it is to be evaluated against criteria such asdisease surveillance, alternative power, triagingtraining etc. (16)

• Access to warning i.e. what percentage ofpopulation can access warning through radio andtelevision or through any informal warningnetworks (14)

• Emergency plan for ensuring drinking water supplyduring disaster (12)

• Condition of the approach road to the community(12)

• Response time for Fire station i.e. time required toreach the community (10)

Recovery ability: Three indicators have beenselected to capture the parameter.

• Per capita income per month (16)

• Access to insurance services i.e. percentage ofpopulation with insurance coverage such aspersonal, crop, house etc. (14)

• Access to formal credit services i.e. percentage ofpopulation with access to banks and other suchinstitutions e.g. cooperative societies for gettingcredits (10)

Social capital: Three indicators are selected formeasuring social capital and communitycohesiveness.

• Number of operational community�basedorganizations (16)

• Joint hazard mitigation efforts undertaken by thecommunity in last five years (14)

• Cooperation among community members i.e.percentage of population who expect to get helpand cooperation from other members whenneeded (13)

Physical environment: The parameter beinghazard dependant, selection of indicators for it wouldbe based on specific hazards applicable for thecommunity.

• Drought:– Awareness among farmers about sustainable

farming practices (14)

– Condition of the sub�surface aquifers (9)

• Flood: Area under vegetation cover i.e. as apercentage of total area (10)

• Epidemics: Availability and quality of drinkingwater, level of sanitation (7)

Psychological preparedness: Two indicators havebeen selected to measure psychological preparednessof community members.

• Perceived preparedness i.e. percentage ofpopulation who feel they are prepared (10)

• Recognition and provision for psycho�counsellingin the disaster plan (7)

Cultural capital: The parameter is represented byonly one indicator i.e.

• Percentage of population aware about traditionalcoping mechanism and recognize its usefulness. Ittries to measure how such traditions arerecognized and promoted. (12)

Household preparedness: The parameter is alsomeasured by one indicator i.e.

• Percentage of households in the community whohave family plans for disaster or have takenmeasures such as food grain reserve, purchase ofinsurance or any other. (12)

Stage-III

Indexing Method

The indicators described above when analyzed can beseen to be both quantitative and qualitative in natureand for making maximum use of such indicators; theyneed to be indexed together. The method suggestedhere for indexing while emphasizing internal andexternal consistency also tries to keep it simple andmake minimum use of mathematical or statisticaltools so that such method can ultimately be expectedto be used by community member themselves. Themethod involves three stage of indexing for a para�meter. In the first stage, all the indicators are ratedon a four point scale (Bad, average, good and verygood). The quantitative indicators (twenty�threeselected indicators are quantitative) are rated on thebasis of eight designed scales2. The remaining twelveindicators are qualitative and have been rated direct�ly on the basis of specific criteria. In the secondphase, rating of these indicators (made in the firstphase) are converted into numerical scores asfollows; bad: 0, average: 1, good: 2 and very good: 3.In the third phase, the arithmetic mean of theseindicator scores are calculated for each parameterand converted into percentages for getting parameter

88

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

score. The average of all the ten parameters (all para�meters are given equal weightage) provides theoverall preparedness for the community measured inpercentages.

Stage-IV

Field-testing of the instrument

The set of indicators selected and the indexingtechnique described above (henceforth theinstrument) was put into practical application forassessment of community preparedness. To ensurethat the instrument is robust and reliable underdifferent condition, it was field�tested in six mosthighly disaster prone communities located in threestates of India (Gujarat, Orissa and Andhra Pradesh).The selection of these communities was made on thebasis of the following criteria; a) the communitiesshould be facing different set of hazards so thatfindings can be generalized for all the six hazards b)they should differ from each other as much possiblee.g. in their level of development, past disasterexperience, cultural practices etc. c) one of thesecommunities need to be relatively better prepared(control group) so that assessment results from othercommunities can be compared with it for validation.Taking the above factors into consideration, thefollowing six communities (panchayat 3) wereselected.

• Valadia Bitta (East) in Anjar Taluka, Kutch district,Gujarat

• Manginapudi in Machlipatnam Mandal, Krishnadistrict, Andhra Pradesh

• Mundpadar in Bangomunda block, Bolangirdistrict, Orissa

• Gupti in Rajnagar block, Jagatsinghpur district,Orissa

• Rampar in Anjar Taluka, Kutch district, Gujarat

• Kallana in Rasulpur block, Jajpur district, Orissa

Hazard assessment

To assess community preparedness for multi�hazardsrequires addressing the complicated issue of hazardassessment. There are two major problems faced herei.e. a) giving weightage to probable hazards e.g. howto differentiate between a dominant recurring hazardand a hazard which is infrequent but with reasonableprobability of occurrence and b) how to selecthazards for preparedness�assessment given thatthere may exist difference between scientific assess�ment and the community assessment of hazards and

89

the necessity to include them. The approach in thiswork to resolve the above problems has been a) togive equal weights to all hazards on the basis thatall probable hazards need to be adequately preparedfor and should not be discriminated on the basis oftheir frequency of occurrence and b) hazard selectionfor a community should be based both on thescientific assessment as well as that of thecommunities. Thus hazards selected (i.e. out of thesix hazards which are under consideration) for eachcommunity in this study is based on taking bothviews i.e. opinion of concerned scientific agenciesand that of community members.

Data collection

Data required for using the instrument werecollected from different sources and throughdifferent techniques such as survey, interviews andfocused group discussions. Two kinds of question�naire surveys were conducted in each of the sixcommunities i.e. a) among head of households andb) among general population. The second survey wasnecessitated for data relating to hazard awarenesswhich required equal representation of women andchildren in the sample. Data were also collected fromkey informants such as senior citizens; opinionmakers etc. and from secondary source such ascensus hand books, government reports, localnewspapers etc. Data after being collected were putinto the designed instrument and indexed andresults obtained for the six communities are givenhere in brief.

Case-I: Valadia Bitta (East), Gujarat

Hazards considered for the community: Earthquake,drought, cyclone, epidemics and fire

The community is one village with a total populationof 779 i.e. 169 households. Literacy level is around25% and agriculture is the major occupation. Theoverall preparedness of the community as assessedthrough the instrument is 33% and the scores forindividual parameters are as follows. Physical safety:42%, hazard awareness: 23%, organizationalpreparedness: 39%, infrastructures and services:47%, recovery ability: 23%, social capital: 56%,physical environment: 23%, psychologicalpreparedness: 50%, cultural capital: 16% andhousehold preparedness: 8%.

Case-II: Manginapudi, Andhra Pradesh

Hazards considered for the community: Cyclone,drought, fire and epidemics

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

The community is one village located near thepopular tourist destination Manginapudi beach witha total population of 1940 i.e. 473 occupiedhouseholds. The community suffered heavy lossesduring earlier cyclones of 1971 and 2003. Drought isalso a recurring hazard here due to the ingress ofsaline water. The overall preparedness for thecommunity is measured to be 37% and the scoresfor individual parameters are as follows; physicalsafety: 45%, hazard awareness: 27%, organizationalpreparedness: 34%, infrastructures and services:60%, recovery ability: 45%, social capital: 56%,physical environment: 23%, psychologicalpreparedness: 34%, cultural capital: 18%, householdpreparedness: 28%.

Case-III: Mundpadar, Orissa

Hazards considered for the community: Drought, fireand epidemics

The community consists of eleven villages with atotal population of 4838 i.e. 1046 households. It islocated in economically poor region of KBK (i.e.Kalahandi�Bolangir� Koraput districts of Orissa) fromwhere several cases of starvation deaths have beenreported earlier. Literacy level in the community is25%. Agriculture (i.e. mostly rain�fed, 7% of the totalagricultural land is irrigated) is the major occupationfor community members. Drought is thus a recurringhazard here in addition to fire and cases ofepidemics. The overall preparedness assessed for thecommunity is 31% and the individual parameterscores are as follows; physical safety: 50%, hazardawareness: 34%, organizational preparedness: 17%,infrastructures and services: 27%, recovery ability:34%, social capital: 45%, physical environment:34%, psychological preparedness: 34%, culturalcapital: 12% and household preparedness: 22%.

Case-IV: Gupti, Orissa

Hazard considered for the community: Cyclone, fireand epidemics

The community is spread over twelve villages with atotal population of 8266 i.e. 1266 households. It islocated close to the Bhitarakanika National Park andagriculture is the major occupation for thecommunity members. The general infrastructuralcondition such as road, communication, health careetc. is very poor. The community had suffered heavylosses during the 1971 cyclone which caused more

than 10,000 human deaths and since then anembankment (popularly known as saline embank�ment) has been constructed around the area as ameasure of protection against sea water surge. Theoverall preparedness for the community is assessedto be 29%. The individual scores for the parametersare as follows; physical safety: 34%, hazardawareness: 42%, organizational preparedness: 34%,infrastructures and services: 27%, recovery ability:12%, social capital: 45%, physical environment:34%, psychological preparedness: 17%, culturalcapital: 28%, household preparedness: 9%.

Case-V: Rampar, Gujarat

Hazard considered for the community: Earthquake,cyclone, drought and epidemics

The community is one village with a total populationof 627 i.e. 124 households. Literacy level in thecommunity is 35% and agriculture and fishing arethe major occupations. The overall communitypreparedness is measured to be 36% and theindividual parameter scores are as follows; physicalsafety: 34%, hazard awareness: 56%, infrastructuresand services: 54%, organizational preparedness:39%, recovery ability: 23%, social capital: 45%,physical environment: 23%, psychologicalpreparedness: 50%, cultural capital: 16%, householdpreparedness: 12%.

Case-VI: Kallana, Orissa (Control Group)

Hazard considered for the community: Flood,earthquake, fire and epidemics

The community was taken as the control group i.e. considered for being relatively better prepared.It consists of six villages with a total population of

5782 i.e. 902 households. Business and agricultureare the major occupation for the communitymembers and available infrastructures are reasonablygood. The community is located on the bank of theriver (Brahmani) and an embankment runs through it.The embankment had breached on an earlieroccasion causing substantial losses to the commu�nity. An interesting feature is that approximately 350 households (i.e. 40% of total) are located inbetween the river and the embankment i.e. in theflood plain itself and it provides an opportunity tostudy how people here live with the risk of annualflooding. The overall preparedness for the communityis 44%. The individual parameter scores are as

90

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

91

follows; physical safety: 45%, hazardawareness: 20%, organizationalpreparedness: 23%, infrastructures andservices: 54%, recovery ability: 56%, socialcapital: 34%, physical environment: 78%,psychological preparedness: 50%, culturalcapital: 47%, household preparedness: 27%.

Analysis of findings

The overall preparedness (as assessed usingthe instrument) of the six communities whencompared shows that it varies between 29 to

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

0%

50%

Measures

Case-communities

COMMUNITY PREPAREDNESS

Series1 33% 37% 31% 29% 36% 44%

I II III IV V VI

Figure 2. Graph showing overall preparedness of the sixcommunities.

0%

10 %

20 %

30 %

40 %

50 %

60 %

I II III IV V VI

Mea

sure

sM

easu

res

Case-communities

Case-communities

0%

10 %

20 %

30 %

40 %

50 %

60 %

I II III IV V VI

0%5%

10%15%20%25%30%35%40%45%

I II III IV V VICase-communities

Mea

sure

s

0%

10%

20%

30%

40%

50%

60%

70%


Mea

sure

s

0%

10%

20%

30%

40%

50%

60%


Mea

sure

s

0%

10%

20%

30%

40%

50%

60%

I II III IV V VI

Case-communities

Mea

sure

s

Figure 3. Physical Safety

Figure 4. Hazard Awareness

Figure 5. Organizational Preparedness

Figure 6. Infrastructures and Services

Figure 7. Recovery Ability

Figure 8. Social Capital

44%. The lowest score of 29% was for community IV(Gupti, Orissa) and the highest was recorded forcommunity VI (Kallana, Orissa) which was also thecontrol group or the community that was considered for being relatively better prepared. The individualscores of the parameters for the six communitieswhen compared (Figures 3�12) show high variance ineach case.

Analysis of these graphs show that overall prepared�ness does not follow any uniform pattern and highpreparedness in one or more parameters for acommunity can be neutralized if it does not scorewell in other parameters. For example, in community�V hazard awareness is 56% which is the highestamong the six but the same community scores poorlyin other parameters such as physical environment(23%), cultural capital (16%) and recovery ability(23%) and as a result its overall preparednessremains low at 36%. Similarly the control group(case�VI) scores highly in four parameters i.e.recovery ability (56%), physical environment (78%),cultural capital (47%) and psychological prepared�ness (50%) but its overall preparedness remainslimited to 44% mainly because of its poor scores inparameter such as hazard awareness, organizationalpreparedness and social capital. This also substan�

tiates the view that high per capita or better recoveryability by itself does not ensure better hazard aware�ness or organizational preparedness and thereforesuch parameters also need attention. One of themost important finding from this study is that thelowest value for overall community preparednessamong the six communities was found to be 29%which possibly indicates that every community carriescertain amount of inherent preparedness and what is important therefore is to identify the weak areas(or the relevant parameters) which need immediateattention. At the same time existing level ofpreparedness in other areas need to be supple�mented and raise it to an acceptable level so as toensure minimum of losses and efficient recovery from the occurrence of these hazards.

Stage-V

Dissemination of project findings: In the final stage of the project, assessment findings were tried to bedisseminated among the same communities wherethe instrument was field�tested. Such disseminationwas considered useful for the following reasons a) it provides an opportunity to get feedback fromthe community members about assessment made

92

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

0%10%20%30%40%50%60%70%80%90%

I II III IV V VI

Mea

sure

s

0%

10%

20%

30%

40%

50%

60%

I II III IV V VI

Mea

sure

s

0%

10%

20%

30%

40%

50%


Mea

sure

s

0%

5%

10%

15%

20%

25%

30%


Case-communities

Case-communities

Mea

sure

s

Figure 9. Physical Environment

Figure 10. Psychological Preparedness

Figure 11. Cultural Capital

Figure 12. Household Preparedness

b) it helps to create awareness among people inthese communities about their own state ofpreparedness c) such assessment can facilitate action from the concerned agencies and the localadministration. The strategy for dissemination focusedon three major groups; a) senior citizens, localleaders and opinion makers b) school children andc) local administration. Interviews and discussionswere conducted with the administration and localleaders for getting their opinions while disseminatingthe project findings. For creating awareness, postersin local languages giving essential information aboutprobable hazards were pasted at strategic places inthe community such as school, market places,community centres and other such places.

Conclusions

The field testing of the instrument in six communitiesand in different conditions show that the instrumentis highly robust and reliable for its use. The highvariances seen in parameter scores (Fig. 3�12), theanalysis subsequently conducted and the generalopinion received from community members about theassessment further substantiates the instrument’sreliability and thus needs to be considered forpractical use.

Limitations and scope for further research: In spite of the instrument’s demonstrated robustness, thereare however several limitations in this work e.g. theselected parameters are not the only ones possible.These parameters have also not been weighted foroverall preparedness and future work must considerthe weighting aspects of it. Similarly the panel ofexperts in the Delphi Analysis if changed, it may also bring in change in the selected set of indicators.There may be biases introduced during data collec�tion and translation. The instrument also needs to befurther tested for other hazards before generalizingto all hazards and for all developing countries.However, it is expected that with progressive use of it, the instrument gets further refined and morereliable for its use.

Reference

Cottrell, A., Cunlitte, S., King, D. and Anderson�Berry,L. (2001) “Awareness and Preparedness for NaturalHazards in a Remote Community: Bloomfield RiverRegion and Rossville” Centre for Disaster Studies,School of Tropical Environment Studies andGeography, James Cook University.

93

End Notes1 Gillespie, D.F. and Streeter, C.L. (1987) “Conceptualizing

and Measuring Disaster Preparedness” InternationalJournal of Mass Emergencies and Disasters,Vol.5(2):155�76

2 Kirschenbaum, A. (2002) “Disaster Preparedness: AConceptual and Empirical Re�evaluation” InternationalJournal of Mass Emergency and Disasters, Vol. 20(1): 5�28

3 e.g. see Dynes, R.L. (1994) “Community EmergencyPlanning: False Assumptions and Inappropriate Analogies”International Journal of Mass Emergencies and Disasters,Vol. 12: 141�58

also see Possekel, A.K. (1999) “Living with theUnexpected” Springer Verlag, Berlin: 187

4 see Enders, J. (2000) “Measuring community Awarenessand Preparedness for Emergencies” Australian Journal ofEmergency Management, Vol. 16 (3): 52�58

Asgary, A. and Willis, K.G. (1997) “Household behaviour inresponse to earthquake risk: an assessment alternatives”Disasters, 21: 354�65

5 Mileti, D. (1999) “Disasters by Design” Joseph HenryPress, Washington, D.C.: 215

6 For measuring household preparedness see Kirshenbaum(2002): End note�5,

For assessment of organizational preparedness seeGillespie and Streeter (1987): End note�4,

For assessment at country level see CDERA (2001) ‘Statusof Disaster Preparedness of CDERA Participating States”ID number: MIPR#OHCHS60120, Caribbean DisasterEmergency Response Agency

Also see FEMA (1997) “Capability Assessment forReadiness” CAR Report, http://www.fema.gov/rrr/carnew.shtm

7 e.g. see Tierney, K.J., Lindell, M.K. and Perry, R.W. (2001)“Facing the Unexpected” Joseph Henry Press, Washington,D.C.:27, see also Mileti (1999) End note�5

8 see Das, K. (2002) “Social Mobilization for Rehabilitation,Relief work in Cyclone affected Orissa” Economic andPolitical Weekly, Nov. 30, also see Narasimhan, S. (2003)“Lessons from Latur: A Decade after the Earthquake”Economic and Political Weekly, Vol. XXXVIII (48): 8�14

9 see Banerjee, M.M. and Gillespie, D.F. (1994) “LinkingDisaster Preparedness and Organizational ResponseEffectiveness” Journal of Community Practice, Vol. 1(3):129�142 also see,

Lewis, J. (1999) “Development in Disaster Prone Places”Intermediate Technology Publications, London

10 RusselL, L.A., Goltz, J.D. and Bourque, L.B. (1995)“Preparedness and hazard mitigation action before andafter two earthquakes” Environment and Behaviour, Vol.27 (6): 744�70

11 Rhodes, A. (2003) “Understanding CommunityPreparedness and Response to Wildfire Risk” paperpresented at Australian Disaster Conference, Canberra,http://www.ema.gov.au/

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

12 Manning, F.J. and Goldfrank, L. (2002) “Preparing forTerrorism: Tools for Evaluating the Medical ResponseSystem Programs” National Academy Press, Washington,D.C.

13 Twelve of these twenty experts are academicians, two are from government agencies one is from UN and theremaining five are from international and national levelnon�government agencies.

14 The scales were designed in consultation with experts inthe field. Twenty�one out of the total twenty�eightquantitative indicators have been rated using only onescale and the remaining using specific scales.

15 Panchayat is considered as a community. It is the lowestadministrative unit in India and may consist of one orseveral villages.

94

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

95

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Strengthening regional capacity

for disaster management

TEAM LEADER

Chipo Muvezwa<[email protected]>

TEAM MEMBERS

Patricia MamvotoChenai Mufanawejingo

PROJECT ADVISOR

Clever Mafuta

Zimbabwe

SYMPOS IUM NOTES

96

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

INTRODUCTION

Southern Africa Research And Documentation Centre(SARDC) is an independent regional informationresource centre which seeks to enhance the effective�ness of key development processes in the regionthrough the collection, production and disseminationof information, and enabling the capacity to generateand use information. There is a wealth of informationabout how disasters have been handled in the past and SARDC has collected such literature andcompiled a Disaster Management BibliographicDatabase. The World Bank Disaster Risk Reductionproject was undertaken to in an effort to consolidateinitial effort.

This document presents the context within whichdisasters occur in the African region as well as thesouthern Africa sub�region. It presents the regionaland sub�regional policy frameworks as they relate todisaster management. The document seeks tohighlight the lessons learned and makes recommen�dations based on experiences. This report suggestsareas for further intervention based on experiences in Africa rather than import ideas.

97

BACKGROUND TO DISASTER RISK IN AFRICA

The African continent is vulnerable to a wide rangeof natural and other disasters which have adverseeffects on societies, national economies anddevelopment objectives, as well as on critical humanand material resources. In this context, communitiesat risk across Africa find themselves even morevulnerable because of several aggravating factors,including poverty, environmental degradation,inadequate exchange of data and informationamong countries and organizations, and inadequateco�ordination of information flow from various levels.

Africa faces serious humanitarian crises with severelong�term consequences affecting all regions. Erraticrainfall, floods, cyclones, poverty, unsustainable debt,failing agricultural policies, household vulnerability,unfair international trade regimes, civil conflicts andinternal and cross�border displacement of peoplehave all contributed to the current situation. The HIVand AIDS pandemic has exacerbated the situation.

The continent is prone to various recurring climaticevents, including drought, floods, storms, cyclonesand resultant impacts on health, education andinfrastructure. Some of these need not be humandisasters if appropriate information is available forplanning and preparation. Events such as cyclones inMadagascar and Mozambique, floods across easternand southern Africa, and droughts in the horn ofAfrica and the Sahel underline the vulnerability ofcommunities across Africa and the need for longterm risk reduction activities including early warning,contingency planning, information sharing andtechnology transfer as well as disaster preparednessand response systems.

Apart from their immediate consequences on theenvironment, disasters have long term impacts onsociety and economic advancements which in manycases provoke serious setbacks in national develop�ment programmes. Poverty contributes significantlyto the vulnerability of communities at risk, and isalso made worse by the effects of disasters.

A well�designed system for disaster management iscrucial to Africa and to the success of its develop�ment plan, the New Partnership for Africa’sDevelopment (NEPAD). This is recognized among thekey responsibilities assigned to the new Commissionof the African Union in its Statutes. The continentlacks a central institution responsible for capacity

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Lessons Learned

building in disaster risk management, although thereare a wide network of organizations involved indisaster management initiatives. There is lack ofcoordination of activities and dissemination of infor�mation and best practices among organizationsinvolved in disaster management work, and yetinformation is an essential tool in disastermanagement.

Cooperation among African countries in the domainof disaster prevention and risk reduction can bestrengthened by adopting regional, sub�regional andnational mechanisms to improve the exchange ofinformation, sharing of experiences and knowledge,and technology transfer.

Disaster reduction is an essential element of regionaland government policy, and must be included indevelopment plans and strategies at all levels due toits multi�disciplinary and inter�sectoral nature. Thisrequires information access and sensitization workabout the impact of disasters and disaster reductionon the results and achievements of all sectors.

Further work on disaster management should reviewthe continent’s response to crises, prospects for therainfall season as it may affect the food securitysituation, as well as the continent’s general state ofpreparedness to face any other potential disastersfrom a multi�sectoral perspective. Such efforts in turn,enhance sustainable use of natural resources andenvironmental protection, and promote povertyreduction and sound governance practices in Africa.

A demographic analysis of worst affected commu�nities would reveal that women and children, beingthe most vulnerable in African society, bear thegreatest burden, and their needs must be an essen�tial ingredient in disaster management and reductionplanning. This should be supported by an appropriateinformation base.

Furthermore, lack of information about the multi�sectoral impact leads among other things to short�sighted policies and misdirected media coveragewhich in turn leads to a vicious cycle of lack ofappropriate information and policies, and fear.

Due to the immediate humanitarian needs associatedwith disasters in Africa, a wide range of organiza�tions get involved, especially at the national level,and efforts have inadvertently been duplicated.Similarly, a lot of conflicting information, includingstatistics, have been published with varying inten�tions and meaning. All these issues point to the needfor a coordinated regional network of organizations

that are involved in the generation, analysis anddissemination of information on disaster manage�ment, using parameters that have relevance to theregion. Capacities need to be built, strengthened andmaintained for this purpose, especially at sub�regional level.

One of the major problems encountered in disastermitigation world�wide is lack of coordinated infor�mation flow and lack of access to the right kind ofinformation at the right time, by those who can useit, for instance in policy planning. Harmonizedinformation, benchmarks and indicators used byNGOs, media, parliamentarians, governments and theprivate sector would help stakeholders involved inthe same area to work towards common solutions,common policies and ultimately common objectives.

There is a need to:

• better reflect and incorporate national perspectivesand priorities in sub�regional international policysettings;

• strengthen institutional capacities of sub�regionalorganizations to carry out policy relevantintegrated assessments on disaster management;

• strengthen sub�regional institutional capacities for the generation and dissemination of a widerange of response�related products to support the interpretation and comprehensive assessmentof disaster management at national and sub�regional levels;

• work in partnerships at all levels to achieve this.

Disaster Management and theAfrican Regional DevelopmentContext

NEPAD is a pledge by African leaders, based on acommon vision and a shared conviction, that theyhave a pressing duty to eradicate poverty and toplace their countries, both individually andcollectively, on a path of sustainable growth anddevelopment. NEPAD, as an agreed agenda of theAfrican Union (AU), is anchored on the determinationof Africans to extricate themselves and the continentfrom the malaise of underdevelopment and is rootedin the belief that a historic opportunity exists for asustained effort to advance human development andpoverty eradication.

What stands in the way of poverty and prosperity,and creates challenges for the NEPAD initiative, is thefact that the African continent is prone to disasters,

98

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

natural and human�made. This can negatively impacton the expected economic growth, setting backeducation and health goals, and especially theplanned advances through the development ofinfrastructure. Three cross�cutting, multi�sectoralchallenges must be confronted and overcome if theMillenium Development Goals (MDGs) contained inthe NEPAD strategy are to be met. These are thechallenges of conflict, of HIV/AIDS and of disastermanagement.

This project document addresses the challenges ofdisaster reduction through information managementand access.

The SADC Sub-Regional Context

The Southern African Development Community(SADC) recently came up with a long term visioningdocument the Regional Indicative Strategic Develop�ment Plan (RISDP). The RISDP, whose core focus is onpoverty reduction, is a regional integration anddevelopment framework setting the priorities, policiesand strategies for achieving the long�term goals ofSADC. It is intended to guide member states, SADCinstitutions, regional stakeholders and internationalcooperating partners in the process of deepeningintegration to turn the community’s vision into areality.

In line with commitments made by member statesunder the Millenium Development Goals (MDGs) andNEPAD, the RISDP identifies the following priorityintervention areas for sectoral cooperation andintegration.• Trade, economic liberalization and development;• Infrastructure support for regional integration and

poverty eradication;• Sustainable food security; and • Human and social development.

The SADC Policy Framework for Health provides acomprehensive coverage of all the key aspects ofhealth and health services delivery in the region andproposes policies, strategies and priorities in areassuch as health research and surveillance; healthinformation system; health promotion and education;HIV and AIDS and disaster management.

Article 25 of the SADC’s Health protocol proposesthat member states shall

a) co�operate and assist each other in the co�ordination and management of disaster andemergency situations;

99

b) collaborate and facilitate regional efforts indeveloping awareness, risk reduction,preparedness and management plans fornatural and man�made disasters; and

c) develop mechanisms for co�operation andassistance with emergency services.

Southern Africa is a region prone to variousdisasters, the commonest of which are drought,floods, storms, cyclones and epidemics. In the past the region has been subject to high levels ofboth internal and cross�border displacements ofpeople, mainly as a result of conflicts in Angola, TheDemocratic Republic of Congo (DRC), Mozambique,and former apartheid South Africa.

There is a wealth of information about how disastershave been handled in the past and SARDC hascollected such literature and compiled a DisasterManagement Bibliographic Database.

A well�designed system for disaster management is crucial for disaster prone southern Africa. The sub�region lacks a central institution responsible for disaster risk management capacity�building,although there is a wide network of organizationsinvolved in disaster management initiatives. There islack of coordination of activities and disseminationof information among organizations involved indisaster management work.

Questions are being raised whether the disastermanagement strategies being used by governmentsand NGOs ought to be overhauled. Further work on disaster management should review the sub�regional response to the humanitarian crisis, and the prospects for the rainfall season as it may affectthe food security situation, as well as the region’sgeneral state of preparedness to face any otherpotential disasters from a multi�sectoral perspective.

Such efforts should enhance sustainable use ofnatural resources and environmental protection, andpromote poverty reduction and sound governancepractices in the region.

Recommendations for the SADC Region

• Improvement of national and regional disasterpreparedness & mitigation systems

• Setting up an early warning system for disastermanagement

• Coordination of activities amongst policy makersand organizations

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

• Dissemination of information amongstorganizations involved in disaster management

• Establishment of a Central organizationresponsible for disaster risk management capacity building

• Support for the establishment of operationalrelationships among relief agencies, policyplanners, governments, development agencies,academics and researchers

• Harmonization of national, sub�regional,inter�sectoral and institutional responses todisaster management

IMPACT OF DISASTERS ONDEVELOPMENT: SOME INSIGHTS

Droughts and floods are endemic to southern Africa,and often trigger serious hydrological imbalances,causing loss or damage to human life, crops, live�stock and wildlife, infrastructure, a shortage of waterfor the people, as well as causing famine anddisease. Drought exerts a severe impact on a widerange of environmental and economic activities.

As a result of drought during the 1994�1995 season,cereal harvests in southern Africa declined by 35percent compared to the previous season, with maizeharvests falling by 42% (SADC, 1996). During the1991�2 drought, cereal production in Namibiadropped by 70 percent. Due to the massive cropfailure the SADC sub�region spent about US$2 billionon drought relief.

Over the last two decades, heavy floods havedevastated parts of the SADC sub�region, resulting inmassive damage to physical infrastructure, crops andlivestock, loss of lives, and public health hazards dueto water related diseases. In early 2000 and 2001,massive flooding and a cyclone caused severedamage with loss of life and property in Mozam�bique, South Africa, Zambia and Zimbabwe. Thegovernment of Mozambique reported that GDP grewby only 3 percent compared to 6 percent predictedbefore the floods. Agriculture production grew by 2 percent compared with 9 percent in 1999, andlivestock production fell to 4.3 percent compared tothe 21.3 percent in 1999.

100

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

1967�73This six�year period was dry across the entireregion. Some records show a severe drought in1967.

1974�80

This period was relatively moist over much ofsouthern Africa. In 1974 the mean annualrainfall was 100 percent above normalthroughout the region.

1981�82 Drought in most parts of the southern Africa.

1982 Most of subtropical Africa experienced drought.

1983A particular bad drought year for all parts of thecontinent

1984�85Near normal seasons, but drought strains fromthe previous three years were still felt in mostparts of the region.

1986�87 Drought conditions returned to the region

1988�90 Near normal season.

1991�92Severe drought in southern Africa, excludingNamibia.

1993�94 Conditions improved.

1994�95

Many SADC countries were hit by the worstdrought in living memory, surpassing effects ofthe 1991�92 drought in some parts of theregion.

1995�96Widespread rains in most parts of the SADCregion prompted forecast of a bumperagricultural yields.

1996�97 Normal rainfall for most of the region.

1997�98Normal rainfall throughout the region includingthe northeast, although impacts of the El Ninowere significant.

1999�2000

Cyclone Eline hit the region and widespreadfloods devastated large parts of the Limpopobasin (southern and central Mozambique, south�eastern parts of Zimbabwe, parts of South Africaand Botswana)

2002�03Significant drought in the SADC region,particularly from Zimbabwe northwards

Rainfall trends in southern Africa 1967 – 2003

Sources: Hirji et al 2002, Division of Water Environment and ForestryTechnology CSIR, Final Report: Protection and Strategic Uses ofGroundwater Resources in Drought Prone Areas of the SADC Region,Environmentek CSIR, Pretoria, 2003.

The socio�economic impact includes damage toinfrastructure, displacement and death of people,particularly among adults in the prime working agesand among children under five, economic decline,increasing number of orphans and lower humandevelopment. The impact on households isdevastating, with growing numbers of child�headedhouseholds and impoverished families.

While the impact of disasters is immediately felt inthe humanitarian sector, its tentacles reach far and

101

wide, cutting across all sectors, from agriculture toeducation, industry and commerce to investment,environment to gender. For instance, high mortalityrates among children and the productive sector ofthe population do not only mask demographic datasuch as life expectancy, but paint a negative pictureof issues such as economic and human developmentresulting in reduced investments in the region. Inaddition, damage to rural communities threatenfuture food security and by extension the health and education sectors.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s Land degradation, Zambia

Flooding, Zimbabwe

Ebola, Democratic Republic of Congo

Food distribution, Malawi

Flooding, Mozambique

Dead elephant, South Africa

DISASTER MANAGEMENTINFORMATION: SARDC INITIATIVE

In 1995, SARDC in partnership with the InternationalFederation of Red Cross and Red Crescent Societiesstarted a Disaster Management Information Project(DMIP). Its objective was to provide accessible andaccurate information on disasters and disaster relatedvulnerability in the southern African sub�region. FromDecember 2002 to March 2003 SARDC incollaboration with the UN�Institute of DisasterReduction (ISDR) resuscitated the project throughcollection building and updating the bibliographicdatabase. The bibliographic database containingmore than 2700 records can be accessed insearchable format on the following link:http://databases.sardc.net/wwwisis/resources.htm

To complement these initiatives, SARDC hasproduced a directory of organizations involved indisaster management work in southern Africa.

Strengthening Regional Capacity forDisaster Management Project

Under the ProVention Consortium Disaster RiskReduction project, SARDC�Information ResourceCentre (IRC) implemented the project StrengtheningRegional Capacity for Disaster Management coveringthe SADC countries i.e. Angola, Botswana,Democratic Republic of the Congo (DRC), Lesotho,Mauritius, Malawi, Mozambique, Namibia, Seychelles,South Africa, Swaziland, Tanzania, Zambia andZimbabwe.

The project produced a directory of organizationsinvolved in disaster management work in southernAfrica which will assist policymakers, researchers andorganizations involved in disaster related work to beaware of initiatives elsewhere.

Activities Undertaken

• Reviewed disaster management contactsdatabases, directories and websites.

• Listed organizations involved in disaster�related work.

• Designed and distributed questionnaire to disaster�related organizations.

• Designed database for disaster�relatedorganizations in SADC.

• Checked Websites and directories for profiles and followed up by distributing questionnaire to obtain accurate records.

• Data input, edit and compilation ofdatabase/directory.

How to access the directory

The information contained in the directory isarranged in alphabetical order first by the countryand then by the name of the organization.

The following shows the database layout, sampleinformation supplied by one of the respondents andis a reflection of the information as it appears in thedirectory.

Name of Organization Disaster Management & Mitigation UnitOffice of the Vice President

Acronym DMM�OVP

Type of Organization Government

Physical Address Plot 25B RoadRhodespark Lusaka, Zambia.

Postal Address P.O. Box 38963, Lusaka, Zambia

Tel.No. 260 1252 692

Fax.No. 260 1255 725 .

Email [email protected]

URL/Internet www.dmmu�ovp.gov.zm

Name of Contact Mwanza, Jones L .

Title of Contact National Coordinator .

Org. Profile The DMMU is the lead DisasterManagement Coordination institution in Zambia, placed withinthe ambit of the second highest office in the land. The unit isstrategically located to coordinate all activities pertaining toDisaster Management . Established in 1994 and became fullyoperational in 1998 the unit has scored a number of achieve�ments including initiating regional disaster management trainingprogrammes. The unit currently has a strength of twenty fullytrained disaster managers having undergone training at CranfieldUniversity’s CDMC.

Disaster Related Initiated training of Disaster Manage�Initiatives ment practitioners in the country

and from neighbouring countries.

Geographical Focus Zambia

Major Keywords Disaster Preparedness; DisasterReduction; Disaster Response;Disaster Recovery; Disaster Prevention.

Other Keywords Agriculture, Floods/Cyclone,Health Related, Refugees, Food Insecurity, HIV/AIDS, Forecasting and Early Warning Drought/Famine,Rehabilitation, Landmines.

The directory is available and can be accessed on thefollowing SARDC website links

http://databases.sardc.net/wwwisis/resources.htm (as a searchable database)

http://databases.sardc.net/dima/disaster_directory_4_s_africa.pdf (in pdf)

102

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

The Directory provides an outline of disaster relatedactivities being undertaken by organizations in theSADC region and provides a profile for the region’sefforts on the world scene.

The directory will form the basis of a morecollaborative framework in the management of riskand disasters in southern Africa. Both thebibliographic database and directory of organizationswill work towards developing a long�term process formonitoring risk and disasters in southern Africa.

Further Disaster InformationManagement Initiatives

Without reinventing the wheel, SARDC proposes toput in place in collaboration with partners asustainable mechanism that coordinates their effortswhile creating a platform for enhancing debate onthis socio�economic development challenge.

SARDC has a well�established Information ResourceCentre with databases on disaster management,state of the environment and water resourcesmanagement, as well as related programmes ongender and development, regional economicdevelopment and governance including conflictprevention, with website access and preparations forexpanding “virtual library” access at advancedstages. SARDC has strong information resources atsub�regional level, and well�established capacity andexperience in facilitating linkages, and establishingand sustaining development partnerships.

SARDC plans to adopt a multi�sectoral approach todisaster information management, with participationfrom social, policy, technical and economic sectors.

A well�coordinated continental response will requirethe development of a collaborative institutional anddata framework to support analysis and interpreta�tion functions, facilitate greater harmonisation atsub�regional and national levels, and create a strongbase for a reporting process for the region. Tostrengthen the coordinated approach and addressthe insufficient regional level comparative data andlack of strong linkages with national assessmentsand reporting processes, there is a need to establisha collaborative and participatory institutionalframework involving a number of institutions andsectors, and a data network infrastructure.

In response to this initiative, and to the fact that nodevelopment agenda can be successfully articulatedon the continent without addressing disaster

103

management, this project proposes a regionalinformation initiative that also seeks to expose andact upon the interface of policy, poverty and infor�mation access in the management and reduction ofdisasters. The harmonization of national and regionalinitiatives, and inter�sectoral and institutionalresponses to disaster management is a key factor inmitigating the effects on people and economies.

An initiative that promotes accessible methods ofinformation collection, analysis and disseminationwhile targeting disaster management organizationsand policy�makers is needed, especially in bridgingthe gap between national, sub�regional and regionalefforts.

SARDC proposes to put in place a sustainablemechanism that supports their efforts throughinformation access while creating a platform forenhancing debate on this socio�economic develop�ment challenge. The ultimate goal is to contributetowards solutions to reduce the burden of disasterson development efforts in Africa. The project will beimplemented in conjunction with sub�regionalorganizations, drawn from governmental institutionsas well as non�governmental and private sector, andwill include outreach to the media and parliaments.The geographical coverage of the project will be theAfrican continent.

There is no region�wide disaster information centrecovering the wide variety of hazard or risk conditionson Africa but there are specialized documentationcentres that are expanding their activities intorelated fields of risk. SARDC is one such highlyregarded centre. (Living with Risk, 2002)

RATIONALE FOR FURTHER PROJECT WORK

Disasters, whether natural or human�made are oneof the greatest threats to the NEPAD goals ofsustainable socio�economic development andpoverty eradication in Africa. Changing land usepatterns and global warming suggest that seriousfloods may recur in the future. The responses to allforms of disasters across the continent should befundamentally informed by the intricate relationshipbetween poverty and disaster management.

Poverty clearly can be caused by and result fromdisasters, and poverty is a significant factorimpacting on disaster mitigation and management.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

In depriving people of access to health facilities,schools and media, poverty limits their access toinformation and preparedness. While access toeducation and information can widen people’scapacity to adopt and implement appropriatepolicies, poverty can hinder that access increasing thevulnerability of households to the impact of disasters.Thus, policies and interventions should take intoconsideration the interface of policy, poverty andinformation access.

AU members states are currently at serious risk of failing to achieve one of their main objectives,“Health for All” in the 21st Century as adopted by the World Health Organisation (WHO). As theenvironmental degradation and management ofwater resources continues to escalate, showing nosigns of relenting in the region, SADC chances ofachieving this goal are also diminishing.

While there seems to be a plethora of informationincluding statistics, it is often conflicting, inaccurateor inaccessible to those who need to use it. Currentinformation is scattered, and generally not accessiblefor use in mainstreaming to other sectors, and isoften raw data or unusable data for the immediatepurpose. Best practices are often not accessiblebeyond local level. SARDC recognizes efforts onrelated issues and initiatives are in place throughoutthe continent. Most of these efforts lack a sub�regional or regional context and exist uncoordinated.

DEVELOPMENT GOALS AND OBJECTIVES

Development Goal

The objectives of socio�economic developmentleading to poverty eradication, and consolidation and maintenance of democracy, peace and securityare advanced through disaster reduction based onwide ownership and knowledge of strategies andbest practices at regional, sub�regional and nationallevels.

Project Objective

Regional, sub�regional and national capacities tomanage disasters in a coordinated manner arestrengthened, through development of an appro�priate institutional collaboration and data frameworkfor information access, monitoring and assessment.

STRATEGIES FOR PROJECTIMPLEMENTATION

A well�coordinated continental response will requirethe development of a collaborative institutional anddata framework to support analysis and interpreta�tion functions, facilitate greater harmonisation atsub�regional and national levels, and create a strongbase for a reporting process for the region. There isError! No index entries found need to establish acollaborative and participatory institutional frame�work involving a number of institutions and sectors,and a data network infrastructure.

In response to this initiative, and to the fact that nodevelopment agenda can be successfully articulatedon the continent without addressing disastermanagement, this project proposes a regional infor�mation initiative that also seeks to expose theinterface of policy, poverty and information access inthe management and reduction of disasters.

Implementation begins with establishment of a smallworking group of selected participants with knowl�edge of specific subject areas or sub�regions. Theseneed not be representative as this is an informalgrouping or think tank for purposes of start�up, andshould be very limited in number. This working groupwill form the basis of a future technical advisorycommittee, and will work on the terms of referencefor establishment of a steering committee of sub�regional representatives.

An institutional framework will be established at sub�regional levels involving the inter�governmentalagency and initially one or two reliable non�governmental partner organizations from each sub�region that could be situated at research institutions,universities or non�governmental agencies that arealready working on disaster�related subject areas,have an interest in information and some infra�structure, and may be at national or sub�regionallevel. The identification of these partner institutionsat sub�regional level is key to building the regionalnetwork, and some time and analysis must be spenton this process.

During this period, a parallel process of informationcollection can begin for sub�regional and regionallevels, based on the existing mechanisms in southernAfrica and elsewhere, with categorization andcataloguing for initial database entry and website

104

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

access. These mechanisms and processes will bereviewed as the project progresses, with participationfrom the key institutions in all sub�regions, and anappropriate regional mechanism will be developedduring the course of the project. This will require adocumentation working group that will develop,among other things, appropriate templates, softwareanalysis, and a thesaurus of keywords. This groupshould meet at least once annually, and establish anactive electronic discussion group as an ongoingmechanism for consultation and sharing of ideas.

Capacity building for sub�regional institutionsincluding intergovernmental agencies can begin onlyin the latter half of the project phase, and isdependent on results of the initial implementation.This will be a beginning and will lay a foundation, ascapacity building is an ongoing exercise, and may notbe uniform across sub�regions as they differ inrequirements and existing capacity.

The strategy envisaged includes a strong process ofimpact monitoring and analysis, after initialimplementation processes have been established. It isessential that this be viewed as an ongoing process,with clear plans for phases, for the sustainability ofthe process, supporting as it does the developmentplans of the region.

105

Annex 1

The Southern African Research And DocumentationCentre (SARDC)

SARDC is an independent regional informationresource centre which seeks to enhance theeffectiveness of key development processes in theSADC region through the collection, production anddissemination of information, and enabling thecapacity to generate and use information.

SARDC was established as a non�profit foundationin 1987, in Harare and Maputo, in response to anexpressed need within southern Africa for greateraccess to information about the region. Its objectiveis to improve the base of knowledge abouteconomic, political, cultural and social developments,and their implications, by making informationaccessible to governments and policy makers, non�governmental organizations (NGOs), the privatesector, regional and international organizations,

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

OUTPUTS/RESULTS

1 Strong regional/sub�regional partnership andnetworks are established and maintained for thepurpose of increasing ownership and participationof stakeholders in disaster reduction.

2 Multi�sectoral information on disaster reduction isaccessible to governments and inter�governmentalagencies, organizations, individuals, private sectorand other stakeholders, including parliaments andmedia.

3 The capacity of sub�regional and nationalinstitutions to collaborate on disaster reductionstrategies through information, access, exchangeand analysis is supported and strengthened forpurposes of reducing vulnerability.

4 Strengthening and institutionalizing partnerships,building on information capacity achieved,developing indicators for sub�regional monitoringand GIS mapping.

development agencies, parliaments, and the media.The Centre has grown to 10 departments, includingfive for specific programming areas and five departments offering professional services supportfor information resources and knowledge management, information technology, publishing,marketing, finance and administration, and projectdevelopment.

Particular areas of interest related to risk reductioninclude the State of the Environment in southernAfrica, disaster management information devotedespecially to drought and other socio�economic and political issues relevant to the developmentprocess and governance that have a direct bearingon matters of risk awareness and managementpractices. In this respect, SARDC hosts the IMusokotwane Environment Resource Centre forSouthern Africa (IMERCSA) which provides userswith current information on environment and

disaster management in southern Africa. IMERCSA isthe leading regional centre for global reporting onthe State of the environment, producing relevantdatabases, reports, books, factsheets a newsletters,including on shared watercourses. SARDC is also wellplaced to facilitate seminars, conduct briefings andundertake consultancies for information exchange onenvironmental issues, human development, gender,electoral processes and related aspects ofinformation networking, including specialised workwith parliaments and media.

SARDC, being a regional organization with offices inMozambique and Zimbabwe, representation inTanzania and an established network of partnerorganizations that include SADC institutions, is wellplaced to implement a project on an issue of cross�cutting nature such as disaster informationmanagement. SARDC has implemented widelyacknowledged projects on issues that includeenvironment and water issues, gender anddemocracy. SARDC’s involvement in the genderarena contributed to the development of genderpolicy for the SADC region and to the SADC genderstrategy approved by Heads of State in 1997.

Our organizational strengths cover the preparation ofaccurate, reliable and accessible information throughworking with a wide range of expertise throughoutthe region, and disseminating the informationthrough relevant networks, organizations,governments, parliaments and media. We do notexpect to have all relevant technical expertise in�house, but have strength in building highly motivatedteams and working groups made up of individualsfrom different countries and disciplines. This has beenused effectively, for example, in producing theMozambique National Human Development Reportfor the past four years; producing a number of Stateof the Environment reports in the SADC region;working with UNEP to contribute to the Africa andGlobal Environment Outlooks as the collaboratingcentre for southern Africa; and working with nationalpartners to produce national gender profiles forSADC member states.

The SARDC staff numbers 50 people including eightin Maputo, of which about 30 are professionals andthe others support staff. All are citizens or residentsof southern Africa, and about half of the staff at alllevels at women. There are a number of students onattachment, and the most recent South�South Fellowwas from Brazil.

SARDC launched its Disaster ManagementInformation Project (DMIP) in 1995, with theInternational Federation of Red Cross and RedCrescent Societies, with financial support from theCanadian International Development Agency. Thefirst phase of the DMIP sought to contribute to theUN�proclaimed International Decade for NaturalDisaster Reduction (1990�1999) by improving theinformation and knowledge base for disastermanagement in southern Africa. This was achievedthrough documentation and dissemination of existingand new technical and non�technical informationrelated to measures for the assessment, predictionand mitigation of both natural and human�causeddisasters.

The project aimed to provide accessible and accurateinformation on disasters in the southern Africanregion. The information is targeted for use by reliefagencies, policy makers, non�governmentalorganizations, the private sector, developmentagencies and the media, in the region andinternationally, as a resource for training and policyplanning so that appropriate and proactive measurescan be taken when a disaster strikes. The first phaseof the DMIP, concluded in 1998, documentedinformation on key disasters in southern Africa,including drought, floods, epidemics (such as cholera,dysentery and HIV/ AIDS), civil conflicts, storms andcyclones, internal and cross�border displacement, andfactors worsening household vulnerability. Throughthe project, a library with computerized bibliographicand contacts databases was set up. The library ’scollection comprises both published and un�publishedinformation dating back to 1960. To date, theproject’s literature holdings number over 2,400books, scientific papers, disaster plans and profiles,etc. The contacts database holds about 500 entries.The results have been highly appreciated and areused by individuals and agencies in disasterreduction and response today.

From December 2002 to February 2003 SARDC hasbeen working on the bridging phase of the ProjectStrengthening Disaster Information Management inAfrica in conjunction with the office of the UN Inter�Agency Secretariat of the International Strategy forDisaster Reduction (UN�ISDR). The project sought toupdate the databases with current information,promote Disaster Information Management andwiden access to disaster information in Africa ingeneral and southern Africa in particular.

106

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Methodology Reliable and accessible informationrelevant to the SADC region is generated through aprocess of participation, networking, wide consulta�tion and ownership, and targeted dissemination,reaching into national, regional and global policyprocesses. SARDC provides information and analysisof current policies and issues through special reportsand news features, books, fact sheets, chronologies,policy analysis, seminars, conferences, specialistservice, and website . This information is well�received because of its scope and accessibility,drawing positive responses about the quality and mixof information, analysis and documentation, whichmakes it available to a cross�section of society, frompolicy makers and parliaments to media and NGOs.

Databases SARDC has an information resourcecentre containing over 12,000 subject files onregional issues, a library of books and periodicals,and computerized databases of selected materialsthat are retrievable through he use of keywords andmaintains specific bibliographic and contactdatabases on primary areas of interest. SARDC hasan extensive website of publications and articles, andis establishing a “virtual library” of internet accessto its data bases and other resources.

107

Institutional Capacity SARDC has 15 years ofinstitutional experience in documenting, analysingand communicating trends in regional development,in publishing and distributing the results, andmonitoring impact, and a range of qualified stafffrom the SADC region as well as an extensivenetwork of regional partner organizations andcontacts. Language capacity in English, French,Portuguese and Swahili. With an extensive networkof partner organizations and contacts in the SADCsub�region and beyond, SARDC is strategicallypositioned to contribute to identifying multi�sectoralsolutions and collecting/disseminating policies andgood practices.

Partnerships SARDC works in close partnershipwith the Southern African Development Community(SADC), with SADC environment sector through aformal Memorandum of Understanding, and with theSADC Gender Unit, SADC Water Sector and othersthrough informal agreements or contracts. SARDCalso has an MOU with the United NationsEnvironment Programme (UNEP).

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

108

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

109

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Multidimensional SNMR

modelling for groundwater

exploration

TEAM LEADER

W. Warsa <[email protected]>

PROJECT ADVISORS

Prof. Dr. Ugur YaramanciDr. Martin Müller

Indonesia

SYMPOS IUM NOTES

Warsa Warsa1,2, Martin Müller and Ugur Yaramanci1 Bandung Institute of Technology, Department of Geophysical Engineering,

Jl. Ganesha 10, Bandung 40132, INDONESIA.2 Technical University Berlin, Department of Applied Geophysics, Ackerstr. 71–76,

D�13355 Berlin, GERMANY

110

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Abstract

Groundwater is an important economic source ofwater supply. It serves as an important source fordrinking water supply and irrigation for agriculture. Inthe relation of managing groundwater resources, it isimportant to forecast of water level, quantity andother changing parameters. Furthermore, the ground�water resources are at risk to contaminants from theland surface as well as from intrusions. In manycountries the land subsidence occurs when largeamounts of groundwater have been withdrawn fromcertain types of rocks, such as fine�grainedsediments.

Surface Nuclear Magnetic Resonance (SNMR) is arelatively new geophysical method that can be usedto determine the presence of culturally and econo�mically important substances, such as the subsurfacewater or hydrocarbon distribution. SNMR is the onlygeophysical method, which allows to determinewater content and pore size distribution directly fromthe surface. In combination with electrical conduc�tivity and permittivity SNMR will be used improve theaccess to the subsurface. The SNMR method isperformed by stimulating an alternating current pulsethrough an antenna at the surface confirms theexistence of water in the sub�surface.

In the future this new method may be applied as anattempt to bring long�term solutions to the watersupply problems and also to protect groundwaterfrom the degradation of aquifers. Furthermore, alsowill be useful as an input to design parameters usedto prevent future environment problems 1. However,in this case of course the use of other geophysicalmethod can provide better answer.

Introduction

Groundwater is an important natural resource allaround the world. To manage and preserve thegroundwater resources quantifying and qualifying ofgroundwater and hydro�geological parameter ofaquifer are provided. Furthermore the availability ofgroundwater resources and how can be exploited areessential for the sustained socio economicdevelopments of any area, whether urban or rural. Itis also critical to forecast the effects of hazard andrisk identification concerning the condition of waterresources and environment.

For this purpose geophysical methods provide a widerange of very useful and powerful tools which, when

111

used correctly and in the right situations, willproduce useful information. As the range applicationof geophysical methods has increased, particularlywith respect to derelict and contaminated landinvestigation, the sub�discipline of environmentalgeophysics has been developed. These measure�ments can then be used to infer the porosity,permeability, chemical constitution, stratigraphy,geologic structure, and various other properties of avolume material near earth’s surface. In applicationof groundwater exploration that can be employed topredict groundwater level, physical parameter ofaquifer, sea water intrusion, contaminants from theland surface, land subsidence caused by humanactivities, mainly from the removal of subsurfacewater.

One of the geophysical method has been used formonitoring subsidence and decreasing water level is4D microgravity 2. Monitoring activity using gravitymethod on a nature phenomenon that relate withphysical change of subsurface body can be visualizeusing its physical parameter (density) of rock if thechange give gravity potential field that sufficientlysignificant to be observed on the surface. Change ofgroundwater level and sea water intrusion areexamples of nature phenomenon that indicateexistence of density change in aquifer.

The other geophysical method which may be appliedfor this purpose is SNMR. The SNMR has been testedon the test site to yield the geometry, water contentand hydraulic permeability of the aquifer3, 4. Besideallows more detailed and reliable assessment ofaquifers the SNMR may also be used to detect thechange of water content and hydrogeologicalparameters caused by subsidence.

To improve the capability of the SNMR method wehave carried out a research to study the response of2D and 3D models, i. e. the SNMR relaxation signalfor various locations of the antenna loop. The mainaim of this research is to improve the ability ofsurface NMR method in determination of hydro�geological parameters of the subsurface. Emphasis ofthe research is the development of a program thatallows to determine the initial amplitude and decaytime of a SNMR signal for 3D water distribution independence of the pulse moment. The amplitudesare directly related to the water content, while thedecay times are linked to pore size, grain size, andtherefore to hydraulic conductivities. At the end wehave developed 2D inversion to increase theinterpretation of SNMR method.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Theoretical Background

The surface nuclear magnetic resonance (SNMR)sounding is a geophysical method that aims atdetermining hydro�geological parameters from mag�netic resonance measurement. This method is basedon the principle of the proton magnetic resonance inthe Earth’s filed of hydrogen 1H atom, which con�tained in the groundwater molecule H2O. An NMRsignal, stimulated by an alternating current pulsethrough an antenna at the surface confirms the exis�tence of water in the sub�surface. The amplitudes ofNMR signal are directly related to the water content,while the decay times are linked to pore size, grainsize and, therefore, to hydraulic conductivities. Thephases are related to the electrical conductivity;however this is only used qualitatitively (Table 1).

Methodology

The new geophysical method of SNMR is based onthe principle of the magnetic resonance of protons ofhydrogen atoms in the Earth’s magnetic field. Analternating current pulse through a wire antenna atthe surface stimulates a NMR signal. Aftertermination of the exciting pulse the response fielddue to the relaxation of the precessing hydrogen ismeasured. The amplitudes of NMR signal are directlyrelated to the water content, while the decay timesare linked to pore size, grain size and, therefore, tohydraulic conductivities 4, 5. The phases are related tothe electrical conductivity, however this is only usedqualitatively.

The geophysical method of surface nuclear magneticresonance (SNMR) allows direct determination ofhydrogeophysical parameters of the subsurface. Themethod of SNMR is based on the principle of themagnetic resonance of protons of hydrogen atoms inthe Earth’s magnetic field. Figure 1a shows spincharacteristic of the proton 1H in the absence of anexternally applied magnetic field. An external staticmagnetic field B0 causes the nuclei to align them�selves in one of two orientations with respect to B0(Figure 1b). Figure 1c shows the precessing of aproton of hydrogen atom. In SNMR an alternatingcurrent pulse through a wire antenna at the surfacestimulates the NMR signal. After termination of the exciting pulse the response field due to therelaxation of the precessing hydrogen protons ismeasured (Figure 1d). The initial amplitude E0 of thesignal corresponds to the water content in the sub�surface 5, 6. The decay time T2* (spin�spin relaxationtime) of the SNMR signal corresponds to pore sizes.The fundamental integral�equation that governs theamplitudes E0 (q) of NMR as a function of the pulsemoment q is given by

( 1 )

in which ω0 is the local Larmor frequency of thehydrogen protons and M0 is the nuclear magneti�zation of the protons. The water content and thedecay time for a unit volume at the point r in thesubsurface are given by f(r) and T2*(r) respectively.Β⊥ (r) states the component of the exciting magneticfield perpendicular to the Earth’s geomagnetic field.The tilt angle of the protons is given by θ(r) =0.5γΒ⊥ (r)q. Increasing the pulse moment q (q = Ι0τ,where Ι0 and t are the amplitude and duration timeof the current pulse, respectively) increases the depthof penetration of the method.

112

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Measured quantity (vs. pulse moment)

Physical parameter (vs. depth)

Amplitude of the PMR signal Eo (q) Water content

Decay time constant of the signal T2* Mean Pore size

Phase shift between signal and current ϕ Rock layer resistivity

Table 1. Physical parameters determined with SNMR

Figure 1: Spin characteristic of the proton. (a) in the absence of an external static magnetic field (b) in anexternal static magnetic field Β0 (c) the precessing hydrogen proton at the Larmor frequency ω0 (d) the timediagram of the SNMR signal measurement.

, )(sin)()(

)(),( )(/00

)(/0

*2

*2

dVBf

eMeqEqtEV

TtqTt

rrr

r

θ

ω

⋅⋅

== ∫⊥

−−

Numerical modeling

Interpreting SNMR data consists in determining thewater content of each layer, in the hypothesis wherethe underground is stratified at the scale of the loopdimensions. The inversion consists in processing theraw data for the whole set of pulse momentscorresponding to the various depths of investigation.However, some parameters are well defined such asthe product of the thickness of a thin layer by itswater content. The development in a 3�D forwardmodelling of SNMR initial amplitudes and decaytimes are based on the work of Eikam7, that can beused for two and three�dimensional interpretation ofSNMR surveys. The formulation is reduced to a finitedimensional matrix problem by considering a finitenumber of cells with constant spin density (watercontent and decay time). The 3D forward modelling isthen calculated by generating synthetic data forseveral of spin density distribution. This allows toanalyse the spatial signal sensitivity of the methodand shows the limits and problems of the 1D inver�sion and interpretation of 2D and 3D structures.

We introduce a 3�D forward modelling of SNMRinitial amplitudes and decay times 7, 8, 9. A prismaticthree�dimensional body model is divided into smallcubic cells of dimension DV = DxDyDz. The watercontent f (x,y,z) and the decay times T2*(x,y,z) areassumed to be constant in each cell. Then the

113

integral equation is approximated by the finitesummation

( 2 )

From this relation, the complete SNMR signal can becalculated as a function of a three dimensionaldistribution of the water content f(x,y,z) and decaytime T2*(x,y,z) in the subsurface.

Figure 2 represents the 3D model�discretization forwhich SNMR model have been calculated using oneturn circular loop of radius 50m in geomagnetic fieldof 48000 nT at an inclination of 60° and declinationof 00 in a low conductive half�space. The initialamplitude for t = t 0, the start of the record, are thengiven by

.

( 3 )

The inner part of the integral is commonly written as the product of a kernel function and the watercontent

.

( 4 )

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Tx/Rxloop

mN

1 2 Ll

mpm2

y (East)v1

z

v2

h

vl

m1

mi∈ fi(y,z), T2j*(y, z)

x (North)

.),,(sin),,(

),,(),( ),,(/00

*2

zyxzyxzyxB

zyxfeMqtEz y x

zyxTt

∆∆∆

= ∑∑∑⊥

−

θ

ω

zyxzyxzyxB

zyxfMqEz y x

∆∆∆

⋅= ∑∑∑⊥ ),,(sin),,(

),,( )( 000

θ

ω

∑∑∆∆

⋅=z y

D zyzyqK

zyfMqE

),;(

),( )(

2

000 ω

Figure 2: 3-D modeling of surface NMR measurement for l = 1,.., L sounding points.

For a study of 2D we assume the water content to bein 2D distributed. We obtain

.( 5 )

Sensitivity of SNMR

The slices of a 2D kernel (K2D(q;y,z) in equation 5) in west�east and south�north directions for pulsemoment q ranging from 1 to 20 [A.s] are presentedin Figure 2�3. The kernels are compiled for a circularloop (D=100 m, 1 turn) in a low conductive media.Increasing the pulse moment q increases the depthof penetration of the method. Whereas sensitivities of west�east sections increase symmetrically, they arefocused to the north in south�north sections. Thiseffect is caused by the non�symmetrical contributionof SNMR signal9, 10.

To study spatial sensitivity for SNMR surveys wemodeled 2D sensitivities for three different fieldlayouts. The antenna (1 turn circular loop, radius R = 50 m) is shifted at the surface for intervals of 50 m, 100 m and 150 m between sounding pointseach for a west�east and south�north profile. The sumof the kernels (magnitudes) of a set of measurementsalong a profile now gives the 2�D sensitivities to thewater distribution. Fig. 2 presents the distribution ofsensitivities for pulse moments q = 1, 10, 20 A.s.,respectively. To evaluate the lateral resolution of eachsurvey the 2�D kernels are compiled.

Increasing pulse moment q increases the depth ofpenetration as well as the lateral extension of thesensitive region. A sounding interval of 50 m (1R)with 50 m overlap yields the best lateral coverage.The consistency of lateral coverage rapidly decreasesfor 100 m (2R) interval for of smaller section than the1R interval. For 150 m (3R) intervals, which would bethe fastest survey progress, the sections displayalmost no consistent lateral coverage. Therefore onlysmaller intervals would be favourable for a 2Dinversion of field data. However, the lateral resolutioncan generally not exceed the sounding intervals.

For South to North profile direction a northward shiftof sensitivities can be observed. Therefore, regions oflow sensitivity occur in the southern part of theprofile whereas regions of high sensitivity can beobserved beyond the northern profile limits.

2D Inversion

The first task of research is extending the initialamplitude 3D forward modeling code to decay times(T2*) for SNMR. This scheme can be used for two andthree�dimensional interpretation of SNMR surveys.For a better understanding and insight of the capa�bility of the method we calculate the SNMR responseof 2D and 3D models, i.e. the SNMR relaxation signalfor various locations of the antenna loop.

The 3D forward modelling is then calculated bygenerating synthetic data for several of spin densitydistribution. This allows to analyse the spatial signal

sensitivity of the method andshows the limits and problems ofthe 1D inversion and interpreta�tion of 2D and 3D structures 9.

At the end of this research areformulation and implementationof 2D inversion. One of the criticalproblems in inversion of geophysi�cal data is developing a stableinverse problem solution, which at the same time can resolvecomplicated geological structures.Traditional geophysical inversionmethods are usually based onTikhonov regularization theory,and they provide a stable solutionof the inverse problem. For inversion of SNMR the Tikhonovregularization method was usedalso to minimize the number ofmeasurements 6 without a loss of the accuracy.

114

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

∑∑∆∆

⋅=z y

D zyzyqK

zyfMqE

),;(

),( )(

2

000 ω

Figure 3: 2-D kernel (magnitudes) for single SNMR sounding at in west-east(left) and south-north (right) direction for pulse moments q = 1, 5, 10, 15,20 [A.s]; earth magnetic field B0 = 48000 nT; I = 600; circular loop with 100m diameter, 1 turn.

115

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Figure 4: 2-D sensitivities for SNMR sections in west-east (top) and south-north (bottom) for soundingpoint intervals: 50 m (left), 100 m (center) and 150 m (right); earth magnetic field B0 = 48000 nT; I = 60°; circular loop of radius 50 m, 1 turn.

116

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

This section describes the inversion routine whichtries to find the smoothest model that fits the data to an a priori value of χ2. The steps of process are

• a forward model, sensitivity matrix, and data misfit are calculated,

• the sensitivity matrix and data misfit are used to estimate a Lagrange multiplier a that controlsthe relative weighting of smoothness versus datafitting in the objective function, and

• modified Levenberg�Marquardt algorithm is used to find the parameters that minimize theobjective function.

The Levenberg�Marquardt algorithm

where α is the damping parameter and Ι is theidentity matrix.

As a first example, we invert the initial amplitude &decay time of SNMR data, which was produced by asingle water lens having water content of 40% anddecay time 150 ms (Figure 5). There are 11 soundingpoint data of measurement and 1% gaussian noisewas added to the data prior. The recovered waterlens model, shown here, is a good representation of

dGIGGm )( 1 TT −+= α

Figure 5: (a) 2-D waterlens model having awater content of 40 vol.percent and decay timeof 150 ms. Thesurrounding material is associated with awater content of 5 vol. %and decay time of 50 ms.

(b) Synthetic SNMRamplitudes;

(c) 2-D inversion result;earth magnetic field B0 = 48000 nT; I = 60°; circular loop ofradius 50 m, 1 turn.

a

b

c

117

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

the true model. The recovered value is slightly lowerthan the true value.

Figure 6a shows a 2D model of a dipping bearinglayer (dipping angle = 7°) striking from north tosouth, e.g. an aquifer underneath a hillside. Themodel has been computed using a grid size of _x =_y = _z = 0.5 m. A water content (WC) of 40 vol. %and a decay time (T2*) of 100 ms is assigned to theaquifer. The surrounding material has a water contentof 5 vol. % and decay time 50 ms.

The 2D inversions (water content and decay time) ofthe dipping layer model using a smooth inversionscheme are presented in Figure 6b & c. The smoothinversion was performed using Levenberg�Marquardtalgorithm up to a maximum depth of 100 m. At adistance of 50 m to the west from the outcrop of thedipping layer no signal contribution of the aquifer isobserved. Corresponding to the model the inversionsshow constant water content of approximately 5 vol.% and decay times of about 30 ms up to 100 mdepth. At the outcrop of the aquifer (profile 150 m)the inversions indicate a slight increase of both watercontent (<20 vol. %) and decay times (<90 ms)near the surface. The water content and the decaytimes of the aquifer as well as its thickness the depth

of the aquifer are in good agreement with the model.Naturally, the sensitivity of the SNMR method getssmaller with increasing depths.

Conclusions

3D modeling of SNMR data is necessary to predictthe SNMR signal response of aquifers. Use thismethod for optimisation of field layouts forappropriate measurement on given 2D situation. 2DSNMR sensitivities are necessary to evaluate lateraland vertical resolution of 2D. We have developed areliable technique for 2�D inversion of SNMR.

Improved inversion algorithm will be investigatedconsist of model parameterizations regularizationschemes. This inversion will be applied for the SNMRfield data hydrogeology. At the end, the SNMR canbe applied to dectect subsurface water in suitablegeological formation to a depth of 100 m and moredepending on the presence of natural and culturalelectromagnetic noise. In the future this new methodmay be applied as an attempt to bring long�termsolutions to the water supply problems and also toprotect groundwater from the degradation ofaquifers.

Figure 6: (a) 2-D dipping layer model having a water content of 40 vol.% and decay time of 150ms. (b) & (c) 2-D inversion result water content and decay time

a

b

c

118

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Outlooks

It was realized that the method passed experimentalstage and is on a good way to be an establishedgeophysical technique. Furthermore, it was realizedthat there is broad on�going focused research indifferent group’s worldwide and increased interestby the end users prospecting for groundwater. As aresearcher in developing country, I will continue thecollaboration with Interpretational programimprovement, and applied some field data. Designfor new field measurement.

From the 2nd MRS/SNMR workshop in Orleans�France could be considered as a meeting not only tothose involved in SNMR development andapplication, but also for geophysicists and hydro�geologists seeking new tools for groundwaterinvestigation. The workshop was attended more than70 participants from 15 countries (Algeria, Australia,Burkina Faso, Chine, Denmark, France, Germany,Great Brittany, India, Indonesia, Jordan, TheNetherlands, Nigeria, Russia, Spain and USA). Thereclearly should be something for everyone interestedin aquifer characterization and localization.

What the components of learning in our project areto find out the relationship between aquifercharacterizations (water content and permeability)using SNMR and disaster risk. Finally, a combinationof methods is proposed by which near�surfacegeophysics can contribute regard to the watercontent and the water distribution.

References

1 Steeples, D.W. 2001. Engineering and EnvironmentalGeophysics at The Millennium. Geophysics, vol. 66, No.1,31�35.

2 Abdulkadir, W.G. et.al. 2002. Studi amblesan permukaantanah dan dinamika air tanah di dataran alluvialSemarang – Jawa Tengah dengan menggunakan metodeMicrogravity 4D. Report of Research. Institute ofTechnology Bandung.

3 Yaramanci, U., Lange, G. and Hertrich, M. 2002. Aquifercharacterisation using surface NMR jointly with othergeophysical techniques at the Nauen/Berlin test site.Journal of Applied Geophysics 50, 47�65.

4 Yaramanci, U., Lange, G. and Knödel, K. 1999.Surface NMR within a geophysical study of an aquifer at Haldensleben (Germany). Geophysical prospecting 47,923�943.

5 Schirov, M. and Legchenko, A. 1991. A New Non�invasiveGroundwater Detection Technology for Australia.Exploration Geophysics, vol. 22, 333�338.

6 Legchenko, A.V. and Shushakov, O.A. 1998. Inversion ofSurface NMR Data. Geophysics 63, 75�84.

7 Eikam, A. 1999. Modellierung von SNMR�Anfangsamplituden. MSc Thesis, Technical UniversityBerlin.

8 Warsa, W., Mohnke, O. and Yaramanci, U. 2002. 3�DModelling of Surface NMR Amplitudes and Decay Times.International Water Resources and EnvironmentalReasearch ICWRER 2002, Vol. III, Eigenverlag des Forumsfür Abfallwirtschaft und Atlasten e.V., Dresden, 209�212.

9 Warsa, W., Mohnke, O., Hertrich, M. and Yaramanci, U.2003. Sensitivity study of 3�D modelling for 2�D inversionof Surface NMR. Proceedings of the 9th EuropeanMeeting of EEGS, Prague.

119

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Action advocacy to ensure right

to livelihood risk reduction and

beyond

TEAM LEADER

Vishnu Prabhir <[email protected]>

PROJECT ADVISOR

Mihir R. Bhatt

India

SYMPOS IUM NOTES

120

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Prabhir Vishnu Poruthiyil,

Disaster Mitigation Institute, Ahmedabad, India

E�mail: <[email protected]>

Introduction

The livelihood security of the urban poor, whichunprotected in normal times are undermined bydisasters and uncompensated in post disasterrecovery efforts. The urban poor perceive provision offood, water and shelter as a source of protection andsupport of their income earning activity. Better foodmeans better nourishment that is reflects in theirproductivity. Proximity and regular supply of goodwater means less time spent on fetching it from adistance or waiting for it. This means more time forwork. Safer shelter means safer place to work, saferplace for keeping raw materials and safety of theirfamilies when they are away. Protection of livelihoodsof the urban poor is a prime mover for achievingsecurity for their future

What happens to the livelihoods of the urban poor when a disaster strikes?

Their economic and social condition renders themdefenseless from the onslaught of disasters and thesustained after effects. Source, support andsustenance of their work are affected adversely. Theysuffer loss of income and assets. Handcarts, sewingmachines and wheelbarrows are damaged andburied under fallen roofs and walls. Shelters, whichdouble as workplaces for the poor, are destroyed.Economic links are damaged and local markets aredestroyed. Worse, in the relief process their long�termneeds are ignored. These sections are not able towithstand the impacts of disasters or recover like thebetter�endowed sections of the society. More oftenthan not, without outside support, this section ofvictims is unable regenerate their livelihoods tosecure their lives again. Even though the poor areknown to take rational decisions in the face ofdisasters and are ready to invest in disaster�proofingtechniques that will make their lives safe, every stepthey take is hinged on their livelihoods.

This paper delineates the findings of a researchstudy, supported by the ProVention Consortium ofthe World Bank, and conducted by the DisasterMitigation Institute (DMI), Ahmedabad among theslum dwellers in Bhuj, (Western India), who are stillrecovering from the January 2001 earthquake. Theresults indicate that overall social development of aregion is in itself an effective risk reduction measureand outlines a roadmap for investment in humancapital, social or economic, as an aid to livelihoodrisk reduction. This approach interlinks recovery and mitigation of disasters to existing efforts indeveloping countries focusing on improving its socialindicators like literacy, and initiatives to create safetynets like micro –credits and micro�insurance.

121

Background of the study

The Stimulus

The internal and external drivers that stimulated theconcept of the original research study —‘ ActionAdvocacy for Livelihood Risk Reduction andBeyond’— have been summarized into a set of threequestions:.How can we develop a framework of sustainablelivelihoods in a disaster prone pocket?

What role does a need-based livelihood relief haveon the lives of disaster prone urban poor?

What role does reducing risks to livelihoods haveon the development of a state, country and region?What works for risk reduction and why?

Most of them are direct. The study has attempted toget the answers from the ground reality in thefourteen earthquake affected slums in Bhuj, stillrecovering from the January 2001 earthquake.

The target population

The approach to the research study was bothquantitative and qualitative, and the interest is tomake generalizations concerning livelihood recoverypattern of a group of slum�dwellers in Bhuj.

For this the study focused on 1100 livelihoodbeneficiaries of Disaster Mitigation Institute (DMI).in Bhuj. A brief introduction to organization is givenin Box 1.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Box 1: Disaster Mitigation Institute andHuman Security

Disaster Mitigation Institute (DMI) is a community basedaction research, action planning and action advocacyorganisation that bridges the gap between policy, practice,and research from the community to the national level.Established after the 1987�89 repeat drought in Gujarat, ithas four programmes: livelihood security; food security;shelter security; and water security. Its activities are organ�ised around its activity centres: Livelihood Relief Fund(LRF); Emergency Food Security Network (EFSN); WaterSecurity Programme (WSP); Emergency Health Unit (EHU);Organisational Resources (OR); Learning Resources (LR) andAction Research and Review Services (ARRS). DMI’smission is to reduce the vulnerability of poor communitiesby increasing mitigation efforts through learning andaction.

DMI’s LRF was created as an activity centre after thecyclone that hit the western port of Kandla situated in thewestern coast of Gujarat in June 1998. Since then, LRF hasgrown steadily to reach out to victims of the Kandla

slum�dwellers were again consulted and their inputsincorporated. A team of slum�dwellers associatedwith the Bhuj Reconstruction Project (BRP) conductedthe actual survey with inputs from two internationalstudents associated with DMI through its internshipprogramme. 246 livelihood beneficiaries were inter�viewed. A set of case studies provided a qualitativeperspective to the findings. The data is fresh and theperspective provided is from the community andinside out. The main findings and analyses have beenseparated and distilled into this paper.

Livelihood Risk Reduction: The links3

How can livelihoods of urban poor that fuel theirdevelopment poor be protected? The scope and therange of areas covered by the research studyincluded the study of different influences andcombinations of a wide range factors and impactsthat jointly affects the livelihood security of thedisaster�prone urban slums. These include a completeanalysis of age, gender, livelihood types on one handand their individual status with respect to literacy,economic recovery, savings capacity and vulnerabilityon the other. As explanation of each an every findingwould be beyond the scope of this paper the mostsignificant among them is summarized below.

Literacy reduces risks

The respondents were divided into five age groups as16�25, 26�35, 36�45,46�55 and 56�75. The educa�tional level of the slum dwellers is quite low. Lessthan 35% of the sample had gained some level ofeducation. Out of these, barely 26% had completedthe primary school, 8% the secondary school, lesswhen 2% entered the high secondary school.

122

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Cyclone, 2000 drought, Bhuj earthquake in 2001 and thecommunal riots that rocked Ahmedabad in 2002. It hasgiven livelihood support to more than 10,000 victims of theearthquake in the three districts of Kutch, Patan andSurendranagar in Gujarat.

LRF in partnership with international aid agencies such asAmerican Jewish World Service (AJWS), Action Aid andEuropean Union (EU) has reached out to the disastervictims and worked for providing them long�term needs. InBhuj, LRF focused its attention in slums that were ignoredeven a year after the earthquake and in�spite of abundanceof relief material.

DMI accepts the wisdom that disasters are no longer atemporary dip in the development graph but a threat to theprocess itself. For mitigating disasters, DMI has four criticalhuman security programmes: livelihood security; foodsecurity; shelter security, and water security. DMI’sconceptual model of human security is shown in theadjoining figure. Livelihood security is consciously placed ina prominent position, as it is linked to food, shelter andwater security. This model has come out of DMI’s ongoingexperience with regenerating livelihoods of the urban poorvictims of floods, droughts, earthquakes and so on.Beneficiaries of LRF and their families are consuming moreand nutritious food. They are ready and capable of investingin disaster�proofing their shelters. They can buy enoughwater for their daily needs. In short, they are surer of asecure future. Livelihood security, DMI has seen, is the hubaround which human security of the urban poor revolves.

Recovery/Literacy

$ 0-20 21-40 41-60 60-above

Literate before Literate after

Recovery/Illiteracy

$ 0-20 21-40 41-60 60-above

lliterate afterIlliterate before

0

25

50

75

0%

25%

50%

75%

DMI's human security model

Sample

LRF’s beneficiary group in Bhuj consists of a diversegroup of livelihoods, which can be classified into fourdifferent groups: small businessmen, small�scalevendors, laborers and home�based workers. Severaldifferent professions can be categorized under thesesections.1

The initial list of issues was prioritized in consulta�tion with a selected group of LRF beneficiaries.2

After, the entire questionnaire was designed, the

There is a marked difference in recovery patterns ofthe literate and illiterate among the slum dwellers.In the graph ‘income /illiteracy’ it can be seenthat there is an increase in the number of illiterate

people in the $21 to $40 category and a decrease inilliterates in higher income categories. Consequently,the average income of the illiterates has shrunk from$39 per month before the disaster to $30 after. Onthe other hand, there is insignificant difference inaverage income of the literate. The average income is almost same ($35). As seen in the graph there isminimum impact of the earthquake on their income.

123

terms) helps in making the right decisions, under�standing markets, awareness of rights, etc. Theliterate, after a disaster, can better link up with the changed and emerging market and align thelivelihood to benefit from it. The illiterate are notable to do so.

Hence literacy of an urban area is an importantindicator for assessing the capacity of the region to withstand and recover from disasters.

Credit is crucial

Livelihoods that require less capital have recoveredbetter compared to others. There is an increase inaverage income of the home�based category andsmall vendors compared to small businesses. In thecategory of small businesses, more than 77% of therespondents were men. The recovery here is cruciallydependent on availability of economic capital. Morethan 87% of them felt that their daily needs werenot covered. The adverse impacts of the relocationfrom the Walled City1 to the outlying areas havebeen most felt by the small businesses. Around 60% of them had to relocate, losing their customers.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Literacy/Age

Recovery/Age

16-25 26-35 36-45 46-55 56-75

Literate Illiterate

Before After

0%

25%

50%

75%

100%

0

20

40

60

16-25 26-35 36-45 46-55 56-75

Recovery/Livelihood types

Before After

HB Vendor Labourer Small Business

0

20

40

60

Another finding that supports the link betweenliteracy and livelihood risk reduction is the analysisof the recovery across age groups. The youngergroups also had majority of the literates from thesample. The recovery of the younger two groups wasalso better than the other three groups. It can beargued that the income of the youngest age groupwould always increase irrespective of the disaster.Therefore in comparing the above graphs the focusshould be on 26�35 and the 36�45 age groups. Thewithstanding capacity of younger group, which ismore literate, is more than the older category.

This is evidence of the significant role that literacyplays in recovery after disasters and is a crucial linkbetween poverty and disaster risk reduction. Amajority of them does not pursue education due tolack of resources and the need to earn to live. Whatis more striking is that even in the poorest categoryof the society, literacy (not only in school going

There is a reduction in income of laborers in spite ofthe massive post earthquake reconstruction work inBhuj. This can be attributed to the impact of largenumber of migrant workers that have entered Bhujfrom the poorer regions of the country like Bihar,Uttar Pradesh and Rajasthan. These migrant arepreferred by contractors undertaking the reconstruc�tion as they can be expected to work for as low asone third of the daily wage that a laborer from Bhujwould demand.

A related finding here is the average income ofwomen to have increased post disaster. The averageincome of women had increased from $26 beforethe earthquake $27 after, while the correspondingincomes of men had reduced from $43 to $37. Thenumber of women who earned no income is lessnow than before and this section has contributed to

the increase in average income of women. Womenconstituted a majority (73.2%) of home�basedworkers. These women produce food items at home,make ready�made garments, do embroidery andmanage in�house stores

Surprisingly twenty percent more women now havework inside their community as compared to beforethe earthquake. Coupling this finding with anincrease in average income there could be aconclusion that livelihood relief targeted at womenhave given dividends.4

This is the second link between poverty and riskreduction. To restart their earning mechanisms thepoorer sections need capital, credit and insurance inmicro quantities. They have limited capacities to saveand hence methods have to be devised to addressthis gap.

Availability and access to low interest credit orinsurance and presence of institutions andgovernment mechanisms in urban areas is anotherindicator of the dimensions of the risks that poorersections face in event of a disaster.

Intervene to empower

The last section of the questionnaire dealt with themarket linkages at the micro level and the impact ithas on the four major livelihood types in the slums. Avast majority felt that the competition in the marketshas increased. The respondents stated two reasonsfor this increase.

Relocation again seems to be the main culprit in thisaspect. Of the respondents, 80% of the respondentssaid that in their profession regulars and passers byare their main customers. Since they are now stayingfurther away from the Walled City, (the central areain the city that has the maximum market activity, andwas also badly destroyed) their customer base hasshrunk. The average customers per day, whencalculated from the responses, have shrunk from 15per day to around 10 per day. The second and moresignificant is their claim that indiscriminate provision

of similar livelihood relief materials, like sewingmachines, in a limited area has tampered with theirmarkets. Thus, humanitarian relief and reconstruction,when without coordination, and market�sterile,dampens the local markets of the poor.

Availability of reliable information of the region withregard to its livelihood pattern and the marketlinkages is an indicator of risk management capacityof the area.

Existing vulnerability of the poor iscompounded by disasters

Among the respondents, more than 80% of therespondents felt that they were not able to invest inany saving mechanism. This number has increased to85% with the laborers showing no capacity to save.Among genders, women have also shown willingnessmore than men to save and improve their condition.

The average saving amount also has reduced from$11.5 to $4.5 per month. Not surprisingly, more than87% of all the respondents said that their extra costswere not covered.

One of the most important questions asked duringthe survey was: ‘Will your family would survive ifhe/she fell ill?’ An overwhelming number, 63%,replied in negative.

For most of the urban poor, security is never possibleas their economic and social condition renders themdefenseless from the onslaught of natural disastersand their sustained after effects. Provision of safetynets by government and humanitarian agencies andadoption of saving habits by the residents of slums isan indicator of disaster preparedness of and riskminimization of the region.

124

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Competition

70.2%

29.8%

Not increased Increased

Impact of Illness

Savings Account

36.3%

63.7%

Yes, my family will manage No, my family will not manage

0

5

10

15

Avg. Savings ($.)

Before earthquake After earthhquake

Reducing livelihood risk: the humancapital approach

Even during normal conditions, the urban poor live ina dysfunctional society with their livelihoods underconstant threat, daily food intake in an imbalance,and limited access to health facilities and safe water.Disasters attack the precarious security or asset baseof the poor and increase its rate of erosion. Liveli�hoods of the urban poor feed them, keep them cleanand healthy and provide them a shelter and educa�tion for their children. Unless their livelihoods areprotected, their security will continue to get repeat�edly erased and discarded when a disaster strikes.

The study unearthed some aspects that influencerecovery of livelihoods of the urban poor in Bhujthree years after a disaster. The study shows thathuman capital of the urban poor, as a result of gooddevelopment indicators like literacy, awareness andavailability of micro�credits, can govern the postdisaster recovery and reduction of risks to urbanlivelihoods. Post disaster recovery in Bhuj showedclear differences in recovery patterns with those withhuman capital clearly at an advantage (see graph).Human capital is therefore an important and cruciallink to reduction of livelihood risk.

Conversely, literacy rate of the region, availability ofmicro�credit/micro�insurance, market linkages forproducts that the urban poor makes and savingpatterns can be used as indicators for riskassessment of the region.

On this perspective, action steps can effectivelyprotect livelihoods from disaster risks, andsimultaneously speed up recovery is suggestedbelow. In succeeding boxes, examples of someattempts made by organizations in the samedirection are provided.5

125

1) Literacy is a crucial link between poverty anddisaster risk reduction. When people in acommunity have the freedom to develop andthe capacity to grow in the direction they chosethen we say that the society is empowered. Theawareness of rights and the confidence in theirability to manage crisis in their stride is thegreatest contribution that literacy, and theassociated skills and confidence that literacygives, can provide to poor people affected bysudden disruptions to their lives. Initiatives forskill and versatility development must bedeveloped for those whose occupations andlivelihoods remain unprotected, uncompensatedand vulnerable in low�income areas of cities.Spread of education across low�income and atrisk population can reduce recovery time andefforts. The demographic profile, geography anddiversity of the people demand a need basedcapacity building. Providing better living andworking conditions for the slum dwellers andbetter chances for education should be apriority for most urban rehabilitation agendaactions.

2) Relief agencies should take effort to find outwhat the region wants and take account oftheir skills prior to any disaster and support andaugment them accordingly. When livelihoodtools are given, market links should not bemissed out. When incomes are generatedsustainability beyond project life should bethought of. When livelihood needs aresupported the measures should be down top,inside out, sustainable, with links to the market,timely and in suitable scale. Impact of relief onlocal markets of the poor among the victimshas to be better understood. To ensure thatrelief and rehabilitation should not strain thelocal livelihoods a strong database oflivelihoods of the region should be maintainedwhich will included comprehensive informationon the livelihood type, resources availablelocally to support them and market linkagesthat exist during normal times from and to thelivelihood location. This information should bemade available to the intervening agencieswhich should take care in strengtheningweaken or damaged links rather than creatingnew unsustainable links.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Recovery with human capital

Recoverylacking human capital

TIME

SHO

CK

4) Disasters can be a spur for massive investmentin reducing risk to infrastructure. The plans forreconstruction of Bhuj are futuristic, advancedand earthquake� proof. Four lane roads and anupcoming new airport are converting Bhuj intoa prominent town in Gujarat. The benefits to the poorer sections, however, still have to filterdown. During the post disaster scenario, theslum dwellers have to be empowered to exer�cise their rights and realize opportunities in therehabilitation, relocation and poverty alleviationschemes. The spatial and locational needs oftheir livelihoods have to be given maximumimportance. Special work and business areas for livelihoods of the weaker sections must bedeveloped within the framework of the urbanplanning and reconstruction plans of disaster�prone urban pockets. Also, any relief, rehabilita�tion and mitigation efforts that reach citiesshould reach slums.

5) Build not use local resources. Any risk reductionmeasure when reaching the grassroots of thecommunity it works for has to simultaneously beable to empower them. Fortification of humancapital has to be undertaken in consultation,

126

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Box 2: LRF and Micro-Mitigation

LRF provides services specifically targeting disastermitigation and recovery through livelihoods such aslivelihood recovery planning and other key trainings.Examples of such activities are the literacy and basiceducation services sponsored by LRF in Bhuj.(Outgrowths of the study have indicated the direct linkbetween literacy and disaster mitigation for livelihoods.)These literacy activities include early education throughChild Care and Education Centers (Anganwadis) andadult literacy training. In addition, eleven capacitybuilding cycles were conducted in Bhuj, focusing ondisaster preparedness, SPHERE standards, emergencyfood and nutrition, safer shelter construction, amongothers.

LRF has a very clear and direct approach to address theissue of basing relief on need. Provision of relief isdemand�based, balanced by the demonstrated ability ofthe beneficiary to use what he or she gets to recoverand develop. For example in Bhuj, The range and varietyof livelihood support that DMI has given is long. Thebeneficiaries were given sewing machines, shops, drums,donkeys, cows, horses, bangles, teashops, hand�pushedcarts and wheelbarrows. GIS maps are being developedfor low�income habitats in Bhuj that will include all theresources available, security�wise and facility�wise withinBhuj and those that are required from outside. Maps ofthree slum areas are ready for use.

LRF beneficiaries have joined hands with DMI to createthe Chamber of Commerce and Industry of SmallBusinesses (CCISB), which will further assist them andother qualified relief�applicants in their continuedlivelihood recovery. By providing well�targeted financialand non�financial services, CCISB assists in stabilizingand strengthening the livelihoods of the disaster�affected, vulnerable poor of the Bhuj slums. The creationof the CCISB’s is DMI’s investment in the slumdwellers’future.

Box 3: Linking women

Self�Employed Women’s Association (SEWA) is amembership�based trade union working for incomesecurity of women for nearly three decades. Three ofSEWA’s initiatives in livelihood risk reduction are trulyinnovative and deserves mention. These include SEWATrade Facilitation Centre (TFC), Livelihood Security Fund(LSF) and Core Sale. (Refer www.sewa.org).

• SEWA TFC: SEWA has taken several initiativestowards facilitating marketing efforts andestablishing marketing linkages for poor women.STFC is a major initiative with a thrust on marketingof craft products made by them.

• Livelihood Security Fund (LSF): The main objective ofthis fund is to ensure that positive experiences inusing new and innovative economic activities as toolsto combat disasters are extended to all the disasteraffected areas as measures in disaster proofing andensuring livelihood security.

• Core Sale: In order to strengthen the livelihood ofartisans, the textile workers and the handloomworkers and garment workers of Gujarat, STC,Government of Gujarat and National Institute ofFashion Technology have come together to form the‘Core Sale’� an enterprise to upgrade skills, providesustained work and employment opportunities to theworkers in the informal sector. (Gujarat has a richtradition of handicraft)

3) Access to low interest credit and insurance tourban poor can raise their asset base andimprove their coping mechanisms in the shortand long term. Increase in livelihoodopportunities with matching access toinfrastructural resources in urban slums must bea priority. Saving mechanisms that incorporatethe earning and expenditure patterns of theslum dwellers have to be developed and thehabit of saving have to be promoted amongthem. Donors should support experimentalactivities through venture funds. There are manyreasons in favor of initiatives focusing on micro�credit/micro finance. It can achieve much interms of smooth recovery of poor livelihoodsand protecting them from future disasters.

communication and co�operation between thelocal micro and macro agencies, in the processexpanding partnerships and synergy for effectiveand efficient implementation. There is a need to develop a symbiotic relationship between the community and outside actors. The targetcommunity should be the agents of prepared�ness and focal points of information dissemina�tion to the community. Every step that is takenin the selected area should be in consultationwith the community that the intervening agencyseeks to serve. Once this is achieved, thecommunity feels sense of ownership of the

127

process, which helps the three partners(community, the local agencies and theintervening agency) to maintain a direct andtransparent relationship.

6) Urban intervention of humanitarian agencieshas to be up�scaled. Compared to their activities in rural disaster mitigation, the workby humanitarian agencies in urban areas (inGujarat) is insignificant. While there aresuccessful interventions by local organizations,mostly the urban vision is focused, sectoral andsmall�scale. In India, with the poverty fluxclearly shifting to urban areas from rural, thereis a definite need to shift from a minuteintervention approach to a more wide�scaleapproach in urban pockets.

Conclusion

The poorest in the society can benefit from theadvantages of overall economic development onlywhen their social development is in sync with thedevelopment of the country. Conversely, theimproved development indicators can (a) mitigatedisaster risks and (b) speed up recovery). Protectionof livelihoods of the disaster prone urban poordelivers them from the debilitating impacts of thethreat�hazard�disaster continuum and is a point forconvergence for social development and riskreduction initiatives in urban areas in developingcountries. Investment in human capital will work.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Box 4: Protect assets

UNDP has initiated livelihood approaches, dovetailedthem into rehabilitation programmes for effectivedrought mitigation and consequent disaster riskreduction. These are classified into three areas thatinclude:

• Structural Measures – UNDP in partnership withKutch Nava Nirman Abhiyan (KNNA) has supporteddrought proofing with watershed developmentprojects and repair and reconstruction of structuresthat were damaged.

• Non� structural measures – UNDP in association withthe Government of Gujarat, and local NGO networkshas set up the Kutch Ecology Fund (KEF) aimed atsupporting and facilitating the planning andimplementation of initiatives towards long�termrecovery and drought proofing of the region.

1 A partial list consists of:Small businessmen: pan shop, tea stall, garage, barbershop, etc.Small�scale vendors: vegetable hand lorry, chocolate and biscuit selling, etc.Home�based workers: stitching work, embroidery, milk producing, cowherds, etcLabourers: masonry, carpentry, scrap collection, etc

2 The objective of DMI’s Bhuj Reconstruction Project (BRP) set up in January 2002 is to reach out to a large section of theslum dwellers, in 36 slums of Bhuj, people who needed immediate help but were ignored during relief.At present, DMI works in 18 slums in Bhuj in which is home to around 32,000 people. BRP has enabled the creation of agroup of motivated slum dwellers to form a network — called ‘volunteers‘ — to work for their community. They haveformed Slum Area Committees (SAC) in their respective slums to organize and work to improve their living conditions. Thisgroup played a central role in the survey.

3 The graphs in this section are adapted from the presentation ‘Economic Recovery Survey’ by Jeremy Drucker, ArchitectPlanner with Bhuj Reconstruction Project (BRP).

4 The sample was comprised of livelihood relief beneficiaries of a Livelihood Relief Fund (LRF), which has a policy ofbeneficiary selection that is positively biased in favor of women.

5 The examples given are certain effective strategies and action steps that contribute to livelihood risk reduction (and haveworked on the ground). To preempt the risk of limiting the boundary of the study to LRF, and to provide a holistic vies of thelivelihood approaches in Gujarat, this section has embedded into it, innovative measures that other organizations haveundertaken in other regions of the state. These include a few references to rural livelihood risk reduction initiatives as well.

128

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

129

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Strengthening rural community

bonds as a means of reducing

vulnerability to landslides

TEAM LEADER

Rutere Salome Kagendo<[email protected]>

PROJECT ADVISORS

Dr. Robinson M. OcharoProf. Octavian Gakuru

Kenya

SYMPOS IUM NOTES

130

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

The aim of this study was to investigate into theexisting rural community bonds in the landslideaffected areas in Kenya. This was deemed necessaryby the fact that when disasters occur, communitymembers in the disaster prone areas are affected inone way or another. The researcher found it necessaryto pay attention to the important role played by theexisting community bonds when a disaster strikes.

The study gathered information from the respondentsby use of questionnaires and tape recorders from twoadministrative landslide prone districts in Kenyanamely Murang’a and Meru. These districts werepurposively sampled due to their repeated occurrenceof landslides causing deaths and destruction ofproperty in the recent past.

The study anticipated data on the community’sperception of what landslides are, their causes, theavailable resources for landslide hazards preventionin the community and the existing network fordisaster management in the community.

Data gathered reveals that although the communityis aware of the causes of landslide, they are stillvulnerable to landslides because they remain in thelandslide risk areas.

Further they have continued with developmentactivities in the areas they mapped as high landsliderisk areas. Another important finding is that NGOsand the Government have not played any vital role inreducing vulnerability to the victims of landslide andthe community at large. The community expressedtheir concern that landslides are on the increase evenin areas that were not prone to landslides earlier on.It has also been established that the communitybonds have weakened with time and therefore thereis need for the community/government/NGOs tointervene by strengthening them.

It is hoped that the community members in alllandslide prone areas will one day be at peace to beable to carry out their daily activities without beingvulnerable to landslides. It is also hoped that thecommunity will one day be able to manage landsliderisks by reducing vulnerability.

Introduction

The main objective of this study was to find out howcommunity bonds amongst the rural people can bestrengthened with an aim of reducing vulnerability tolandslides. Emphasis was on how to maximize theexisting community bonds in landslide risk reduction.

131

Landslides are not a new phenomenon in Kenya.They have been occurring in different parts of Kenyasince time immemorial, but in the recent years theyhave been on the increase in some parts of thecountry. Their destruction to human life and propertyhas also been on the increase. The most vulnerableareas in kenya are: Murang’a, Kirinyaga, Kisii,Mombasa island, Nyeri and Meru. These areas enjoyhigh rainfall of over 1200 mm annually. Most ofthese areas are very hilly with steep slopes.

In the recent past, Landslides have been recurring inboth Murang’a and Meru districts claiming severallives and property. Murang,a and Meru are adminis�trative districts in Kenya. Murang’a is a district incentral province of Kenya while Meru is in theeastern province of Kenya. For the purpose of thisstudy, murang’a and Meru districts were sampled.

In this study, community members in both sites weresampled and mobilized to identify landslide hazards.This was done with the help of local leaders (chiefs,headmen, church leaders and group leaders) in thesaid areas. These respondents identified actualelements at risk, which included life and property.In as far as life is concerned; several people havelost their lives while others have maintained injuries.Others have to live with the trauma of the landslideexperience, having been victims of the landslide orhaving witnessed their loved ones die in thelandslides.

The community is still at risk and vulnerable tolandslides. No changes or efforts have been done toimprove the situation. For instance, they carry outtheir daily activities in the same places wherelandslides have occurred. Worse even is the fact thatone can actually see deep cracks on the earthsurface that were caused by the landslides thatoccurred recently.

The community members are aware of the risksinvolved in these cracks but claims that they have nochoice but to live with the fear of landslidesoccurring there again. Actually, they have carried outtheir farm activities even on the spots where theactual landslide took place; only leaving out placeswhere cracks are visible. All this suggests that thesepeople are still at risk and very vulnerable tolandslides. Landslides that have been occurring inthese areas have destroyed a lot of property andmore is at risk. In fact the community membersdisclosed that some of their neighbors are poorbecause of the impacts of the landslides, whichdestroy their property. Some of the property

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

identified as being at risk include, houses, domesticanimals, crops, roads, schools and arable land.

This study was carried out during a rainy season. Thismeans that farm activities were being carried out onthe farms. The researcher found people on their farmseven were landslides had occurred and actually leftsome lose soil. Some respondents argued that it isfate to die in a landslide. That if one is destined todie in a landslide, then there is nothing one can do.Further they argued that they did not have otherplaces safe from landslides where they could migrateto. All this talk can be interpreted to express signs ofdespair and hopelessness. The fact that they haveseen landslides occur and claim lives and propertyand nothing has been done to salvage the situationhas made them give up.

Landslides have been on the increase in Murang’adistrict, which is hilly and highly populated. Humandevelopment activities have activated this hazardthat has frequently caused intensive damages tothose that have been affected. The communitymembers are vulnerable and continue to lose theirlives and property.

Problem statement

Landslides are a major threat each year to humansettlements and infrastructure all over the world.Kenya is not an exception of this hazard and datacollected reveals that landslides are on the increasein the country. Landslides have occurred in severaladministrative districts and the fatal ones occurring inMurang’a and Meru districts. This study is guided bythe following research questions:

• What are the causes of landslides in Kenya?

• Why are landslides on the increase in the country?

• What is the level of community participation inlandslide risk reduction?

• Are the existing community bonds strong enoughto help reduce vulnerability to landslides?

• What resources are available in the community tohelp reduce vulnerability to landslides?

Goals and objectives of the study

• Community mobilization for hazard identifi-cation. This includes definition of landslides by the community, establishing historical causes andeffects of landslides, and hazard mapping.

• Establishing the resources available for hazardprevention in the community. This includesminimizing vulnerability, mitigation and recovery.

• Establishing network for disastermanagement. This includes intra community,NGOs and government.

Justification of the study

It is in the interest of all well wishers to help savelives and also protect property from any kind ofdestruction. Whenever a landslide has occurred in thenamed places it has left a trail of death and massiveproperty destruction. To rescue the communitymembers from landslides vulnerability, there wasneed to carry out this kind of study.

Other than trying to bring in external forces toaddress issues of how to address landslides inlandslide areas, it is important to involve thecommunity itself. This way they will take the projectas their own and sustain it.

Whenever a landslide occurs people loss their livesand property is damaged. In order to curb this therewas need to investigate into the research topic. Thisway lives and property will be saved and probablylandslide risk and vulnerability be reduced tomanageable heights.

Scope and limitation

Although Kenya experiences several hazardsincluding famines, floods, and fires both in rural andurban centers, the study concentrated on landslides.This was necessitated by the available resources andthe time frame within which the project should havebeen completed. However there is need for intensiveresearch in disaster management to establishpossible ways of reducing vulnerability to hazards.

Research methodology

Site of the study

As mentioned earlier, this study was carried out intwo administrative districts in Kenya. These districtswere selected purposively because they had experi�enced the worst landslide damages (which includedloss of lives and massive loss of property) at the timethe study was being undertaken. Also people are stillvulnerable to landslides in these areas.

132

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Sources of data

Data was gathered from primary and secondarysources. Primary data was gathered by use ofquestionnaires and tape recorders through individualinterviews and focused group discussions. Secondarydata was gathered from any relevant works donerelated to this study through literature review.

Sample size

As stated earlier, this study interviewed respondentsthrough focused group discussions, Key informantsand also individual interviews. There were sevenfocused group discussions, seven key informants, andtwenty individual interviews per site. In total therewere fourteen focused group discussions, fourteenkey informants and forty individual interviews. Allthese were randomly sampled to create arepresentative sample.

Community mobilization for hazardidentification

Community definition of landslides

Landslides are not a new phenomenon in Kenya. Theyhave been occurring in different parts of the countrysince time immemorial. However, landslides disasterswere not as destructive as they are today. The highdestruction of property, lives and livelihood can beattributed to population increase and the weakeningof community participation in the prevention,management and response to disasters.

As far as landslides are concerned today, the mostvulnerable areas in Kenya are: Murang’a, Kirinyaga,Nyeri, Meru, Kisii, and Mombasa. These places enjoyhigh rainfall of over 1200 mm annually. Most ofthese areas are very hilly with steep slopes. In recentpast, landslides have been recurring in bothMurang’a and Meru (administrative) districtsclaiming several lives and property. The two districtsinvolved in this study had the following definitionsfor a landslide:

• It is movement of lose soil that has no support (no vegetation to hold it).

• It is soil erosion.

• It is movement of soil under the influence of toomuch under ground water.

• It is a place with too much underground waterthat has no out let and therefore forces its way out to the surface and hence a landslide occurs.

• It is massive movement of soil.

133

Although most of the community members could notdefine a landslide using the academic language, itcame out very clearly that they had full knowledgethat it involved movement of lose soil and rockdownhill. They were fully aware that this occurredmostly during rainy season or occasionally whenpeople mine sand under the hills. With this, thecommunity members ended up identifying thehistorical causes of landslides as being:

• Too much underground water

• An earthquake

• A curse/revenge from a supernatural power

• Intensive human activities on weak earth surface

• Heavy rainfall

• The gradient of the landscape

The respondents complained that there is a lot ofunderground water in their land as a result of heavyrainfall in the area. This water has no outlet andforces its way out and in the process causes a land�slide. They attributed this to the fact that their land�scape is hilly and with a lot of underground water.

Some had other explanations. For instance landslidesare blamed on supernatural powers. Some peopleargued that landslides are as a result of humaninterference with sacred grounds. According to them,sacred grounds should be left fallow for divinepurposes. Interfering with them meant annoying thegods and this is why the gods revenge throughlandslides. They believe that landslides are caused bygods who are on the verge of revenge for havingbeen interfered with. This category of people needsto be educated on the causes of landslides andprobably how they can be managed.

The study also established that landslides are causedby intensive human development activities on losegrounds. These activities further the weakness of thesoil causing a landslide.

Hazard mapping

In this study, community members in both study siteswere mobilized in order to do risk mapping,identification of elements at risk, and planning onhow to reduce vulnerability with maximum relianceon the community resources. This was done with thehelp local leaders (chiefs, headmen, church leaders,and other group leaders in the said areas.

The respondents proved to be fully aware of theplaces where landslides had occurred and wherethey could occur in future. In places they predicted

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

occurrence of landslides in the near future, therewere big crack on the ground threatening to cutanytime. The respondents argued that it is only amatter of time and a landslide would occur in such aplace. They are so vulnerable that one can almost tellfor sure a landslide will occur any time. They areaware of the danger but have no other alternativeother than taking refuge in safer areas when alandslide occurs in their place and return when theyassume they are out of danger.

The community members are forced to relocate tosafer grounds whenever it rains because their landsbecome impassible, and landslides occur. The safergrounds in most cases are a distance from theirhomes. They move to the near by safe towns untilwhen they think they are safe. However their safetyon return is questionable because a landslide canoccur at anytime regardless of whether it hasstopped raining or not. These people are neveradvised by experts to return to their homes. They justassume that since the rains have subsided, thereforeit is safe to return home. This means that they couldstill be at risk and this is why some have beenvictims of landslides even after predicting that it wasgoing to take place.

The questions to ask are: When are these people atpeace to carry out there daily activities without fear?What is the social, economic and political cost of alandslide to the victims and the country at large? Therespondents argued that they are hardly at peace;any loud bang or unfamiliar sound sends fearthroughout their bodies. This means that they arenever at ease and more so during the rainy seasons.It was also established that landslides escalateduring rainy seasons.

The knowledge held by the community membersabout the areas that can be affected by landslides isincredible. The community seems to have learnedfrom history about occurrence and reoccurrence oflandslides. Some community members warn withcertainty that a landslide will occur at a specificplace. However all this knowledge has not been putinto good use. Although they can map all landsliderisk areas in their district, and seem to be aware ofthe causes of landslides, they are still quitevulnerable. They live and carry out their developmentactivities (construct houses, carry on extensive/unplanned farming activities) in those risk areas.

One can easily deduce that these communitymembers are ignorant or probably are not aware ofthe consequences of a landslide. The truth is that thepeople have no option but to live in these areas. In

fact some respondents argued that if they could berelocated to other safer grounds they would be morethan willing to settle there. When interviewing themone sees/feels a resigned attitude towards landsliderisk reduction amongst the respondents. Somerespondents argued that this is fate, while othersblamed the government for deserting them. All thisboils down to lack of technical know how andresources to reduce their vulnerability.

Although the community members are aware of alandslide, their causes and effects no seriousprecautions have been taken to reduce theirvulnerability to landslides. For instance they identifiedplaces where landslides could occur in future and yetpeople are still living there and carrying on with theirdevelopment activities in the risk marked areas.

Effects of landslides

Landslides have had several adverse effects on thevictims as follows:

• Loss of life (Many lives have been lost throughlandslides where people have been buried alive forexample in January 2001, five members of thesame family were buried alive. The two membersof the family who survived the ordeal are stilltraumatized to talk about it.

•Loss and destruction of property (there has beenmassive lose and destruction of property whenevera landslide occurs. These loses includes, housesand belongings therein, destruction of crops,farming grounds, and infrastructure).

•Traumatizing experiences (Victims of landslides arealways traumatized by the ordeal. In most casesthe victims are buried alive and those who survivelive in fear of going through the same again.

•In the landslide prone areas, the infrastructure ispoor. There are no roads. What this means is thatwhenever a landslide occurs the victims/survivorscannot be rushed to hospitals. For instance in the2001 landslide that claimed five members of thesame family, two were pulled out of the wreckagealive but one died on the way to the hospital. Thisis because there are no roads and therefore nomeans of transport to the hospital. The other — a young boy then — survived to tell the tale.

•Disruption of normal life. Whenever a landslideoccurs, life can never be the same again and moreso for the victims. For example they have to learnto live without their loved ones, lucky survivorsmay relocate to other safer places if they are luckyto have some, and try to come to terms withreality.

134

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

This study was carried out during a rainy season. Thismeans that farm activities were being carried out onthe farms. The researcher found people on theirfarms even were landslides had occurred and actuallyleft some lose soil. Some respondents argued that itis fate/destiny to die in a landslide. That if one isdestined to die in a landslide then there is nothingone can do. Further they argued that they did nothave other places safe from landslides where theycould migrate to. All this talk can be interpreted toexpress signs of hopelessness. The fact that they haveseen landslides occur and claim lives and propertyand no efforts have been done to salvage thesituation has made them give up.

Establishing the resources availablefor hazard prevention in thecommunity.

It was deemed necessary to establish the resourcesavailable in the community for hazard reduction/prevention. This study investigated what the commu�nity had to enable them minimize vulnerability. Thiswas done by asking the participants to name all thepossible resources in their community that can bechanneled to prevention of landslide hazards.

It was established that there are no funds set asidefor reduction/prevention of landslide hazards. Furtherthere are no formal plans to fight this hazard thathas claimed lives and property severally in thesecommunities.

The community members however, try to help theirneighbors by giving handouts to them whenever theyare affected by a landslide. For instance they givethem accommodation, food and clothing until things get to normal again. They also give thempsychological support whenever necessary.

As stated earlier, infrastructure in these areas is verypoor. For instance there are no roads where land�slides have been occurring. The dry weather pathsthat are there are impassible during the rainyseasons. It is during the rainy seasons that landslideshave been occurring. When landslide occurs, it is verydifficult to rescue the victims and more so to rushthem to the hospital. The place is quite hilly, and withno proper roads. The community members havefound it unnecessary to own vehicles and this makesit more difficult to save lives when a landslide occurs.People are forced to walk for long distances withvictims in order to get any medical help. This is tosuggest that the hospitals are very far and there areno quick means of transport.

135

The community members actually confirmed thatvictims of 2001 landslides died because of lack ofmedical care. It took the neighbors quite some timeto pull them out of the rumbles and carry them tothe nearest hospital.

It is important to note here that the rural society justlike its urban counterpart has lost its traditionalcommunity bonds that tied it together. These bondswere very helpful in times of hazards because peoplecould help one another and also look for solutionstogether. This study established that these bondshave weakened over time and with further frustra�tions from landslides one can almost conclude thatthey are too strained to be of great help. What thismeans is that community members cannot for surecount on one another for help in case a hazardstrikes. This is not to dismiss totally the importanceof the role played by these bonds. These bonds arevery important and the study highly recommendsthat there is an urgent need to strengthen them anduse them to reduce hazard risks.

The respondents also express their disappointmentswith the government. According to them, thegovernment has not been of help and especially sowhere deaths have not occurred. For instance, thisstudy established that the only thing the governmentdid during the 2001 landslide that claimed severallives in Murang’a, was to hold a fundraising in thearea. The proceeds that were realized were used tobuild a house for the surviving victims and supportthem financially, which they claimed lasted for onlytwo months. They also received help from theirneighbors for a short while and now they are ontheir own. The problem is that all they had wasdestroyed by the landslides leaving them poor.

The community blames the whole issue on thegovernment, which they argue has neglected them.According to them, the solution to this problemwould be to be relocated them to other placeswhere they’re no landslides. Others feel that theycan continue living in these landslide prone areassince they have lived there for many years so long asthere is a good infrastructure. Others would like tolive else where but be able to access their farms asneed be.

In fact the community added that the governmentshould be in apposition to know when a landslidewill occur and warn them in advance to move out ofthe risk areas. They added that the governmentshould be able to warn them early enough and alsoshare any research findings about landslides with

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

the community. This way the community is keptinformed of any possibilities of risk and vulnerability.

All these opinions do not provide a solution to thisproblem of landslides. There is need for allstakeholders; the community, the government, NGOs,researchers and all other well wishers to come upwith a good strategy to address this problem asdiscussed later.

Establishing network for disastermanagement

This includes intra community, NGOs and govern�ment.This study established that although landslideshave been destructive and fatal there is no net�working between the stakeholders. For example asdiscussed earlier, community bonds are weak, thegovernment seems to have ignored its role in disastermanagement, and the non governmental organiza�tions have also not taken ant action in risk reduction.The result of this has been deaths and propertydestruction. There is need to bring all the stake�holders together to strategize on the way forward.

Lessons learned

Preparation of a landslide hazard map

• It is important to locate areas prone to landslides.This permits planners to determine the level of riskand to make decisions regarding avoidance,prevention or mitigation of existing and futurelandslide hazards. These techniques rely on pasthistory, topographic maps, bedrock data and aerialphotographs.

• The most effective way to reduce damage causedby landslides is to locate development on stableground and to utilize landslide susceptible areas asopen space or for low intensity activities. Land usecontrol can be enacted to prevent hazardous areasfrom being used for settlements or as sites forimportant structures. The controls may also involverelocation away from the hazardous areaparticularly if alternative sites exist.

• Public education programs will help people under�stand the causes and the effects of landslides,identify unstable areas and avoid settling orhaving ay developments on them.

Monitoring, warning and evacuatingsystems

• Areas susceptible to landslides should bemonitored to allow timely warning and evacuation.

• Monitoring and warning systems should placeinhabitants on alert when heavy rains occur or ifground water levels rise.

Mitigation efforts

• Mitigation involves not only saving lives and injuryand reducing property loses, but also reducingeconomic activities and social institution.

• To reduce physical vulnerability, weak elementsmay be protected or strengthened. To reduce thevulnerability of social institutions and economicactivities, infrastructure may need to be modifiedor strengthened or institutional arrangementsmodified.

• The focus of mitigation policies must be onreducing the vulnerability of the elements andactivities at risk.

• All members of the community should be aware ofthe hazards they face, know how to protectthemselves and should support the protectionefforts of others and of the community as a whole.

• It is important to empower the community bypromoting, planning and management of its owndefenses and obtaining outside assistance onlywhere needed.

Summary, conclusions andrecommendations

Landslides are one of the many hazards that occur inKenya. Other hazards include; frequent droughts andfamines, floods, fires, environmental pollution, anddeforestation. As discussed in the text, landslideshave been occurring since time immemorial but arereported to be in the increase and are more fatal anddestructive than any other time in the Kenyan history.The study also revealed that community bonds haveweakened hence the weakening of strength inreducing community’s risk reduction strategy.

Conclusions

There is need to strengthen rural community bondsas a means of reducing vulnerability to any hazard inKenya.

136

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Recommendations

• The community should participate in all efforts toreduce risks and vulnerability to all hazards in theirsociety. This way they contribute in helping themget helped and at the same time a level ofownership of the project by the community iscreated. This goes along way in making the projectsustainable because the community already ownsit and will take care of it.

• In order to fight hazards successfully, there is needto involve all the stakeholders in these projects.This is in line with the fact that resources are fewand sometimes we see replication of projects thatare done haphazardly. There is need for seriousconsultations amongst the stakeholders.

• There is need for proper infrastructure (roads,telephone, hospitals etc) in hazard prone areas.

• Research findings should be disseminated to thestakeholders. This may help them understand thehazard more and learn how to deal with it. Therespondents requested that the findings of thisstudy should be disseminated to them.

• There is need for more research in disastermanagement area. The data collected could helpto come up with good policies which ifimplemented well, could help reduce vulnerability.

Areas of further research

Hazards in Kenya are on the increase and moredestructive than any other time in the history ofcountry. For example, landslides, floods, famine andfires have caused and continue to cause a lot ofdamages to property and more often cause deaths.

137

This year alone there has been a wave of firebreakouts in the main towns of Kenya causingmassive destruction of property and life. Fires in theslums of Nairobi are almost a daily routine. Thissuggests that those living in these areas are alwaysvulnerable. Floods in the country are also an area ofconcern. For instance ‘Budalangi’ floods are worthsome attention to establish a lasting solution to theproblem. Every year a lot of property is destroyed inthis district whenever it rains and people are lefthomeless and poor. Famines strike most of the northeastern province and some parts of the easternprovince causing deaths of both people andlivestock. More often victims of hazards are givenhandouts in form of food and later left on their own.These hazards recur almost every year and nopermanent solutions have been found. This is tosuggest that there are victims of hazards every yearin Kenya and there are no viable plans in place toreduce this vulnerability and therefore managedisasters. As a result of this, many people in the riskareas continue suffering and becoming more andmore poorer and this hinders development.

One of the greatest obstacles in this area of disastermanagement is the dearth of information regardingthe extent of the damages caused by hazards,vulnerability and the possible solutions to thesehazards that have caused people to remain poor.Further there is the problem of uncoordinatedactivities between those involved in this field andtherefore information is not shared. This calls forintensive research in the area of disastermanagement to establish the causes and solutionsto the increasing risks and vulnerability to hazards.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

138

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

139

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

GIS and remote sensing for

flood disaster identification:

a case study of the Koshi River

basin in Nepal

TEAM LEADER

Archana Pradan <[email protected]>

PROJECT ADVISORS

Dr. Vishnu DangolD. Prakash Chandra Adhikari

Nepal

SYMPOS IUM NOTES

140

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

This research is conducted for flood hazardassessment and risk identification around the Koshiriver area. This research is a Provention Consortiumfunded project for the applied research grants fordisaster risk reduction program.

Floods and landslides are the most destructive typesof water�induced disaster in Nepal. In July 1993,Nepal experienced a devastating flood in the Terairegion, which took the lives of 1,336 people and left487,534 people homeless. In 1999, floods and land�slides killed 113 while 47 people were reportedmissing and 91 seriously injured. A total of 8,844families were affected, 3,507 houses and cattlesheds were destroyed and 177,32 hectares of landand agricultural crops were ruined in the year’sfloods and landslides (HMG/N, 2000). The disastercaused a total loss of NRs. 3.6 million. In July andAugust 2002, Nepal experienced numerous land�slides and floods in the Eastern and Central regions,which took the lives of 451 people and affectednearly 55,000 families. The estimated loss of propertyfrom natural disaster is about NRs. 10,000 millionannually, which is about 20% of GDP (DPTC 1995).

Study area

The Koshi River, one of the largest rivers of the world,is also a major tributary of the River Ganges.Reoccurring flood events in the Koshi River causesheavy loss of human lives and physical properties inNepal and Bihar of India every year (Pradhan, 2000).During the monsoon season, the problem of floodingand water logging occurs throughout the area.

The present study area is located in the southeasternpart of Nepal. The study area extends from Chatarato Bhantabari, and it covers three districts, namelySaptari, Sunsari, and the southern tip of Udaypur(Fig. 1). The villages that fall within the project areaare Chatara, Byarban, Rajabas, Prakashpur,Chakraghatti, Jhumka, Madhuban, etc. all of whichborder the bank of the Koshi river.

141

Objectives of the study

The main objective of the project was focused on theflood hazard assessment and risk identification usingtools such as GIS and Remote Sensing with detailedfield survey. An important component in this studywas the socio�economic data analysis whichassessed local people’s perception and indigenousknowledge for risk identification from floods. Theseobjectives can be listed as:

• To identify the perception and consequence offlood disaster in local people.

• To evaluate the socio�economic impact of flood.

• To prepare detail flood hazard map and to identifythe frequency, magnitude, and extent of flooddisasters experienced.

• To make recommendation regarding the decisionfor investment and design for sustainable projectsthat will withstand the impacts of potentialhazard events.

Regional geology and climate

The Terai plain lies in the northern part of the Indo�gangetic plain. Tectonically, the area is situatedsouth of the Siwalik zone, separated by the MainFrontal Thrust (MFT) and extends up to the banks ofthe Ganges River, over which the Terai plain isassumed to be thrusted.

The study area represents most of the Terai plain ofthe Sunsari district. Quaternary geological study ofthe area revealed us five lithostratigraphic units.Sediments of the area are in general fining towardssouth. The upper surface is made up of boulder andpebble beds whereas the lower surface is predomi�nantly fine clay and silt. The surficial sediments ofthe area can be divided into the five lithostrati�graphic units (Pradhan, 2000), described in Table 1.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Figure 1: Location of the study area

Stratigraphic unit Composition Distribution

BayarbanPebble with sand andfew clay

Around Bayarbanvillage

Madhuban Fine sand, silt and clayAround Madhubanvillage

Rajabas Fine to coarse sandAround Rajabas,Prakashpur villages

ChakraghattiCoarse to mediumsand with clay veneers

Around Chakraghattivillage

ChataraBoulder with coarsesand, silt and clay

Exposed aroundChatara village inChatara Dharan road

Table 1: Lithostratigraphy of the study area

Climate

The study area has a subtropical climate. Theestimate of the Koshi alluvial fan can be described asa transistion between the dry and moderatelyextreme climate of the lower hilly area and the Teraiflood plain. The average rainfall in winter (Decemberto Febuary) is approximately 32 mm, whereas in themonsoon it can be as high as 550�725 mm. Themosoon contributes 70 to 80 percent of the totalannual rainfall in the area. The rainy days aremaximum in July and August.

Flood analysis

Estimation of flood volume/rate is an important taskfor planning and design of flood regulation work andmeasures to be used for river protection. In order todesign the flood control, irrigation and hydropowerprojects, the knowledge of Probable Maximum floodvolume and its corresponding stage are essential.

The Probable Maximum Rainfall and the ProbableMaximum Flood were computed by using Hazen’smethod, California’s method and Weibul’s method.Among these three methods the Hazen’s methodgives the least value and the California’s methodgives the highest value. The Probable MaximumRainfall for the recurrences interval of 10, 100, 1000,10,000 years by using California’s method are 2650mm, 3500 mm, 4350 mm, 52000 mm respectively.Again, the Probable Maximum Flood by California’smethod for the recurrence interval of 10, 100, 1000,10,000 years are 10, 100 cumec, 15,900 cumec,21,100 cumec, 26,300 cumec respectively.

Flood hazard assessment

The flood hazard mapping and assessment startedwith the study of aerial photograph and LandsatThematic Map (7 bands) of Koshi River Basin. The useof satellite remote sensing method is useful inpreliminary assessments during the early stages of adevelopment planning study because of the small�to�intermediate scale of the information produced andthe ability to meet cost and time constraints. Remotesensing imagery was used to assess the currentstatus of floodplain and flood�prone areas. Thedelineation and characterization of the flooded areaswas based on relationships between physicalparameters such as reflectance and emittance fromfeatures located on the surface. Reflectance andemittance decreases as the water content or the soilmoisture on the ground increases. Accordingly, theflood plain and water bodies were delineated. Theimportant bands of the Landsat TM imagery thatwere used in the analysis are given below:

Band 1 – useful in determining concentration ofwater bodies

Band 4 – useful in delineation of water bodies

Band 6 – This band is especially useful in studyingfloodplain and soil moisture anomalies

The floodplain and the flood�prone areas that weredelineated from Remote Sensing imagery and theaerial photographs were geo�referenced anddigitized. This information was used in the productionof the final maps as well as for the GIS analysis. Theinitial information, as obtained from Remote Sensingand Aerial photograph were later verified during thefield visits to those sites. Any new and missinginformation was later added and amended. The newupdate of data from the field was later on added tothe initial database to produce that final hazard map.Also, the method of participatory GIS was used,incorporating local people’s knowledge of historicalflood events and their impact. Accordingly, the finalflood hazard map was prepared and relationshipsbetween the floodplain and the spatial features mostvulnerable to flooding were analyzed which isshown in Figure 3.

As seen from the filed survey and in the flood hazardmap, the flood casualties are mostly confined alongthe riverbanks of the Koshi River, starting fromChatara (Where the river starts to Fan outward) andnear up to the Koshi barrage. The impact of the floodis more severe along the east bank that is moredensely populated with high agricultural practices.

142

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Hydrograph of Saptakoshi at Chatara-Khotu Mean monthly discharges

JAN FEB MAR APR MAY JUNE JULY AUG SEP OCT NOV DEC

Months (1977-1995)

Series1

0

Dis

char

ges

1000

2000

3000

4000

5000

Figure 2: Mean monthly discharges hydrograph ofSapta-Koshi at Shatara-Khotu (Mean monthly discharges).

The bottleneck structure formed by the Koshi barrage(Koshi dam) near the Nepal�India border is alsofacilitating the river to accumulate large volume ofwater which ultimately discharges water laterallyovertopping the fertile land despite the constructeddikes and leeves. Mostly Bardaha and Bhantabariarea on the southern part of the map, as shown inFigure 3, frequently suffers from such inundation.

The study showed that the area could be divided into three flood probable zone, low, moderate andhigh hazard zone. The highhazard zone is characterizedby recurrence interval of 4�7 and 10�12 years inmoderate hazard area yearsalthough inundation of landand soil erosion is frequentin every monsoon season.The high flood hazard zoneencompasses an area of16.21 sq km, moderate12.15 sq km and the lowhazard zone comprises an area of 32.11 sq kmrespectively. Mostly theflood�vulnerable area islocated on the eastern bank of the river, fromChatara to Bhantabari area.The settlements particular of the Prakashpur VDC and

143

Varahkshetra VDC are in potential threat from flooding.Within these VDCs the most vulnerable area is in thePrakashpur VDC, around the Rajabas and Madhuvan. Thepeople residing in the Rajabas areas fall within the probableflood high hazard zone category and are in potential riskfrom flooding. These areas frequently experience flooding ofrecurrence interval of 4�7 years although inundation duringmonsoon season is common. The affected settlementswithin those localities are Maratol, Joginia, Bhawarahitol,Koiriyonitol, Lavatolia and Bithtol. The affected settlementswithin the study area are shown in Figure 4.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Figure 3. Flood hazard map, showing the low, moderate and high hazard zones

Figure 4. Affected settlements from flooding

The people in Rajabas have suffered already a largeloss of life and property in 2037 B.S. The infrastruc�tures such as dikes and spur constructed to protectfrom flood and rive bank erosion have also beendegraded from the sedimentation of the river. Flashfloods are mostly seen around the Bhantabari,Bhardaha and Hanuman Nagar area which inundatesthe land from few hours to 5�7 days. This hasdegraded the fertility of the cultivated land and even have made crop plantation vulnerable duringmonsoon seasons. Some protective engineeringstructures such as dikes are also in danger from the river sedimentation and bank erosion.

Socioeconomic analysis

The study area involves 14 Village DevelopmentCommittees (VDC) of Sunsari (6 VDCs), Saptari (7 VDCs) and Udaypur (1 VDC) Districts. The projectarea only covers a small portion of the southern tipof Udaypur. The total population of three districts isestimated to be 1,195,915, settled in 221,436households (2001 Census). These two districtsaccounts for 30.1 percent of the total population ofthe eastern region and 5.9 percent of the country’spopulation. Population density of the area isestimated at 460 persons/km2, which is extremelyhigh compared to the national average (157 person/km2), regional average (188 persons/km2) and teraiaverage (330 persons/km2).

Of the 90 respondents interviewed, 84.4% were menand 15.6% women. Some of the female respondentswere de facto heads of households in the absence ofhusbands. Age of the respondents ranges from 26 to87 years with an average of 52 years. Over half ofrespondents are between 41 and 60 years with28.9% over 60 years. Among the respondents inter�viewed 8 of them are landless. Initially, most of thepeople of the Koshi flood plain area are subsistencefarmers/agriculturists.

The flood events of various years have done signifi�cant loss of land and property. Table 2 shows themajor changes in land ownership due to floodevents. Though the loss of land has not taken placeat once in most of the cases it has resulted into lossof land and ultimately the livelihood of the people ofthat area. Due to the loss of land the people arecompelled to do labor work instead of farming atnearby sub�urban areas. Other activities include jobsin herding, petty commerce, governmental/non�governmental services etc. Sixteen of the respondentshave salaried employment. The 26.66% (24) of therespondents have only agriculture/farming as theiroccupation.

Household survey and mass meeting in the VDC wasdone to understand local people’s perception of theflood and their traditional practices to cope with it.Among 90 peoples interviewed, most of them wereignorant about the causes of the floods and thinkmostly as the intense rain for the primary cause. Therespondents were asked: whether the flood disasterin lowland (their area) is due to the mismanagementof watershed in upland areas? Most of them (50%)did not mention that as the reason for the heavyflood. While 30% were not sure about this and theremaining 20% told that the flood disaster inlowland, among other reasons, is also due tomismanagement of watershed at upland.

Most of the respondents could not remember thedates of the event (flooding). The respondents couldnot trace the dates of the flood events with greaterdamage out. So the precise dates could not be tracedout because of the categorization of the extent ofloss by the flood differed from one individual toother. The most repeated dates of flooding eventswere the events of 2011, 2022, 2025, 2037, 2039,2044, and 2052 B.S. Due to heavy rainfall duringevery monsoon land mass wasting/flooding is experi�enced in the area. The majority of respondents notedthat the loss due to flood increase at every 4�7 yearinterval. However, the devastating effect due to theflood and loss and damage of property experiencedevery 10�12 years time. The main problemsmentioned by the respondents were problems inlivelihood, loss of land and houses and properties.

These people understand that they are living in theflood risk prone area but can’t move to safer placessince they don’t have sufficient money to migrate/invest to other places. Very few don’t believe thatthey won’t get problems from flood in their area.

144

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Initial landownership

Land ownershipafter damage dueto flood event(s)

% changein land

ownership

12 Bigaha 2 Bigaha 83%

8.5 Bigaha 2.5 Bigaha 70.5%

7 Bigaha 6 dhur 99.7%

5 Bigaha No land100%

(no land)

3 Bigaha 6 kattha 93.3%

1 Bigaha 10 kattha 50%

Table 2. Major changes in land ownership fromflooding. Source: Field Survey, 2003

The other reason they mentioned for living in thoseflood vulnerable area is the social attachment. “Noalternative! What to do and where to go?” said oneof the respondents. They do not have land at othersafer places. They told that they have no other goodalternative, so it’s their compulsion and are forcedto live there. Some view that the problem due toflood will be reduced or managed after some period.Some told that “Even though we suffer due to hard�ship, we do not want to leave this place because atleast small piece of land is here with us and therelatives are at nearby places”.

These peoples are ignorant of the early symptoms offlood and rainfall. Few respondents stated that theearly symptom of the flood is the crawling of ants tothe tips of the grass found in the riverbanks of theKoshi. Few mentioned that the deepening of thecolor of the cloud to blackness results into heavyrainfall and subsequently to flood in the River. Otherfew respondents told that the abnormal movementof birds is also one of the symptoms of heavy rainfall.

It has also been seen that very few people aregrowing bamboo at gullies and Dalbergia sisoo(Sisoo) at the Koshi flooded area to reduce or lessenthe damage from flooding. However, most of thepeople believe that they can not control the floodingfrom the Koshi using their traditional methodsbecause of very high discharge in Koshi during themonsoon season.

Local people perception for the control of flood

Almost all of the respondents believe that the effectof flood disaster can be reduced. While asked for themethods to minimize the negative effect of flooddisaster they think that the following methods couldminimize the extent of flooding:

• Spur and Gabion Wall construction at various riskprone areas

• The problem is due to Koshi Barrage. So, the IndianGovernment should provide appropriatedcompensation to the people and develop plans toincrease the benefits from the Irrigation Canal

• Only government can do, we can not attempt tocontrol Koshi

• Should block Koshi from both the sides (river rightand left) and allow it to flow at one specificdirection

• Should build gabion walls at both sides

• Should stop to stone and sand dredging practice

145

• Should build notch at various points where theriver meanders

• Should distribute the water to 3�4 channels alongthe flow

• Build another Canal from Chatara towards east

• Should conserve forest along the upland areas

• Open all the gates (52) of the Koshi Barrageduring heavy rainfall. This will result in reductionof the water level and the damage will beobviously reduced.

Conclusion and discussion

Flood disaster is one of the major disasters thatoccur every year particularly in the southern part ofNepal. The flood hazard study of the Koshi area,focused particularly around the Prakashpur, Rajabas,Bayarban, Chatara, Madhuban, Chakkraghatti, isperformed with a view of risk identification anddisaster preparedness for the future events. Thestudy utilized an integrated approach of GIS, remotesensing and socio�economic data analysis as well ashydrological and meteorological data analysis. Adetailed field study was also carried to verifyinformation obtained from the RS imagery and GIS.

The Koshi River basin corresponds to a fluvialerosional system dominated powerful river which isconsidered to be its early mature stage. Active riverbank erosion can also be seen mostly on the easternbank of the river. The river has been migratinglaterally from west to east direction eroding theeastern bank and making it more vulnerable.

Flooding in Koshi area can be attributed mainly tothe prolonged rainfall over an extensive catchmentarea that generates high volumes of run�off, whichspill out onto the river’s natural flood plains andovertops the cultivated land and settlements. Suchoverflowed water inundates large areas for weeks at a time owing to the long response time of thecatchment, and subsequent slow rise and fall of theflood hydrograph. The bottleneck appearance due tothe construction of large dam, Koshi barrage down�stream near the Nepal India border, has furtherfacilitated the water to move laterally and inundat�ing the fertile lands. The excessive sedimentationnear the Koshi barrage is also considered to be thesecondary cause for the lateral migration of waterovertopping the fertile lands. The local people alsoblame the Koshi barrage for flooding their area,claiming the flooding was not as intense or frequentbefore the construction of the Koshi barrage.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Flooding in the study area has also been from theflash floods resulting from intense rainfall of arelatively short duration, which is also a commonphenomenon in the hills and mountainous part of the area that has steep slope. The maximum annualprecipitation during the monsoon season is 93% ofthe total precipitation and 37% of the total occur ina single event within 24 hours which generates avery high discharge rate and ultimately flooding theadjacent areas to the river bank. Eastern bank of thearea is often impacted by the flash floods creatingloss of land and property.

The probable maximum flood calculated usingCalifornia’s method for the recurrence interval of 10,100, 1000 and 10000 years are 10 cumec, 100cumec, 15,900 cumec, 21,100 cumec and 26,300cumec respectively. Similarly the probable maximumrainfall for the recurrence interval of 10, 100, 1000,10000 years are 2650 mm, 3500 mm, 4350 mm and5200 mm respectively.

Participatory GIS method together with combinationof remote sensing and field survey data interpreta�tions were used for the flood hazard mapping of thearea. Regional information about the area such asmapping water bodies and flood plain were accom�plished using satellite imagery and aerial photograph.The study showed that the area could be divided intothree flood probable zone, low, moderate and highhazard zone. The high hazard zone shows therecurrence interval of 4� 7 years although inundationof land and soil erosion is frequent in every monsoonseason and 10�12 years in moderate hazard area.Mostly the flood vulnerable area is located on theeastern bank of the river, from Chatara to Bhantabariarea. The high hazard area encompasses 16.21 sq.km and people from six different settlements areliving in this area despite the threat from flooding.Similarly, the moderate and low hazard area occupies12.15 sq.km and 32.11 sq.km respectively. The mostvulnerable area is in the Prakashpur VDC, around theRajabas and Madhuvan. The people in Rajabas havesuffered already a large loss of life and property in2037 B.S. The infrastructures such as dikes and spurconstructed to protect from flood and rive bankerosion have also been degraded from the sedimen�tation of the river. Flash floods are mostly seenaround the Bhantabari, Bhardaha and HanumanNagar area that inundates the land from few hoursto 5�7 days. This has degraded the fertility of thecultivated land and even have made crop plantationvulnerable during monsoon seasons.

The flood has made many people homeless in thepast fifty years even with loss of life and property.This has also done a significant loss of the land andhas made major changes in land ownership. Thoughthe loss of land has not taken place at once in mostof the cases it has resulted into loss of land andultimately the livelihood of the people of that area.Due to the loss of land the people are compelled todo labor work instead of farming at nearby sub�urban areas.

Records showed that the flood victims are mostlypoor and homeless people. Interview with peoplesdepicted that these peoples are aware of the floodbut lack of sufficient capital to migrate to safe placehave forced these peoples to live in the risk proneareas. Some peoples are also socially attached withthe areas and do not want to leave their traditionalhomes. Also, the existence of fertile land next to theriverbank has made people to live in the nearbyareas despite the potential threat to the floodingwhich makes these people more vulnerable duringflooding.

The flood hazard mapping done from this study willbe very much fruitful for proper planning andresource location for disaster preparedness for theconcerned line agencies and organization. Theresulting flood hazard map, as provided to the VDCwill also be very helpful for the local people todistinguish the safe and vulnerable areas from flood.The map will be helpful even for the uneducatedpeople to understand about the existing flood andland condition, as visual display is the easiest andexpressive ways for the people to comprehend andunderstand their existing conditions from theflooding.

The involvement of the local people for themodification and changes in the flood hazard mapafter its completion made the villagers to think thework as their own work and for their own benefitand has raised a feeling of ownership in them. Theirwillingness and readiness to form a flood fightingcommittee substantiates their positive feelingtowards the project. Also, the commitment of theVDC committee to incorporate the flood studyrecommendation in their periodic plan shows theirreadiness for the flood disaster reduction.

146

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Lessons Learned

1) Natural disaster has a significant impact inpoverty and has always hindered economy of adeveloping country such as Nepal. There is asignificant role between the poverty alleviationand the disaster risk reduction. Frequentlyoccurring flood disaster is one of the maindrawbacks for the poverty in the study area.Whenever the disaster occurs, the poor andhomeless people are affected the most. Thefrequent flooding has caused a significantamount of economic loss to these peoplesburdening them with more financial debts forreforms and seed money to restart theagriculture.

2) Flood victims are mostly poor and homelesspeople. Interview with peoples depicted thatthese peoples are aware of the flood but lack ofsufficient capital to migrate to safe place haveforced these peoples to live in the risk proneareas.

3) Greed to exploit the river flood plain properties,mostly seen around the fertile plains of the river,is one of the major factor as seen during manycasualties. Some people with moderate incomewere seen to be living in the flood risk proneareas. Although these people have sufficientfund to migrate to safer places they still insist onliving there because of the greed to occupy theland with good yield.

4) Existence of fertile land next to the river bankhas made people to live in the nearby areasdespite the potential threat to the flooding.

147

5) A proper planning for any program aimed atdisaster risk reduction is successful only if it isdone through active participation of the localcommunity people, as it awares the people andcreates a feeling of ownership in them. Also, it isimportant during the implementation of theproject to involve these local people in theproject and respect their thoughts.

6) Visual displays are the most effective means forinformation dissemination. Both the educatedand uneducated people understand it. In ourcontext, the flood hazard map was very informa�tive and people from all the age showed intereston it which would be very helpful for thesepeople to understand their local environmentand threats from flooding. The different colorrated for low, moderate and high hazard, forrivers and landuse types made people clearabout their surrounding area in a larger scale,which they have not done before.

7) Initial assessment flood hazard and riskidentification is very important to lessen theeffects of flooding. Initial assessment of theflood hazard and the flood hazard map was veryuseful for the proper planning of flood riskreduction and mitigation. The study highlightedthe risk prone areas from flooding. It has servedas a guide map for the policymakers and stakeholders working in the study area to implementany program targeted for the lessening theeffects of flooding.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

References

Chheetri and Bhattarai D., (2001), Mitigation andManagement of Floods in Nepal, Kathmandu, HMG�MOHA

Disaster Annual Report (1995), Water Induced DisasterPrevention Technical Centre (DPTC)

Final Annual Report (2002), Nepal Red Cross Society (NRCS)

IDNR, (1995). Disaster: Threat to Social Development, WorldSummit for Social Development, UN, Copenhagen, Denmark,6�12, March, 1995

Lammerink M.P., Bury P, Bolt E., (1999) An Introduction toParticipatory Action Development (PAD), pla notes Bulletinvol.35

Panday, R.K., 1995. Development Disorders in the HimalayanHeights, Challenges and Strategies for Environment and

Development. Panday’s publication, Kathmandu,NepalPokhrel, Lekhnath Nath (2002), Water Induced Disasters inNepal and Its Management, Disaster Review Annual Report2002, DWIDP

Poudal K. P. (2001), Mountain Environment and GIS Modelsand Application

Pradhan, A., (2000), Sedimentological and Environmentalinvestigation of the Koshi Alluvial Fan, Unpublished Mastersthesis, Central Department of Geology, Kirtipur

Regmi, A., (2000), Hydrological study of the Koshi Alluvial Fan,Unpublished Masters thesis, Central Department of Geology,Kirtipur

Victoria, L.P.(2002) Community Based Disaster Managementin the Philippines: Making a Difference in People’s Lives

148

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

149

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Local and popular folklore and

culture on hazard and

vulnerability meets geographical

information system for the risk

reduction preparedness of the

people of the Barrio San Antonio

of Naiguatá, Estado Vargas

TEAM LEADER

Gabriela I. Muñoz Rodríguez <[email protected]>

PROJECT ADVISOR

Diana C. Valera

Venezuela

SYMPOS IUM NOTES

The inteview model and interview matrix for this project weremade by the team. Project photos were taken by the team.

150

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

This Project is the offshoot of the team’s participa�tion in the community preparation program for riskmanagement which took place from 2000 to 2002 inmarginal urban areas throughout Venezuela, andmore specifically within the framework of the Projectbegun in 2000 with Mercy Corps International, whichwas subsequently continued, in 2002, with thesupport of the Corporation for the Recovery andDevelopment of Vargas State (CORPOVARGAS), aswell as with the Fire Fighters and the Civil ProtectionInstitute for Vargas State.

From our knowledge of the region and revision ofavailable documentation our conclusion is that inVargas State as well as in the rest of the country riskmanagement has been limited to attending to“unforeseen” emergencies and disasters, withoutthe adequate use of tools to enable the developmentof local risk management policies, or that emphasizeprevention, mitigation, preparation and alertregarding probable adverse events of a natural man�made origin. On the other hand, the risk handlingmodels that have been implemented to date byspecialized organizations lack familiarity with thelocal culture and history (cultural experience of risk)and with local response systems to adverse eventsthat could be used to confront this type of event.

All the preceding has made it difficult to approachand establish an efficient relationship among theFirst Response Organizations (OPRs) and affectedcommunities, whose knowledge and experience have

151

been underestimated and underrated. Consequently,we consider it necessary to work to increaseintegration among the communities and organiza�tions responsible for preparing the community tohandle risks, in order to generate effective disasterprevention and mitigation strategies; participationthat not only includes their coordination with respectto response strategies, but, perhaps, will influencethe contents or certain programs and data sources.

The SIG is a practical tool for integrating this knowl�edge which will allow OPRs to analyze and easilypresent information on maps that is necessary forgreater integration of scientific know�how and localknowledge, to thus achieve effective risk manage�ment whether by rescuing useful knowledge orproviding data of interest to help understand andperhaps predict behavior in the community.

Goal

It is for the preceding reasons that the aim of thisstudy was to compile, in the Geographic InformationSystem (hereinafter SIG), the specialized technicalknow�how along with the wealth of local knowledgethat exists among the inhabitants of Naiquatá, inorder to establish two�way communications amongthe actors on both ends of Risk management(Experts – Community) so as to generate moreefficient strategies and practices to prevent andreduce disasters.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Town Naiguatá before the torrential flow of 1999 Town Naiguatá after the torrential flow of 1999

From: Misión 030602 / Year: 1999�2000 / Scale: 1:5000 / Instituto Geográfico de Venezuela Simón Bolívar

Geographic location

In the original version of the Project presented toProvention Consortium the idea was to workexclusively with the Barrio San Antonio in Naiguatá,Vargas State, Venezuela.

However, in subsequent meetings with the firefighters from the Naiguatá station the idea was born to expand geographic coverage to Naiguatá,because of the increased usefulness of SIG in theireveryday work and in community preparation in thearea. This recommendation seemed reasonable to usbecause the Barrio San Antonio is functionally andculturally integrated with the rest of the town ofNaiguatá, so that any information used will also beapplied to this sector as well as to other aspects. Forexample, with respect to the complex issue of riskmanagement, the inhabitants of this “barrio” have ashared risk experience with that of the town.

Naiguata parish it is appraised that the zones of the coast constituted in his majority by cones ofdejection and accumulated materials recently, what itmakes prone to has a long history of disastersgenerated by diverse natural phenomena, and inDecember 1999 it was affected in an unequalmanner, being among the sectors that suffered themost severe consequences (mudslides and floods inthose areas closes to river banks), while others werealmost unaffected. The project is interested inevaluating such dissimilar occurrences in space. Onthe other hand, we presume that there may be inter�sector diversity in one and the same community withrespect to oral tradition regarding the issue of riskand disasters, differences that could eventually betranslated into different ways of confronting suchcontingencies, a vital topic for OPRs.

Beneficed population

There are five (5) direct participants who aremembers of First Response Organizations (OPR):

• Two (2 ) from the Naiquatá Fire Fighters.

• Two (2) from the Community Preparation Unit forVargas State.

• One (1) from Naiguatá Civil Protection.

This group of fire fighters was selected because of itsinterest in the project and because, since 2000, it hasparticipated in a series of workshop within theframework of the Corpovargas project carried out bySOCSAL on community preparedness and riskmanagement. The understanding is that thesepersons will then act as “multiplying agents” wherethey work, as well as among other entities andinterested parties.

According to the National Census carried out in 1990by the National Institute of Statistics (Spanishacronym INE), there are 9446 habitants in the townof Naiguatá, which would be the indirectparticipating population.

Theoretical-methodologicalframework

The project’s theoretical�methodological frameworkwas adapted for which we basically used certainworks with an ethno�methodological approach and,above all, two working documents generated forSOCSAL by the anthropologist, Carmen Luisa Ferris,titled “Culture and risk: common sense knowledgeof risk in daily life. The Naiguatá Case,” and“[Abstract] SOCSAL�PROVENTION CONSORTIUMNaiquatá Study”.

152

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Bolivarian School after the 1999 torrential flow House affected by the 1999 torrential flow

This is based on an “emic” approach, characteristicof ethno�methodology which takes into account thepoint of view –whether or not traditional� of theinhabitants or their interpretations and explanationsof characteristics, genesis or recurrence, forecastingpossibilities, and behavior considered to be the mostappropriate to follow with respect to risks. Thispersonal point of view, frequently traditional or“folkloric” is called “Common Sense Knowledge”,and may have its own internal coherent complexityand profitability, and may or may not coincide withconventional “Technical�Scientific” knowledge, thatis, with the external “ethical” point of view imposedby specialists with academic influence or formation.

The starting point is the idea that OPRs’ familiaritywith local common sense knowledge will enablethem to better understand motivations and forms ofresponse among the communities with which theyinteract, and gives them the aptitude to converse andeven negotiate in their own terms, as well as to—assuggested by certain studies—obtain valid, empiricalknowledge that is useful and exact to complementthe technical�scientific information that can, there�fore, be used in local risk management.

While this approach has been used for other types of research (for example, that related with health orecological issues), it is relatively new in the countrywith respect to risk management, although analog�ous or similar approximations have been documentedabroad, from the sociological, anthropological orsocial psychology viewpoint when initially docu�menting this project. Even less common seems to be the link between this type of data of a qualitativenature and the use of GIS technology.

From the methodological point of view, CommonSense knowledge can be classified into three types:sensory (related with the capacity and ability topredict or perceive threats), the practical (routine,regular, frequently traditional behavior when

153

confronting threats), and the one that includesformulas (linguistic acts and genre, frequentlyritualized or that remit to ritual components, towhich the power to influence the development ofthreats is attributed).

These variations of Common Sense knowledge alsoremit us to the concepts of culture, tradition, values,belief system, standards and motivation and areperhaps understandable as regards their possibleorigins in view of the relatively isolated and Creolenature of most of the inhabitants of Naiquatá, aswell as the repeated exposure of this population todifferent types of threats.

In practical terms, each of these types of CommonSense knowledge, analogously to Scientific�technicalknowledge, may be divided into topics thatcorrespond to specific layers of information in theSIG and then related to variables as such or to the“Scientific Layers” that can also be visuallyexpressed in maps or in the form of summaryreports.

Given the nature of the data, beginning with thesecond phase of the Project, the idea is to adopt aqualitative, methodological viewpoint, with markedlyethnographic influence based one method:

Contact and research interviews applied to a non�random sample of inhabitants from different walksof life and genders, focusing on older inhabitantswho have lived for many years in the town and whowork in areas that presuppose the handling ofimportant volumes of locally accumulated, empiricalknowledge, since it is assumed – but will have to beverified – that this type of knowledge responds totraditions, inter�generational lines of transmission, orperhaps to the memory�appropriation of knowledgederived from having confronted such events before.Included also were local leaders, educators andmembers of religious fraternities (see Annex 1).

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Workshop about the GIS ofNaiguatá

Interview with an older member of the cofradia of the Devils ofNaiguata and an older member of the community

Data was systematized in a matrix made by the team (see below), which includedthe selected variables. The items that were more repeated were drawn in the map.

GIS description

As was mentioned previously, the productof this study will be a SIG in which it ishoped to combine, in addition to thespecialized technical knowledge of risk, thewealth of local knowledge of the inhabi�tants of Naiquatá, in order to facilitate thedesign of more efficient strategies andpractices for prevention and mitigation.

The work is being done with Mapinfo ver. 6.5software because its functions allow quick andsimple use of maps that can, in turn, be used to prepare reports and illustrations that bring to light the true meaning of the informationunderlying the lines and columns of data tables,contributing efficiency in decision making.

Design of the GIS

GIS will cover five (5 ) thematic layers describedbelow, and efforts are being made to ensurethat its design is as simple as possible in orderto optimize ease of handling of the product.

154

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Interview matrix

Tema 1: Conceptos asociados a riesgo

Tema 2: Memoria histórica de desastres o eventos adversos

Tema 3: Susceptibilidad flujos torrenciales

Tema 4: Susceptibilidad deslizamiento

Tema 5: Susceptibilidad sísmica

Entrevistado(ENT)

Riesgo Amenaza Peligro VulnerabilidadDesastreNatural

AmenazaNatural

Amenazas enla Parroquia

Amenazas enNaiguata

Ent Tipo de desastre Natural (P ò CP) Dónde Cuándo Víctimas Infraestructura Observaciones

1

Ent Inundación DeslavesFlujostorrenciales

Ubicación Frecuencia Origen/ causaGrado deseveridad

Otra característicaresaltante

1

Ent Derrumbe Deslizamiento Ubicación Frecuencia Origen/causaGrado deseveridad

Otra característica resaltante

1

Ent Sismo Temblor Terremoto Ubicación Frecuencia Origen/causaGrado deseveridad

Otra característicaresaltante

1

Another interviewwith an elderly manfrom the community

Characterization

Contains fundamental aspects of the study area inorder to have a clear, updated and actual view of the zone that can be handled by First ResponseOrganizations (FRO) when preparing community riskmanagement.

Based on statistical data from 1881 to date; is sub�divided into topics of health, education, housing anddemographic data.

Technical–Scientific

Contains exact, conventional, technical�scientific dataregarding natural threats present in the study zone.

Is a guide for community work, when preparingemergency prevention and mitigation plans.

At parish level 3, natural threats are beingconsidered:• Susceptibility to torrential floods • Massive movements • Seismicity

These three (3),are based on:• Spatial location• Degree of severity• Origin

Natural threat from torrential floods is beingconsidered in the town, based on:• Spatial location• Degree of severity• Probability of occurrence• Affected surface

Historical events

This refers to all the adverse events that haveoccurred in the Naiguatá from 1900 to 1999, whichwill enable the FRO to have a historical record of the threats that have affected the zone, which isfundamental for working with the community.

OPR resources

This refers to both technical data of the FRO as wellas to the behavior during and after an event in thetown of Naiguatá, in order to facilitate the work ofthe FRO in different areas. To date, the only fields in

155

which these are available are current use, restrictionsin use, and location of life lines.

Common sense

Refers to local knowledge among the people ofNaiquatá, regarding technical�scientific information,either their own or traditional, adverse events thathave occurred in the area, and certain aspectsindicated in the layer known as FRO Resources.

GIS applications

• In the Naiguatá parish,

Determine: location, degree of severity andperiod when the three natural threats occurred:torrential flows, massive movements, and tremors, based on information from the Simón Bolíva Venezuelan Geographic Institute(our “experts” who, in this case, synthesizetechnical�scientific data)

Determine: location, degree of severity andperiod when the three natural threats occurred:torrential flows, massive movements, and tremors(this time based on opinions of the communityand local residents familiar with these issues).

• In the town,

Identifies shelter zones, vital life lines and userestrictions for soil according to experts and thecommunity.

Know the location of basic, public institutions(education, health), as well as other entities,stressing the coverage capacity of the differentinstitutions or entities (in the town).

Determine location, severity and period ofoccurrence with respect to the natural threat of torrential flows, as per the Simón BolívarVenezuelan Geographic Institute.

Determine location, severity and period ofoccurrence with respect to the natural threat of torrential flows, as per community opinion.

• Compare the opinion of community expertsregarding the previously mentioned threats on different levels.

• Become acquainted with the historical record on adverse events that have occurred in the parish (from 1900 to 1999); type of event,places affected, date occurred, number of victims and infrastructures affected.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Overall balance of project programexecution

• Revision and analysis of similar GIS experiences.

• Description of the most salient physical�naturaland social aspects of the study area.

• Preparation of a theoretical�conceptual frameworkof common sense and risk.

• Design of the GIS.

• Implementation of a sensitivity workshop for OPRs:the purpose of the meeting was to outline Projectobjectives and receive feedback from theseorganizations in the form of information neededfor interviews.

• Interviewing the different players from thecommunity.

• Establishment of agreements with the FROs aboutthe jeans of collecting information in the field andover time.

• Awareness building workshops at the ColegioDiego Osorio, regarding the Project and the needto manage risk on the school level.

• Request for donations of computers from differentinstitutions.

• Preparation of a GIS user’s manual.

• December 15, 2003, was the 4th anniversary ofthe tragic mudslides in the Vargas State whichoccurred in 1999. A commemorative activity wasprepared with the Naiguatá Parish Board andfirefighters from the Vargas State which included:presentation to the community of the results of theProject, simulation by the Naiguatá firemen of anevacuation due to fire in the Colegio Diego Osorio,and preparation by the students of a risk map ofthe Pueblo Arriba sector of Naiguatá.

Results of the Project

• An easy to handle SIG was designed that isadapted to the needs of the community and of the first response organizations.

• Full commitment was obtained from three of theinstitutions in the participating areas—Henry Iriarte,Alejandra Da Silva and Felipe Quintero—tocontinue with the Project, to update existingthematic layers and add new information, and totransmit this information to the area’s schoolsand the community as a whole.

• The two directors of the San Rafael and DiegoOsorio schools are impressed by the Project andwant to have risk management workshops in theseschools.

• Both students and teachers have been motivatedto do research similar to that involved in thisProject.

• Other first response institutions, such as trafficcontrol have shown an interest in having access tothe GIS and have proposed adding informationfrom the Naiguatá Parish, such as number ofvictims of traffic accidents.

• The Naiguatá firefighters have seen the SIG inaction and how easily it can be used to solve theirtechnical problems, particularly everythingpertaining to permits, zoning and restrictionsregarding the use of space.

• Two computers will be donated in January 2004.One will be located in the Naiguatá station andthe other in the central station in the office ofcommunity preparation. The donation comes fromthe TOTAL oil company through thenongovernmental organization V ía Tecnológica.

• Projects in schools need to be emphasized as oneof the most important achievements because thisis a group that is motivated to receive informationand it can transmit that information not only to itspeers, but also to family members.

Lessons Learned

• Common sense with respect to adverse eventsdoes not tie in actions as a set of systematicallylinked opinions thereby allowing them to beprecisely explained. They acquire validity byresisting ongoing comparison with experience andtherefore can be modified at any given moment.

• Lack of systematic explanations for adverse eventsis a condition which adds simplicity to informationto be transmitted.

• Because natural events study area do not occurfrequently, people rely on the same knowledgethat has been transmitted from generation togeneration, and this knowledge acquired by theirancestors has not been modified.

• The inhabitants have been accumulatingobservations about and knowledge of theirgeographic space, which has allowed them tohandle themselves and/or to survive in that space.

156

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

This knowledge is the foundation forrecommendations or actions that the community,either as a whole or individually, has beenadopting with respect to the environment.

• We believe two basic conditions are needed inorder to generate effective strategies forpreventing or reducing natural disasters:

– A constant exchange of information among thepeople (among different generations) regardingadverse events that have occurred in the area, alocal early warning system and local responsesystems, in other words, a cultural lifeexperience of risk.

– Greater integration among communities andorganizations responsibly for preparing thecommunity in risk management, because if thespecialized organization have knowledge of thecultural life experience of risk they will be betterable to understand response motivations andforms among the communities with which theyinteract. That is, they will have a better attitudewhen it comes to conversing and evennegotiating with communities in the sameterms. When we speak of “Terms,” we arereferring to the vocabulary to be used, the siteswhere meetings are to be held, length oftraining, among others.

• The community’s threat priorities need to beidentified. For example, in the case of Naiguatá,many of its inhabitants (those interviewed) aremore concerned with the day�to�day threats ofcrime, unemployment, drugs, among others, thanby natural threats, such as earthquakes, torrentialfloods and landslides. In the case of earthquakes,they believe these cannot be predicted and theymust necessarily suffer the consequences. In thecase of torrential floods (mudslides) the inhabi�tants believe these will occur every 50 years, andbecause the last one occurred in 1999 they don’tconsider this to be a problem and feel that futuregenerations and not the current generation shouldworry about them. The consequences of cave�insor landslides haven’t been all that serious in thearea and, therefore, the inhabitants do not assignto them the required importance. Consequently, amethodology is needed to prepare the communityto face threats of Nature by using examples fromdaily life.

157

Links between poverty and disasterrisk reduction in developingcountries

The GIS is a planning tool this Project transversespoverty and disaster risk reduction programs, notonly by scientifically identifying areas not suitable forhabitation or construction, but also it gives institu�tions in charge of community preparedness in thearea a tool to become acquainted with the beliefsand cultural experiences of risk in different areas.

With precise knowledge of the natural and anthropicrisks in an area, not only from the scientific stand�point but also from that of the community itself, itwill be possible to carry out projects that are bothmore viable and sustainable over time. Therefore, thisProject helps to plan the investment to be made inan area that is more in keeping with that area’sgeographic situation and the vision of thecommunity with respect to all its risks. This will helpto reduce poverty levels in underdeveloped countries.

The proposal was planned to obtain, more than aGeographic Information System, a tool so that anyorganized group for preparing community riskmanagement can collect basic information not onlyin Venezuela, but in any other country by making theappropriate adaptations.

Visualization of the map makes it dynamic and thecommunity feels the need to participate in ensuringthe sustainability of the project.

The proposal will have continuity in time because theVargas State firefighters have seen that there is aneed to continue to improve and perfect theGIS–Naiguatá. This group is in the process ofobtaining financing from State Government andother organizations such as Corpovargas.

On our part, because the GIS is a useful tool for theFRO we plan to replicate this experience in twocommunities in two other states (Monagas andAnzoategui) that have been affected by differentthreats of anthropic origin.

Moreover, the proposal will have continuity in timebecause it is linked to the “Promotion of RiskManagement in Vargas State Parishes” project thatSOCSAL has been developing since 2001 withfinancial support initially from Mercy Corps

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

I Identification of the interview

II Identificatin of the person interviewed

International, and currently from CORPOVARGAS(Corporación para el Desarrollo del Estado Vargas–Corporation for the Development of Vargas State) inalliance with the firefighters from Vargas State andthe Instituto de Protección Civil y Administración deDesastres de Vargas (Vargas State Institute for CivilProtection and Disaster Management).

Since the product of the research is a geographicinformation system it can be applied by institutionswhose mission is community preparedness and riskmanagement to any population located in marginal�urban areas of developing countries, because all thematerial for collecting information is in simple andpractical language thereby facilitating its use andunderstanding by persons without expertise in thesocial sciences.

This is explicitly geared to key informants in thecommunities, both men and women and includes:

• Natural community leaders

• Social actors with different occupations (hunters,fishermen, farmers, masons, healers, fire fighters,members of communities, drivers, teachers andother professionals)

• Community promoters (inhabitants of the sectorsselected in the town of Naiguatá)

158

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

The components of learning in your project

Identification of areas of threat by specializedinstitutions and the community in general.

Awareness building and training of OPRs in handlingthe GIS and its advantages when it comes toplanning.

Combination of two types of knowledge, scientificand common sense (formal and informal)

We would like to thank some people who help to finish thisproject, as: Carmen Ferris, Virginia Jimenez, Freddy Colina,Pedro Rivas, Miguel Ángel Ortega and all the team of SOCSALand Fire Departament of Vargas.

Especially, but not exclusively mature people, orthose with accumulated experience who –it isconsidered� could contribute useful information toprepare a description of the community, experiencesin local organization and attention to risks, datafrom their own traditional wealth of knowledgerelative to the topic of risks, which must be takeninto account when planning social development,prevention and/or response programs.

ANNEX 1 — INTERVIEW MODEL

Date: Time: Duration:

Name of interviewer:

Context/circumstances/environment of the interview:

Name:

Place of birth: Sex: Age:

Date from which theindividual has residedin this locality

Places in Naiguatáwhere the personhas lived

Sector where theperson currently livesin Naiquatá:

Address:

TEL: How person can be reached:

Religion: Occupation:

Highest gradecompleted:

Role in thecommunity:

III TOPIC 1: Concepts associated with risk

3.1 What is your understanding of the word“risk”?

3.2 What is your understanding of the word“threat”?

3.3 What is your understanding of the word“danger”?

3.4 What is your understanding of the word“vulnerability”?

3.5 What is your understanding of the term“natural disaster”?

3.6 What is your understanding of the word“natural threat”?

3.7 Enumerate all the threats present inNaiquatá parish.

3.8 Enumerate all the threats present in thetown of Naiquatá

IV TOPIC 2: Historical recollection ofdisasters or adverse events:

4.1 Has Naiguatá parish experienced any typeof natural disaster?

4.2 What type of natural disasters haveoccurred?

4.3 Where? When?

4.5 Number of victims and infrastructureaffected?

V TOPIC 3: Community’s perception ofnatural threats. (COMMON SENSE)

IN THE PARISH:

Seismic susceptibility:

5.1 What is the difference between a “tremor”and an “earthquake”?

5.2 Generally speaking, would you say thatthere are sectors in Naiquatá parish wheretremors occur more often? Yes? Forexample, in which sectors? (LOCATE ONMAP)

5.3 How often would you say a tremor occurs?Do they always occur with the sameintensity or are there sectors of the parishwhere they are felt more strongly?

5.4 Have you heard of any thing or factor thatmight contribute to or promote tremors?What? Is this something new or havepeople remarked about this before? F

159

or example, what have you heard about thisamong people from other eras?

5.5 Whenever there is a tremor or earthquake,have you noticed that some homes orstructures such as walls, bridges etc., aremore or less resistant to these movements?What have you seen or do you know aboutthis? In your opinion which materials are“good” and withstand tremors and whichare “worse” and do not? Sometimes itseems that the “people from before” knewvery well how to build because there arevery old buildings that are still standing,despite the passage of centuries. What dothe “older” people say about this? Whatmaterials did they recommend as being thebest to withstand tremors? And what typeof construction “tricks” enabled them toresist more, or at least protect people frominjuries or damage?

5.6 What types of signs do you recognize asoccurring before a tremor? What have youheard about people who have a “kind ofpremonition” that something will happen?For example, in town, who is famousbecause he or she knew or felt that atremor or some other threatening eventwas going to occur and it did? What didthat person feel? What else can you add?

Susceptibility to torrential flows:

5.7 What difference is there between a“flood,” a “mudslide,” a “torrentialflow” and a “torrential avalanche”?

5.8 Generally speaking, would you say there aresectors of Naiquatá parish where mudslidesare more frequent, yes? For example, inwhat sectors? (LOCATE ON MAP).

5.9 How often would you say a mudslide occur?Do they always occur with the sameintensity, or are there sectors of the parishwhere they are stronger?

5.10 Have you heard of anything or of any factorthat contributes to or promotes mudslides?What? And are people just saying this nowor did others say this before? For example,what have you heard about this from thepeople from other times?

5.11 When a mudslide occurs are some housesor structures such as walls, bridges etc.,more or less resistant? What have youseen or do you know about this? In your

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

opinion, what materials are “good”and resistant and what materials are“bad” and not resistant? Sometimesit seems that the “people from before”knew very well how to build becausethere are very old buildings that are stillstanding, despite the passage ofcenturies. What do the “older” peoplesay about this? What materials do theyrecommend as being more resistant tomudslides? And what type ofconstruction “tricks” enabled them toresist more, or at least protect peoplefrom injuries or damage?

5.12 What types of signs do you recognizeas occurring before a mudslide? Whathave you heard about people who havea “kind of premonition” thatsomething will happen? For example, intown, who is famous because he or sheknew or felt that a mudslide or someother threatening event was going tooccur and it did? What did that personfeel? What else can you add?

5.13 Knowledgeable people can sometimes“gauge the intensity of rain”, whetherthere will be a lot or not, or if it willlast a long time or not, but looking atthe changes in the environment, as forexample in the “plants” either here intown or on the mountain. They caneven do this from a distance. Whattypes of “signs” in vegetation haveyou heard talk about in this area? (Ifthe person says he/she doesn’trecognize any signs…) Nothing? Doesthe mountain always look the same orhave people talked about it lookingdifferent? What kind of differences arewe speaking of? And nothing can benoticed not even in the type of“plants” or “branches” or sediment,earth that comes down the river? Forexample, if it rains a lot and the“plants” are uprooted or loosened,what is it that you most often see“come down”?

5.14 And what about the animals? Do theybehave the same way? No? Forexample, the insects and things likethat? Are any of these indications ofrain or “things” associated with rain?Yes? Of what? (If the person being

interviewed is giving short answers, trysome other type of animal …)

5.15 Farmers often can distinguish betweentypes of storms and rain and they evengive them different names. Bearing inmind that Naiquatá is still “verytraditional,” what type of names orclasses of rains have you heard about?How do they differ? Are there specifictypes of rains that are more closelyrelated with landslides or “mudslides,”for example?

Susceptibility to massive movements:

5.16 What difference is there between a“rock fall” and a “rock slide”?

5.17 Generally speaking, would you saythere are sectors of Naiquatá parishwhere rock slides are more frequent?Yes? Where, for example? (INDICATEON MAP)

5.18 Every how often would you say that arock slide occurs? Do they always occurwith the same intensity or are theresectors of the parish where these arestronger?

5.19 And can it happen “just like this, atany time,” or does it seem to you thatthis occurrence “seems to repeatitself”? And does it happen “everyonce in a while” or is it “random”? Ifit happens “every once in a while,” atwhat intervals is it repeated?

5.20 Have you heard of anything or of anyfactor that contributes to or promotesrock slides? What? And are people justsaying this now or did others say thisbefore? For example, what have youheard about this from the people fromother times?

5.21 When a rock slide occurs are somehouses or structures such as walls,bridges etc., more or less resistant?What have you seen or do you knowabout this? In your opinion, whatmaterials are “good” and resistantand what materials are “bad” and notresistant? Sometimes it seems that the“people from before” knew very wellhow to build because there are very oldbuildings that are still standing, despitethe passage of centuries. What do the

160

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

“older” people say about this? Whatmaterials do they recommend as beingmore resistant to mudslides? And whattype of construction “tricks” enabledthem to resist more, or at least protectpeople from injuries or damage?

5.22 What types of signs do you recognize asoccurring before a rock slide? What haveyou heard about people who have a“kind of premonition” that somethingwill happen? For example, in town, whois famous because he or she knew or feltthat a rock slide or some other threaten�ing event was going to occur and it did?What did that person feel? Who else?

IN THE TOWN:

Torrential flows:

5.23 In your opinion, are there sectors of thetown of Naiquatá where mudslides aremore frequent? Yes? For example,where? (LOCATE ON MAP)

5.24 What produces these mudslides? Whathave you heard about this in yourcommunity? Has this always been thecase? What did the people from othereras say about this? And do the peopleof today continue to agree?

5.25 In your opinion, every how often does amudslide occur? Do they always occurwith the same intensity or are theresectors of town where they are moreforceful?

5.26 What type of signs do you recognize asoccurring before a rock slide? What haveyou heard about people who have a“kind of premonition” that somethingwill happen? For example, in town, whois famous because he or she knew or feltthat a rock slide or some other threaten�ing event was going to occur and it did?What did that person feel? Who else?

5.27 As to prevention, what have you heardwas done “in the old days”? Whatmaterials did they recommend as beingmost resistant to mudslides? And whatkind of construction “tricks” allowedthem to resist or at least to keep thewater from injuring people and causingdamage?

161

VI TOPIC 3: Useful information forOPRs:

6.1. During the most recent tragedy inVargas, in ‘99, were you in Naiquatá(the town)? What did you do? Wheredid you take shelter? How did you findout about what was happening? Andthe people? What did they do? Wheredid they take shelter (Indicate on map).How did the community organizeitself? How did the people act toprevent or reduce damage, loss or lifeand materials?

6.2 With respect to other types of threatsin the town of Naiquatá, how dopeople find out what is happening?What do they do? Where do they takeshelter? (Indicate on map). How dothey organize themselves? What dothey do to prevent or reduce damage,loss or life and materials?

6.3 At this time, do you consider that theNaiquatá community (the town) isprepared to confront a disaster similarto that of ‘99, or a similar typedisaster? Yes? Why? Are thereinformation systems of early warningsystems in the zone? Are thereinstitutions, organizations, groups inthe zone that might help in case ofdisasters?

6.4 In your opinion, are there differenceswith respect to safety in different sitesdepending on the type of threat? Yes?For example…What have you heardabout them in your community?

6.5 What areas, according to thecommunity, might be best suited forliving, business and houses in the townif pertinent measures are taken?(Indicate on map).

6.6 According to the community, whatareas are best suited for living,business and houses in the town,without taking any measures? (Indicate on map)

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

6.7 You will have seen that after a disasterit is important to both cure bodilyinjuries, as well as vital to give moraland spiritual support to the people.Another way is to find strength inreligion. In fact, Naiguatá is very wellknown in Vargas state as being acommunity that respects religion...When things like this occur, what dopeople tend to do? For example, dothey pray to a specific saint? [DO NOTCONTINUE WITH THESE QUESTIONS IFTHE PERSON YOU ARE INTERVIEWINGIS AN EVANGELIST OR JEHOVAH’SWITNESS] To which one do they pray?What do they pray for, that is, againstwhat kind of threat is that saint said toprotect people? Who else? What dopeople pray for? In your family, forexample, whom do you pray to? And ifyou make it through the danger allright, how did you or do you show yourthanks? And the other saints, how didyou or how do you show yourappreciation? Sometimes, those mostrespectful are the people who belong

to religious communities. Do you or anymember of your family belong to one?Yes? To which? Both the priest orpastor, for example, or brothers andnuns tend to give a lot of help indisasters. They give, for example,spiritual help or attend to people intheir residences or churches. That iswhy it is important for you to knowwhere these places are. Where is thechurch or meeting place for thereligious communities you frequent?What other places of this type do youknow? [DO NOT CONTINUE WITHTHESE QUESTIONS IF THE PERSON YOUARE INTERVIEWING IS AN EVANGELISTOR JEHOVAH’S WITNESS] In additionto the churches, is there another placewhere people meet to pray to a saintand is there a family that “takes care”of that place? Where are these placeslocated? In places such as Los Coralesand in the Ermita (Hermitage) delCarmen, in La Guaira, during the ’99disaster, believers saw “certain signs orchanges in the saints”. Have you heardof any such thing?

162

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

163

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Communication strategies for

industrial disaster risk reduction

TEAM LEADER

Avanish Kumar <[email protected]>

PROJECT ADVISOR

Ms. Meena Raghunathan

India

SYMPOS IUM NOTES

164

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Background

Industrial disasters are no less devastating thannatural ones. In India, this became poignantly evidentduring the Bhopal Gas Tragedy in 1984. The disastertook more than 2,500 lives and left more than500000 with lifelong health effects. Following Bhopallaws and regulations have been made morestringent, but the state of industrial safety andmanagement measures on the ground is such thatthere could be disasters, major and minor, in almostevery corner of the country. And it would becommunities near the industries which wouldprobably bear the brunt of the damages. Even withinthe community, the poor are the most vulnerable andmost effected by a disaster.

While damages due to such disasters, includingdeaths, can be avoided by tightening industrial safetyregulations within industrial set�ups, anotheressential input is preparing the community nearindustrial locations for a disaster event. Even in thecase of Bhopal, if the community around had thesimple knowledge of how to avoid Methyl Isocyanate(MIC) fumes from entering the lungs, thousands oflives could have been saved. If doctors in localhospitals had known what the chemical was, andhow to handle it, things might have been better.

It is in this background that the project looked atevolving a model for community disasterpreparedness. Schools in the vicinity of an industrialestate were seen as the one of the most potentdriver for such a programme. Providing informationand capacity�building the students to takeappropriate action during an emergency situationand channelizing this information to the communitythrough them was the aim of the project.

Following activities were undertaken to fulfill theobjectives of the project:

1 – Linking with an industrial estate and schoolsin the vicinity: The project began with the task ofidentifying the industrial estate where this projectcould be carried out in way to best meet theobjectives of the project.

• Identification of industrial estate There are 11industrial estates in the vicinity of Ahmedabadpromoted by the Gujarat Industrial DevelopmentCorporation. Out of these, the Vatva industrialestate, one of the oldest and largest estate ofAhmedabad was identified for the projectactivities. The estate was suited for the project as ithas many hazardous industrial units, with human

165

settlements quite close to the estate. It is spreadover an area of more than 491 hectares, andhaving over 1800 small to medium scale industrialunits. The estate gives employment to about80,000 people. Most of these workers are settledin the vicinity of the estate. All these reasonsalong with a fairly responsive IndustriesAssociation at the estate made us choose theVatva Industrial Estate for our project activities.

The Industries Association was approached withthe outline of the project activities and the needof their cooperation. We found that although theoffice�bearers of the Association were responsive,they were skeptical whether the activities mightproject the estate and its industries as unsafe orenvironment�unfriendly. After a few round ofdiscussions to convince them about theimportance of the project, the necessarycooperation was extended by the Association.

• Linking with schools Vatva industrial estate hasmany settlements around it, mainly occupied byindustrial workers and their families. It wasenvisaged in the project that schools would be a medium to communicate to the communityabout industrial disaster preparedness. The schoolsselected for the project were the ones which werenear the estate and which had students comingfrom houses near the estate. The area has about10 schools in the periphery of 3�4 km of theindustrial estate; all ‘private’ , excluding onewhich is run by the Ahmedabad MunicipalCorporation. After a survey of these schools, finally6 schools were selected for the project. All theseschools are at an average distance of 2�3 km fromthe industrial estate. Mostly the students comingto these schools are from the housing settlementsnear the industries and many of them have theparents working in the near by industries.

2 – Survey work with industries and dataanalysis (through and with oriented teachers andstudents): This activity of the project involved severalsteps, including teacher/ student orientation, linkingwith individual industries for visits, school industrialvisits, surveys etc. The activities in detail are:

• Orientation of teachers and students The 9th Std.(one section with about 50 students) of each classwas selected to carry out the project activities. Thepurpose of selecting 9th standard students waswith the premise that they will have betterunderstanding of industry related issues than anystandard below them and also comparatively lesspressure of studies than 10th (when in India there

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

is a Board examination), and hence could take partin project activities more conveniently. Only fromthe Municipal school the 7th standard studentswere involved as that school was only till 7thstandard.

To brief and orient the students about the project,and to understand their knowledge andperceptions about industry, an orientation sessionwas conducted in each school with the students.Emphasis of this first session was to get an idea ofthe perception of students had about industries.For this purpose, two exercises were carried out inwhich they were asked to fill a brief questionnaireand to draw a picture of the industrial estate asthey see it.

Around 250 students from 6 schools participated inthis exercise. The findings of these exercise arepresented graphically below and important conclu�sions from the questionnaire are also mentioned:

• The benefits of the industries are understood by amajority of the students and they answered thatindustrial estate is important in their vicinity(53.77 %) as it gives employment to many peopleand gives us various items of daily use.

• About 30 % students answered that they haveheard about industrial accidents in the estate.Some of them had heard of gas leak relatedmishap, some of fire and others of injuries relatedto individual accidents. Thus a good percentage ofchildren are already exposed to such information,which could apply that these occurrences arecommon in the estate.

• The students also answered that having industriesnearby has its harmful effects (71.22 %) as itcreates heavy pollution in the area which in turnaffects the human health.

• The students are aware that industrial accidentscan damage the environment (23.11%) and alsoharm employees working there (18.7%) but itseems that students are unaware about the effectsof industrial disasters on the community residingnear the estate.

From the above survey/findings we were able togather a good perception of the present level ofawareness, understanding and perspectives ofstudents regarding industries and industrialaccidents. One of the important findings that cameout of the survey was that the students are notaware about implications of industrial accidents ontheir own safety and that of the community livingnearby.

Some survey questions and responses are presentedin diagrams below:

166

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Have you ever been inside an industrial estate?

Do you have any family member working in the industry?

8.66

4.33

8.26

2.4

0

2

4

6

8

10

What happened in the accident?

12.59

25.59

37.79

19.7

05

10152025303540

An accident in an industry can cause damage to:

Only

work

ers

of i

ndus

try

Peop

le liv

ing

near

by th

e in

dustr

y

Envir

onm

ent

All

Accid

ents

relat

ed to

the g

as le

akag

e,

chem

ical b

urn

Occu

patio

nal

accid

ents Fir

eac

ciden

tsAc

ciden

ts in

which

perso

n inj

ured

due

to th

e fall

of s

tructu

re

YES44.88%

NO55.12%

NO49.6%

YES48.42%

The students were also asked to draw a picture ofthe industrial estate as they have seen it or as theyperceive it to be. In this exercise most of the studentsdrew houses, factory gates, chimneys with smokeetc.. Some of the students who have been inside theindustries drew whatever they remembered of theindustries. This helped to gain an insight intochildren’s perception about industries.

• Linking with individual industries For the projectactivities, it was important that the students andteachers become familiar with typical industriesthat exist in the estate and have a general ideaabout how industrial operations are carried on,what kind of dangers exist and what precautionscan be taken to avert any accident. Thus it wasthought that the schools would link up withrepresentative industries of the estate which willhelp them in better understanding and ‘feel’ ofthe industries.

To identify and finalize the industries which couldbe a part of the project, the Vatva IndustriesAssociation was approached. It was felt that theAssociation would be able to guide us as to whichindustries would be most suited, and their influ�ence would also be useful in having an industry bea part of our project. The involvement of theAssociation was also felt necessary as they are amajor stakeholder in the issue and necessary forany mainstreaming. When we approached theAssociation office�bearers, they were very coopera�tive but had the apprehension whether theprogramme would focus on industries being pollut�ing and dangerous. But we were able to convincethem that our programme would be more to dispelsuch doubts and would result in better preparationamong the community and the workers for anyaccident or disaster in the industry and the estate.We had several meetings with the Association andfinally the industries were finalized for the project.

The five industries finalized were: Coates of IndiaLtd.—manufacturer of printing inks, J.B. Packagingand Jyoti Plastic Industries—manufacturer of plasticjars, ice�cream cups, etc., Matangi Industries—manufacturer of Vinyl Sulphone Ester, OshoPharmaceuticals —manufacturer of pharmaceuticalformulations.

• School visits to industries: Students and teachersoriented in the initial session were taken for visitsto industries. Each school visited one industry. Eachstudent was given a survey form for the visit. Theyhad to fill up this form at the industry through thebriefings given by the industry managers and by

167

asking questions. The survey format designed forthe visit emphasized on information like accidentpotential of the industry and type of damage likelyto result, and the working, environmental andsafety norms prevailing there. The students filledthe forms through interaction with the industrysupervisors/managers.

During the visit it was realized that some industrypeople, especially those dealing with hazardouschemicals were not open about their practices andshied away from sharing much information.

As this was the first�of�its kind visit for schoolchildren, they not only enjoyed the outing but alsothe type of exposure and information they gotfrom the visit. The students were keen and eagerto learn during the visits and asked manyquestions of the concerned persons in theindustries. For many of them this was the firsttime that they went inside an industry and sawthe industrial processes and safety equipmentsinside the industry. Through the survey they wereable to gather the relevant information about theindustrial processes and safety mechanisms in theindustries. They were also able to assess whetherthe industry had the potential of causing harmdue to pollution or due to an industrial accident.

After the industrial visits, a debriefing session wasorganized where children shared their experiences.They also made drawings based on what they sawduring the visit and what struck them most.

Another survey form was given to students basedon which they interviewed their family members,neighbours etc. regarding their knowledge aboutindustrial accidents in the estate, safety andpreparedness measures etc. The results revealedthat 70 per cent of those surveyed felt that therecould be major accidents in the area, while closeto 35 per cent had direct or indirect experiencesof accidents in their work area. Only around 10per cent of those surveyed had even a preliminaryidea of what to do in disasters like gas leak orfire. Thus the relevance and need for the projectwas strengthened with these survey results.

3 – Collection of background material: One of theimportant activities that was carried on all throughthe project was to gather relevant backgroundmaterial for the project. Many libraries inAhmedabad and offices of Factory Inspectorate,GIDC, Vatva Industries Association, Fire Brigadeoffice etc. were visited to gather more information.A visit was also made to Gujarat Safety Council,Vadodara which is an institution involved with

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

developing and conducting many safety relatedprogrammes with the industries in the state.Meetings were also held with Safety Managers atGujarat State Fertilizer Corporation (GSFC) andGujarat Alkalis and Chemicals Ltd., Vadodara. Theseindustries have carried out community awarenessprogrammes related to industrial accidents. Manyother local, state and national level organizations likeNational Safety Council, Loss Prevention Associationetc. were contacted for collecting backgroundmaterial, especially related to school and communityawareness programmes. Extensive web search wasalso carried out to gather relevant information.

4 – Development of teacher booklet and posterAfter completing the visits, surveys and collection ofbackground material, work on developing aTeacher’s Manual and poster was started. TheManual adopts an activity�oriented approach forconveying the concepts as well as preparednessmeasures detailed in the manual. The Manualaddresses the links of industry and environment,industrial estates etc, in general and goes on to givespecific information about Vatva estate, its hazardousindustries etc., potential threats and responsemechanisms for each. A list of Annexure providesuseful contacts. The draft of Teacher Manual wasshared not only among the teachers of theparticipating schools but also with relevant expertsand other key stakeholders for the project.

The poster is targeted to the general communityliving in the vicinity of the estate. It depicts thesequence of actions to be undertaken in case ofthree specific disasters—gas leak, fire and chemicalspill. Most of the information is communicatedvisually so that illiterate community members canalso become aware of the responses. The posters arebeing put up at common public places and also atindividual schools and some industries.

5 – Teacher orientation workshop A TeacherOrientation Workshop was organized which had avery good response from all the schools. From someschools, three teachers participated (we hadrequested for at least one teacher to participate). Theworkshop oriented participants to the content andapproach of the Manual. Some of the activities fromthe Manual were performed by the teachers to getan idea of how the activity approach helps in easiertransmission of information. The teachers participatedenthusiastically and made several suggestions. Thetopics for student Elocution, Drawing Competitionswere suggested by the teachers themselves.

6 – School and community programmes forinformation dissemination The schoolprogrammes for dissemination involved carrying outactivities through which the students themselvesbecame aware and then further passed this messageto other in the schools. The teachers oriented throughthe workshop undertook various activities with thestudents. Activities like Bulletin Board featuring newsrelated to industrial accidents was started in schools.Drawing competition was also organized for thestudents.

A poster was prepared which had all the steps to befollowed in case of three main industrial accidentsviz. gas leak, fire and chemical spill. The instructionswere mostly visual with some in words. These wereput up at important community places like hospital,bus stop, post office, association building, shoppingcomplexes etc. were they can be seen by thecommunity. The posters had a good response withmany people thronging to see it as soon as theywere put.

A skit on the subject of industrial disaster prepared�ness based on traditional mythological charactershad been developed which was first performed byteachers at the Orientation Workshop. The teachershave then oriented the students for the skit in theirrespective schools. The students would be performingthis skit at the school for everyone and subsequentlyat other community places and also in factories.

7 – Concluding event A big concluding event wasorganized at the premises of the Ind�German ToolRoom at the estate. All the stakeholders viz. theexecutive of the industrial association, AhmedabadFire Brigade personnel, representative from FactoryInspectorate, local NGOs working in this field and theschools were present for the event. The event hadperformance of the skit which all the schools hadprepared related to measures to be undertaken incase of an accident. There was elocution contest ontopics related to industrial disaster and safety. Theprizes for both these categories were given on thespot by stakeholder dignitaries. This was followed byan elaborate drill by the fire brigade personnel. Theydemonstrated different kinds of methods and techni�ques to deal with accidents, especially those relatedto industrial accidents. Later as a mock drill, they alsoused smoke bombs to simulate for gas leak andasked the children to perform the right action for thiseventuality. While most of the children covered theirnose and mouth, they were undecided to the direc�tion in which they should run. But soon they figured

168

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

it out. But this was a good exercise to demonstratethe preparedness of the children and the community.

The best part of the event was not only theenthusiasm shown by the schools but also theparticipation of other stakeholders which cameforward to take ownership of the activities.

8 – Feedback The feedback process for some of thecomponents has been completed while for some theprocess is on. The feedback from the teachers afterthe orientation workshop was very positive. Theyliked the activity�approach and their experience ofdoing the activities and skit themselves. Ideas likeBulletin Board and slogan writing on kites weremuch appreciated and most of the schools havestarted implementing these in their schools. Theresponse of students after each session, industrialvisits and debriefing by teachers has also shownpositive increase in their understanding of the issueand the objectives of the project. Most students haveparticipated in the competitions organized in theirschools. The feedback process for posters and othercommunity programmes is on. The schools haveassured that they would be carrying on thisprogramme as a regular activity for their 9thstandard students and in doing the communityprogrammes whenever possible.

The Manual prepared for the project has been mailedwidely and similar estates are being looked into forcarrying out similar awareness and preparednessactivities through partnerships.

Important elements with an impacton disaster risk reduction

Important elements having an impact on disaster riskreduction in India and worldwide which came outthrough this project are as follows:

• Identifying urban poor, especially those living nearfactories as an important vulnerable community.When we talk of poor and vulnerability reduction,most commonly it is the rural poor which come tomind. Identifying the vulnerability of an urban poorcommunity and working towards their riskreduction has been one of the primary goals ofthis project.

• Utilizing education and communication as aneffective tool for disaster risk reduction. Educationand communication as demonstrated through thisproject is one of the most non�interventionist, low�resource and with widely reaching approach forvulnerability reduction.

169

• Demonstrating schools as an effective low�resource communication tool to widely reach outto communities.

• Involvement of schools in a large measure ensureslonger sustainability of the project activities thanany other approach, as schools are an alreadyexisting institutionalized structure.

• A consensus�building or enabling approach helpensuring greater stakeholder participation thanone of confrontation or activism. The cooperationextended by the Industrial Association, govern�ment institutions was possible due to the friendlystance adopted by us rather than one of activistkind—ready to expose them. Taking them intoconfidence had enabled us to achieve much nowand much more in future.

Components of learning

Learnings from stakeholder participation The execution of the project and its outcomes heavi�ly depended upon the participation of the key stake�holders. Some key learnings from the nature anddegree of participation from different stakeholdersare given below:

• Schools: Schools the prime vehicle for communica�tion displayed a very positive response for all theproject activities. At the beginning of the project,one of the apprehensions was also about the roleand participation of the schools. It was felt thatschools might take this as an additional burdenon their already stressed curricula processes,might also be wary of sending children for indus�trial visits or organizing any exercises or drills inschools which affects their regular schedule. Butthere was a tremendously positive response fromall the schools that were approached. They carriedout all the activities and are keen for furtherprocesses to continue this trend. Their interest can be gauged by the fact that for the teacherorientation workshop each school sent two orthree teachers, whereas we had requested for at least one.

• Industries Association: The response fromIndustries Association was a little shaky at thebeginning but improved with time. This wasprimarily due to their fear of their estate beingprojected as an unsafe or environment�unfriendlyestate. Assurances and their involvement at everystage of the project ensured that this notion waserased. The draft Teacher’s Manual was sharedwith them so that they can be assured thatnothing in the manual paints a false picture about the estate.

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

• Government institutions: The response fromgovernment institutions was more than positivethan what we had expected. The FactoryInspectorate, GIDC, Fire Brigade sharedinformation and provided whatever help we askedwith commitments towards the future also. Thusour common notion that it is difficult to interactwith state�institutions was somewhat erodedduring the project activities.

A threat-aware but ill-prepared community A significant learning from the project was that thecommunity is rather aware of the kind of threats theyhave due to their close proximity to the estate, butextremely unprepared from any kind of responsemechanism during a disaster event. Through thesurveys done with the students where there werealso questions for their family members, it came outthat people knew of what kind of accidents canoccur in the industries, but what were hardly awareof what they should do in such events. There was afelt need for preparedness and thus the messagesfrom the project were well taken. There is still a lotmore scope of preparedness which reaches out to allthe sections of the society vulnerable to suchdisasters.

It was also noticed that not only industrial disaster,but the community is also unaware of responses inother disasters like earthquake, etc. Gujarat faced adevastating earthquake in 2001 and parts ofAhmedabad were severely effected. The schools werealso not aware of preparedness mechanisms tohandle such disasters. Thus the scope of riskreduction can be increased to include thepreparedness measures for other disasters also.

Consensus-building approach helps more than aconfrontationist one We felt that during our initialvisits to Vatva, that there is lot that has not beenaddressed in the estate with regard to safety mainlybecause safety has not been a priority issue amongthe industry management. The fear of the IndustriesAssociation was also to some part associated withthis. Our approach was not to confront them forthese issues but try to develop a rapport with them,so that not only the project activities can take placebut also we can slowly convince them aboutnecessity of safety and preparedness. With time,the Association was completely convinced that ourpurpose was community preparedness and notexposing their lack of measures to address safety.

Sustainability, ownership and way forward

• Addressing the link between poverty and disaster risk reduction in developingcountries

The project aims at industrial disaster riskreduction for communities living in the vicinity ofan industrial estate. In a developing country likeIndia, the human settlements in the vicinity of theestate are mostly of workers employed in thefactories inside the estate. These communities arealso the most vulnerable to any accidenthappening in the estate due to their closeproximity, as well as lack of access to informationabout responses and access to medical andinstitutional help. The project targeted one of themost vulnerable, low�income level communitywhich has not been part of any formal or non�formal programme towards risk reduction.

It is important that communities like these becomeaware of disasters that can occur in the estate andappropriate responses to cope with them.Awareness and preparedness measures by theemployers or other formal institutions are not thescale or depth required. Most only prepare theworker for any accident inside the factory. Thisleaves the community extremely vulnerable to anydisaster event, as was witnessed during BhopalGas Tragedy. It is also important to note that urbanpoor are not the ones to be addressed firstwhenever the links between poverty and disasterrisk reduction are spoken of.

This project has attempted to fill this informationgap and prepare communities through a uniquechannel of information. The schools in the vicinityof the estate have more than 50 per cent of theirstudents coming from settlements close to estateand of workers employed in the factory. Thefamilies of a large percentage of the students,belong to the lower income level of the society.

The project at the first level has oriented thestudents and teachers towards awareness andpreparedness exercises for different kinds ofdisasters that can occur in the estate. The wholeschool is then exposed to the knowledge throughthese oriented students. This message, through thechildren gets passed on to their individual familiesand further to the community through the familiesand other community programmes. There areposters developed in the project which pictorially

170

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

depict what actions have to be taken in eachdisaster. These have been displayed at publicplaces, reaffirming the message of preparedness.Drama based on traditional characters to conveypreparedness messages, kite flying with safety and preparedness slogans and many such similarexercises helped in wide dissemination of themessage through a variety of approaches. Theseare extremely low�resource approaches and those which are very popular in the lower strata of the place.

Thus by using school children to create a multipliereffect, the project has been able to reach the mostvulnerable section in the most cost�effective way.Similar situations exist in industrial estates notonly in India, but also in many developingcountries. There are reasons to believe that thiscommunication strategy might work as an hithertountried but a useful approach towards industrialdisaster risk reduction.

• Sustainability of the project beyond the six months

The programme would be sustainable as themanual and other information developed would beavailable with the schools and it will have trainedteachers to sustain the effort. Hence, every newbatch of students would get trained in thispreparedness activity and spread the messageacross. This will also make the material developedas part of the project very cost�effective as it willbe repeatedly utilized. The teachers and studentspropagating this already feel a sense of ownershipto the whole exercise and would act as leaders incase of an event. While only the teachers havebeen directly oriented, the students and thecommunity thereafter have been the indirectlearners and beneficiaries of the project, whowould sustain this preparedness exercised beyondthe period.

The Vatva Industries Association which has beenan active partner in the whole project, has alsoshown interest in sustaining the activities. Thereare also chances of scaling up the project as thereare more than 200 industrial estates in Gujaratwith very similar conditions. The options forcarrying out similar activities in these estates arebeing explored.

171

• Ownership and sharing with stakeholders

The ownership of the project has slowly with theend of the six months shifted to one of the mostimportant players in the project—the schools.Through the orientation sessions, the teacher’sworkshop and finally the awareness programmesby students in schools and further to community,the school management, teachers and thestudents have really begun to take ownership ofactivities of the project. This has come about dueto our efforts to involve the schools not just acomponent of the project but taking their advise,suggestions and framing ideas for school activitiesthrough them. Schools started some of theactivities like Bulletin Board, drawing competitionfrom the very next few days after the teacherorientation workshop. They are enthusiastic aboutthe school projects and the drama activity by thechildren. Although the project has ended itsstipulated period, they are eagerly awaitingparticipation in a estate�level elocution, drawing,drama etc competition which is planned to beorganized during end�January.

Other key institutions related to the project goalsand activities have also started to develop a senseof ownership for the project rather than beinganother actor for project activities. One of themost stakeholder involved is the Vatva IndustriesAssociation, who have started feeling that thisactivity is beneficial for the welfare of their estateand can bring credit to them for being a safety�friendly industrial estate. The Association hasproposed that an event be organized where allthe key stakeholders are invited. This would helpin sharing of results of the project, in planning forsustenance of the project activities and to buildupon future programmes related to communitypreparedness. Other important institutions like theFactory Inspectorate, Gujarat Industrial Develop�ment Corporation, Fire Brigade Department, NGOsinvolved in such work have all participated andcontributed enthusiastically in carrying out projectactivities. The Factory Inspectorate has shared theinformation about the functioning of the LocalCrisis Group at Vatva, GIDC has given all relevantinformation needed time and again for executionof project activities and development of theManual. The Ahmedabad Fire Brigade organized a massive drill and demonstration at Vatva and

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

NGOs like Unnati working in the field have shared their experiences and communicationmaterial with us.

The draft of the Teacher’s Manual was sharedamong all of the above mentioned institutionswhich provided relevant inputs. Their inputs have been well�recognized throughout the project and they will also serve as channels ofdissemination for the communication materialdeveloped in the project.

The event held in February served as a goodplatform for sharing of results, experiences andlooking at ways of moving the project activitiesforward. The event had major ownership from

within the estate. The whole premises, includingthe auditorium was provided free by themanagement of the Indo�German Tool Room. TheIndustrial Association came forward to bear theexpenses of refreshments provided for the kids.Thus it seemed that with this a good beginninghas been made.

At the state and national level, institutions such asGujarat Safety Council and National Safety Councilhave provided meaningful inputs in the project.The National Safety Council has also asked if CEEcould serve as a Centre to disseminate theircommunication material.

172

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

173

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

Risk reduction through first

aid capacity building of youth

emergency teams

TEAM LEADER

Nadejda Petrova <[email protected]>

TEAM MEMBERS

Mariana PanayotovaAtanas TsvetanskiTeodora SakaliiskaIvaylo PastuhovAleksander Manushev

PROJECT ADVISOR

Jivka Dimitrora

Bulgaria

SYMPOS IUM NOTES

174

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

175

WO

RK

ING

PA

PE

RS

Mu

lti-

Ha

zard

s

In the second half of 2003 Bulgarian Red CrossYouth implemented a project related to transfer ofknowledge and skills and awareness raising aboutways of reducing disaster risk under the name ‘RiskReduction Through First Aid Capacity Building ofYouth Emergency Teams’.

The project was initially planned to be implementedin 14 out of the 28 administrative regions ofBulgaria, in which exist the so�called YouthEmergency Teams.

The Youth Emergency Teams (YET) are voluntaryteams of young people (aged between 16 and 25),members of the Bulgarian Red Cross Youth, whichare active on the territory of particular administrativeregions of the country. They have some supportivefunctions, related to disaster response, but their main tasks are in disaster preparedness. The YouthEmergency Teams (which became 17 in the course ofthe project) disseminate materials, related to disasterrisk among the population, they carry out lecturesand field drills in schools and different public orbusiness organizations.

The idea behind the project was that by building thecapacity of the existing Youth Emergency Teams inBulgaria in the field of first aid, their work withintheir local communities will become more efficientand they could take a leading role in risk sharing.This was planned to be done by demonstrating theimportance of first aid as primary step in reducingrisk in emergencies and by training people indelivering first aid.

At the beginning of the project was established ateam of advanced members of YET, whose role wasto build the capacity of the others in the field of firstaid. In July 2003 the team took part in the largestfirst aid competition in Europe – the First AidConvention of Europe (FACE) – in Prague, the CzechRepublic. This was the first occasion for the last fewyears that a Bulgarian youth team was able toparticipate in FACE and it was used by the projectteam for obtaining the latest standards in first aid. AtFACE the team checked its level of training and got11th position among all the 26 teams presented inthe competition.

In the following months the team started carryingout subsequently one�day trainings of each the YETsfrom the different parts of the country. Because ofthe limited time and some budget considerations, the

team was divided into sub�groups of 2 trainers each.On a rotating principle each couple of trainerstravelled to several different destinations to workwith the YETs based there.

Meanwhile 3 new YETs were established in thetowns of Ruse, Montana and Targovishte. This wasan evidence for strengthening of the YET network towhich the project contributed to a large extent.

Project links between poverty anddisaster risk reduction in developingcountries

The project we implemented is not directlyconcerned with the issue of poverty, thought there isa certain link between the process of socialtransition in Bulgaria as a post�communist country(one of which marks is poverty) and the disaster riskreduction. As the state is experiencing a difficultperiod, the government’s efforts are directed atmeeting mainly the primary needs of the population.Unfortunately, disaster risk sharing is not amongthem and especially in the small communities peoplelack information and skills, which would help forreducing the disaster risk.

Through the current project, Bulgarian Red CrossYouth concentrated particularly on smallcommunities, disseminating information and trainingparticular groups of the population (mainly youth) infirst aid skills.

Although the project could not fill in the gap in thepopulation’s knowledge, related to disaster riskreduction and first aid, it demonstrated a goodpractice in disaster transfer, through a voluntary�based approach.

Project sustainablity

The project is based on the work of 17 voluntaryteams of young people, who are disseminatingknowledge and skills, related to disaster riskreduction among the population. Through the firstsix months the capacity of these youth emergencyteams was build up, so that they can continue doingtheir important work within the communities.

Even though these structures already exist and thereis no need that funds are allocated for developingsuch, they still need proper equipment and funds forcovering their subsistence costs.

176

WO

RK

ING

PAP

ER

S

Mu

lti-Ha

zard

s

Project ownership among a broadcommunity and institutional base

The project is recognized by the Bulgarian Red Cross as an important activity in the field of disasterpreparedness.

Currently, within the Bulgarian Red Cross, the YouthEmergency Teams are the only functional structure,engaged in disaster risk sharing, through direct workwith the population.

Components of learning

The project as a whole is based on the concept oflearning by doing. People are trained in practicalskills for delivering first aid in case of disaster,which could prevent or reduce the disaster risk.

The success of the learning process is ensuredthrough, training in small groups, taken at the same format at which they exist naturally in thecommunity. This helps for strengthening the relations within the communities and developingfeelings of mutual trust and belongingness.

LESSONS LEARNED

After receiving their first aid training, the YETs had toput the lessons learnt into practice. Each of the YETdeveloped an action plan for work within the localcommunity for building its capacity for disaster riskreduction.

In each of the project regions were carried outexercises in schools, based on an agreement with theMinistry of education and Science in Bulgaria thatRed Cross youth volunteer have access to each publicschool in the country. Till December 2003 in the

events, related to first aid, participatedapproximately 1600 secondary school students in allthe project regions.

The First Aid Day, 13th of September 2003, was usedfor raising the awareness of the local communities inthe project regions about disaster risk reduction andfirst aid measurements in cases of small�scaledisasters or accidents.

In 7 towns were organized in the open simulationsof first aid measurements in case of emergency.

We estimate the direct beneficiaries from the projectto be about 1700 people, who attended the differentpublic events (different than those who werereached in the schools).

The project had an impact on the standardization ofthe concept (and the practice) of first aid in Bulgaria.

In September 2003 Bulgarian Red Cross issued afirst aid manual for the public. The project team wasinvited to take part in the in the editorial work onthe manual and contributed a lot for its successfulfinalization.

Members of the team also participated in their roleof advanced first aid trainers in a training film forfirst aid, produced after the manual, which will bedisseminated for free among the population.

The project implemented by the Bulgarian Red CrossYouth team played a significant role in the field ofdisaster preparedness in Bulgaria. It was particularlyvaluable because of the community based workwhich was done, which is a new approach for ourcountry.

177

65

RE

SE

AR

CH

GR

AN

TS

SYMPOS IUM NOTES

THE SIXTY-FIVE RESEARCH GRANTS

For the ProVention program, Applied Research Grants for Disaster Risk Reduction,there were sixty�five proposals selected to receive grants of up to US$5,000. Theprojects were. Arranged by country, here are those projects (completed between June 2003 and January 2004) with contact information for the Team Leader, a listing of team members, and advisors.

Argentina

Armenia

Bangladesh

Barbados

Bhutan

Bulgaria

China

Colombia

Costa Rica

Georgia

India

Indonesia

Kenya

Mexico

Nepal

Nigeria

Pakistan

Philippines

Senegal

South Africa

Sudan

Tajikistan

Turkey

Uzbekistan

Venezuela

Vietnam

Zimbabwe

178

65

RE

SE

AR

CH

GR

AN

TS

ARGENTINA

❚

TEAM LEADER

Pablo Suarez <[email protected]>

PROJECT T ITLE

Reducing vulnerability to climate variability andchange: cognitive and behavioral obstacles torational use of information

PROJECT ADVISOR

William P. Anderson

❚

TEAM LEADER

María Alejandra del Campo <[email protected]>

PROJECT T ITLE

Study of the awareness of Earthquake risk in the population of Mendoza

PROJECT ADVISOR

Engels German Cortés Turjillo

TEAM MEMBERS

Andrea ConaHumberto Enrique Marin UribeAndres Martina Cervan

179

ARMENIA

❚

TEAM LEADER

Sos Margaryan <[email protected]>

PROJECT T ITLE

New methodology for seismic microzonation ofthe cities at the highest risk

PROJECT ADVISORS

Hector BabyanAvetis Arakelyan

TEAM MEMBERS

Mr. Raffi Durgaryan Mr. Samvel BabayanMrs. Lilit Asatryan Mr. Artak Simonyan

❚

TEAM LEADER

Hayk Ktunyan <[email protected]>

PROJECT T ITLE

The main principles of earthquake early warningsystem creation around critical facilities inseismic active zones

PROJECT ADVISORS

Dr. Valery ArzumanyanDr. Zaven Khlghatyan

TEAM MEMBERS

Aroyan Gevorg Marine Simonyan

65

RE

SE

AR

CH

GR

AN

TS

❚

TEAM LEADER

Valery Manusagyan <[email protected]>

PROJECT T ITLE

Preliminary vulnerability assessment method ofresidential stone buildings

PROJECT ADVISOR

Dr. Zaven Khlghatyan

TEAM MEMBERS

A Baldryan A Martirosyan

BANGLADESH

❚

TEAM LEADER

Marina Parvin Juthi <[email protected]>

PROJECT T ITLE

Reducing vulnerability to floods in southwestborder districts of Bangladesh

PROJECT ADVISORS

Mr. Anwar Firoze Mr. Ashraf-ul-Alam Tutu

❚

TEAM LEADER

Syed Ashraf ul Islam <[email protected]>

PROJECT T ITLE

Technological & chemical hazards and the risk ofbulk transportation of dangerous goods alongthe ship-breaking yard of Chittagong, the portand adjacent area: a qualitative analysis of thecombination of risk factors

PROJECT ADVISOR

M. Safiur Rahman

BARBADOS

❚

TEAM LEADER

Kerry Hinds <[email protected]>

PROJECT T ITLE

Community risk management

PROJECT ADVISOR

Miss Judy Thomas

BHUTAN

❚

TEAM LEADER

Jigme Dorji <[email protected]>

PROJECT T ITLE

Demonstration to the Bhutan building industry:Better safety of life and cost saving by avoidingun-reinforced brick infill in RCC frames designedfor seismic attack

PROJECT ADVISOR

J. Spencer Nicholls

180

65

RE

SE

AR

CH

GR

AN

TS

BULGARIA

❚

TEAM LEADER

Nadejda Petrova <[email protected]>

PROJECT T ITLE

Risk reduction through first aid capacity buildingof youth emergency teams

PROJECT ADVISOR

Jivka Dimitrora

TEAM MEMBERS

Mariana Panayotova Atanas TsvetanskiTeodora Sakaliiska Ivaylo PastuhovAleksander Manushev

CHINA

❚

TEAM LEADER

He Xiaoyan <[email protected]>

PROJECT T ITLE

Vulnerability of a community’s structures, peopleand property based on flood risk assessment

PROJECT ADVISOR

Dr. Cheng Xiaotao

TEAM MEMBERS

Wang Yanyan Hu Changwei

181

❚

TEAM LEADER

Honglei Wang <[email protected]>

PROJECT T ITLE

The disaster awareness and risk management offlooded Yangtze River: A case study of 1998Dyke Burst in Jiujiang

PROJECT ADVISOR

Guo Qiang

TEAM MEMBER

Zeng Huaide

❚

TEAM LEADER

Yi Tinghua <[email protected]>

PROJECT T ITLE

Near source strong ground motion estimation

PROJECT ADVISOR

Asst. Prof. Dr. Wang Guoxin

❚

TEAM LEADER

Shen Lixin Shen Lixin <[email protected]>

PROJECT T ITLE

Social-forestry based disaster preparedness andmanagement in rural community of MenkongCatchment, Yunnan Province, Southwest China

PROJECT ADVISOR

Dr. Hu Gang

TEAM MEMBERS

Zhan Wen Jiao Xiaoxu6

5 R

ES

EA

RC

HG

RA

NTS

COLOMBIA

❚

TEAM LEADER

Martha Liliana Carreño T.<[email protected]>

PROJECT T ITLE

Expert system for post earthquake buildingdamage evaluation and massive risk occupancy

PROJECT ADVISOR

Omar D. Cardona

COSTA RICA

❚

TEAM LEADER

Milena Berrocal Vargas<[email protected]>

PROJECT T ITLE

Risk identification and reduction in NW CostaRica: Soil liquefaction at Nicoya Peninsula andvolcanic vulnerability at Arenal volcano

PROJECT ADVISORS

Prof. Eduardo Malavassi Ph.D.Lic. Vilma Barboza

GEORGIA

❚

TEAM LEADER

Jumber Mamasakhlisi<[email protected]>

PROJECT T ITLE

Research on establishment of general concepton interaction of different disaster data

PROJECT ADVISOR

Mr. Otar Tavelishvilli

INDIA

❚

TEAM LEADER

Rashmi Tewari<[email protected]>

PROJECT T ITLE

Rehabilitation including basic health care for 35,000 Riang displaced population atKanchanpur, North Tripura District, India

PROJECT ADVISOR

Dr. P. R. Trivedi

182

65

RE

SE

AR

CH

GR

AN

TS

❚

TEAM LEADER

Amit Kumar <[email protected]>

PROJECT T ITLE

Urban multiple disaster scenario and decisionmaking system for vulnerable area of MadhyaPradesh, India

PROJECT ADVISOR

Prof. Amit Bose

TEAM MEMBER

Dilip Kumar Singh

❚

TEAM LEADER

Dash Biswanath <[email protected]>

PROJECT T ITLE

Indicators for disaster preparedness

PROJECT ADVISOR

Prof. Vinod Kumar Sharma

❚

TEAM LEADER

Nalini Sankararamakrishnan <[email protected]>

PROJECT T ITLE

Environmental monitoring of pesticide residuesand heavy metals in and around Kanpur, UttarPradesh, India

PROJECT ADVISOR

Prof. Rabindranath Mukherjee

TEAM MEMBER

Rashmi Sanghi

❚

TEAM LEADER

Vishnu Prabhir <[email protected]>

PROJECT T ITLE

Action advocary to ensure right to livelihood riskreduction and beyond

PROJECT ADVISOR

Mihir R. Bhatt

183

❚

TEAM LEADER

Shivani Khanna <[email protected]>

PROJECT T ITLE

Organisational learning for community baseddisaster risk reduction: lessons from DisasterMitigation Institute.

PROJECT ADVISOR

Mihir R. Bhatt

TEAM MEMBERS

Mr. Mehul Pandaya Mr. Himanshu Kirani

❚

TEAM LEADER

Aurobindo Ogra <[email protected]>

PROJECT T ITLE

Disaster management information systems

PROJECT ADVISOR

Mr. Vivek Ogra

TEAM MEMBER

Deepak Kaul

❚

TEAM LEADER

Avanish Kumar <[email protected]>

PROJECT T ITLE

Testing communication strategies for industrialdisaster risk reduction

PROJECT ADVISOR

Ms. Meena Raghunathan

❚

TEAM LEADER

Kishor Subhash Jaiswal<[email protected]>

PROJECT T ITLE

Earthquake hazard assessment and informationdissemination in Mumbai, India

PROJECT ADVISOR

Dr. Ravi Sinha

65

RE

SE

AR

CH

GR

AN

TS

❚

TEAM LEADER

Shagun Mehrotra <[email protected]>

PROJECT T ITLE

Institutionalising Innovative Local Practices andPartnerships for risk reduction in earthquake-prone settlements of the poor: A case study inPorbandar District, Gujarat, India

PROJECT ADVISOR

Dr. Ranjith Perera

❚

TEAM LEADER

Rawat Pushplata <[email protected]>

PROJECT T ITLE

Developing & disseminating a model ofcommunity-based, disaster-mitigatingdevelopmental planning

PROJECT ADVISORS

Dr. Dinesh Sati Mr. Madhur Darmora

TEAM MEMBERS

Mr. RS Negi Ms. Ranjani Jain

❚

TEAM LEADER

Jojo Sunil Kumar Dr. Bipin K.<[email protected]>

PROJECT T ITLE

Community based advanced risk sharingprogramme

PROJECT ADVISORS

Dr. Bipin Kr Jojo Mr. KML Sastry

❚

TEAM LEADER

Arpana Das <[email protected]>

PROJECT T ITLE

Planning for risk reduction through sustainablesiting of embankments in the Sundarbans

PROJECT ADVISORS

Dr. Sekhar Chandra Nilanjan Sengupta

TEAM MEMBERS

Meghamala Basu (Roy) Soumen MitraAmitava Mallick Parama BhattacharyaSanmarga Mitra Suman De (Sen)

❚

TEAM LEADER

Upasna Sharma <[email protected]>

PROJECT T ITLE

Improving the methodology for assessing naturalhazard impacts

PROJECT ADVISOR

Anand Patwardhan

❚

TEAM LEADER

Pratima Singh <[email protected]>

PROJECT T ITLE

Post disaster risk identification: A case ofrehabilitation process of rural Kutch in Gujurat,India

PROJECT ADVISOR

Dr. Greg Bankoff

TEAM MEMBERS

Prof. Saswat BandyopadhyayMr. Alok Das Ms. Nikeeta Singh

❚

TEAM LEADER

P.V. Krishnan <[email protected]>

PROJECT T ITLE

Critical Study into the government policies ondisaster management in India

PROJECT ADVISOR

Mr. D. K. Mishra

❚

TEAM LEADER

Sonalini Khetrapal <[email protected]>

PROJECT T ITLE

Hazard resistant health delivery system

PROJECT ADVISOR

Dr. Bipin kumar Verma

184

65

RE

SE

AR

CH

GR

AN

TS

INDONESIA

❚

TEAM LEADER

W. Warsa <[email protected]>

PROJECT T ITLE

Multidimensional SNMR modeling forgroundwater exploration

PROJECT ADVISORS

Prof. Dr. Ugur Yaramanci Dr. Martin Müller

KENYA

❚

TEAM LEADER

Rutere Salome Kagendo <[email protected]>

PROJECT T ITLE

Stregnthening rural community bonds as ameans of reducing vunerability to landslides

PROJECT ADVISORS

Dr. Robinson M. Ocharo Prof. Octavian Gakuru

185

MEXICO

❚

TEAM LEADER

Fernando Briones Gamboa<[email protected]>

PROJECT T ITLE

El Niño phenomenon: social construction of risk in the Isthmus of Tehuantepec: A multidisciplinary approach

PROJECT ADVISOR

Virginia García Acosta PhD

TEAM MEMBERS

Fercia Adelaida Angulo FernándezMartha Leticia González Álvarez

❚

TEAM LEADER

Sergio Omar Saldana Zorilla <[email protected]>

PROJECT T ITLE

Reducing vulnerability from natural disasters in Mexican agriculture

PROJECT ADVISORS

Dr. Joanne Linnerooth-BayerDr. Reinhard Mechler 6

5 R

ES

EA

RC

HG

RA

NTS

NEPAL

❚

TEAM LEADER

Komal R. Aryal <[email protected]>

PROJECT T ITLE

Community disaster management with GIS in Nepal

PROJECT ADVISORS

Dr. Andrew Collins Dr. Robert MacFarlane

❚

TEAM LEADER

Keshav Prasad Paudel <[email protected]>

PROJECT T ITLE

Mapping and assessing risk and vulnerability ofwater induced disaster in the Tinau Watershed,Western Nepal

PROJECT ADVISORS

Prof. Dr. Vidya Bir Singh KansakarNarendra Raj Khanal

❚

TEAM LEADER

Madhab Prasad Bhusal<[email protected]>

PROJECT T ITLE

Flood disaster reduction in Karnali Watershed, Western Nepal

PROJECT ADVISOR

Dr. Pushkar Kumar Pradhan

❚

TEAM LEADER

Ripendra Awal <[email protected]>

PROJECT T ITLE

Floodplain analysis and risk assessment ofLakhandei RIver

PROJECT ADVISOR

Dr. Narendra Man Shakya

❚

TEAM LEADER

Archana Pradan <[email protected]>

PROJECT T ITLE

GIS and remote sensing for flood disasteridentification: A case study of the Koshi RiverBasin in Nepal

PROJECT ADVISORS

Dr. Vishnu DangolD. Prakash Chandra Adhikari

NIGERIA

❚

TEAM LEADER

Jimoh Ismaila Adetunji <[email protected]>

PROJECT T ITLE

Potential hazards of landfill sites on surface andground water resources in selected cities in SWNigeria

PROJECT ADVISORS

R.O. Majolagbe A. Adenrele

186

65

RE

SE

AR

CH

GR

AN

TS

❚

TEAM LEADER

Ahmed Adedoyin Balogun<[email protected]>

PROJECT T ITLE

Development of a national wildland fireinventory and fire disaster management action plan for Nigeria

PROJECT ADVISORS

Prof. J. A. Omotosho Dr. O.O. Jegede

PAKISTAN

❚

TEAM LEADER

Sarwat Naz <[email protected]>

PROJECT T ITLE

Effectiveness of policies for the reduction of flood hazard in Pakistan: A case study of Nalaleh, Rawalpindi City

PROJECT ADVISORS

Prof. Dr. Shafique-ur-RehmanMr. Falak Nawaz

TEAM MEMBERS

Mohammed Ishfaq

187

❚

TEAM LEADER

S. Jalal us din Shah <[email protected]>

PROJECT T ITLE

Ghariat potential landslide: demonstratinghazard risk assessment tools and techniques inmountain environments

PROJECT ADVISOR

Hadi Husani

TEAM MEMBERS

Mujeeb Alam Ejaz KarimZahra Shaikh

PHILIPPINES

❚

TEAM LEADER

Rose Berdin <[email protected]>

PROJECT T ITLE

Coastal erosion vulnerability mapping along thesouthern coast of La Union, Philippines

PROJECT ADVISOR

Dr. Fernando P. Siringan

TEAM MEMBERS

Christina T. Remotigue Peter ZamoraYvainne Y. Yacat

65

RE

SE

AR

CH

GR

AN

TS

SENEGAL

❚

TEAM LEADER

Aliou Mamadou Dia <[email protected]>

PROJECT T ITLE

Earth observation & GIS-based flood monitoringin the Senegal River Estuary and Valley

PROJECT ADVISOR

Dr. Alioune Kane

SOUTH AFRICA

❚

TEAM LEADER

Dewald van Niekirk <[email protected]>

PROJECT T ITLE

The development of uniform datasets formeasuring hazards in Southern Africa

PROJECT ADVISOR

Prof. Gerrit van der Waldt

❚

TEAM LEADER

Nogenga Faith Yolisa <[email protected]>

PROJECT T ITLE

Research in forest fires for South African disastermanagement

PROJECT ADVISOR

Dr. Krisno Nimpuno

❚

TEAM LEADER

Bernadette Wanjiru Kamani<[email protected]>

PROJECT T ITLE

Droughts in East and South Africa

PROJECT ADVISORS

Dr. Krisno Nimpuno Dr. R.M.Ocharo

❚

TEAM LEADER

Tobias Landmann <[email protected]>

PROJECT T ITLE

A Geographical Information System (GIS) on fuelbiomass derived from validated fuel models

PROJECT ADVISOR

Dr. Bob Scholes

❚

TEAM LEADER

Scott A. Ristow <[email protected]>

PROJECT T ITLE

Honey-hunters: Entrepreneurship or a symptomof social and economic poverty?

PROJECT ADVISORS

Tiaan Pool Josua Louw

188

65

RE

SE

AR

CH

GR

AN

TS

SUDAN

❚

TEAM LEADER

Eshraga Abdulla Ali Fadlalla<[email protected]>

PROJECT T ITLE

Disaster Laws in the Sudan

PROJECT ADVISOR

Mr. Mohamed Omer Mukhier

TAJIKISTAN

❚

TEAM LEADER

Firuza Nazirova<[email protected]>

PROJECT T ITLE

Safety of children in earthquakes

PROJECT ADVISORS

Mrs. Goulsara PulatovaMr. Sobit Negmatoullaev

TEAM MEMBER

Zoulfia Mirbaeva

189

❚

TEAM LEADER

Timur Idrisov <[email protected]>

PROJECT T ITLE

Emergencies Resource Center

PROJECT ADVISOR

Mrs. Goulsara Pulatova

TEAM MEMBERS

Galina Sokolova Anastasiya Idrisova

TURKEY

❚

TEAM LEADER

Bilgen Sungay <[email protected]>

PROJECT T ITLE

Seismic conservation of historical and culturaltreasures of a world city: Sizing the need andformulating an action plan for the museums ofIstanbul, Turkey

PROJECT ADVISOR

Marla Petal

TEAM MEMBER

Nevra Erturk

❚

TEAM LEADER

Ebru A. Gencer <[email protected]>

PROJECT T ITLE

Sustainable planning for disaster mitigation in Istanbul

PROJECT ADVISORS

Arthur Lerner-Lam Elliott D. Sclar

65

RE

SE

AR

CH

GR

AN

TS

UZBEKISTAN

❚

TEAM LEADER

Timur Abduvakhidov <[email protected]>

PROJECT T ITLE

Development of Album of technical decisions forreinforcement of self constructed residentialbuildings (existing and new) from local materialsin earthquake prone areas of Central Asia

PROJECT ADVISOR

Dr. Shamil Khakimov

VENEZUELA

❚

TEAM LEADER

Gabriela I. Muñoz Rodríguez <[email protected]>

PROJECT T ITLE

Local and popular folklore and culture on hazardand vulnerability meets geographicalinformation system for the risk reductionpreparedness of the people of the Barrio SanAntonio of Naiguatá, Estado Vargas

PROJECT ADVISOR

Diana C. Valera

VIETNAM

❚

TEAM LEADER

Hoang Nam <[email protected]>

PROJECT T ITLE

Fire-safety assesment for tall buildings in large cities of Vietnam

PROJECT ADVISOR

Dr. Pham Huy Giao

❚

TEAM LEADER

Nhu Nguyen Hong Cuong <[email protected]>

PROJECT T ITLE

Seismic microzonation in ancient quarter of Ha Noi based on microtremor observations

PROJECT ADVISOR

Phan Huy Giao

190

65

RE

SE

AR

CH

GR

AN

TS

ZIMBABWE

❚

TEAM LEADER

Chipo Muvezwa <[email protected]

PROJECT T ITLE

Strengthening regional capacity for disastermanagement

PROJECT ADVISOR

Clever Mafuta

TEAM MEMBERS

Patricia MamvotoChenai Mufanawejingo

191

65

RE

SE

AR

CH

GR

AN

TS

192

65

RE

SE

AR

CH

GR

AN

TS

193

CO

LLA

BO

RA

TIN

GC

EN

TER

S

SYMPOS IUM NOTES

THE COLLABORATING CENTERS

In order to coordinate and manage the program of Applied Grants for Disaster Risk Reduction,World Bank’s Hazard Management Unit (HMU) and the ProVention Consortium selected threewell�respected organizations working in international disaster/emergency management:

Asian Disaster Preparedness Center (ADPC)is a regional resource center working towards disaster reduction for safer communities andsustainable development in Asia and the Pacific. Established in 1986, the Center is recognizedas an important neutral focal point in Asia and the Pacific for promoting disaster awareness andthe development of local capabilities to foster institutionalized disaster management andmitigation policies.

Cranfield Disaster Management Centre (CDMC)was founded in 1985. Its aim is to save lives and livelihoods at risk from disaster impactthrough the promotion of risk and vulnerability reduction, preparedness and effective disasterresponse. The CDMC believes that disaster risks and vulnerabilities can be reduced through theapplication of sound management principles and practice.

University of Wisconsin–Disaster Management Center (UW–DMC)has served the learning needs of disaster/emergency management professionals in thedeveloping world since 1982, working closely with experts recognized for their field experienceto develop disaster management training activities with a practical emphasis. The center’sgoal is to help improve the emergency management performance of non�governmentalorganizations, local and national governments, and international organizations, through itscomprehensive professional development program in disaster management.

The World BankHazard Management Unit1818 H Street, NWWashington, DC 20433USAwww.worldbank.org/hazards/

ProVention Consortium SecretariatP.O. Box 37217, chemin des CrêtsCH-1211 Geneva 19SWITZERLAND

http://www.proventionconsortium.org

WORKING PAPERS

global symposium for hazard risk reduction - gfdrr · pdf filewebsite — ......

Documents