UNITED NATIONS

Global Survey of Early Warning Systems

An assessment of capacities, gaps and opportunities towards building a comprehensive global early warning system for all natural hazards

Final Version A report prepared at the request of the Secretary-General of the United Nations

United Nations Inter-Agency Secretariatof the International Strategy forDisaster Reduction (UN/ISDR)International Environment House II, 7-9 Chemin de Balexert, CH 1219 Chatelaine, Geneva 10, SwitzerlandNew phone and fax numbers:Tel: +41 22 917 8908/8907Fax: +41 22 917 8964isdr@un.orgwww.unisdr.org

ISDR Platform for the Promotion of Early Warning (PPEW)Hermann-Ehlers-Strasse 10D-53113 Bonn, GermanyTel.: 0049 228 815 0300Fax: 0049 228 815 0399isdr-ppew@un.orgwww.unisdr-earlywarning.org

UNITED NATIONS

Global Survey of Early Warning Systems An assessment of capacities, gaps and opportunitiestoward building a comprehensive global early warning system for all natural hazards

A report prepared at the request of the Secretary-General of the United Nations

THE SSECRETARY-GGENERAL--

FOREWORD TTO TTHE GGLOBAL SSURVEY OOF EEARLY WWARNING SSYSTEMSSeptember 22006

If an early warning system had been in place when the tsunami of 26 December 2004 struck the IndianOcean region, many thousands of lives could have been saved. That catastrophe was a wake-up call forGovernments and many others about the role early warning can play in avoiding and reducing the human andphysical impacts of natural hazards.

In the days following the tsunami, I called for the development of a global early warning system for allnatural hazards and all communities. Such a system would build upon existing national and regional capacities,and complement broader initiatives aimed at disaster preparedness and mitigation. As a further step, in March2005, I asked the Secretariat of the United Nations International Strategy for Disaster Reduction (UNISDR), inconsultation with relevant United Nations organizations, to undertake a global survey to identify existing capacities and gaps in earlywarning systems. This report is a culmination of that process.

The report highlights the significant global progress that is being made in our ability to assess risks and togenerate and communicate predictions and warnings. These gains are mainly a result of growing scientificunderstanding and the use of modern information and communication technologies.

However, considerable shortcomings and gaps remain, especially in developing countries, where basiccapacities, equipment and resources are often not available. And at the human level, we are still failing to raisepublic awareness of risks, disseminate timely and understandable warnings, and strengthen communitypreparedness and resilience. We must address these issues if the growing toll of disasters is to be arrested andreversed.

I encourage all partners -- from decision-makers in Governments to organizations involved in early warningand preparedness -- to study and act on the information and recommendations contained in this report. Inparticular, I urge all relevant actors to develop and support comprehensive strategies and follow-up actionsthrough the UNISDR and its International Early Warning Programme. Natural hazards will always challengeus, but people-centred early warning systems can be a potent weapon in ensuring that natural hazards do not turninto unmanageable disasters.

Kofi A. Annan

i

U NN II TT EE DD NN AA TT II OO NN SS NN AA TT II OO NN SS UU NN II EE SS

ii

iii

The Global Survey of Early Warning Systems was called for by the Secretary-General of the United Nations inhis report of progress towards the implementation of the Millennium Development Goals1 . Its preparation wascoordinated by the International Strategy for Disaster Reduction (ISDR) secretariat through the ISDR Platformfor the Promotion of Early Warning (PPEW) in consultation with relevant United Nations organisations.

The Inter-Agency Task Force for Disaster Reduction (IATF/DR), at its 11th session, May 2005, established aWorking Group to support the preparation of the survey. The Working Group was co-chaired by the WorldMeteorological Organization (WMO) and the Office for the Coordination of Humanitarian Affairs of theUnited Nations Secretariat (OCHA) and included the International Telecommunication Union (ITU), theUnited Nations Development Programme (UNDP), the United Nations Environment Programme (UNEP),the United Nations Educational, Scientific and Cultural Organisation (UNESCO), the United Nations HumanSettlements Programme (UN-HABITAT), the United Nations Institute for Training and Research (UNITAR),the United Nations University Institute for Environment and Security (UNU-EHS), the Asian DisasterPreparedness Center (ADPC), the IGAD Climate Prediction and Applications Centre (ICPAC), the Global FireMonitoring Center (GFMC), and the International Federation of Red Cross and Red Crescent Societies(IFRC).

A number of other organisations contributed to the survey, including: United Nations Children's Fund(UNICEF), the United Nations Convention to Combat Desertification (UNCCD), World Food Programme(WFP), Food and Agriculture Organisation (FAO), European Commission Joint Research Centre (EC-JRC),the Asian Disaster Reduction Center (ADRC), ProVention Consortium and South Pacific Applied GeoscienceCommission (SOPAC).

A meeting of international experts was held in Bonn, Germany, 12-13 December 2005 to review a draft and toprovide support for the preparation of the report.

The ISDR secretariat gratefully acknowledges the many contributions received from individuals, institutions andgovernments. In particular:Working Group Co-Chairs Maryam Golnaraghi and Ricardo Mena;Consultants Seth Vordzorgbe and Piero Calvi-Parisetti; Expert-workshop participants Alessandro Annunziato, Douglas Pattie, David Rogers, John Scott and Karl-OttoZentel; Expert reviewers Robyn Betts, Fouad Bendimerad, Michael Glantz, Chip Guard, John Handmer, Terry Jeggle,Ilan Kelman, Sir David King, Horst Letz, Marcus Oxley, LeHuu Ti, Gordon McBean, Michael Coughlan,Juan Carlos Villagran, Ben Wisner and John Zillman.

Editorial and production team: Reid Basher, Mario Barrantes, John Harding, Silvia Llosa, Karen Moubarak,Carolin Schärpf and Parsa Zarian.

© United Nations, 2006.

Acknowledgements

1 In larger freedom: towards development, security and human rights for all, Report of the Secretary-General, 2005,http://www.un.org/largerfreedom/

iv

v

If an effective tsunami early warning system had been in place in the Indian Ocean region on 26December 2004, thousands of lives would have been saved. The same stark lesson can be drawn fromother disasters that have killed tens of thousands of people in the past few years. Effective earlywarning systems not only save lives but also help protect livelihoods and national development gains.Over the last thirty years, deaths from disasters have been declining2 , in part thanks to the role of earlywarning systems and associated preparedness and response systems.

To be effective, early warning systems must be people-centred and must integrate four elements - (i)knowledge of the risks faced; (ii) technical monitoring and warning service; (iii) dissemination ofmeaningful warnings to those at risk; and (iv) public awareness and preparedness to act. Failure in anyone of these elements can mean failure of the whole early warning system.

With a view to establish a “worldwide early warning system for all natural hazards building on existingnational and regional capacity”, United Nations Secretary-General Mr. Kofi Annan requested inMarch 2005 that a global survey of capacities and gaps for early warning systems be undertaken3 .The present report synthesises the findings of this survey, which was carried out by the ISDRsecretariat in collaboration with a multi-party working group established at the 11th session of theInter-Agency Task Force on Disaster Reduction (IATF/DR) in May 2005.

Information for the survey was gathered from existing sources, including from reports submitted by122 countries for the World Conference on Disaster Reduction, in 2005, regional reports prepared forthe Second International Conference on Early Warning, in 2003 and ISDR publications. Inputsobtained specifically for the survey included updated information from 23 countries and 20international agencies and early warning system reports and surveys undertaken by other agencies.

The survey finds that considerable progress has been made in developing the knowledge and technicaltools required to assess risks and to generate and communicate predictions and warnings, particularlyas a result of growing scientific understanding and the use of modern information and communicationtechnologies. Early warning system technologies are now available for almost all types of hazards andare in operation in at least some parts of the world.

However, the experiences of the Indian Ocean tsunami, the hurricanes in the Gulf of Mexico andmany other recent events such as heat waves, droughts, famine, wildfires, floods and mudflows, pointto significant inadequacies in existing early warning systems. In many cases, especially in developingcountries, warning systems lack the basic capacities of equipment, skills and resources. Systems forsome hazards, such as tsunamis and landslides, are often absent. The survey concludes that there are

Executive Summary

2 Centre for Research on the Epidemiology of Disasters (CRED), “Thirty Years of Natural Disasters 1974-2003: The Numbers”, PressesUniversitaires de Louvain, 2004.

3 In Larger Freedom: towards development, security and human rights for all (A/59/2005, paragraph 66), http://www.un.org/largerfreedom.

vi

many gaps and shortcomings and the world is far from having the global system for all hazards and all communities that theUnited Nations Secretary-General first called for in January 2005.

Progress on each of the above four components of people-centred early warning systems is mixed. Even where the capabilityexists to reliably generate and issue warnings, the other three components are too often absent or weak. Among bothdeveloped and developing nations, the weakest elements concern warning dissemination and preparedness to act. Warningsmay fail to reach those who must take action and may not be understood or address their concerns. Root causes appear to beinadequate political commitment, weak coordination among the various actors and lacks of public awareness and publicparticipation in the development and operation of early warning systems.

However, and more positively, there are many and great capacities and strengths already available upon which a trulyeffective globally comprehensive early warning system can be built – not as a monolithic centralised system, but as a networkof systems, drawing on the expertise and technical capacities of the different hazard fields and the knowledge and insight ofrelevant social and economic fields. Moreover, what needs to be done to address the shortcomings is not a mystery, but hasbeen already laid out in general terms in a succession of documents and meetings over the last decade, such as theInternational Conference on Early Warning Systems (EWC’98), Potsdam, Germany, September 1998; Second InternationalConference on Early Warning (EWC II), Bonn, Germany, October 2003; and the Third International Conference on EarlyWarning (EWC III), Bonn, Germany, March 2006 (see www.unisdr-earlywarning.org).

The survey makes five main recommendations, as follows:

1. DDevelop aa gglobally ccomprehensive eearly wwarning ssystem, rrooted iin eexisting eearly wwarning ssystems aand ccapacities A global early warning system will require long-term sustained action by diverse players, strong political commitment toengender public action and to make early warning a core task of national policy and disaster risk reduction strategy,strong international support and coordination, with clear roles and responsibilities and wide participation of NGO,private sector and regional organisations.

2. BBuild nnational ppeople-ccentred eearly wwarning ssystems Country-based early warning systems are needed for the protection of citizens and also provide the building blocks of theglobal early warning system. The recommendation includes calls for a national multi-party roundtable on early warning,a national plan based on a survey of capabilities, a warning dissemination strategy, community-based approaches, publiceducation and mock exercises.

3. FFill tthe mmain ggaps iin gglobal eearly wwarning ccapacities The recommendation highlights gaps and opportunities that deserve immediate concerted action, including for tropicalcyclones, floods and tsunamis for the most ill-protected populations, agreements and networking for drought, foodsecurity and wildland fire, a global survey and mobile monitoring facility for volcanoes and a major early warning projectin each least-developed country.

4. SStrengthen tthe sscientific aand ddata ffoundations ffor eearly wwarning The scientific and technical recommendation seeks action on a long-term global data plan, upgradedtelecommunications, an agreement on basin-wide data exchange for floods, a pan-African project to fill major data gaps,improved hazard and vulnerability mapping, an early warning science and technology agenda and an internet portal fornatural hazards, risks and warnings.

5. DDevelop tthe iinstitutional ffoundations ffor aa gglobal eearly wwarning ssystem This recommendation addresses the needs for underpinning mechanisms of international and regional governance,coordination and support, starting with a call for the UN system to affirm the goal of a global early warning system andincluding requests for various UN and international agencies in technical, humanitarian and development fields toundertake specific governance and support roles.

vii

Background to the Survey and to Early Warning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

1.1 Survey Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11.2 People-Centred Early Warning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3 Early Warning - An Evolving Agenda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Effectiveness of Early Warning Components: Capacities and Gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.1 Risk Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72.1.1 Practice and Capacities in Risk Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . 72.1.2 Major Gaps in Risk Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.2 Monitoring and Warning Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.2.1 Practice and Capacities in Monitoring and Warning Services . . . . . . . . . . . 92.2.2 Major Gaps in Monitoring and Warning Service . . . . . . . . . . . . . . . . . . . . . 16

2.3 Dissemination and Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172.3.1 Practice and Capacities in Dissemination and Communication . . . . . . . . . . 172.3.2 Major Gaps and Challenges in Dissemination and Communication . . . . . . 19

2.4 Response Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212.4.1 Practice and Capacities in Response Capability . . . . . . . . . . . . . . . . . . . . . . . 212.4.2 Gaps and Challenges in Respect to Response Capability . . . . . . . . . . . . . . . 22

2.5 Cross-Cutting Issues and Gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Towards a Comprehensive Global Early Warning System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.1 Overall Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253.2 Recommendations for Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Acronyms List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31Annex I: Methodology for the Global Survey of Early Warning Systems . . . . . . . . . . . . . . . 33Annex II: Matrix of International Organisations Involved in Early Warning Systems . . . . . 35

Contents

1

2

3

1

1.1 SSurvey BBackground

In his 2005 report on the implementation of theMillennium Declaration, In Larger Freedom:towards development, security and human rights forall, United Nations Secretary-General Mr. KofiAnnan recommended “the establishment of aworldwide early warning system for all naturalhazards, building on existing national andregional capacity” to complement broader disasterpreparedness and mitigation activities(A/59/2005, paragraph 66). To assist in thesystem’s establishment, Mr. Annan requested theISDR secretariat to coordinate a survey ofexisting capacities and gaps in early warningsystems, in cooperation with all United Nationssystem entities concerned.

The survey and the preparation of the report wereundertaken by the ISDR secretariat incollaboration with the Working Group on theGlobal Survey of Early Warning Systems, whichwas established during the 11th session of theInter-Agency Task Force on Disaster Reduction(IATF/DR) in May 2005, to provide guidance

Background to the Survey andto Early Warning Systems

for the effective implementation of thesurvey and particularly to incorporateinformation, knowledge and contributionsfrom IATF/DR member agencies. TheWorking Group was co-chaired by WMOand OCHA.

Information for the survey was gatheredfrom existing sources, as recommended bythe IATF/DR Working Group. Suchsources included reports submitted bycountries to the ISDR secretariat for theWorld Conference on Disaster Reductionand regional reports prepared for the SecondInternational Conference on Early Warning.Information gathered specifically for thesurvey included updated information from23 countries and 20 international agenciesinvolved in early warning systems. Otherinputs included ISDR published materials,early warning system reports and surveysundertaken by other agencies such as theCommonwealth Secretariat and SOPAC.Annex I provides a full description ofsources and methods for the survey.

1

2

Global Survey of Early Warning Systems

The survey report is concerned with early warning systemsfor natural hazards4 . These include hydrometeorologicalhazards, such as floods, mudflows, tropical cyclones,storms, extreme temperatures and droughts; geologicalhazards, such as earthquakes, tsunamis, volcanic activityand emissions and mass movements including landslides; aswell as biological hazards, such as epidemics, plant andanimal contagion and infestations. Environmentaldegradation processes that contribute to an increase invulnerability and frequency and intensity of naturalhazards, such as desertification and wildland fires, areconsidered and also near-earth space objects. ISDR definesearly warning as “the provision of timely and effectiveinformation, through identified institutions, that allowsindividuals exposed to a hazard to take action to avoid orreduce their risk and prepare for effective response”.

1.2 PPeople-CCentred EEarly WWarning SSystems

The objective of people-centred early warning systems is toempower individuals and communities threatened byhazards to act in sufficient time and in anappropriate manner so as to reduce thepossibility of personal injury, loss of life,damage to property and the environment andloss of livelihoods.

A complete and effective early warning systemcomprises four inter-related elements: riskknowledge, monitoring and warning service,dissemination and communication and responsecapability (see Figure 1). A weakness or failurein any one part could result in failure of thewhole system.

Risk knowledgeRisks arise from the combination of the hazardsand the vulnerabilities to hazards that arepresent. Assessments of risk require systematiccollection and analysis of data and should takeinto account the dynamics and variability ofhazards and vulnerabilities that arise fromprocesses such as urbanisation, rural land-usechange, environmental degradation and climatechange. Risk assessments and risk maps help tomotivate people, prioritise early warning systemneeds and guide preparations for response anddisaster prevention activities.

Monitoring and warning serviceWarning services lie at the core of the system. They musthave a sound scientific basis for predicting and forecastingand must reliably operate twenty-four hours a day.Continuous monitoring of hazard parameters andprecursors is necessary to generate accurate warnings in atimely fashion. Warning services for the different hazardsshould be coordinated where possible to gain the benefit ofshared institutional, procedural and communicationnetworks.

Dissemination and communicationWarnings must get to those at risk. For people tounderstand warnings, they must contain clear, usefulinformation that enables proper responses. Regional,national and community-level communication channels andtools must be pre-identified and one authoritative voiceestablished. The use of multiple communication channels isnecessary to ensure everyone is reached and to avoid thefailure of any one channel, as well as to reinforce thewarning message.

4 See http://www.unisdr.org/eng/library/lib-terminology-eng.htm for definition of key terms. ISDR defines a hazard as “a potentially damaging physical event, phenomenon orhuman activity that may cause the loss of life or injury, property damage, social and economic disruption or environmental degradation, and vulnerability as “the conditionsdetermined by physical, social, economic, and environmental factors or processes, which increase the susceptibility of a community to the impact of hazards.” Together,hazards and vulnerability give rise to risk: “The probability of harmful consequences, or expected losses (deaths, injuries, property, livelihoods, economic activity disruptedor environment damaged) resulting from interactions between natural or human-induced hazards and vulnerable conditions.”

Figure 11: The ffour eelements oof ppeople-ccentred eearly wwarning ssystems

Source: ISDR Platform for the Promotion of Early Warning.

1

3

Disaster context and background to the survey

Response capability Communities must also respect the warning service andknow how to react to warnings. This requires systematiceducation and preparedness programmes led by disastermanagement authorities. It is essential that disastermanagement plans are in place and are well practisedand tested. The community should be well informed onoptions for safe behaviour and on means to avoiddamage and loss of property.

Strong inter-linkages are required between all of theelements, underpinned by effective governance andinstitutional arrangements, including goodcommunication practices. This requires the involvementof a broad range of actors, some not traditionallyidentified as dealing with the subject. It also requireslinking the subject, perceived as predominantlytechnical, to sustainable development and communitydevelopment agendas and to disaster risk reductionagendas.

An important context to this discussion is the recentinternational agreement to take concrete action toreduce disasters, namely the Hyogo Framework forAction 2005-2015: Building the Resilience of Nationsand Communities to Disasters, which was concluded bynegotiation among states and organisations at theWorld Conference on Disaster Reduction (WCDR)in Kobe, Hyogo, Japan, 18-22 January 20055 . Itsimplementation is now the core goal in disasterreduction for the United Nations and the ISDRsystem. Risk assessment and early warning form oneof the Framework’s five areas for priority action, witha focus on developing early warning systems that arecentred on the needs of people.

Roles aand rresponsibilitiesDeveloping and implementing an effective early warningsystem requires the contribution and coordination of awide range of individuals and institutions. Each has aparticular function for which it should be responsibleand accountable.

Communities, particularly those most vulnerable, arecentral to people-centred early warning systems. Theirinput to system design and their ability to respondultimately determines the extent of risk associated withnatural hazards. They should be aware of the hazardsand the related effects to which they are exposed and be

able to take specific actions to minimise the threat of loss ordamage.

Local ggovernments usually have direct responsibilities forcitizen safety and considerable knowledge of the hazards towhich their communities are exposed. They must beactively involved in the design and maintenance of earlywarning systems and understand advisory informationreceived to be able to advise, instruct or engage the localpopulation in a manner that increases their safety andreduces the possible loss of resources on which thecommunity depends.

National ggovernments are responsible for policies andframeworks that facilitate early warning and usually also forthe technical systems for preparing and issuing timelywarnings. They have responsibility to ensure that warningsand related responses address all of the population,particularly the most vulnerable. They also provide supportto local governments and communities to develop theiroperational capabilities and to translate early warningknowledge into local risk reduction practices.

Regional iinstitutions aand oorganisations provide specialisedknowledge and advice in support of national efforts todevelop or sustain operational capabilities of countries thatshare a common geographical environment. Regionalorganisations are crucial to linking international capabilitiesto the particular needs of individual countries and infacilitating effective early warning practices among adjacentcountries.

International bbodies provide support for national earlywarning activities and foster the exchange of data andknowledge between individual countries6 . Support mayinclude the provision of advisory information, technicalcooperation and policy and organisational support necessaryto ensure the development and operational capabilities ofnational authorities or agencies responsible for earlywarning practice.

Non-ggovernmental oorganisations (NGOs) includingvolunteers involving organisations play a critical role inraising awareness among individuals and organisationsinvolved in early warning and in the implementation ofearly warning systems, particularly at the community level.In addition, they play an important advocacy role to helpensure that early warning stays on the agenda ofgovernment policy makers.

5 The Hyogo Framework and the full report of the WCDR are available in several languages at http://www.unisdr.org/eng/hfa/hfa.htm6 Annex II provides a partial listing of organisations active at the international level.

4

Global Survey of Early Warning Systems

The pprivate ssector has a diverse role to play in earlywarning, including developing early warning capabilitiesin their own organisations. The media plays an importantrole in improving the disaster consciousness of the generalpopulation and by disseminating early warnings. Inaddition, the private sector has a large untapped potentialto help provide skilled services in the form of technicalmanpower, know-how or donations (in-kind and cash) ofgoods or services, especially for the communication,dissemination and response elements of early warning.

The sscience ccommunity has a central and critical role inproviding specialised scientific and technical input toassist governments and communities in developingeffective early warning systems. Their expertise isfundamental to analysing natural hazard risks facingcommunities, supporting the design of scientific andsystematic monitoring and warning services, supportingdata exchange, translating scientific or technicalinformation to comprehensible messages and to thedissemination of understandable warnings to those at risk.

Natural hazard monitoring and forecasting are carried outby specialised scientific agencies. At national level,government agencies, research organisations anduniversities are usually the main parties involved. At theinternational level, specialised agencies of the UnitedNations work extensively with relevant national agenciesto coordinate the development of technical capacities formonitoring, detecting and warnings for a wide range ofhazards and their impacts.

The responsibility for organising appropriate responses towarnings usually lies with national disaster managementorganisations. In developing countries, significant rolesare played by emergency response agencies, emergencycommittees and organisations, NGOs and UN agencies.Integral parts of this task are the preparation of hazardand vulnerability maps and disaster preparednessstrategies. Often there are specific political responsibilitiesfor ensuring that warnings are issued and acted upon andthat evacuation plans are effective.

Links tto ddisaster rrisk rreduction aand ddevelopmentEarly warning systems are recognised in the HyogoFramework as an important element of disaster riskreduction and hence to the achievement of sustainabledevelopment and sustainable livelihoods. Disasteroccurrences and impacts (economic losses and number of

people affected) are increasing, mainly owing to anincrease in the size and vulnerability of exposedpopulations, but also possibly to increases in the frequencyand severity of certain hydrometeorological hazards,perhaps as a result of climate change. Globally, naturaldisasters affected an average of 200 million people andclaimed 62,000 lives annually during the 1990s. Theeconomic cost of natural disasters rose 14 fold since the1950s, reaching USD 1 trillion in the last 15 years withsuch losses doubling every 10 years7 . Economic lossesfrom disasters can greatly set back hard-won developmentgains, particularly among the poorest countries.

Regionally, the Asia-Pacific region is the most afflicted bydisasters, particularly by flood-related disasters thatcaused estimated USD 110 billion damages over the lastdecade8 . Other major hazards include cyclones,earthquakes, landslides, tsunamis, droughts, mudflowsand tornadoes. But perhaps the situation is graver inAfrica: it is the only continent whose share of reporteddisasters in the world increased over the last decade, as analready low level of development progress is furtherhampered by major disasters of epidemic, flood, landdegradation and drought. Among small island states,natural disasters, especially from hydrometeorological andgeological sources, threaten the very existence of nations.Floods, windstorms, heat waves and landslides dominatein Europe, while the major natural disasters in theAmericas and Caribbean regions include tropical cyclonesand storm surges, floods, windstorms, tornados, wildfires,earthquakes, volcanic activity and landslides.

Early warning systems need to be accompanied by effortsto reduce the fundamental sources of vulnerability. In theUNDP report Reducing Disaster Risk: A Challenge forDevelopment, developing countries tend to have highervulnerabilities, which are aggravated by poverty,environmental degradation, population growth anddisplacement, urban growth, conflicts and weakinstitutional capacities. Such social vulnerabilities as wellas emerging climate variability and change are also threatsto developed countries. The recent cases of HurricaneKatrina in the US in 2005 and Hanshin-AwajiEarthquake in Japan in 1994 showed that developedcountries are also at risk of large disasters.

Early warning helps to reduce economic losses byallowing people to better protect their assets andlivelihoods. For example, they can safeguard homes, sell

7 Source: Centre for Research on the Epidemiology of Disasters (CRED) Emergency Events Database (EM-DAT), http://www.em-dat.net/.8 Source: See CRED EM-DAT, http://www.em-dat.net/

1

5

Disaster context and background to the survey

livestock or select appropriate crops for a drought, thuslimiting not only the immediate impact of a disaster butalso the knock-on effects on assets that can reduceeconomic well being and increase poverty. Early warninginformation allows people to make decisions thatcontribute to their own economic self-sufficiency and theircountries’ sustainable development. If well integrated intoa systematic framework of risk reduction, early warningsystems can provide many development benefits.

The establishment of people-centred early warningsystems and associated preparedness and response systemshave been an important contributor to the progressivereduction in the number of deaths from disasters over thelast several decades9 . This applies to drought and famine-affected regions, as well as for developed countries whereearly warning systems and preparedness, mitigation andrisk-transfer measures are generally well developed.

Early warning systems also advance the development andthe application of scientific knowledge, includingimproved science and technology informationdissemination. They contribute to the creation of dataarchives and information bases that are essential tomedium and long-term economic development planningand decision-making, such as in urban development,transport infrastructure and agriculture and watermanagement and to research on hazards and warningsystems. People-centred early warning systems enhancecommunity capacities through participation processes,public-private partnerships and recognition of traditionalknowledge and values.

Parallel to developments in early warning for naturaldisasters, early warning systems for human-made complexpolitical emergencies have gained ground in the pastyears. The two types of disasters are interlinked. Theexistence of a political crisis or armed conflict in a countrywill often indirectly amplify the impact of a naturaldisaster by exhausting coping mechanisms and responsecapacities. The occurrence of natural disasters amidcomplex political emergencies is a widespreadphenomenon: over 140 natural disasters have occurred incountries with complex political emergencies in the pastfive years10 . Early warning analysis, preparedness andresponse measures related to natural disasters must takeinto account the socio-political environment and theresponse capacity of local populations and institutions.

1.3 EEarly WWarning - AAn EEvolving AAgenda

One of the most fundamental and widely accepted rolesof governments through the ages has been the protectionof their people from external threats including thosederiving from natural hazards. Over the past century,many countries have developed and integrated earlywarning capabilities, in particular for meteorological andhydrological events, as an important tool for avertingdisasters. The need for international coordination ofthese efforts is well recognised and is generally providedthrough specialised United Nations agencies responsiblefor geological hazards, food security, environmentalprotection and humanitarian response.

Effective early warning systems for natural hazards arenow increasingly perceived as an integral component ofdisaster risk reduction programmes - involving a broadspectrum of actors. The International Decade for NaturalDisaster Reduction (IDNDR, 1990-1999) promoted thisconcept and worked to raise the profile of early warningaccordingly, resulting in the acknowledgement of itscrucial importance in the Yokohama Strategy for a SaferWorld, endorsed at the World Conference on NaturalDisaster Reduction in 1994. The International Strategyfor Disaster Reduction, the successor to the IDNDR, hasintroduced a stronger focus on vulnerabilities and hasemphasised the need to integrate disaster risk reductioninto sustainable development.

The World Conference on Disaster Reduction11 adoptedthe Hyogo Framework for Action 2005-2015: Building theresilience of nations and communities to disasters in whichrisk assessment and early warning is one of the fivethemes of disaster reduction. Specific recommendationsinclude the call for countries to develop people-centredearly warning systems.

Other relevant development frameworks such as theAgenda 21, the multilateral environmental agreements,the Barbados Plan of Action for Small IslandDeveloping States and the Johannesburg Plan ofImplementation have called for actions to expand,deepen and strengthen local, national and internationalinitiatives to develop early warning in particular anddisaster reduction in general, as critical tools forpromoting sustainable development and povertyreduction.

9 CRED, “Thirty Years of Natural Disasters 1974-2003: The Numbers”, Presses Universitaires de Louvain, 2004. According to CRED, in the second half of the 1970s, theabsence of catastrophic famines reduced by half the number of deaths.

10 “Natural Disasters in Complex Political Emergencies”, a report on a DfID-funded seminar hosted by the British Red Cross.11 See Footnote 3.

6

Global Survey of Early Warning Systems

Three international conferences on early warning, in 1998,2003 and 200612 produced a set of internationally agreedguiding principles for effective early warning systems aswell as the outline of an international programme on earlywarning to reduce disasters13 . The conferences addressedtechnical considerations, strategic issues and institutionalrequirements and made specific recommendations for

strengthening early warning systems, including theincorporation of early warning into policy anddevelopment frameworks, a greater emphasis on the sociafactors in early warning systems and mechanismssustaining dialogue and collaborative action among keystakeholders.

12 International Conference on Early Warning Systems for the Reduction of Natural Disaster (EWC'98), Potsdam, Germany, September 1998. Second InternationalConference on Early Warning (EWC-II) - 'integrating natural disaster early warning into public policy', Bonn, Germany, October 2003. The 3rd International Conference onEarly Warning (EWC-III) is to be held in Bonn, Germany. 27-29 March 2006. See www.unisdr-earlywarning.org and www.ewc3.org.

13 The International Early Warning Programme entitled "Effective Early Warning to Reduce Disasters: The Need for More Coherent International Action" was launched at theWorld Conference on Disaster Reduction.

2

7

2.1 RRisk KKnowledge

2.1.1 PPractice aand CCapacities iin RRisk KKnowledge

The development of effective warnings dependson the generation of accurate risk scenariosshowing the potential impacts of hazards onvulnerable groups. Authorities of WarningCentres need to define acceptable levels of risk tocommunities ditto determine whether and whento warn. Making this determination requirescapabilities to analyse not only the hazards, butalso the vulnerabilities to the hazards and theconsequential risks.

Risk assessment is a recent practice in disasterrisk reduction in many countries. Assessmenttechniques are well developed for some hazards,such as floods, earthquakes and volcanic hazardsand can be applied by using standard computersand readily available software. In addition, theapplication of geographic information systems hasexpanded the possibilities for risk assessment ofsingle and multiple hazards. The characteristics ofrisk are usually presented through risk mapping,frequency distributions, scenario plans andexercises, annualised risk mapping and qualitative

measures. The use of risk maps is the mostwidespread.

Many countries, especially high-incomecountries, develop risk maps for selectedgeographical areas and hazards. Harmonisedrisk maps of national coverage are rare, butparticular Europe, Japan and the UnitedStates develop more and more harmonisedhazard maps.

Risk assessment should also consider thecumulative effects of multiple hazards andrelated vulnerability. At present, there arerelatively few truly comprehensive multi-hazard assessments including all potentiallydamaging natural hazards in a givenlocation. A few countries, including Turkeyand Montserrat, have developed multi-hazard maps or expect to have total nationalcoverage of hazard maps as planned byAustria by 2008. In Switzerland, multi-hazard assessments are a requirement for allcantons. The general lack of multi-hazardassessments arises partly because thepreparation of hazard maps is rarely a legalrequirement.

Effectiveness of Early Warning Components:Capacities and Gaps Section 2 describes the existing capacities and gaps revealed by the survey, summarised for each of the fourelements of early warning systems. The details of each hazard are captured in specific subsections in the sectionon monitoring and warning services.

8

Global Survey of Early Warning Systems

National capacities for generating knowledge on hazardsare determined by country capabilities in science,technology and research and the availability andsustainability of observation networks. Such capacities aregenerally rather limited in developing countries. Countriesneed to have access to high-quality data on the magnitude,duration, location and arrival time of hazard events and tobe able to extract information on hazard frequency andseverity from observational data sets which requires:

• On-going, systematic and consistent observations ofhazard-relevant parameters;

• Quality assurance and proper archiving of the data intotemporally and geographically referenced andconsistently catalogued observational data sets;

• Capacities to locate and retrieve needed data and tofreely disseminate data to public users;

• Sufficient dedicated resources to support these activities.

In most countries, institutional arrangements for riskassessment are ad hoc and tend to be developed only after adisaster occurs. Some countries have specialised researchinstitutions with the capacity and resources to undertakespecific risk assessments. National meteorological andhydrological services usually can provide detailedinformation, including maps and frequency distributions,on weather related hazards such as extreme temperatures,heavy rainfall, storms and cyclone occurrence.

Risk mapping appears to be least advanced in Africa whereonly a few countries have prepared risk maps, mainly ondrought and floods. The information base published by theLand Degradation Assessment in Drylands programme hasproduced and compiled a number of vulnerability mapscovering risk and desertification in the developing world14 .

International capacities to support risk assessment arelargely limited to the hazard component. Major globalhazards have been identified and generally welldocumented through research publications, based onscientific observation programmes and through hazardimpact assessments. Scientific knowledge of hazards andthe technological means of managing them have expandedgreatly over the last three decades. The role of UNagencies such as UNESCO, WMO, UNEP, UNDP andinternational scientific organisations, such as theInternational Association of Seismology and Physics of theEarth’s Interior (IASPEI) and the InternationalAssociation of Volcanology and Chemistry of the Earth’sInterior (IAVCEI), can be noted in this respect. IFRCpromotes through its many National Societies vulnerabilityand capacity assessments as a basic process to identify the

strengths and weaknesses of households, communities,institutions and nations to a specific hazard. A number ofregional and national organisations concerned with hazardsand vulnerability mitigation are also very activeinternationally.

To facilitate global disaster reduction efforts, it is importantto identify countries at risk based on composite indices ofdisaster risk. Approaches to measuring country disaster riskinclude the World Bank supported Natural HazardApparent Vulnerability Indicator (NHAVI) forbenchmarking countries on a global vulnerability indexscale and the UNDP Disaster Risk Index, which indexescountries for each hazard according to their degree ofexposure, relative vulnerability and risk.

International capacities to collect, document and analysedisaster data and trends include the Global EmergencyEvents Database (EM-DAT) supported by the Centre forResearch on the Epidemiology of Disaster (CRED) andthe World Health Organisation (WHO), the NatCatservice of Munich Reinsurance, ReliefWeb operated byOCHA and the multi-partner Global Disaster InformationNetwork (GDIN). Several organisations have adopted theGlobal Disaster Identifier Number (GLIDE), whichidentifies each disaster with a unique number. Hazardspecific networks include the Global SeismographicNetwork that facilitates global detection of significantearthquakes and the PREVIEW Global EarthquakesDataset developed by UNEP.

2.1.2 MMajor GGaps iin RRisk KKnowledge

Inadequate eemphasis oon ssocial, eeconomic aand eenvironmentalvulnerabilityRisk assessment has been predominantly concerned withhazards, for which there are relatively good data resourcesand considerable progress and has been inadequate for thesocial, economic and environmental factors that increasevulnerability. Social science data can be difficult to obtainand the data that is available remains underutilised forvarious reasons. More work is needed to make qualitativedata and information about vulnerability accessible anduseable to engineers, planners and policy makers.

Partly as a consequence, populations are often unaware oftheir vulnerability to specific hazards and how theirvulnerability is changing and influenced by policy andpractices such as environmental degradation orurbanisation. However, some early warning systems arestarting to integrate vulnerability analysis and monitoring, a

14 Please see www.fao.org/ag/agl/agll/drylands/mapsglobal.htm

9

Effectiveness of Early Warning Components: Capacities and Gaps 2good example being the famine and food security earlywarning systems that have introduced an emphasis onvulnerability analysis and livelihood sustainability. Somesystems are also integrating information on drought, flood,desertification, famine and food security to provide aknowledge basis for decision making. NGOs and volunteerinvolving organisations have taken an important role in thisregard.

Need ffor iinclusiveness aand ppublic pparticipation iin vvulnerabilityassessmentIn addition to gathering statistics and mapping populations’risk factors, risk assessments should involve the communityto ascertain their perceived risks and concerns, as well asexisting coping strategies. People-centred assessmentensures that actions taken by authorities, aid organisationsand communities will be more relevant to people’s needsand available resources. Moreover, participatory riskassessment allows for the formal integration of valuabletraditional knowledge into risk assessments and earlywarning systems.

Data ggapsAlthough significant progress has been made in somecountries and long historical records do exist in many cases,particularly for hazards, in others data is scarce and thereare significant variations in data quality. Inconsistencies inthe historical records across national boundaries and overtime are common. In many cases data is accessible in non-digital paper form only.

At the national level, the main challenges include:

• Establishing and maintaining observing systems anddata management systems;

• Maintaining archives, including quality control anddigitisation of historical data;

• Obtaining systematic social and environmental data forvulnerability analysis;

• Securing institutional mandates for collection andanalysis of vulnerability data.

Difficulty iin aaccessing iinformationIt is human nature to dislike being considered vulnerableand monitored, which makes it difficult to collect accuratedata, particularly for variables such as health. Moreover, insome countries, as a result of tightened security,information is increasingly restricted. A particular issue isthe unwillingness of some countries to share hazardinformation with neighbouring countries that are exposedto these hazards, for example rainfall and flood data formajor river basins. International and regional policies andagreements are needed to deal with this problem.

Some countries appear to be recognising the importance ofinvesting in hazard monitoring networks and data systemsas a national resource, as an investment towards enhancedrisk management and socioeconomic development indisaster prone countries. Unfortunately, data is oftenrestricted by national or institutional policies that seek togenerate revenue from the sale of publicly funded dataresources. In the meteorological community, the basis forthe exchange of data is set out in Resolution 40 (WMOCongress-XII, May-June 1995, “WMO Policy andPractice for the Exchange of Meteorological and RelatedData and Products including Guidelines on Relationshipsin Commercial Meteorological Activities”).

Lack oof eearly wwarning iindicatorsRisk assessment practitioners require indicators that areinternationally agreed and locally referenced to measuresuccess and failure of early warning systems and thusimprove the basis for collecting and analysing risk data.

Danger oof ssocietal mmemory lloss aabout hhazardsIn parts of the world threatened with the loss of a wholegeneration through HIV/AIDS, there is real danger inlosing societal memory of past hazards, particularly forinfrequent hazards, such as tsunamis. In addition to losingknowledge of hazards, young communities face losingknowledge about how to reduce vulnerability and how torespond to warnings. Disaster museums, such as theearthquake museum in Kobe, Japan and annual warning-preparedness rehearsals can help maintain awareness.

2.2 MMonitoring aand WWarning SServices

2.2.1 PPractice aand CCapacities iin MMonitoring aand WWarningServices

There have been marked improvements in the quality,timeliness and lead time of hazard warnings, largely drivenby scientific and technological advances, particularly incomputer systems and communications technology. Therehave been continuous improvements in the accuracy andreliability of monitoring instrumentation and in integratedobservation networks particularly through the use of remotesensing techniques. In turn these have supported researchon hazard phenomena, modelling and forecasting methodsand warning systems.

However, capacities in monitoring and prediction ofhazards vary considerably by hazard and by country andeven within countries depending on prevailingsocioeconomic and political conditions. An overview ofcurrent capacities for the main hazards follows:

10

Global Survey of Early Warning Systems

Hydrometeorological hhazards

Disasters of hydrometeorological origin cover a wide rangeof atmospheric, hydrological and oceanographic phenomenaand comprise the majority of disasters. Certain types ofextreme events, notably droughts and high rainfall, areprojected by the Intergovernmental Panel on ClimateChange to increase over time due to the impact of climatechange. Reflecting this wide range, early warning systemsexist for various hydrometeorological hazards but theirstatus and stage of development vary across hazards andregions.

Capacities to monitor and predict hydrometeorologicalhazards are relatively more developed than for other typesof hazards. National Meteorological and HydrologicalServices (NMHSs) are responsible for continuouslyobserving, monitoring, detecting, forecasting anddeveloping hazard warnings for a wide range of weather-,climate- and water-related hazards. Internationalcoordination is very well developed through the WMO andits 187 National Members, its 10 Scientific and TechnicalProgrammes, three World Meteorological Centres and 40Regional Specialised Meteorological Centres (all operatedor supported by NMHSs).

The WMO global operational network supports observing,monitoring, detecting and forecasting hazards and theissuing of early warnings for weather-, climate- and water-related hazards, such as extreme temperature, severestorms, tropical cyclones, floods and droughts. Theintegrated Global Observing System (GOS) enables thesystematic observation and collection of weather, climateand water information from around the globe, while theGlobal Telecommunications System (GTS) provides anetwork of continuously operating telecommunicationfacilities and centres connecting countries through theirNMHSs. The GTS also supports the global exchange ofinformation for non-hydrometeorological hazards,including tsunami-related ocean data, information andwarnings, seismic data and messages related to volcanic ashcloud. Finally, through this network, WMO has developeda Global Data Processing and Forecasting System(GDPFS) providing technical support, analysis, forecasts,alerts and bulletins to the NMHSs of all countries,particularly those with the least resources. Furtherstrengthening of the system, particularly for certain hazards(climate and weather related) and capacities (e.g. upgradingof the GTS where needed) could significantly enhancenational early warning capabilities.

Major challenges include: 1) availability of these technicalcapabilities in a sustainable fashion in all countries,particularly those with the least resources; 2) integration of

the existing technical capacities into the disaster riskreduction decision process in a more effective and proactivefashion; and 3) need for improvement of technical warningcapabilities for many hazards such as droughts and floods.

FloodsSome countries monitor flash floods through theirNMHSs, while others monitor flash floods and river floodsthrough environmental agencies and hydrology servicesseparately. It is not possible to forecast flash floods but theycan be detected as they occur by weather radars if these arepresent. Most flash floods occur in countries or districtswithout radar coverage. They are a major killer duringtropical cyclone events.

Dedicated systems to monitor and forecast river basinfloods are well established in developed countries, wherethey are operated by a variety of technical organisations.However, such systems are much less widespread indeveloping countries particularly in Africa, Asia and theCaribbean. In many tropical areas, such as in the IndianOcean Commission (COI) region, flood monitoring andwarning systems are closely linked to tropical cyclonewarning systems.

Operational global flood forecasts from specialised warningsystems provide three-day warnings but several initiativesare underway to extend the warning range. Most floodwarning systems are stand-alone national operations, butwarning systems have been developed covering severalinternational rivers, such as for the Rhine, Danube, Elbeand Mosel in Europe, the Mekong, Indus and Ganges-Brahmaputra-Meghna basins in Asia and the Zambezi inSouthern Africa. Globally, the Dartmouth FloodObservatory in the United States detects, maps, measuresand analyses extreme flood events worldwide. The NationalOceanographic & Atmospheric Administration (NOAA) inthe United States can provide river-flooding guidance sixmonths in advance based on seasonal forecasts andknowledge of major river catchments.

UNESCO and WMO currently coordinate an operationalflood warning system with NMHSs for river flooding. TheInternational Flood Initiative/Programme (IFI/P),launched during the World Conference on DisasterReduction in January 2005, is a joint programme ofUNESCO and WMO to be operated by the InternationalCentre on Water Hazard and Risk Management(ICHARM), which is hosted by the Public WorksResearch Institute (PWRI) in Japan. The InternationalFlood Network, through the Global Flood Alert System,provides information on precipitation based on satellite datato global subscribers for free. Such initiatives enhance theservices provided by national authorities.

11

Effectiveness of Early Warning Components: Capacities and Gaps 2Tropical ccyclonesTropical cyclones, also popularly known as hurricanes ortyphoons, are globally monitored and forecasted on a dailybasis through the WMO Global Tropical Cyclone WarningSystem. This is a global network for observations, dataexchange and regional forecasting and analysis capabilities,operated by NMHSs and includes six Regional SpecialisedMeteorological Centres (RSMCs) that provide around-the-clock forecasts, alerts and bulletins on the severity, projectpath and estimated land fall to the National MeteorologicalServices of countries at risk, which then issue the warning fortheir countries with lead times of at least 24 hours and up toseveral days. These lead times are sufficient to achieveeffective mass evacuations and thereby avoid widespread lossof life. Five Tropical Cyclone Regional Committees(comprising experts in tropical cyclone modelling andforecasting) provide regional coordination including trainingsupport.

Severe sstormsSevere storms comprise several phenomenon types includingtornadoes, hailstorms, lightning, flash flood and sand anddust storms. Tornado warning lead time is inherently veryshort and warning systems are only operational in a fewcountries at risk, such as the USA, where warning lead timeis up to 15-20 minutes. Tornado warnings are most effectivein enabling people to seek shelter when they are preceded byalerts, called “tornado watches”, which warn the communitiesat risk of the possibility of a tornado strike many hours inadvance. The number of tornado deaths significantly droppedin the USA during the last century mainly as a result of theDoppler Radar Network. Similarly, operational systems forhailstorm monitoring are limited at country level withwarning lead time of a few hours. The forecasting of severestorms is challenging as the hazards affect discrete locations.Warnings for sand and dust storms, with a lead time of up to3 days, are issued through the NMHSs but operationalwarning systems for many of these hazards are lacking inmost countries and regions.

DroughtEarly warning systems for drought are more complex thanthose for other hydrometeorological hazards and are,consequently, relatively less developed globally. They areheavily reliant on monitoring of observed patterns of monthlyand seasonal rainfall, streamflow, ground water levels, snowpack and other parameters and the use of historical statisticaldata. Global Circulation Models (GCMs) and associatedstatistical ensemble methods are being routinely used toprovide predictions of upcoming climate anomalies and offerpromise for increasingly useful forecasts of the onset, severityand duration of drought for large geographic regions onmonthly and seasonal timescales. Requirements for earlywarning range from a few weeks to several months.

Nonetheless, many countries have developed droughtearly warning systems capable of integrating informationfrom various sources and providing warnings of theimminent onset of drought. In Africa, regional centressuch as the IGAD Climate Prediction and ApplicationsCentre (ICPAC) and the Drought Monitoring Centre(DMC) in Harare, supported by WMO and regionaleconomic commissions and the Sahara and SahelianObservatory provide current data, develop climateoutlooks and issue warnings to NMHSs. They organiseRegional Climate Outlook Forums comprising national,regional and international experts to review conditionsand develop climate outlooks. User representatives fromdifferent sectors often participate in the forums. Suchforums are also organised in other regions of the world.

When there is evidence of developing El Niño SouthernOscillation (ENSO) conditions, WMO coordinates thedevelopment of a global scientific consensus, involving acollaborative process to review best available evidencesand predictions. The outcome is the El Niño Update, aunified global statement on the expected evolution ofENSO for months ahead, which is issued to NMHSsand to the world at large.

Traditional forecasting remains an important source ofclimate information in many rural communities. There isgrowing appreciation that traditional observations andoutlook methods may have scientific validity andincreased interest in harmonising traditional and modernscientific methods of climate prediction. Studies havebeen initiated in some countries, such as Zimbabwe andKenya, to gain further understanding of traditionalforecasting.

Extreme ttemperaturesExtreme temperatures, including both heat waves andcold waves, are a threat to vulnerable populations in bothdeveloping and developed countries. In many temperatecountries, cold conditions in winter are responsible forexacerbating cardiovascular, cerebrovascular, circulatoryand respiratory diseases, which lead to increased mortality,particularly amongst the elderly and those living in poorhousing or without shelter. Cold conditions affect thehuman population indirectly as well, when livestock, avital part of the human food chain, perish because of theweather. In the dzud (bitter cold and snow) of 2000 inMongolia, livestock froze by the thousands, creating greathardship and food shortages. The extent of this crisis waspartially determined by such factors as over-concentrationof stock and overgrasing of pastures in some areas.

The 2003 heat wave that affected Europe killed manyelderly residents in urban areas, and according to CRED,

12

Global Survey of Early Warning Systems

the impact of extreme temperatures has been rising (during1995-2004), especially in Europe, where temperatureextremes accounted for over 50 percent of total deaths fromdisasters. The two heat wave events during the 1990s inChicago, USA, affected mostly poor elderly people livingin the inner city.

In some parts of the world, mainly in the USA and inEurope, many cities are served by a new Heat-HealthWarning System (HHWS) that provides warnings ofthermal and other conditions (night-time humidity, windpatterns, etc.) that can lead to excess mortality. The serviceis organised as a collaboration between meteorological,health and social sector partners. Normally the health sectorissues the warning, based on the model run by themeteorological agency. The programme includes outreachand education and interventions by the health and socialsectors to follow up with the most vulnerable groups to helpachieve appropriate responses to the warnings. WMO isdeveloping guidelines to help the meteorological and healthcommunities develop such services.

Air ppollution, hhaze aand ssmokeAir pollution, haze and smoke are common in bothdeveloping and developed countries. Photochemical smogis commonly produced in high economic-activity areas by acombination of automobile use and industry. During theWestern European heat wave of 2003, 25 to 40 percent offatalities were caused by pollution that was exacerbated bythe heat. Warning lead time for the risk of air pollutiondepends on the pollutant and the prediction method andcan be about 24 to 48 hours. Regional movements of largepollution plumes such as those associated with wildfires,land-use fires and/or dust storms can be up to 3 to 5 days.In South East Asia observations reveal that smoke-hazegenerated by land-use fires regularly cover the whole regionfor many weeks or months. Haze and smoke fromvolcanoes and wildland fires are major threats to aviation,limiting visibility and leading to engine failure (seeadditional information in the following section onvolcanoes). Satellite observation has been very useful in thisrespect.

Dust aand ssandstormsDust and sandstorms are a consequence of soil erosion bywind. Warning lead times can be up to 3 days and arebased on satellite and ground observational data but thesecapabilities are not yet operational. A number ofinternational efforts including climate modelling andprediction work in China, Japan, Australia and the USAhave indicated the importance of up-to-date information onglobal climate patterns for monitoring of weather conditionsconducive to dust and sandstorms and drought (often aprecursor to dust events). There is a need for the

development of dust and sandstorm warning systems inmost countries facing this risk.

Snow aavalanches aand wwinter wweather hhazards:Winter hazard warning services are well developed inmountainous countries, such as in Switzerland where theInter-Cantonal Early Warning and Crisis InformationSystem provides comprehensive warning and responsecapability development services that were inspired bylessons from the 1999 avalanches. Winter weather, such asblizzards and ice storms, is forecasted as part of operationalweather forecasting and warning services. In large cities,blizzards and ice storms can be a major threat particularly ifthey also lead to failure of national power systems.

FamineA common outcome of drought, and sometimes of otherweather hazards, is the risk of famine or extreme foodinsecurity. Early warning systems for food security in manydeveloping countries make use of information from themajor international food security monitoring systems. TheFAO Global Information and Early Warning System onFood and Agriculture (GIEWS) is the most globallycomplete system but other systems, including the FoodInsecurity and Vulnerability Information and MappingSystems (FIVIMS), the WFP Vulnerability Analysis andMapping (VAM) system and USAID-sponsored FamineEarly Warning Systems Network (FEWS NET), are alsoimportant. FEWS NET is mainly focused on Africa,where the majority of food security-warning systemsoperate, but it also covers parts of central Asia, of CentralAmerica and of the Caribbean.

Geological hhazards

Geological hazards pose a significant threat to manycountries and communities. Great strides have been madethrough advances in satellite-based observing systems,computing and communications and fundamental scientificdiscoveries in earth science and are helping to understandthe physics of hazards and promote integrated observationand modelling of the Earth’s oceans, landmass, atmosphere,ice shield, glaciers and of the upper mantle of the Earth’sinterior. The National Aeronautics and SpaceAdministration (NASA) is partnering with several agencies,including UNESCO, United States Geological Survey(USGS), Committee on Earth Observation Satellites(CEOS) and the space agencies of Japan, Canada andEurope to develop a robust geohazards mitigationprogramme through application of these developments thatsupports the goals of the Global Earth Observing Systemof Systems (GEOSS).

13

Effectiveness of Early Warning Components: Capacities and Gaps 2While capabilities exist at the global level to identify areasof occurrence of geological hazards, there is much lesscertainty in predicting when hazardous events are likely tooccur. The signs of an impending volcanic eruption or alandslide can often be detected at an early stage and usedfor warnings, but the recognition of earthquake precursorsis difficult and routine predictions have not been possible.

EarthquakesThe world’s earthquake-prone areas and plate boundarieshave been identified and extensively studied. Regionalearthquake monitoring systems have been installed in mostearthquake-prone regions. Monitoring at the global levelstarted with the installation of the World WideStandardised Seismograph Network (WWSSN) in the late50s. Today, USGS operates the Global Seismic Networks(GSN) consisting of more than 100 stations. TheGEOFON (network of the GeoForschungsZentrumPotsdam) operates seismic stations at all continents and is amulti-national initiative.

Prediction capability for earthquakes remains elusive, however,and it must be reiterated that the location, magnitude and timeof occurrence of earthquakes cannot be forecasted. In someareas, such as Mexico City and Southern California, technicalsystems exist to identify the first seismic wave following thestart of an earthquake that may have happened 100 kilometresaway or more. With this information, critical systems such asgas supply lines can be shut down and fire station andambulance doors opened, before the most severe shakingreaches the location. The warning depends on the distancefrom the epicentre and the depth of the earthquake and the leadtime may be just a few seconds. The Japan MeteorologicalAgency (JMA) is planning to establish a network to issue“Nowcast Earthquake Information,” to provide this type ofearly warning to transportation systems. It is possible to act onan initial earthquake shock as a “warning” of possible furthershocks and consequential hazard, such as fire or buildingcollapse, for example by quickly shutting off a gas stove andmoving to a safer place. A rough estimate of the probability of atsunami can be made within 10-15 minutes when thehypocentre of the earthquake is located offshore.

TsunamisA tsunami is a series of ocean surface waves that aregenerated by submarine earthquakes, submarine volcaniceruptions and submarine or coastal landslides. Following anoffshore earthquake, a rough estimate of the probability of atsunami can be made within 15 minutes based on theestimation of the earthquake hypocentre using global andregional seismic observations. Once a tsunami is observed,through ocean observing systems, its size, movement andlikely arrival time can be calculated accurately enough forwarning purposes.

The deadly Indian Ocean tsunami of 26 December 2004reminded the world of the destructive nature of tsunamis.Although the massive damages to infrastructure werelargely unavoidable, thousands of lives might have beensaved if a tsunami early warning system had been in placein the region. Currently, there is no global tsunami earlywarning system, although an ocean-wide warning systemunder the auspices of the IntergovernmentalOceanographic Commission (IOC) of UNESCO has beenoperational in the Pacific region for more than 40 years. InJune 2005 the IOC secretariat was mandated by itsMember States to coordinate the establishment of tsunamiwarning systems for the Indian Ocean, the North EastAtlantic and Mediterranean, as well as for the CaribbeanSea. Intergovernmental Coordination Groups (ICGs) foreach of these regional systems were established in 2005.

The Pacific Tsunami Warning System (PTWS) iscomprised of the Pacific Tsunami Warning Center(PTWC) in Hawaii, the Tsunami Warning Center of theJMA and the Alaska Tsunami Warning Center, which havebeen providing ocean-wide timely alerts and warnings toIOC Pacific Member States since 1965. Due to the closeproximity of a large potential source region, JMA issuestsunami warnings within three minutes of the detection of asubmarine earthquake near Japan, based on seismicinformation only. Experimental warning systems forEurope (in Portugal, France and Greece), national systems,such as the Yuzhno-Sakhalinsk Tsunami Warning Center ofRussia and various local systems exist. The initial IndianOcean Tsunami Warning System (IOTWS) is scheduled tobe in place by July 2006 while a system for the North EastAtlantic and Mediterranean is scheduled to be set up by theend of 2007.

The PTWC, IOC’s International Tsunami InformationCentre in Hawaii and JMA are currently providing interimtsunami advisory information to Indian Ocean countries.While technical monitoring capacity for tsunami in theregion has been improved significantly during the last year,there is still a need to complete the system and to improveassociated disaster management capacities, including bothstructural and non-structural countermeasures fortsunamis. Tsunami risk assessment must be implementedregionally in tsunami endangered coastal areas, based onsolid scientific research and analysis.

During its XXIIIrd session in June 2005 the IOC GeneralAssembly adopted a resolution to establish a globalframework for tsunami and other ocean-related hazardsearly warning system, based on the four regionalintergovernmental coordination groups, existing oceanobserving and data and information exchange systems, withparticipation of related UN agencies including WMO and

14

Global Survey of Early Warning Systems

ISDR. At the present time all regional intergovernmentalcoordination groups on tsunami early warning are in place.The working group 6 on Mitigation, Preparedness andResponse was formed at the UN/ISDR Workshop onDisaster Risk Reduction: Context for Indian OceanTsunami Early Warning Systems (TEWS) which was heldin Bali, Indonesia on 30 July 2006.

VolcanoesAbout 50 to 60 terrestrial volcanoes erupt every year andabout 3,000 volcanoes worldwide are capable of futureeruptions. Prediction of the timing of volcanic eruptionshas been accomplished, but the size, duration and climax ofa volcanic eruption cannot be predicted. Examples ofsuccessful forecasts and effective evacuations include theeruptions of Mount Pinatubo in the Philippines in 1991,Vulcan and Tavurvur in Papua New Guinea in 1994,Montserrat in West Indies in 1995 and 1997, Mt. Usu andMikaye Island in Japan in 2000 and Merapi in Indonesiain 2001 and 2006. These successes from around the globehave been largely driven by extensive progress inmonitoring changes in seismic activity.

Satellite-based systems for global monitoring of volcanoeshas been established jointly by the International CivilAviation Organisation (ICAO) and WMO for the nearreal-time prediction of the risks and the movement of ash-clouds that can affect air safety, as in the case of theeruptions of the Aleutian Islands and Southern Alaska.

The capacity for monitoring volcanoes is variable globally.Some areas are well served by strong national systems (suchas USA and Japan) or cooperative regional arrangements(such as for the North West Pacific) but large areas of thedeveloping world (such as Africa, central and South Asia)lack the resources to operate seismic monitoring systems.

LandslidesThe timing of landslides can be predicted provided slopesare being monitored, which is often not the case. The 1995landslide along the Yellow River in China was predictedone day ahead of its occurrence, for example. Advances inmonitoring technology, particularly in real-timetransmission of information, have enhanced techniques forshort-term prediction, but they are only available in a few ofthe areas subject to landslide risk in the world. Nonetheless,capacities for landslides early warning are expanding. InCentral America, where countries are among the mostexposed to landslide risks, lessons from hurricane Mitchhave contributed to the development of early warningsystems for Nicaragua and El Salvador. In China, the earlywarning system for the Yangtze River valley has so farsuccessfully forecasted 217 landslides, avoiding significanteconomic losses. In Africa efforts are underway to establish

community-based landslide warning systems in Kenya andUganda.

The International Consortium on Landslides (ICL) wascreated in 2002 with support from UNESCO, WMO,FAO, UN/ISDR and the Government of Japan, as well asKyoto University. It aims at internationally coordinatedactions towards landslide risk reduction through projectimplementation. The February 2006 landslide in Leyte,Philippines, which killed more than 1800 people, is areminder of the need for further improvement of landslideearly warning systems. Landslide early warning systemsneed to take into account observations of uphillenvironmental degradation due to deforestation, land-usesystems and wildland fires.

Near-eearth oobjectsNatural rock objects (meteorites) in space are very rare buteven small objects can have powerful effects on the Earthand its oceans and atmosphere, owing to their high speedand momentum relative to the Earth. Past impacts can beassessed from geological evidence, particularly cratercharacteristics. If the location and trajectory of a distanceobject is known, it is readily possible using physics laws toaccurately calculate its path and therefore predict if andwhen it would hit the Earth. Currently a programme isunderway using telescopes to map all large objects and tocalculate their paths. To date none of the objects measuredhave been found to be on a path that is dangerous to Earth,although the trajectory of the object NEA Apophis (320metres in diameter) is reported likely to be very close in2035.

Biological hhazards

The definition of natural hazard used in the ISDR systemincludes biological hazards in addition to the morecommonly understood meteorological, hydrological andgeological hazards. In practice, the biological hazards thatare disease-related are typically handled independently bythe public health community, with rather limited interactionwith the institutions and coordination mechanismsassociated with the other natural hazards. To the degreethat this arrangement works satisfactorily, it can be acceptedwithout concern, but at the same time it should be notedthat some governments and institutions are taking a moreintegrated view of all hazards, irrespective of type, and aredeveloping policies that simultaneously address all types ofpossible risks in a single common framework.

EpidemicsEpidemics pose significant disaster risk worldwide andundermine global development and security. National

15

Effectiveness of Early Warning Components: Capacities and Gaps 2surveillance systems for human and animal epidemics andpest infestations exist globally at various stages ofdevelopment and effectiveness. Comprehensive epidemicearly warning systems are undeveloped in severaldeveloping countries, although health information systemsare in use in epidemiological monitoring and detection.

WHO coordinates international efforts to reduce the riskof transmission of infectious diseases and epidemics.WHO’s integrated disease surveillance strategy, based ondeveloping national capacities for participatory andpeople-targeted surveillance systems is being adapted inthe Africa region and applied to monitoring and alertactivities in the Eastern Mediterranean region, Europeand South East Asia. Worldwide, WHO coordinates theGlobal Outbreak Alert and Response Network(GOARN), a technical collaboration of existinginstitutions and networks, to undertake rapididentification and confirmation of epidemics. Thisnetwork is supported by databases, including the GlobalPublic Health Intelligence Network that scans globalsources for outbreak-related information. Internationalcooperation also includes the European Centre forDisease Prevention and Control, which promotescooperation among national disease control agencies andproduces the weekly Eurosurveillance journal. TheUnited States’ Centres for Disease Control detect andinvestigate health problems, as well as provide informationon epidemics and diseases around the world. In addition,the WMO works closely with WHO to developmonitoring capacities throughout NMHSs formonitoring and detecting meteorological andclimatological conditions that favour the development ofinfectious diseases and epidemics, particularly vector-borne diseases such as malaria.

The need for malaria early warning systems is mostpressing in Africa, South East Asia and South America.In Africa, a framework for designing malaria earlywarning systems has been developed and disseminated aspart of the global roll back malaria campaign. Otherinitiatives include the joint WHO-UNICEF HealthMapinitiative, which is being assessed for use in malariasurveillance and the continent-wide Mapping MalariaRisk in Africa (MARA) project for malaria early warningusing Geographical Information System (GIS), which isunderway.

Globalisation has widened the scope for early warning onthe spread of diseases with disastrous potential, such asHIV/AIDS and SARS. Countries are now on the alert forpotential outbreaks of a lethal strain of bird flu, highlypathogenic avian influenza (HPAI), particularly of theH5N1 subtype. Currently, the exact timing, infectious

potential and extent of a pandemic are not predictable. Aworldwide coordinated early warning and alert system isunder development for avian influenza. Monitoringprogrammes aimed at determining the prevalence of avianviruses in domesticated and wild flocks are beingundertaken in several countries and several countries havenational avian influenza centres. Currently, it takes severaldays to characterise avian viruses at the subtype level.Several initiatives to develop more comprehensive andstructured early warning systems are underway, includingthe Task Force on Avian Influenza and Wild Birds andproposals for cooperation between WHO, FAO, WorldBank and the World Organisation for Animal Health(OIE).

FAO provides early warning of migratory pests and otheragricultural pests and diseases through the EmergencyPrevention System (EMPRES) for Transboundary Animaland Plant Pests and Diseases.

Locust sswarmsLocust swarms affect Africa, South West Asia and regionsof Australia. Warnings are based on biological models,locust field observations and meteorological data. Warninglead times range from less than one month for an initialoutbreak to 3 to 6 months for a plague. General warningbulletins and advisories are produced mainly throughNational Locust Control Centres. At the international level,the Desert Locust Information Service of FAO preparesmedium and long-term forecasts for all countries andregions. WMO works with NMHSs to enhance theirmonitoring capacities in support of locust early warningsystems.

Environmental ddegradation

DesertificationDesertification develops slowly from mismanagement ofland and through the interaction of the natural ecosystemand the human social system. Its assessment and predictiondepend on the availability of physical, biological, social andeconomic information from different sources.Desertification is difficult to predict because of thecomplexity of the interaction of the multiple driving forcesand its long-term nature. There are still knowledge gapsand pending questions on translating broadly acceptedprinciples of early warning into action-oriented modalitiesfor UNCCD’s National and Regional Action Programs. Asof June 2005, a total of 81 countries affected bydesertification, of which 21 are in Africa, had drawn upNational Action Programmes. WMO and UNCCDcollaborate on these issues.

16

Global Survey of Early Warning Systems

Wildland ffireWildland fire early warning involves use of fire dangerrating to identify in advance critical time periods of extremefire danger. Fire danger rating (FDR) has long been usedas a tool to provide early warning of the potential forserious wildfires and can provide a 4 to 6 hour warningusing basic daily weather data. These lead times can beextended to 14 days through the use of forecasted weatherdata and up to 30 days when calibrated with local data, asdeveloped for South East Asia to indicate the potential fordisaster-level haze events from peatland fires. FDR toolsfor early warning are highly adaptable and have beenapplied by a wide range of users. Consequently, there arenumerous examples of current operational systems,particularly in most industrial countries. However, nointernationally accepted fire warning system exists butglobal and regional collaborative efforts are underway, suchas the joint efforts of the World Weather ResearchProgramme (WWRP), the Canadian Forest Service andGlobal Observation of Forest and Land Cover Dynamics(GOFC-GOLD), coordinated by the Global FireMonitoring Center (GFMC), to develop, adapt,standardise and apply fire early warning systems worldwide.

2.2.2 MMajor GGaps iin MMonitoring aand WWarning SService

While very significant progress has been made in manycountries on the technical aspects of monitoring andforecasting natural hazards, many major overall gaps exist,particularly in the developing and least developed countries.Key issues include:

• Inadequate coverage and sustainability of observingsystems for monitoring of hydro-meteorologicalhazards;

• Inadequate level of technical capabilities (resources,expertise and operational warning services) in theoperational technical agencies responsible formonitoring and forecasting of severe events, such as theNMHSs;

• Lack of systems for many hazards such as dust andsand storms, severe storms, flash floods and stormsurges, particularly for at-risk developing and least-developed countries;

• Lack of internationally negotiated data-exchangepolicies and procedures to share essential data in atimely fashion among countries for the development ofmodelling and for operational forecasting and warningsystems, such as for tsunami and earthquake;

• Inadequate access to information (forecasts andinterpreted data) from countries outside of the regionaffected;

• Insufficient multi-disciplinary, multi-agencycoordination and collaboration for improvingforecasting tools such as for storm surge and floodforecasting and for integrating warnings into thedisaster risk reduction decision processes in a moreeffective and proactive fashion;

• Inadequate communication systems to provide timely,accurate and meaningful forecasting and early warninginformation down to the level of communities.

Overall, systems exist to provide hazard forecasts andwarning against impending disasters induced byhydrometeorological hazards, but the scope of hazardcoverage at the country level is highly variable and reflectscountries’ economic development level. The globalgeographical distribution of hazard forecasting systems isuneven. Developed countries and disaster-exposed areas ofthe developing world operate more hazard forecasting andobserving systems than African countries and otherdeveloping countries with historically less disaster exposure.For many developing and least developed countries thesustainability of warning systems is a major challenge.

Effective monitoring and forecasting systems are availablefor most hazards, including for complex hazards likedrought, El Nino, food security and desertification.Significant improvements have occurred in tropical cycloneand windstorm warning systems and sub-regional floodwarning systems. Systems are less well developed fortsunami, landslides, wildland fires and volcano-relatedhazards (eruptions and lahars). No scientific basis iscurrently available for routine earthquake prediction.

Most countries have hazard monitoring and forecastingsystems for the dominant hazards that affect them, but inmany cases the warning systems do not cover all hazardsand all parts of the national territory. Good examples ofeffective forecasting systems include those for cyclones inMauritius and storm surges in Bangladesh and the newapproach to weather warning (Carte de Vigilance) inFrance. There is a need for continued development ofoperational forecasting and warning systems within theNMHSs, including providing more effective warnings ofsevere storms.

Globally, the fastest onset hazard (earthquake) and theslowest (drought) are the most difficult to predict and posesignificant challenges to development of early warningsystems worldwide. Overall, most early warning systemsfocus on hazard monitoring and forecasting, excludingvulnerability assessment. Emerging systems forenvironmental hazards, such as desertification, arebeginning to integrate the vulnerability component.

17

Effectiveness of Early Warning Components: Capacities and Gaps 2Overall, a major need across the board is the integration ofrisk information into hazard warning messages. This wouldrequire close collaboration between technical operationalagencies such as the NMHSs and national agenciesresponsible for vulnerability and risk assessment. Capacitiesfor risk assessment need to be developed at national andlocal levels, on methodologies, hazard and varioussocioeconomic data. To this end, countries should considertheir hazard observing networks as an investment andensure sustainability of these systems over time.

There is a widespread need for closer collaboration amongthe meteor