analytical-empirical model for the assessment of ... · pdf fileanalytical-empirical model for...
TRANSCRIPT
1
Analytical-Empirical Model for the Assessment of
Earthquake Casualties and Injuries in a Major City in
Israel – The Case of Tiberias
Principal Investigators:
Igal M. Shohet (coordinator), Limor Aharonson-Daniel and Tsafrir Levi
Investigators:
Robert Levy, Amos Salamon, Oren Vilnay, David Ornai,
Bruria Adini and Yaron Bar-Dayan
Research Assistants:
Hsi-Hsien Wei, Stav Shapira and Ohad Levi
Contract 3-9618
Ministry of National Infrastructures
Energy and Water Resources
Geological Survey of Israel
Report No. GSI/21/2016 Jerusalem, July 2016
Ben-Gurion University
of the Negev
Analytical-Empirical Model for the Assessment of
Earthquake Casualties and Injuries in a Major City in
Israel - The Case of Tiberias
Principal Investigators:
Igal M. Shohet (coordinator) 1, Limor Aharonson-Daniel 2,3 and Tsafrir Levi4
Investigators:
Robert Levy 1, Amos Salamon 4, Oren Vilnay 1, David Ornai 1,
Bruria Adini 2,3 and Yaron Bar-Dayan2
Research Assistants:
Hsi-Hsien Wei 1, 5, Stav Shapira 2 and Ohad Levi 1,4
1 Department of Structural Engineering, Ben-Gurion University of the Negev.
2 Department of Emergency Medicine, Recanati School for Community Health Professions,
Faculty of Health Sciences, Ben-Gurion University of the Negev.
3 PREPARED Center for Emergency Response Research, Ben-Gurion University of the Negev.
4 Geological Survey of Israel.
5 Department of Civil and Environmental Engineering, University of Maryland, College Park, U.S.A.
Contract 3-9618
Report No. GSI/21/2016 Jerusalem, July 2016
Ben-Gurion University
of the Negev
Ministry of National Infrastructures
Energy and Water Resources
Geological Survey of Israel
3
Table of Contents
Chapter 1: Introduction ..............................................................................................................8
1.1 Background ......................................................................................................................8
1.2 Objectives of the research ................................................................................................9
1.3 Structure of the report.................................................................................................... 10
1.4 Contribution of the research .......................................................................................... 11
Chapter 2: Literature Review – Risk Factors that Influence Human Vulnerability in Earthquakes and
Earthquake Casualty Assessment Models ............................................................................... 13
2.1 Agent Characteristics ................................................................................................ 16
2.2 Environmental Characteristics................................................................................... 17
2.3 Host Characteristics ................................................................................................... 17
2.4 Interaction between the host, agent an environment ................................................. 20
Chapter 3: Research Methods for Assessment of Earthquake Casualties ....................... 22
3.1 Literature reviews .......................................................................................................... 22
3.2 Mapping population distribution ................................................................................... 22
3.3 Estimating casualty rates ............................................................................................... 23
3.4 Performing a casualty estimation process for the city of Tiberias ................................ 25
Chapter 4: Casualty estimation models to predict seismic mortality and morbidity hazards 26
4.1 Common approaches for Predicting casualties in an Earthquake .................................. 26
4.2 The HAZUS-Multi-Hazard Casualty Estimation Model ............................................... 27
4.2.1 Time frames ............................................................................................................ 27
4.2.2 Population distribution ........................................................................................... 28
4.2.3 Casualty rates ......................................................................................................... 28
4.2.4 Output ..................................................................................................................... 29
4.3 Applicability of HAZUS-MH Casualty Estimation Methodology in Israel .................. 30
Chapter 5. Development of Fragility Curves .......................................................................... 31
5.1. Background .................................................................................................................. 31
Chapter 6 Risk Assessment Model ........................................................................................ 34
4
6.1 Introduction ................................................................................................................... 34
6.1.1 Hazard Module ....................................................................................................... 35
6.1.2 Inventory Module ................................................................................................... 38
6.1.3 Vulnerability Module ............................................................................................. 40
6.1.4 Loss Module ........................................................................................................... 44
6.2 The Geology and tectonics of Tiberias .................................................................... 44
6.3 Ground motion ...................................................................................................... 46
6.4 Demographic loss ................................................................................................... 47
6.5 The classification of the building.......................................................................... 48
6.6. Earthquake scenarios along the DSF system ...................................................... 51
6.7 Earthquake scenarios for testing the sensitivity of the casualties' matrix ....... 51
6.7.1 Casualties' matrix results ................................................................................ 53
6.8 Simulations results ................................................................................................ 57
6.8.1. Jordan fault scenarios ..................................................................................... 57
6.8.3. Poria fault scenarios ....................................................................................... 59
6.8.4. Almagor fault scenarios .................................................................................. 59
6.8.5. Beit HaKerem fault scenario ........................................................................... 60
Chapter 7: Discussion and Conclusion ................................................................................ 63
7.1 A survey of Earthquake casualty matrices used worldwide .......................................... 63
7.2 Tiberias’ population characteristics and distribution .................................................... 65
7.3 Estimating casualty rates – the results of the modified ‘Delphi procedure’ .................. 70
7.4 Estimating casualties – the case study of Tiberias ........................................................ 74
7.5 Discussion and Conclusions .......................................................................................... 76
Appendix ................................................................................................................................. 80
References ............................................................................................................................... 82
5
Abstract
Mitigating the results of potential earthquakes requires an estimation of the casualties that
may incur, and accordingly an appropriate response model can be developed. Based on an
extensive literature review of the consequences of earthquakes, the following parameters were
identified as significant in estimating human casualties:
1. The earthquake hazard in the designated area, such as active faults, liquefactions, landslides
and ground motion amplifications;
2. Building and structural vulnerabilities to seismic hazards, assessed by an empirical or
analytical approach combining simulation of seismic events with the data bases of structures,
geological data, seismic data;
3. Assessment of the socio-economic conditions in the designated area, in order to estimate the
potential population that was present in different building types at the time of the shake-out.
Israel is situated in an area prone to earthquakes. To present, the estimation of casualties that
might be incurred during an earthquake in the region has not taken into consideration the
different types of buildings and the presence of the population in the different structures at
different time-frames.
The research employs the semi-empirical approach to estimate the surviving casualties and
fatalities that might be caused by a potential earthquake. The city of Tiberius and its
surroundings are used as the case-study to develop the model that will then serve as the basis
for estimating the overall number of casualties that might be caused in the country.
The goal of the research is to develop and implement a semi-empirical model for casualty
estimation that will enable to forecast the extent, types and severities of casualties that will
result in Tiberius and its surroundings following several scenario of given earthquakes. This
will enable the research group to assess the risk envelope caused by potential earthquake
scenarios and develop strategies for preparedness of the structures and the population for the
given scenarios.
6
The following milestones have been achieved in this research:
1. Literature review – A comprehensive literature review of epidemiological, social,
geological and structural aspects of seismic hazards (Chapter 2).
2. Classification of types of buildings – A detailed field survey was carried out in six census
tracts and focused on data collection and engineering documentation of the building inventory.
Based on the preliminary findings and the fact that the Israel Standard (413) for seismic design
of structures was implemented only in some buildings constructed between 1985 and 1995 and
in all of the buildings constructed since 1995, it appears that most of the building in these tracts
are residential and that about 3/4 (73%) have no or low earthquake resistance (Chapter 6).
3. Mapping present population in different time-frames - A detailed map of the population in
the city of Tiberias (demographic and socioeconomic characteristics) and its occupancy in
different structures across the city according to the time frames that have been detailed above
was constructed (Chapter 3;
4. Adaptation of semi-empirical model to the Israeli structure and characteristics- A
semi-empirical approach that combines casualty matrices of similar structures in Turkey along
a Delphi-study of experts in Israel was introduced and implemented (Chapter 6);
5. Modified Delphi process to achieve consensus – This milestone has been fully
accomplished through a 2-iteration Delphi study with participation of 20 seismic hazards
experts. The results of this study reinforce the high vulnerability of the study area (city of
Tiberias) to seismic hazards. The exposed population characteristics that are documented in
the literature as increasing the risk for earthquake-related injury and death.
6. Estimating damage to buildings – A detailed risk assessment that includes building damage
assessment was developed: it is composed of four basic modules of a risk assessment model:
hazard classification, building inventory, vulnerability, and loss assessment.
7. Estimating casualties and injuries that will be caused – Detailed assessment of the injury
and mortality was accomplished using the Casualty matrices that were the outcome of the
7
modified Delphi-study. It was observed that the Tiberias city area is highly vulnerable area for
the given seismic events.
8. Validating the results of the model with reference to former earthquakes – Results of the
present research were compared with seismic events such as the north Anatolian Fault event
in Izmit 1999.
This research offers a multidisciplinary approach to a problem that has previously been
studied mostly intra-disciplinarily. As earthquakes cause injuries through many mechanisms, a
clear understanding of these pathways can facilitate emergency response planning and
mitigation strategies. A comprehensive Analytical-Empirical model is developed that
integrates an interdisciplinary research team that combines: geological, structural, emergency
medicine, and risk assessment approaches together to establish a unique comprehensive and
robust methodology that is based in the well validated. The research outcomes will provide
decision makers and public policy makers with risk-based tool for rationale policy making
and resource allocation decision support tool.
8
Chapter 1: Introduction
1.1 Background
Mitigating the results of potential earthquakes requires an estimation of the casualties that
may incur, and accordingly an appropriate response model can be developed. Based on an
extensive literature review of the consequences of earthquakes, the following parameters were
identified as significant in estimating human casualties:
4. The earthquake hazard in the designated area, such as active faults, liquefactions, landslides
and ground motion amplifications;
5. Building and structural vulnerabilities to seismic hazards, assessed by an empirical or
analytical approach combining simulation of seismic events with the data bases of structures,
geological data, seismic data;
6. Assessment of the socio-economic conditions in the designated area, in order to estimate the
potential population that was present in different building types at the time of the shake-out.
Many studies have focused on the extent of fatalities incurred from the numerous earthquakes
that occurred in different parts of the world. In contrast, only a limited number of studies
focused on surviving casualties depicting different extent, types and severities of injuries.
Generally, the number of casualties and the level of injury are not easily attainable due to the
limited quality and lack of information in earthquake casualty data. However, several studies
that established casualty rates with respect to various building types and damage levels were
published during the last two decades.
There are four common approaches that have often been utilized to estimate casualties that
might be caused by earthquakes: 1) an empirical approach consists of direct correlation
between ground motions and the population that was present in the area at time of earthquake
occurrence, based on historic earthquake statistics; 2) a semi-empirical approach takes into
account the types of buildings that characterize the area and estimates damage rates according
to the different structures; 3) pure analytic approach predicts behavior of buildings and their
9
effects on individuals that are inside them, based on seismic hazard analysis; and, 4) hybrid
approach consists of estimating the fraction of the population killed due to collapse of
different types of buildings, considering macro-seismic intensities.
The empirical analysis ignores many relevant variables, such as building and structural
vulnerabilities, as well as the presence of the population inside the structures during the event.
The hybrid and analytic approaches do not address behavior of the numerous types of
buildings that might be situated in the area struck by an earthquake and necessitate
accumulation of data that are usually difficult to obtain due to inconsistent and poorly
characterized historical earthquake casualty data. Therefore, a semi-empirical approach can
more effectively estimate the damage and casualties that might be caused by a given
earthquake, as it includes the identification of different types of buildings, the presence of the
population during different time-frames, and the types of injuries that might be caused taking
into account different socio-economic condition.
Israel is situated in an area prone to earthquakes. To present, the estimation of casualties that
might be incurred during an earthquake in the region has not taken into consideration the
different types of buildings and the presence of the population in the different structures at
different time-frames.
The research employs the semi-empirical approach to estimate the surviving casualties and
fatalities that might be caused by a potential earthquake. The city of Tiberius and its
surroundings are used as the case-study to develop the model that will then serve as the basis
for estimating the overall number of casualties that might be caused in the country.
1.2 Objectives of the research
The goal of the research is to develop and implement a semi-empirical model for casualty
estimation that will enable to forecast the extent, types and severities of casualties that will
result in Tiberius and its surroundings following several scenario of given earthquakes. This
will enable the research group to assess the risk envelope caused by potential earthquake
10
scenarios and develop strategies for preparedness of the structures and the population for the
given scenarios.
The objectives of the sub-research team are as follows:
- Assess the connection between Structural and Non-Structural Damage Due to
Earthquakes and Human Vulnerability;
- Develop a comprehensive risk assessment model for assessment of the total loss as a
result of the consequences of given earthquake scenarios;
- Review and quantify the effect of factors that Influence Injury and Death in the Event of
an Earthquake;
- Develop a method to characterize the geological hazards and review the historical
background of the city studied.
- Integrate all the above-mentioned modules into a holistic mortality and morbidity risk
assessment analytical-empirical model.
1.3 Structure of the report
The main body of report is composed of the following six chapters: Chapter 2 – Reviews the
state-of-the-art in the four interdisciplinary domains of the research: the Connection between
Structural and Non-Structural Damage Due to Earthquakes and Human Vulnerability; review
of risk assessment models; review of the factors that influence injury and death in the event of
an earthquake, and review of geological hazards and historical background of the region. This
review clearly indicates that there is a strong correlation between the level of preparedness
and the rate of casualties in earthquake event, and this is relevant to all aspects of the event:
the structural preparedness the population, and the medical services.
Chapter 3 reviews the research methods, it emphasizes the scientific adaptation of the
HAZUS methodology by adapting the fragility curves using push-over analysis on local
structures schemes, validated method for casualty analysis using the Delphi study method,
11
classification of geological, structural, and environmental hazards that affect the vulnerability
of the population to an earthquake scenario, the analytical-empirical approach that will
provide the total risk envelope as a results of the analysis of the given earthquake scenario.
The next stage of the data gathering is described in details as well as the four different
earthquake scenario that will be analyzed in year 2 of the research.
Chapter 4 reviews the methods developed and used to predict the consequences of seismic
events in terms of mortality and morbidity from geological, structural, and emergency
medicine perspectives.
Chapter 5 reviews state-of-the-art in Fragility curves development and methodologies for the
development of fragility curves.
Chapter 6 reviews the risk assessment model design with details of the four basic modules of
a risk assessment model: hazard, inventory, vulnerability, and loss. The chapter concludes
with the risk management module that will be developed for the establishment of a rationale
public policy for retrofitting and upgrading the existing building stock in light of the given
earthquake scenario.
Chapter 7 summarizes the findings of the research at the end of the second year of the project.
The factors that affect the rate of mortality and morbidity are widely discussed including level
of building stock preparedness, level of preparedness and knowledge of the population. The
structural survey of the existing building stock results are described and show that for the first
3 census tracts surveyed most of the building stock is older than 38 years; it is composed of
Reinforced Concrete framed structures with unreinforced masonry infill walls.
1.4 Contribution of the research
This research offers a multidisciplinary approach to a problem that has previously been
studied mostly intra-disciplinarily. As earthquakes cause injuries through many mechanisms, a
12
clear understanding of these pathways can facilitate emergency response planning and
mitigation strategies. A comprehensive Analytical-Empirical model is developed that
integrates an interdisciplinary research team that combines: geological, structural, emergency
medicine, and risk assessment approaches together to establish a unique comprehensive and
robust methodology that is based in the well validated. The research outcomes will provide
decision makers and public policy makers with risk-based tool for rationale policy making
and resource allocation decision support tool.
13
Chapter 2: Literature Review – Risk Factors that Influence Human Vulnerability
in Earthquakes and Earthquake Casualty Assessment Models
The estimated number of casualties is one of the central factors guiding emergency
preparedness planning for a strong earthquake. Several studies have examined the association
of earthquake casualties with various seismic factors such as the magnitude of the earthquake
(Alexander 1985; Samadjieva & Badel 2002), ground motion parameters such as peak ground
acceleration (Shoaf et al. 1998; Mahue-Giangreco et al. 2001), shaking intensity, and distance
from epicenter (Liang et al. 2001). These factors do not account for the wide variations in
earthquake casualties. For example, magnitude is not strongly associated with the number of
casualties in each event.
Table 2.1 and Figure 2.1 present 13 fatal earthquakes (>1000 fatalities), which occurred from
the year 2000 onwards around the world, according to the magnitude of the quake and the
number of mortalities in each event. The data demonstrate the lack of linear relation between
the magnitude of the earthquake and the number of fatalities in a given event (see Table
2.1and Figure 2.1 below).
14
Table 2.1: Fatal earthquakes (over 1000 fatalities) since the year 2000
Magnitude Local Time death toll injured
1 Bhuj, India 2001 7.7 3:16 20,000 167,000
2 Afghanistan 2002 6.1 14:56 1,000 Several Hundreds
3 Bam, Iran 2003 6.6 5:26 31,000 30,000
4 Algeria 2003 6.8 19:45 2,300 10,000
5 Sumatra 2004 9.1 7:58 280,000
6 Sumatra 2005 8.6 23:09 1,300 300
7 Kashmir,
Pakistan 2005
7.6 8:52 80,000 70,000
8 Indonesia 2006 6.3 5:54 5,700 38,000
9 Wanchuan,
China 2008
8 14.28 87,000 374,000
10 Sumatra 2009 7.5 17:16 1,100 1,200
11 China 2010 6.9 7:49 2,200 12,000
12 Haiti, 2010 7 16:53 316,000 300,000
13 Tohoku, Japan
2011
9 14:46 21,000 5,000
* Based on data from the USGS website
15
Fig. 2.1 Fatal earthquakes (over 1000 fatalities since the year 2000
*Based on data from the USGS website
Data from previous earthquakes also shows that a country’s level of development has an
influence on the amount of damage done as a result of an earthquake. For example, the case
of Japan and China; two countries that have been affected by multiple earthquakes through
the course of their history. China is a developing country, ranked at the 101th
place in the
Human Development Index – HDI1 – in contrast to Japan, which is a developed country,
ranked at the 10th
place using the same index. When comparing the number of casualties in
earthquakes in these two countries since the end of the 19th
century and up until today, one
can note that despite the fact that in Japan, a higher number of earthquakes, at a magnitude
above 6, occurred (46 events vs. 27 events in china), the number of overall casualties in Japan
is significantly lower (219,440 fatalities in Japan, vs. 643,513 fatalities in China). One can
also note the negative trend in the number of Japanese casualties over time (with the
1 The Human Development Index (HDI) is a tool for distinguishing developed from developing countries in the
world; it is used by the United Nations Organization, which submits it as an annual report.
16
exception of the 2011 earthquake, in which most of the fatalities occurred as a result of the
tsunami secondary to the earthquake), while in china this trend does not exist [USGS].
Countries such as Japan and New Zealand, which have implemented strict seismic building
codes and invested resources to enhance community preparedness have demonstrated a
dramatic decrease in the number of earthquake casualties compared to other developed
countries such as China, India or Haiti, which still present high vulnerability to earthquakes
(Paton et al. 2010; Crowley & Elliott 2012). The populations that occupy buildings have
various characteristics in different countries and regions, which may play a crucial part in
enhancing or impeding vulnerability to earthquakes, ultimately influencing the number of
casualties in a given event. These nonstructural variables include individual characteristics,
such as gender, age, physical disabilities and behavioral strategies; household characteristics,
such as socioeconomic status (usually household income); and community characteristics,
such as the existence of search and rescue teams and the medical aid response. Evidence from
other natural events such as floods, suggests that communities with socially vulnerable
populations experience more casualties (Cutter, et al. 2003; Zahran et al. 2008). In light of
recent events such as the 2010 Haiti earthquake, this finding should be investigated in the
specific context of earthquake loss estimation and casualty modeling.
This section reviews the state of knowledge of epidemiological and engineering research
regarding earthquake-induced casualty assessment, according to the classic Epidemiological
Triad model of agent, host, and environment (Parrish et al. 1964).
2.1 Agent Characteristics
The “agent” in this case is the earthquake itself and the parameters that characterize it
(seismological and geological) such as: magnitude and intensity (which are measures of
earthquake strength), epicentral distance which can be used to estimate the localized strength
of shaking; and other factors such as ground composition which influence wave transmission
during seismic events and landslides (Ramirez & Peek-Asa 2005). The time of the day at
17
which it occurred has a significant influence for several reasons: there is a direct influence of
the time on the distribution of the population with respect to location. An earthquake that
occurs during the daytime will “encounter” the exposed population mostly at workplaces,
educational institutions, or, in case that the earthquake occurs during the commuting hours—
on the road. An earthquake that occurs during the nighttime will “encounter” most of the
population at homes, and typically in a state of sleep (Shoaf et al. 1998). One might assume
that both the location of the population in different structures (residential structures as
compared with commercial structures or educational institutions) and the physical state of the
population (awake versus asleep) has an influence on the number of casualties at a certain
event and on the process of assessing the casualties; this aspect will be discussed below.
2.2 Environmental Characteristics
The “environment” in this case relates to the characteristics of the built environment. Based
on existing studies, damage to buildings is considered the most important factor causing
injury and death in earthquake events worldwide (Jaiswal et al. 2010; Peek-Asa et al. 1998;
Noji & Kelen 1990; Spence et al. 2003; Wald et al. 2011; Porter et al. 2008; Jaiswal et al.
2009). Among other, this includes the type of structures, construction methods, the quality of
the building materials, number of floors in the structures, building’s age, etc. Additionally,
transportation infrastructures such as bridges and roads have been identified as being highly
vulnerable in earthquake scenarios. Collapsing of bridges and car accidents at the time of an
earthquake or after it (due to damage done to traffic lights or road signs, for example)
constitute a factor which has potential consequences on the number of casualties in an
earthquake (Peek-Asa et al. 1998; Ramirez & Peek-Asa 2005).
2.3 Host Characteristics
The third aspect in the Epidemiological Triad model is the “host,” or the human factor, which
encompass the demographics, behavioral response and physical resiliency of individuals.
Studies that examined the effects of gender had conflicting results, finding either no
18
difference by gender or a higher rate of casualties among women. Studies from the United
States, Taiwan, Turkey and Haiti found no differences in the risk of injury and death by
gender (Mahue-Giangreco et al. 2001; Liang et al. 2001; Ellidokuz et al. 2005; Doocy et al.
2013). Among the four studies that found women at higher risk for injury and death, two from
the United States (Shoaf et al. 1998; Peek-Asa et al. 2003) had the highest odds ratios (OR) of
1.64 (95% CI: N/A) and, 2.4 (95% CI: 1.2, 5.1), respectively. Studies from Taiwan and Peru
(Chou et al. 2004) (Doocy et al. 2009) found women to be at higher risk, but with smaller
odds ratios: OR=1.2 (95% CI: 1.1, 1.3) and, OR=1.6 (95% CI: 1.0, 2.7), respectively. The
United States and Taiwan had conflicting findings, with studies in both regions reporting
increased risk for females and no significant differences related to gender regarding the same
earthquake events.
Studies that examined age had inconsistencies, but reported some common results. Those
spanning 25 years and a variety of geographic regions found higher rates of casualties among
the elderly and children. One exception was a study from the United States that reported
decreased risk of earthquake-induced injury with increasing age (OR=0.7, 95% CI: N/A)
(Shoaf et al. 1998). Among the seven studies that found the elderly to be at higher risk for
injury and death, Chou et al. from Taiwan (2004) had the highest odds ratio of 5.5 (95% CI:
4.4, 6.8). Three studies from the United States and Haiti (Mahue-Giangreco et al. 2001;
Peek-Asa et al. 2003; Doocy et al. 2013) found lower odds ratios of 2.69 (95% CI: 1.34, 5.4),
2.9 (95% CI: 1.2, 7.4) and 2.8 (95% CI: 1.66, 4.72), respectively. Two other studies from
Taiwan and China (Liang et al. 2001; Dong et al. 2013) also found age to be significantly
associated with higher mortality and reported coefficient value of 0.101 (95% CI: 0.076,
0.125) and B coefficient of 1.539 (95% CI: N/A). A study from Peru (Doocy et al. 2009) did
not report odds ratio or coefficient values, but stated that increases in age increased the risk of
injury by 3% (95% CI: 2, 4).
Studies from the United States reported conflicting findings, with studies regarding the same
19
event reporting that the elderly were at higher risk of injury and death and that increased age
lowered risk. Studies from Haiti (Kolbe et al. 2010) and Taiwan (Chou et al. 2004) found
children to be at higher risk of injury and death, and both reported similar odds ratios of 5.8
(95% CI: 4, 8.3) and 5.5 (95% CI: 4.4, 6.8), respectively.
Although many articles discussed the role of poverty and socioeconomic status on earthquake
casualties, few measured the risks related to these factors (Cannon 1994; Fothergill & Peek
2004; Kahn 2005). Figure 2.1 above demonstrates that most deadly earthquakes occur in
developing countries rather than in developed countries, but even when an earthquake strikes
a well-developed country such as the United States, there is evidence that low-income
households were more affected than high-income households (Fothergill & Peek 2004).
However, studies from Taiwan (Chou et al. 2004) and Peru (Doocy et al. 2009) used
analytical methods to measure the role of socioeconomic status as a risk factor for
earthquake-related injury and death. The Taiwanese study found that the degree of
vulnerability to earthquake-related death increased with decreasing monthly wage (OR=2.2,
95% CI: 1.6, 3.0). Similarly, the study from Peru found that members of households with
lower economic status faced increased risk of injury and that the risk decreased by 3 percent
per additional 100 soles (Peruvian currency) of monthly household income (Doocy et al,
2009).
Only two studies directly assessed the link between physical disability and the risk of
earthquake-induced injury and death; both found a higher rate of casualties among physically
disabled individuals. The Japanese study (Osaki & Minowa 2001) found an odds ratio of 1.9
(95% CI: 1.0, 3.4) and similarly, the study from Taiwan (Chou et al. 2004) found an odds ratio
of 1.7 (95% CI: 1.2, 2.3).
Two studies examined the effect of human behavior during an earthquake on the risk of injury
and death and found that exiting the building decreased the risk of injury and death. The first
study, from the Philippines (Roces et al. 1992), found a lower odds ratio for injury (OR=0.3;
20
95% CI: 0.1, 0.9) among subjects who were outside buildings. Similarly, a study from
Armenia (Armenian et al. 1992), found a higher odds ratio for injury (OR= 4.8; 95% CI: 2.34,
10.0) among subjects who remained inside.
Ethnicity was discussed in American studies and was found a significant risk factor effecting
injury and death in earthquakes (Shoaf et al. 1998; Mahue-Giangreco et al. 2001), although
not consistently. Since ethnic groups and their characteristics vary from region to region, this
issue is not addressed here.
2.4 Interaction between the host, agent an environment
It is important to state that although each factor mentioned above in each group, has an
independent association to earthquake injury and death risk, these factors may also interact in
many situations to increase risk (Ramirez and Peek-Asa 2005). For example, seismically
stable structures located on vulnerable soil types may collapse during an earthquake while
less stable structures located on solid rock may endure; different individuals are less or more
vulnerable to the same earthquake event as a function of their ability to take protective
strategies such as fleeing outside of a collapsing building. Global trends toward urbanization
combined with an increase in population density, poverty, and social gaps, increases the
vulnerability of urban centers in the face of natural disasters, including earthquakes
(Tekeli-Yesil 2006; Donner & Rodriguez 2008).
The factors affecting an earthquake are summarized in Table 2.2.
21
Table 2.2: Parameters that influence the process of assessing the number of people injured in
an earthquake
Agent – The
Earthquake
Host – The Human Factor Built Environment
Magnitude
Intensity
Soil characteristics
Topographical
characteristics
Time of occurrence
Demographic characteristics
Socio-economic characteristics
Behavioral characteristics
Physical resilience
Construction characteristics
Infrastructure Characteristics
Casualty modeling has traditionally been conducted using engineering methods, which focus
only on damage to the built environment. Thus, the major source of information for
community preparedness activities does not consider the exposed population and has rarely
integrated human factors. In light of recent events such as the 2010 Haiti earthquake, these
factors should be investigated in the specific context of earthquake loss estimation and
casualty modeling.
22
Chapter 3: Research Methods for Assessment of Earthquake Casualties
3.1 Literature reviews
A comprehensive literature review was conducted on the epidemiology of injury and
mortality in earthquake events that occurred globally, with emphasis on earthquakes that
occurred in the last two decades. Different trends and factors that are considered to have an
influence on the process of prediction of casualties in a future earthquake were surveyed –
both in respect to characteristics of the population and in respect to the characteristics of the
environment and earthquake itself (detailed in chapter 2). In parallel, a survey of different
approaches used to predict casualties in earthquakes was conducted (detailed in chapter 4) and
casualty rates (matrices) used in different models globally were identified (detailed in chapter
7). The search for these matrices was conducted in four languages: English, Spanish, Chinese
and Hebrew by three research assistants that were instructed to screen local databases with
predefined search words (e.g. earthquake*, estimation*, casualty*, rate*, injury*, death* etc.).
The surveyors were also instructed to gather information regarding the origin of the matrix
(development method) if applicable.
3.2 Mapping population distribution
The mapping of the population distribution was conducted according to data from the Israel
Central Bureau of Statistics, using data from the last census of 2008. A detailed map of the
population in the city of Tiberias (demographic and socioeconomic characteristics) and its
occupancy in different structures across the city according to the time frames that have been
detailed above was constructed; the data was collected for each of the 12 census tracts of
Tiberias. For each tract, the following information was collected: number of residents,
number of households, gender and age distribution, level of education, rate of disabled
population, socioeconomic index rank, income details, form of home-ownership, residents’
line of occupation (including school pupils and students), labor force participation rate, and
commuting rate. These findings were integrated to the HAZUS-MH population distribution
23
data catalogue. Some of the data that is not currently taken into consideration in the casualty
loss estimation of HAZUS (e.g. age, gender, and disability) was used to detect risk areas
according to epidemiological models.
3.3 Estimating casualty rates
A statistical validation procedure of the existing casualty matrix was conducted according to a
Modified Delphi method (Dalkey, Brown, & Cochran, 1969). Local experts from different
fields: structural engineers, physicians, search and rescue team members, risk-management
and disaster-management professionals participated in a two-round modified Delphi
procedure aiming to reach a consensus higher than 70% among all participants, in relation to
the existing casualty matrix suitability.
Delphi procedure – round I and II:
For this purpose, an electronic survey was developed, containing two sections and fourteen
questions: the first section included ten questions, which presented the current casualty rates
according to each of the five damage levels, indoor (five clauses) and outdoor (five clauses).
Appendices with photos describing typical structures in Israel and photos illustrating each
damage level were attached. The experts were asked to rate their level of agreement with the
suitability of these rates to describe the number of casualties expected in case an earthquake
occurs in Israel region, on a five point Likert-type scale (1 = I don’t agree to 5 = highly
agree). If the rating given was 1 = I don’t agree or 2 = agree to a lesser extent, the participant
was asked to explain his answer in free text. The second section collected information
regarding the participants’ characteristics: gender, age, profession and seniority. For survey
example, see figure 3.1. Since the community of earthquake experts in Israel is rather limited,
the experts were sampled using a “snow-ball sampling” method in which each participant was
asked to recommend more possible experts eligible to participate. The survey was sent to each
participant via email and was conducted during February to June, 2014. The results were
24
analyzed to calculate the total agreement percentage between all experts and since it was
lower than 70% a second round of survey was conducted between July and October, 2014. In
the second round, each participant received a second survey and was asked to review an
alternative casualty matrix (Erdik et al, 2003; Kaplan & Yilmaz 2007) and rate his agreement
to its implementation suitability in Israel. This alternative matrix was identified as part of the
literature review process described above and is considered as a viable alternative to use in
Israel since it offers a stricter estimation of expected casualty rates due to an earthquake. The
level of consensus between experts regarding the computability of this alternative matrix to
use in earthquake casualty estimation in Israel was calculated again and was found to be
higher than 70%, the Delphi procedure was concluded.
Figure 3.1: Experts survey, round I. Presents indoor casualty rates for the highest structural
damage level (collapse).
25
3.4 Performing a casualty estimation process for the city of Tiberias
The alternative casualty rates, processed in the Delphi procedure were integrated to the
HAZUS-MH casualty estimation model. In order to assess the sensitivity of the factors in the
casualty estimation model, six simulations were conducted using three earthquake scenarios.
For all six simulations, the casualties were calculated only for reinforced concrete (C3)
structures (as defined by the alternative matrix). The scenarios examined were: a) Mw=6
earthquake, epicenter located 10 Km from Tiberias city; b) Mw=6 earthquake, epicenter
located 3 Km from Tiberias city; and c) Mw=7 earthquake, epicenter located 3 Km from
Tiberias city. For each scenario, two simulations were conducted: once using the HAZUS-MH
casualty rates (FEMA, 2003), and once using the alternative rates (Erdik et al, 2003), a total
of six simulations.
Further analysis was conducted to estimate the sensitivity of the casualty estimation model to
a change in the ‘collapse rate’ factor (probability of collapse given a complete damage state).
In this analysis, five simulations were conducted for one earthquake scenario - Mw=7
earthquake, epicenter located 3 Km from Tiberias city. The collapse rates tested in the
analysis were: a) 13% (multiplication factor=1, rate used in HAZUS); b) 25% (multiplication
factor=2); c) 32% (multiplication factor=2.5); d) 38% (multiplication factor=3); and e) 46%
(multiplication factor=3.5).
For all the simulations, the number of casualties according to four severity levels (i.e. light,
moderate, severe and fatal injuries) were depicted and a range of uncertainty was provided.
26
Chapter 4: Casualty estimation models to predict seismic mortality and
morbidity hazards
4.1 Common approaches for Predicting casualties in an Earthquake
Four approaches are often used to estimate casualties in earthquakes (Maqsood & Schwarz,
2011). The empirical approach consists of directly correlating the magnitude/intensity of the
earthquake and the size of the population in the area at time of the earthquake, based on
historic earthquake statistics (Samadjieva & Badel 2002; Jaiswal et al. 2010). The
semi-empirical approach builds on the empirical approach by additionally accounting for the
types of buildings that characterize the area and damage rates according to the different
structures (Jaiswal & Wald 2010). The purely analytic approach predicts behavior of buildings
and their effects on individuals inside them, based on seismic hazard analysis (Porter et al.
2008; Spence & So 2009). An expansion of the other approaches that includes all the
above-mentioned parameters and considers the population distribution in various occupancies
and time frames e.g. day, night and commute time has been proposed (FEMA 2003).
In regions that have experienced numerous earthquakes with high numbers of fatalities,
typically developing countries with dense populations living in vulnerable structures, enough
data exist to calibrate fatalities from the historical earthquake record alone (FEMA 2008). In
such regions, building inventories are typically lacking, as are systematic analyses of their
vulnerabilities; hence, analytical tools are inadequate for loss estimation. In contrast, in the
most highly developed countries, particularly those with substantive building codes, structural
responses are more easily analyzed and information on their distribution and occupancy are
more readily available (Li et al. 2011; Jaiswal et al. 2010). Due to the success of such building
codes, for the purpose of fatality loss modeling this category of countries typically has had
relatively few fatal earthquakes, making it difficult to use empirical calibration from past
events alone. In such cases, fatality estimates are largely informed from analytically-derived
collapse rates and inferred fatality ratios given a structural collapse (Jaiswal et al. 2011;
27
Jaiswal & Wald 2010). The last approach is applied in the HAZUS-MH loss estimation model
(FEMA 2003), which is an American loss estimation methodology that can be utilized
globally in different countries and regions. In HAZUS-MH, casualties are calculated as a
function of building damage for given ground accelerations. Some demographic data such as
age, ethnicity and household income are entered into the database, but these data are used
only for estimating the expected displaced population and shelter needs and not to calculate
casualties. Factors such as gender, disability and expected behavior during earthquakes are
not taken into account in the current methodology (FEMA 2003).
4.2 The HAZUS-Multi-Hazard Casualty Estimation Model
The model is based on the assumption that there is a high correlation between damage to
structures and the number of casualties as well as the severity of their injury. An additional
assumption is that in “small” earthquakes, damage to non-structural elements will be
responsible for most of the injuries, wherever in stronger earthquakes which result in a higher
degree of structural damage this will be the main cause of casualties. In the existing
methodology, the rate of casualties is calculated solely by means of damage to structures and
bridges. The rate of casualties in the methodology is expressed as the percentage of residents
in a building who are liable to become injured or killed as a function of different degrees of
damage and type of structure.
4.2.1 Time frames
In the existing model there is a reference to three time scenarios:
Night Scenario – an earthquake at 2:00 in the morning
Day Scenario – an earthquake at 14:00 noon
“Commuting” Scenario – an earthquake at 17:00 in the afternoon
28
These scenarios are expected to produce the largest number of casualties at home (night), at
work/school (day), and at the heaviest traffic hour (the “commuting” phenomenon).
4.2.2 Population distribution
The casualties for a given scenario are calculated at the census tract level. The building stock
for each tract is distributed into basic groups (e.g. residential, commercial, educational,
industrial and hotels). Specific information regarding the population characteristics is
gathered to create an accurate representation of the population distribution across the day and
night; such as the number of students, number of residential population and number of the
number of tourist population. An accurate population distribution map is essential for
appropriate casualty assessment and the HAZUS developers recommend that when applying
the model in regions other than the US, modification to the default assumptions should be
made according to local characteristics.
4.2.3 Casualty rates
The HAZUS-MH methodology estimates casualties based on the correlation between the rates
and severity of injuries and the damage level of structures (Levi et al. 2015). The casualty
rates (also referred as casualty matrix) currently used in HAZUS-MH is based on a report
published in 1985 by the ‘Applied Technology Council’ (ATC) and provides experts
estimations regarding earthquake damage and losses in California. The experts based their
estimations on data from historical events and conducted a consensus-based assessment to
compile the final report (ATC-13, 1985). These rates were evaluated and revised based on
comparison with a limited amount of data from the Loma Prieta, Nisqually and Northridge
earthquakes (FEMA 2003). As mentioned above the HAZUS-MH methodology was
developed in the United States and its data catalogues describe the structures typical to this
region. Generally, applying a model developed in one region to another is problematic, not
only because the structures may differ but also because the basic assumptions regarding injury
29
and death mechanisms may not be valid in the “new” area. This research aim was to bridge
this gap and to evaluate and calibrate the default casualty rates currently used in HAZUS-MH
to the Israeli characteristics.
4.2.4 Output
The model offers the following output: 1) the rate of indoor casualties according to five levels
of damage caused to the structure; 2) Rate of casualties Outside of a Structure – also
according to five damage degrees.
The casualties are classified to four severity levels:
1) Light injury – which requires basic medical aid (e.g. sprains, cuts, first or second degree
burns).
2) Moderate injury – requires a greater degree of medical care and the use of medical
technology, but not expected to progress to a life threatening status (e.g. fractured bone).
3) Severe injury – poses an immediate life threatening condition (e.g. uncontrolled bleeding,
crush syndrome).
4) Instantaneously killed or mortally injured.
The above-mentioned scale represents a compromise between the requirements of the
health-care system regarding information about casualties that allows to take proper
preparedness measures and plan a proper response to such an event and the engineers’ ability
to provide the required information. An example of this tension is the health-care system’s
requirement to receive data regarding related morbidity in the event, and not just data
regarding those injured through a direct mechanism – for example, people who have
experienced heart attacks during the earthquake or disaster related anxiety victims. The
current methodology also does not assess people injured in car accidents, falls, electrical
blackouts, injuries that occurred after an earthquake while rescue teams operated in the area,
30
electrocution, tsunami injuries, landslides, liquefactions, fault ruptures, collapsing of damns,
fires, and release of hazardous materials. Likewise, psychological damage to a population is
not calculated in the existing model.
4.3 Applicability of HAZUS-MH Casualty Estimation Methodology in Israel
As stated, the model was developed in the U.S. and the existing data in the program describe
the building stock that exists in the U.S. and not in Israel. Generally, applying a model
developed in one region to another is problematic, not only because the structures may differ
but also because the basic assumptions regarding injury and death mechanisms may not be
valid in the “new” area. As can be understood, since the data in the methodology has
undergone its last calibration, about 20 years have passed and a number of major earthquakes
have occurred in the world (for example the quake in China in 2008, Haiti 2010, and Japan
2011). Likewise, global changes such as accelerated urbanization, economic, social, and
political changes may also have an influence on existing data in the methodology and this
should be examined.
The rate of casualties (which is expressed by the percent of residents in the building who are
liable to get injured and killed as a function of different levels of damage and the type of
structure) that are being offered in the existing model will be tested as well by Israeli content
experts in light of the results of earthquakes that have occurred in the recent period of time
and, as needed, they will be calibrated and updated in the model being offered in the current
study.
31
Chapter 5: Development of Fragility Curves
5.1. Background
Building response is characterized by building capacity curves (Fig. 5.1). These curves
describe the push-over displacement of each building type and seismic design level as a
function of laterally-applied earthquake load. The Methodology uses a technique, to estimate
peak building response as the intersection of the building capacity curve and the response
spectrum of PESH shaking demand at the building’s location (demand spectrum). The
capacity spectrum method is one of the two nonlinear static analysis methods described in the
NEHRP Guidelines for the Seismic Rehabilitation of Buildings [FEMA, 1996a].
The development of Fragility curves which defines the probability of damage levels
depending on the ground motion is highly important process. The curves which are based on
the dynamic behavior of structures, allow us to predict the potential damage and to develop
the subsequent emergency preparedness plans. Most of the seismic risk assessment
methodologies have been developed in the United States over the past two decades, with the
major development being HAZUS (Hazards United States) – a standardized loss assessment
methodology.
Building damage functions are in the form of lognormal fragility curves that relate the
probability of being in, or exceeding, a building damage state to for a given PESH demand
parameter (e.g., response spectrum displacement). Figure 5.2 provides an example of fragility
curves for the four damage states used in this methodology.
Each fragility curve is defined by a median value of the PESH demand parameter (i.e., either
spectral displacement, spectral acceleration, PGA or PGD) that corresponds to the threshold
of the damage state and by the variability associated with that damage state. For example, the
spectral displacement, Sd, that defines the threshold of a particular damage state (ds) is
32
assumed to be distributed by:
Where is the median value of spectral displacement of damage state, ds, and
εds is a lognormal random variable with unit median value and logarithmic standard deviation,
βds
In Hazus, there are many damage (fragility curves) functions that are adapted for predict the
damage in different types of buildings and structures.
The damage functions were defined in hazus for a wide variety of buildings types
combined with different levels of seismic resistance: 1) no resistance; 2) low resistance; 3)
medium resistance; and 4) high resistance level. Thus for example, an old building that was
not built according the standard building will be defined as structure that is not resistant. To
the extent necessary, new damage functions can be defined within the program.
Parallel to the different degrees of resistance are five general levels of damage (FEMA,
2003): 1) no damage; 2) light damage; 3) intermediate damage; 4) advanced damage; and 5)
irreparable damage. Light damage in buildings could be manifested by short cracks in the
upper part of the building; moderate damage could be manifested in diagonal cracks in the
high part of the building; extensive damage, in diagonal cracks throughout the building,
including shear movement; complete (irreparable) damage can be manifested in long diagonal
cracks throughout the building including significant shear movement to the state of partial or
total collapse of the structure.
33
Figure 5.1: Example of capacity curve and demand spectrum. The
intersection point is marked by filed red circle
Figure 5.2: Example of Fragility Curves for Slight, Moderate, Extensive and Complete
damage. Notable, that in large magnitudes the differences between the damage levels
remains almost constant. This leads that the differences in damage in high magnitudes
(generally above MW=7.2) remain constant (e.g., field et al., 2005)
34
Chapter 6: Risk Assessment Model
6.1 Introduction
Risk assessment models have played an important role in engineering design for natural
hazards. For instance, structural engineers design buildings using particular probability of
exceeding of the later forces by winds or earthquakes. Furthermore, these models have been
developed to assess the casualty or economic losses caused by the damage of buildings result
from natural hazards.
The four basic components of a risk assessment model are: hazard, inventory, vulnerability,
and loss, as shown in Figure 6.1. First, the hazard module characterizes the risk of the natural
hazard phenomena itself. For instance, an earthquake hazard is characterized by its relevant
parameters like soil condition, epicenter location and moment magnitude. Meanwhile, the
frequency of certain magnitudes or frequencies of earthquakes are needed to investigate.
Second, the inventory module characterizes the inventory of properties at risk. One essential
parameter to describe the inventory is location of inventory at risk. Moreover, in order for
more accuracy of the estimate, factors describing building attributes can be also considered,
such as structural type, the height, the age of buildings.
Next, in the vulnerability module, the vulnerability of the inventory to damage is calculated
from the result of the hazard and inventory modules. For example, the HAZUS program
classifies structural damage in four damage states: Slight, Moderate, Extensive, and Complete
state. Meanwhile, fragility curve of a building is used to represent its vulnerability to damage.
In order to estimate economical loss, factors like its contents and also time element losses,
such as business interruption loss or relocation expenses, can be added in the model.
Finally, in the loss module, loss could be classed as direct or indirect. On one hand, direct
losses could include human casualties, or cost to repair a building, whose loss can be
calculated directly by the level of damage. On the other hand, examples of indirect losses are
35
business interruption impacts and relocation costs of residents, which can be considered as
consequences due to the damage. In general, indirect losses are more difficult to qualify than
direct ones.
Hazard
Vulnerability
Inventory
Loss
Figure 6.1: Risk assessment model basic components
6.1.1 Hazard Module
A hazard module generally needs to address three basic issues regarding the source
parameters of a hazard: the locations of events, its possibility of occurrence and its severity.
These three elements are usually closely relevant so that the sets of data which is needed to
investigate and validate are shared to one another. Usually, probability distributions are
investigated from historical data for parameters that define the simulated event. Consequently,
the intensity of the event, as the result of this module, is calculated by propagating the
corresponding parameters.
Locations of Events
To attain more accurate estimates of loss, the scope of the model need to be defined. In
conducting a risk assessment for earthquakes in the city of Tiberius in Israel, those faults and
seismic source zones that have measurable impact on the building inventory of interest must
be identified as shown in Figure 6.2: Synthetic earthquake HAZUS scenarios along Dead Sea
Transform (DST)
(Levi et al., 2010. In general, such information is measured and investigated based on
historical data including the epicenter location and moment magnitude of recorded earthquake
36
events. An identified modal domain is essential to further define all required source
parameters of hazard within the boundary.
Figure 6.2: Synthetic earthquake HAZUS scenarios along Dead Sea Transform (DST)
(Levi et al., 2010)
Probability of Occurrence
Estimating the probability of potential earthquakes events is sometimes the most uncertain
aspect of the module, especially for those regions with scarce historical data. In general, the
probability of occurrence is linked to the location of events. The probability of occurrence of
earthquakes and their magnitude, usually represented as Gutenberg-Richter magnitude
distribution, are modeled as characteristics of earthquakes. For instance, as stated by Patel et
al. (2005).
A fault is considered to have a characteristic earthquake when it ruptures at fairly regular
intervals, producing earthquakes of a similar magnitude. Generally, a fault does not rupture
37
with such predicable probability. Characteristic earthquakes can be determined either by a
single magnitude or by a magnitude range with particular distributions.
Intensity of ground motions and Site Effects
In addition to the source of earthquakes, the sites of the investigated building inventory need
to be investigated to estimate the damage associated with the events. In general, depending on
geological attributes of studied areas, site effects include rupture, fault, and soil condition.
The intensity of events changes as the site effects propagate over the affected area. The
seismic waves propagate away from the epicenter through the affected region, as illustrated in
Figure 6.3: Attenuation and local soil/site effects. Patel et al. (2005) indicate the following
parameters are considered to determine the amplitude of those waves when building damage
is estimated: 1) those that are controlled by the earthquake’s source mechanism, 2)
characteristics of the intervening geological materials through which the waves travel and 3)
by the complexities of the local soil materials underlying each affected site. Since it is
practicable to develop a model to simulate an earthquake mechanism with associated
parameters over a region, taking attenuation equations as example, a common way is to
determine empirical relationships and they identify the rate at which the seismic waves
decease or increase when the waves propagate from the source over the affected region.
Bedrock
Rock Stiff Soil Soft Soil Near-source Rock
Focus
RC Commercial
1960Masonry Residential
1920
SRC High-rise
2010
Wooden House
1900
Figure 6.3: Attenuation and local soil/site effects
38
6.1.2 Inventory Module
Building inventory plays a crucial role in estimating the potential losses to building structures
in a risk assessment model. The process of developing a complete building inventory is
usually very time and cost-consuming, and the database within the studied area at postal code
level needs to be updated regularly by local authority or private sources.
The data should include the number of properties, or risks, and their values, broken down by
type of designation (residential, commercial and industrial), by coverage (building,
appurtenant structures, contents and time element, or loss of use) and by occupancy patterns
and construction type (Patel et al., 2005) that shows the main Israeli building types and
illustrates the building stocks in the nation.
Construction types are essential function to their damage. For instance, masonry buildings
usually have weak seismic resistance comparing to steel or concrete reinforcement buildings.
Also, buildings which are built by engineering codes perform better to earthquakes than those
are considered as non-engineered buildings. Local building codes and construction practice
usually differ from one another and reflect regional differences. Meanwhile, to estimate the
damage of contents, both structural damage and occupancy patterns need to be investigated in
the model.
In this research, for particularly important buildings or crucial facilities, on-site inspections is
conducted by experts and we incorporate information provided in actual design documents,
including specifications of the physical dimensions of individual components (beams,
columns, joints, partitions, etc.) and their material properties. The vulnerability component of
the model is then developed by representing as fragility curves to describe the behavior of the
building when they are subjected to earthquakes force.
39
Table 6.1: Main Israeli building types (Levi, 2010)
Figure 6.4: Israeli building stocks (Levi, 2010)
40
6.1.3 Vulnerability Module
The vulnerability module investigates the severity of building damage due to earthquakes.
The link between earthquake intensity and the level of building damage, or more specifically-
of the level of casualty and economical losses, has been investigated by various approaches.
One of the most common approaches is using engineering judgment. This approach combines
experts’ opinion with historical data However, due to its arbitrary in nature, it is difficult to
update the module or database when new information becomes available.
The other common approach is based on theoretical engineering techniques and thus it is
considered as a more sophisticated approach. However, adjustment for specific local
conditions of built environment is needed for more accurate estimate. As indicated by Patel et
al. (2005), it is impractical to directly apply these techniques to portfolio risk assessment in a
city. Therefore, this approach is usually modified to make possible their application to
portfolio risk assessment. For instance, the building stock is categorized into many typical
building classes (e.g., reinforced concrete building) with different characteristics (e.g., three
stories, built between 1960 and 1970).
Each class is then subdivided based on different characteristics to justify details that have
impacts on the building response under earthquakes. One typical building of each building
class is investigated using the techniques mentioned above. The response of a typical building
subjected to different severity of ground motion intensity is then assigned to the property in
the portfolio that belongs to that class.
Patel et al. (2005) indicates that two major steps are conducted in the application of the
engineering-based vulnerability approach to portfolio risk analyses:
1) Identification and definition of typical buildings in the modeled affected region.
41
2) Assessment of the building response to ground motion at different intensities. This will be
referred to here as vulnerability analysis.
Buildings Identification
One of the most time-consuming tasks in identifying the building stocks is evaluating the size
of the populations of different types in the studied area. For more accuracy in classification of
building occupancy types, data usually needs to be divided by different types such as
residential, commercial, industrial, public, and agricultural and so on. Meanwhile, the
building classes are also identified by some crucial factors which affect structural behavior to
earthquakes. For instance, these factors could be building material (e.g., wood or reinforced
concrete), structural system (e.g., moment frame or shear wall), and height (e.g., one to 10
stories). Error! Reference source not found. shows some building structure types used in
the HAZUS programming.
Table 6.2: Building structure types (FEMA and NIBS, 2006)
42
Evaluation of Building Performance
According to structural theory, building performance is the result by calculating the
relationship between the intensity of the subjected force and the expected structural damage.
In general, both structural and non-structural damage are caused by the lateral building
deformation resulted from earthquakes. One approach to investigate the lateral building lateral
deformation is using the maximum inter-story drift to get the level of damage to the
components at that story. In the case of earthquakes, the relationship between the severities of
the external excitation to building damage is identified by a fragility curve for given structural
damage states, for instance, the damage states includes minor, moderate, severe damage, or
collapse. A fragility curve for a given damage state provides the likelihood which the damage
state will be exceeded as a function of the severity of ground motion at the site (Patel et al.,
2005).
In general, for loss analysis, building damage is expressed in terms of a damage ratio. The
damage ratio ranges from 0% to 100% of the total loss. As a general damage function shown
in Figure 6.5, one can see both the intensity of the external force and the level of damage
given the level of force are uncertain quantities. Accordingly, the damage ratio of the building
is uncertain as well. The damage state of the subjected building for a given level of external
force is identified by the aggregated damage of its structural components, for instance, beams
or columns. As mentioned previously, the damage level of a building depends on the
maximum deformation of each story.
Patel et al. (2005) states that the engineering analyses performed to estimate the level of
building deformation for a given level of ground acceleration entails building a computational
model of the structure. The building is then either subjected to historical ground acceleration
records of different intensities or pushed in lateral increments until collapse to simulate the
lateral response of the building during different size earthquakes. Figure 6.6 shows a
43
schematic flow of the damage assessment process. This procedure assesses the damage states
for buildings and also their contents, which indicate the amount of loss associated the function
of buildings.
Figure 6.5: Illustration of a typical damage function (Patel et al., 2005)
Figure 6.6: Flow chart of damage assessment process (Patel et al., 2005)
44
6.1.4 Loss Module
As mentioned in chapter 6.3, the loss assessment approach has been to link ground motion
intensity directly to the level of casualty or monetary loss. In our research, damage functions
are investigated based on experts' judgment and on engineering analysis of building types.
Experts in structural engineering are asked to assess the damage ratio to a typical building of a
specific construction type under a given ground motion intensity. Based on those experts’
judgment, the responses of the building subjected to a certain intensity of ground motion can
be evaluated.
The loss module employs according cost models that translate estimates of structural damage
into economic loss. The model produces estimates of the cost of repair or replacement for
each damaged structural and non-structural component as investigated by the engineering
analysis (Patel et al., 2005). The level of repair depends on the level of damage, and the
building type. For example, the costs of repair could be minor for a masonry wall with a few
small cracks. On the other hand, the building would need to have a major renovation when the
cracks are wide enough to jeopardize its structural system.
6.2 The Geology and tectonics of Tiberias
Tiberias is located along the western border of the Kinneret basin, next to four major normal
faults (Jordan fault, Poria fault, Almagor fault and HaOn fault) which are part of the DSF
system (Fig. 6.7), and approximately ten kilometers from the Beit HaKerem fault which is
considered as an active fault in the Galilee (Sagy et al., 2013). The shallower section below
Tiberias is composed of folded Miocene alluvial and lacustrine sedimentary rocks that are
unconformably overlain by Pliocene–Pleistocene flood basalts. The upper section in the eastern
part of Tiberius area, along the coast of the Sea of Galilee, is composed of thick unconsolidated
sediment sequence and pebbles. Such a geological setting, where soft sediments overlie stiff
45
rocks with higher seismic velocities, generates local ground motion amplification during an
earthquake.
The long documented history of destructive earthquakes in Israel shows that the whole area,
where modern Tiberias is now located, was severely affected with considerable damage and
many casualties. In the second millennium several worth mentioning earthquakes occurred: for
example the 1033 in the Jordan Valley that caused massive destruction at Tiberias, the two 1759
events that caused the collapse of the walls of Tiberias and seiche on the Sea of Galilee, the
1837 "Safed earthquake" where 28% of the population of Tiberias were killed and city walls
destroyed again. Of particular interest is the earthquake of 18 January 749 that almost totally
destroyed the towns near the Sea of Galilee, including Tiberius, Sussita (Hippos), and Bet
Shean. The surface faulting dated to 749 was found in archaeological excavations in Tiberius
(Marco et al. 2003) as well as in paleoseismic trenches north of the Dead Sea (Reches and
Hoexter 1981), indicating a magnitude of at least 7 for that rupture along the Jordan Valley.
Figure 6.7: The main faults in Tiberias area
46
6.3 Ground motion
Reliable geological and seismological information are the basis for an appropriate simulation.
The analyzed earthquake is usually given in broad terms involving the location, magnitude and
sometimes the rupture length. Attenuation models provide the severity of the ground motion at
each point (census tract in the case of HAZUS) in respect with the source magnitude and
mechanism, distance to the epicenter and local soil effects. Deterministic hazard assessments
calculate the spatial distribution of the earthquake ground motion that results from a given
(scenario) earthquake. In Hazus, the potential damage for aggregated building and
demographic information, and the ground motion at the given census tract are calculated at the
centroid point of the tract (Tantala et al. 2008).
The ground motion calculations in present the study were based on the soil amplification
factors provided in Israel Building Code (IS 413) 1995/2004, amendment no. 3 (2009) and by
the attenuation equation for strike-slip faults suggested by Campbell and Bozorgnia, (2008),
which were recommended for use in Israel (Klar et al. 2011). For estimating the site effects in
Tiberias, the present study first considers the site class-map of Katz et al. (2008), which is based
on the classification system recommended by the National Earthquake Hazard Reduction
Program (NEHRP) (FEMA 1997). This method is based on the average shear wave velocity of
the upper 30 m (Vs30) in a given site and has been accepted in Israel (IS 413, 1995/2004,
amendment No. 3, 2009, and 5, 2013) and used as the standard method in many other countries
of the world (Kanl et al. 2006; Rošer and Gosar 2010; Motazedian et al. 2011).
According to Zaslavsky, (2009), in Tiberias area the amplification factor mostly ranges
between 2 to 4.5. Therefore, to increase the accuracy of the ground motion calculations, the
amplification factors and the shear wave velocity of the upper 30 m were modified in
accordance with the acceleration spectra that were established based on H/V
(horizontal-to-vertical) spectral ratio (the Nakamura’s technique) for different sites in Tiberius
47
(Zaslavsky, 2009).
6.4 Demographic loss
Accurate and well organized demographic data is the basis for proper casualty estimations
which is done at the census tract level. The building stock for each tract is distributed into basic
groups of residential, commercial, educational, industrial, and lodging.
Specific information on the demography, such as the number of the residential population
during the day and that at night, enables a better estimate of the casualties. In order to provide
advance estimates of demographic (e.g., number of shelter needs) and economic losses to the
governmental authorities, it is necessary to assign complete demographic data to the census
tracts (FEMA, Hazus MR. 2.1 technical manual, chapter 13). Since the demographic data
available in Tiberias were incomplete (e.g., distribution in the census tracts of those working in
commercial and industrial occupancies based on the time of day), the process of completing the
demographic data was carried out by data collection from the Central Bureau of Statistics and
calculations based on a number of statistical parameters (e.g., population density and
population working in the commercial and the industrial sectors; number of visitors in the
hotels, etc).
The number and severity of casualties are strongly related to the extent of both the structural
and non-structural building damage (Erdik et al. 2005). Therefore, in Hazus and in other
software, one of the major inputs necessary for earthquake casualty estimation is the correlation
between the number and severity of injuries and the damage level of the structures (Coburn and
Spence 1992; Seligson and Shoaf 2003).
The output of casualty estimate breaks down into four severity levels of injury: (1) minor
injuries; (2) serious but non-life threatening injuries; (3) serious and life threatening injuries;
and (4) fatalities.
48
6.5 The classification of the building
The area of Tiberias covers about 10.7 km2 and the estimated number of citizens for 2012 is
about forty-five thousand inhabitants.
Tiberius is divided into 12 census tracts (Figure 6.8) and the data (Survey of Israel, 2008)
included the distribution of buildings by five attributes: 1. construction year; 2. Floor area; 3.
Number of stories; 4. Occupancy (residential, commercial; industrial; religious; governmental;
and educational) (Fig. 6.9).
A detailed field survey was carried out in six census tracts and focused on data collection and
engineering documentation of the building inventory. Based on the preliminary findings and the
fact that the Israel Standard (413) for seismic design of structures was implemented only in
some buildings constructed between 1985 and 1995 and in all of the buildings constructed since
1995, it appears that most of the building in these tracts are residential when (1) ~54% of the
buildings were built before 1964 with no earthquake resistance; (2) ~14% of the buildings were
built between the years 1965 and 1974 and they are also with no earthquake resistance; (3) ~5%
of the buildings were built between 1975 and 1985 with low earthquake resistance; (4) ~17% of
the buildings were built between 1985 and 1994 with moderate earthquake resistance, and (5)
~10% of the buildings were built after 1995 with high earthquake resistance. It is estimated that
~85% of the buildings were constructed of concrete frames with unreinforced masonry infill
walls where multi family dwelling buildings are “soft story buildings” with slight earthquake
resistance.
49
The present study suggests taking a conservative approach as a whole, and takes into
consideration the fact that all the pre-1975 buildings are of weak construction, and thus will not
withstand a strong earthquake event.
To provide a first damage estimation based on the above data and observations, we used the
following group of building distributions:
C3LP-MC (3%), C3MP-MC (1%), C3LP-LC (22%), C3MP-LC (9%), C3LP-PC (45%),
C3MP-PC (20%). We think that In this case the building type distribution would reflect the age
distribution and the strength of the buildings in cities along the DST such as Tiberias.
50
Figure 6.8: The census tracts in Tiberias.
Figure 6.10: The building age distribution in Tiberias.
Figure 6.9: The building occupancy distribution in Tiberias.
0%
10%
20%
30%
40%
50%
60%
70%
Residential Commercial Industrial Religious Governmental Educational Agriculture
66%
22%
2% 3% 2%
6%
0%
51
6.6. Earthquake scenarios along the DSF system
Twelve synthetic earthquake scenarios were designed along the main four faults for
damage and loss estimations in Tiberias (Fig. 1a). The earthquake scenarios are based on the
identified seismogenic zones (Shapira 2002; Shapira et al. 2007; IS 413; more ref) and the map
of ‘Active’ and ‘Potentially Active’ faults in Israel (Sagy et al., 2012). Four scenarios were set
along the Jordan Fault following Mw=6.0, Mw=6.5, Mw=7 and Mw=7.5 earthquakes with an
estimated recurrence times of ~500 years, ~1,000, 5,000 and 12,500 years, respectively. The
repeat time suggested hereby follows roughly the estimates that were given by Begin (2005).
Two scenarios were set along the HaOn Fault following Mw=6.0 and Mw=6.5, quakes with an
estimated recurrence times of ~500 years, ~1,000, 5,000 and 12,500 years. Two scenarios were
set along the Poria Fault following Mw=6.0 and Mw=6.5 events with an estimated recurrence
times of ~2,000, 10,000, respectively. Three additional scenarios were set along the Almagor
Fault according to Mw=5.0, Mw=6 and Mw=6.5 events with an estimated recurrence times of
~200 years, ~2,000 and 10,000 years, respectively. Finally, a single scenario was set along the
Beit HaKerem Fault following Mw=6 earthquake with an estimated recurrence time of ~50,000
years. The fault mechanism for all the twelve scenarios was set as a strike slip at a shallow depth
of 8 km. To test the sensitivity of the new fragility curve (C3L-L) we run the Jordan 6.5= Mw
and the Jordan 7=Mw scenarios with applying its new parameters into HAZUS.
6.7 Earthquake scenarios for testing the sensitivity of the casualties' matrix
The estimates of the number of injured following an earthquake are within a large
uncertainty. Presumably, this uncertainty will grow mainly in countries, such as the State of
Israel, that fortunately have no experience with strong earthquakes accompanied by
significant damage.
In order to test the sensitivity of the factors in the “new casualties matrix” (Kaplan and
Yilmaz, 2007,) six different runs were carried out in the set of scenarios. In all of the runs the
52
structures in Tiberias were determined as type C3, in accordance with the definition of the
new matrix (C3). Arbitrarily, in the first run (#1) the magnitude of the earthquake was
determined at MW=6 and the fault was located around 10 km away from Tiberius (depth 2
km). In the next two runs (#2, #3) the earthquake’s magnitude was set at MW=6 and MW=7,
respectively, and the fault was located around 3 km away from Tiberias (depth 2 k"m). These
three runs were carried out using the casualties matrix suggested by FEMA. In the three
additional runs (#4, #5, #6) the new casualties matrix (Kaplan and Yilmaz, 2007) was used in
the set of scenarios in accord with the different levels of damage. The earthquake magnitudes
in these runs were similar to those in the first three runs.
Although the present study did not focus on the subject of collapse rates of buildings,
four additional runs were prepared especially by multiplying the collapse rates factor of 2, 2.5,
3 and 3.5. The simulation of an earthquake in these runs was similar to runs #3 and #6.
1. In the new casualties' matrix, the factors of slight damage from light injury to killed are: 0.05,
0.005, 0, and 0, respectively. In the FEMA casualties' matrix the factors are: 0.05, 0, 0, and 0,
respectively. In fact, no difference.
2. In the new casualties' matrix of moderate damage the factors from light injury to killed are:
0.2, 0.02, 0, and 0, respectively. In the FEMA matrix the factors are: 0.2, 0.025, 0 and 0. In
fact, no difference.
3. In the new casualties' matrix of extensive damage the factors from light injury to killed are: 1,
0.5, 0.01, and 0.01, respectively. In the FEMA matrix the factors are: 1, 0.1, 0.001, 0. 001.
Changes in the severity 3 and the severity 4 levels injuries by one order.
4. In the new casualties' matrix of complete damage with no collapse the factors from light
injury to killed are: 10, 8, 0.1, and 0.1, respectively. In the FEMA matrix the factors are: 5, 1,
0.01, and 0.01. Changes in the severity 3 and the severity 4 levels injuries by one order.
53
1. In the new casualties' matrix of complete damage with collapse the factors from light injury to
killed are: 50, 15, 10, and 10, respectively. In the FEMA matrix the factors are: 40, 20, 5, and
10, respectively. A “slight” change in the severity 3 injury level.
6.7.1 Casualties' matrix results
Table 6.2: General Analysis
1) As evident in Table 6.2, at severity 1 and severity 4 injuries levels there are no significant
differences between the simulation of the injured based on the FEMA matrix and that based
on the “new” matrix (for example, runs 2 and 5).
2) There are differences in severity 2 and severity 3 injuries levels between the simulation of the
injured based on the FEMA matrix and that based on the “new” matrix but they are less than
an order. It should be mentioned that the difference in the factors at severity 3 level injury
(between the FEMA matrix and that of the ‘new matrix’ for extensive damage) is of one order
although in the final results the differences between the two matrixes were much less
significant
3) The differences between the FEMA and “new” matrixes in severity 2 and severity 3 levels
injuries are blurred when an earthquake and damages are both moderate (Table 6.2, runs 1 and
4, and Fig. 7.7).
Table 6.3
1) The factor of the collapse rate (doubling of the collapse rate) affects strongly the estimate
of the numbers of injured.
2) A change in the factor of the rate of collapse in more than 1.5 leads to injured numbers
increase by an order (Fig. 7.7 below).
General
Calculation of the injured in Hazus is a complex procedure. In general, the calculation is
divided into the following steps:
54
1) Total area of a structure
2) Number of persons in the total area of a structure (density)
3) Dividing of the total area by the damage probabilities factors
4) Number of persons for every structure area by damage level (i)
5) Calculation of the injured for every level of damage (i) with the casualties matrix
6) Calculation of the injured in the complete damage level by the collapse rate
Because of the above complex procedure in Hazus an increase in a specific factor in the
casualties' matrix even by one magnitude does not necessarily bring to an increase or decrease
in the final number of injured by one order or even by a significant change. The calculations
show that a significant change in the injured levels 3 and 4 will be greatly dependent on the
amount of area damaged completely or by the size of the collapse area.
Furthermore, it seems that the changes in the number of injured can stem directly from: 1)
population density; 2) division of the area by level of damage expected (basing on fragility
curves); and 3) percent of structure collapse.
55
Table 6.2: Sensitivity analysis of casualty matrices in casualty estimation
Magnitude
(Mw)
No. of
completely
damaged
structures
Severity 1
(uncertainty)
Severity 2
(uncertainty)
Severity 3
(uncertainty)
Severity 4
(uncertainty)
Matrix
1a 6, epicenter
10 km from
Tiberias
34 94 (50-190 19 (10-40) 3 (<20) 5 (<20) FEMA, 2003
2a 6, epicenter
3 km from
Tiberias
660 1110
(600-2000)
325
(160-700)
53
(30-110)
106
(50-200)
FEMA, 2003
3a 7, epicenter
3 km from
Tiberias
2079 3101
(1600-6000)
1007
(500-2000)
173
(90-300)
344
(170-700)
FEMA,
2003
1b 6, epicenter
10 km from
Tiberias
34 104
(50-200)
46
(20-90)
5 (<20) 5 (<20) Erdik et al,
2003
2b 6, epicenter
3 km from
Tiberias
660 1304
(700-3000)
595
(300-1200)
112
(60-200)
112
(60-200)
Erdik et al,
2003
3b 7, epicenter
3 km from
Tiberias
2079 3874
(1900-8000)
1789
(900-4000)
362
(180-700)
362
(180-700)
Erdik et al,
2003
56
Table 6.3: Sensitivity analysis of casualty estimation model according to ‘collapse rate’
Magnitude
(Mw)
Collapse
rate (%)
Severity 1
(uncertainty)
Severity 2
(uncertainty)
Severity 3
(uncertainty)
Severity 4
(uncertainty)
1 7 13 3874
(1900-8000)
1789
(900-4000)
362
(180-700)
362
(180-700)
Erdik et al,
2003
2 7 25 5238
(3000-10000)
2430
(1300-5000)
700
(400-1400)
700
(400-1400)
Erdik et al,
2003
3 7 32 5967
(3000-12000)
2557
(1300-5000)
880
(400-1800)
880
(400-1800)
Erdik et al,
2003
4 7 38 6601
(3000-13000)
2668
(1300-5000)
1037
(500-2000)
1037
(500-2000)
Erdik et al,
2003
5 7 46 7330
(4000-15000)
2796
(1400-6000)
1218
(600-2000)
1218
(600-2000)
Erdik et al,
2003
General conclusions
1) It is suggested to conceder applying the new casualties' matrix (Kaplan and Yilmaz, 2007)
in Israel since that for an earthquake of large magnitude the injured number of in severity 3
level expected be larger than that of FEMA matrix
2) Following the many uncertainties in the parameters used in calculating the number of
injured people in an earthquake, these estimates should be presented in a wide range of scale
(e.g., 0-99, 100-499, 500-999, 1000-4999, 5000-9999…), as it is already done in a number of
other countries.
3) Sever absolute injured number for specific area may calculate by using duplication factors
(e.g., number of standard deviations).
57
6.8 Simulations results
6.8.1. Jordan fault scenarios
The Jordan fault scenarios (Table 6.4 below) show that the number of buildings expected to be
'extensively and completely damaged' ranges between 423 and 3100, depending on the
earthquake magnitude. In the Jordan fault scenarios the weight of debris ranged between
0.3×106 and 2.9×10
6 tons. Accordingly, the number of people who are expected to sustain
severity 2 and 3 injuries ranges between 4 to 610 and the maximum number of estimated
fatalities range between ~20 up to 600 (Table 6.3). The number of persons requiring shelter
ranges between 1,700 and 15,000. Jordan fault scenarios show that the maximum total direct
economic loss (maximal numbers) ranges between 1.3 and 10 billion dollars (Figs., 6.11, 6.12,
and 6.13)
58
Slight
Moderate
Extensive
Complete
1 dot = 2 Bldg
Figure 6.11: Building damage in Jordan Mw=7 scenario for building type, C3M
1 dot = 10 House holds
1 dot = 2000
ton
Figure 6.12: Weigh of debris distribution in Jordan Mw=7 scenario
Figure 6.13. Displaced households in Jordan Mw=7 scenario
59
6.8.2. HaOn fault scenarios
The HaOn fault scenarios (Table 6.4 below) show that the number of buildings expected to be
'extensively and completely damaged' ranges between 209 and 658, depending on the
earthquake magnitude. In the HaOn fault scenarios the weight of debris ranged between
0.18×106 and 0.53×10
6 tons. Accordingly, the number of people who are expected to sustain
severity 2 and 3 injuries ranges between 1 to 75 and the maximum number of estimated
fatalities range between 20 and up to 30. The number of persons requiring shelter ranges
between 808 and 2663. The maximum total direct economic loss ranges between 0.7 and 1.9
billion dollars.
6.8.3. Poria fault scenarios
The Poria fault scenarios (Table 6.4) show that the number of buildings expected to be
'extensively and completely damaged' ranges between 1336 and 2361, depending on the
earthquake magnitude. In the Poria fault scenarios the weight of debris ranged between 1×106
and 2×106 tons. Accordingly, the number of people who are expected to sustain severity 2 and 3
injuries ranges between 28 to 784 and the maximum number of estimated fatalities range
between 110 and up to 300. The number of persons requiring shelter ranges between 5836 and
10,915. The maximum total direct economic loss ranges between 1.9 and 7.1 billion dollars.
6.8.4. Almagor fault scenarios
The Almagor fault scenarios (Table 6.4 below) show that the number of buildings expected to
be 'extensively and completely damaged' ranges between 29 and 1725, depending on the
earthquake magnitude. In the Almagor fault scenarios the weight of debris ranged between
0.04×106 and 1.44×10
6 tons. Accordingly, the number of people who are expected to sustain
severity 2 and 3 injuries ranges between 0 to 408 and the maximum number of estimated
fatalities range between less than 20 and up to 190 hundreds. The number of persons requiring
60
shelter ranges between 124 and 7834. The maximum total direct economic loss ranges between
0.2 and 5.1 billion dollars.
6.8.5. Beit HaKerem fault scenario
The Beit HaKerem fault scenario (Table 6.4) shows that the number of buildings expected to be
'extensively and completely damaged' is 174 and the weight of debris is 0.15×106 tons.
Accordingly, the number of people who are expected to sustain severity 2 and 3 injuries is 9, the
maximum number of estimated fatalities is less than 20 and the number of persons requiring
shelter is 739. The total direct economic loss expected to be is 0.6 billion dollars.
61
Scenario
Casualty Estimates Expected Building Damage
Shelter
needs
Economic
Loss
Debris
Generation
(People) (No of Bldg) (People)
(Billions of
dollars)
(million
tons)
Level
1
Level
2
Level
3
Level
4
Max Slight Moderate Extensive Complete
Jordan
6.0 Mw
148 29 4 7 < 20 1,086 925 376 47 1,712
0.30 -
1.30
0.35
Jordan
6.5 Mw
332 80 12 23 50 1,073 1,237 659 154 3,328
0.60 -
2.30
0.66
Jordan
7.0 Mw
923 263 43 84 170 794 1,423 1,096 550 7,161
1.20 -
4.80
1.4
Jordan
7.5 Mw
2,675 858 147 291 600 278 979 1,282 1,819 14,783
2.50 -
10.00
2.91
HaOn 6.0
Mw
67 10 1 2 < 20 953 619 196 14 808
0.20 -
0.70
0.18
Table 6.4: Summary results of the damage and loss simulations
62
HaOn 6.5
Mw
250 56 8 15 30 1,100 1,143 554 104 2,663
0.50 -
1.90
0.53
Poria 6.0
Mw
663 179 28 55 110 911 1,432 977 359 5,836
0.50 -
1.90
1.07
Poria 6.5
Mw
1,624 494 83 164 300 514 1,318 1,329 1,031 10,915
1.80 -
7.10
2.02
Almagor
5.0 Mw
12 1 0 0 < 20 456 165 29 0 124
0.10 -
0.20
0.04
Almagor
6.0 Mw
271 61 8 17 30 1,107 1,190 584 104 3,015
0.50 -
2.00
0.54
Almagor
6.5 Mw
1,010 290 47 93 190 759 1,435 1,145 580 7,834
1.30 -
5.10
1.44
Bet
HaKerem
6.0 Mw
56 8 1 1 < 20 942 573 166 8 739
0.10 -
0.60
0.15
63
Chapter 7: Discussion and Conclusion
7.1 A survey of Earthquake casualty matrices used worldwide
The result of the survey, held in four languages (English, Spanish, Chinese and Hebrew), are
summarized in Table 7.1. The casualty rates were sought mainly for reinforced concrete (RC)
structures, which are the common type of structures in Israel (Yankelevsky et al. 2011). Seven
matrices were identified, in addition to the matrix used in HAZUS-MH (Erdik et al. 2003;
Aleskerov et al., 2005; Coburn & Spence, 2002; Zuccaro & Cacace, 2011; Yeh 2003; DGPC
& IGN, 2002; Avirav 2011). The English survey resulted in the highest number of matrices
(n=5) from four countries (USA, Turkey, New Zealand and Italy); the rest of the surveys
resulted each with one matrix only (total, n=3) from Taiwan, Spain and Israel. Most of the
matrices were developed based on historical data (KOERILoss, DSS-DM, Coburn & Spence
2002, and TELES) or with a combination of historical data and experts’ opinion
(HAZUS-MH). The rest of the matrices were developed based on experts’ opinion solely
(SES 2002 and Veronic).
64
Table 7.1: Casualty distribution for collapsed RC structures (%) in the different models
Model Country Method
Severity 1
Severity
2
severity
3
Severity
4
1. HAZUS-MH (FEMA
2003)
USA
Experts opinion (ATC
13) + Limited
Historical data
40 20 10 5
2. KOERILoss (Erdik et
al. 2003) Turkey
Historical data 50 15 10 10
3. DSS-DM (Aleskerov et
al. 2005) Turkey
Historical data 30 11.5 8 6
4. Coburn & Spence 2002
New
Zealand Historical data 70* 9.2 0.8 21
5. Zuccaro & Cacace 2011 Italy N/A 50 30
6. TELES (Yeh 2003) Taiwan Historical data 8 6.4 4.8 4
7. SES 2002 (DGPC &
IGN, 2002)
Spain
Experts opinion
(originally based on
ATC 13)
40 40 20
8.
Veronic (Avirav 2011)
Israel
Experts opinion
(originally based on
ATC 13)
40 20 20 20
*Severity 1 and uninjured
65
7.2 Tiberias’ population characteristics and distribution
Tiberias is originally divided to 15 census tract according to the Israeli Bureau of Statistics
(IBS). Some modification was needed to adjust discrepancies between this division and the
one used in the HAZUS-MH building stock catalogues. For this reason, several tracts were
united to form a new division of 12 census tracts in total that will accommodate both
information catalogues (this will be detailed below). The original distribution is described in
Figure 7.1 below.
Figure 7.1: Map of the city of Tiberias and its census tracts according to IBS
The map shown in Fig 7.1. also demonstrates the Socioeconomic Index (SEI) rank of each
tract. The SEI is published by the Israeli central bureau of statistics and classifies
geographical units (census tracts in municipalities) by elements such as demographics,
education, employment and standard of living. The SEI ranges from 1-20 (when cluster 1
indicates the lowest level of socioeconomic rank). The SEI rank of Tiberias (to each census
tract separately) ranges from 6 to 10. The yellow colored tracts in the map shown in Fig 7.1.
are ranked between 8-10, and the green colored tracts are ranked lower between 6-7.
66
The gender and age distribution in each of the 12 tracts is described in Table 7.2 below.
Table 7.2: Demographic Characteristics of Tiberias’ residents according to census tracts
Census
tract
No. of
Residents
No. of
House-h
olds
Male
Population n
(%)
Female
Population
n (%)
Males
aged < 16
n (%)
Males aged
16-65 n (%)
Males aged
> 65 n (%)
Females aged
< 16 n (%)
Females
aged 16-65
n (%)
Females
aged > 65
n (%)
25 3038 975 1526 (50) 1512 (50) 396 (26) 1020 (67) 110 (7) 379 (25) 996 (66) 137 (9)
12 + 13 7023 2077 3531 (50) 3493 (50) 1293 (37) 1997 (57) 241 (7) 1318 (38) 1824 (52) 351 (10)
11 5926 1806 2932 (49) 2994 (51) 923 (32) 1708 (58) 301 (10) 852 (29) 1743 (58) 399 (13)
14 1205 438 626 (52) 578 (48) 205 (33) 341 (54) 90 (13) 154 (27) 331 (57) 93 (16)
21 5122 1666 2584 (50) 2574 (50) 562 (22) 1730 (68) 256 (10) 586 (23) 1673 (65) 315 (12)
22 + 23 4638 1625 2123 (46) 2515 (54) 379 (18) 1492 (70) 252 (12) 579 (23) 1609 (64) 328 (13)
15 1198 334 662 (55) 536 (45) 173 (26) 400 (60) 89 (13) 62 (12) 377 (70) 96 (18)
31 + 32 5467 1572 2721 (50) 2746 (50) 833 (31) 1651 (61) 237 (9) 841 (31) 1609 (59) 296 (11)
24 2152 637 1032 (48) 1120 (52) 379 (37) 635 (61) 18 (2) 420 (38) 690 (62) 10 (1)
36 2584 784 1316 (51) 1268 (49) 354 (27) 812 (62) 150 (11) 327 (26) 658 (52) 283 (22)
34 1974 589 1031 (52) 943 (48) 494 (48) 479 (46) 58 (6) 378 (40) 530 (56) 36 (4)
33 1752 732 870 (50) 882 (50) 309 (36) 467 (54) 94 (11) 231 (26) 504 (57) 146 (17)
The data in Table 7.2. suggests that the different neighborhoods of Tiberias (represented in
census tracts) vary in terms of age distribution. Some of the tracts are populated with a
significantly higher rate of elderly population (>65 years), for example census tracts 36 and
15. The same observation in demonstrated for other characteristics such as physically disabled
population rate and average household income, as show in Figures 7.2 and 7.3 below.
67
Figure 7.2: Average income of households in Tiberias (yearly) according to census tracts.
Figure 7.3: The physically disabled rate of the population in Tiberias according to census tracts
Education levels were also screened, although currently not taken into consideration in the
casualty estimation model. The results indicated that the educational level in Tiberias are
considerably lower compared to the average rate in Israel. While the rate of individuals with
higher education (academic or tertiary) in Israel stands on 46% (OECD, 2014), the average
Average Household Income in Israel
(Yearly) = 152,000 NIS
The average rate of physically disabled
persons in Israel (17.5%)
68
rate among Tiberians is only 16%. This rate ranges from 7% (census tract no. 36) to 31%
(census tract no. 22+23). The data is presented in Figure 7.4.
Figure 7.4: Rates of residents with higher education in Tiberias according to census tracts.
Data regarding the populations’ form of homeownership was collected and presented in
Figure 7.5.
The rate of residents with higher education in Israel (46%)
69
Figure 7.5: Form of home ownership in Tiberias according to census tracts.
The distribution map of Tiberias population included calculation of the number of residents
present at residential dwellings during day and night. The ‘night residents’ population
(defined in HAZUS as NRES) was defined as the number of individuals residing in the census
tract since it was assumed that all of these residents occupy their dwelling structures during
night. The ‘daytime residents’ (defined in HAZUS as DRES) was calculated as a combination
of the individuals that a) did not belong to the workforce, b) were unemployed or c) working
from their homes. The students population was also defined for each census tract (the number
of residents aged between 6-18 years) since it was assumed that most of the students will
occupy educational structures close to their homes. This data is presented in Figure 7.6.
70
Figure 7.6: Distribution of residents in Tiberias in residential and educational structures
during daytime
Similar information was gathered for commercial and industrial workers (number of workers
in Tiberias), and commuters but since their workplace or presence in commuting time is not
necessarily at the same census tract they reside, this was not presented in this chapter. All of
the data above, was integrated into the social information catalogue in HAZUS.
7.3 Estimating casualty rates – the results of the modified ‘Delphi procedure’
Participants
Of the 20 participants in round 1, 100% were male with a mean age of 56 years (SD=9.1). The
professional classification of the participants is detailed in Table 7.3. The average professional
experience of participants was 28 years (SD=9.3).
71
Table 7.3: Professional classification of Delphi panel experts
Profession n % Details
Engineers 14 70 Structural engineers (n=13); rescue engineers
(n=1)
Physicians 1 5
‘Search and Rescue’ members
(SAR)
1 5 SAR unit of the home front command
(IDF)
Both Engineers and SAR
members
2 10
Other 2 10 Geologist (n=1) and Disaster relief expert
(n=1)
Iteration 1 of Delphi
In this iteration, the participants were asked to evaluate the casualty matrix currently used in
HAZUS-MH. The matrix includes three sections: a) Indoor casualties, b) Outdoor casualties
and c) Casualties due to damaged bridges. Each section in the survey displayed the casualty
rates (on a four level severity scale) for each the three modes and according to five structural
damage levels. The results and the ranking of the items are displayed in Table 7.4.
72
Table 7.4: Ranking of the casualty matrix currently used in HAZUS-MH for its suitability to
use in an earthquake scenario in Israel
Item N Mean item
weight (SD)
Agreement
percentage
(%)
1 Indoor casualties in slight structural damage 20 3.78 (1.17) 76
2
Indoor casualties in moderate structural
damage 20 3.19 (0.98) 64
3
Indoor casualties in extensive structural
damage 20 2.67 (1.11) 53
4
Indoor casualties in complete structural
damage (no collapse) 20 2.57 (1.21) 51
5
Indoor casualties in complete structural
damage (with collapse) 20 3.24 (1.09) 65
6 Outdoor casualties in slight structural damage 20 4.05 (0.80) 81
7
Outdoor casualties in moderate structural
damage 20 3.57 (0.87) 71
8
Outdoor casualties in extensive structural
damage 20 2.90 (1.04) 58
9
Outdoor casualties in complete structural
damage 20 3.05 (1.16) 61
10
Casualties due to major or continuous bride
damage (complete) 20 3.45 (0.94) 69
11
Casualties due to single span bride damage
(complete) 20 3 (0.77) 60
73
As can be seen from the results of round 1, a number of item did not reach the desirable level
of agreement among experts (>70%). The experts stated in their comments that the current
casualty rates may underestimate the potential number of casualties in an earthquake event in
Israel. These results necessitated a continuation of the procedure to a second round.
Iteration 2
In the second iteration, an alternative casualty matrix was selected from the matrices
described in Table 7.5. The selected matrix was a Turkish matrix by Erdik et al. used for
earthquake casualty estimation in the KOERILoss model. The casualty rates in this matrix are
higher compared to HAZUS-MH (see Table 7.5 below) and are based on historical data rather
than on experts’ opinion. The alternative matrix was presented to the experts in round 2 and
again they were asked to rank its suitability to use in earthquake casualty estimations for
Israel on a five point Likert scale. This round was also conducted via an electronic survey, and
completed by 17 experts (response rate = 85%). The mean weight of the experts score was
3.53 (SD=0.94), indicating 71% of agreement between the different experts. Following this
result, the Delphi procedure was concluded and the alternative matrix was accepted.
74
Table 7.5: Casualty matrices used in HAZUS-MH model and KOERILoss model (alternative
matrix) for reinforced concrete structures
Slight
damage
Moderate
damage
Extensive
damage
Complete
damage
(no
collapse)
Complete
damage
(with
collapse)
HAZUS-MH
Severity 1 - light injury 0.05 0.25 1 5 40
Severity 2 - moderate injury 0 0.03 0.1 1 20
Severity 3 - severe injury 0 0 0.001 0.01 5
Severity 4 - fatal injury (fatalities) 0 0 0.001 0.01 10
KOERILoss (Erdik et al, 2003)
Severity 1 - light injury 0.05 0.2 1 10 50
Severity 2 - moderate injury 0.005 0.02 0.5 8 15
Severity 3 - severe injury 0 0 0.01 0.1 10
Severity 4 - fatal injury (fatalities) 0 0 0.01 0.1 10
7.4 Estimating casualties – the case study of Tiberias
In order to assess the number of expected casualties from an earthquake event in the city of
Tiberias and assess the sensitivity of the casualty estimation model to different casualty rates,
six simulations were conducted. The results of the simulations are described in Table 6.2.
75
The results depicted in Table 6.2 indicate that the largest change in casualty numbers between
the two matrices is attributed to severity 2 and 3 casualties (moderate and severe injuries).
This effect is weakened when the earthquake is moderate which causes relatively moderate
damage levels (for example, simulations 1a and 1b). This is also demonstrated in Figure 7.7.
Figure 7.7: Results of three earthquake scenarios (x axis), showing the number of casualties
expected (severity 2 and 3) (y axis) based on the casualty rates of HAZUS-MH (Blue) and
Erdik et al (Brown), 2003
The sensitivity analysis of the casualty estimation model according to ‘collapse rate’ factor are
described in Table 6.3. The results depicted in table 6.3 demonstrate high sensitivity in
regards to structures ‘collapse rate’. This is demonstrated in all four simulations (2 to 5) and
attributed to all severity levels of injury.
76
7.5 Discussion and Conclusions
The results of this study reinforce the high vulnerability of the study area (city of Tiberias) to
seismic hazards. The exposed population characteristics that are documented in the literature
as increasing the risk for earthquake-related injury and death. Additionally, several parts of the
city (i.e. census tracts) were more vulnerable compared with others (e.g. census tract 36 with
high rates of elderly and disabled population and low levels of household income), this
information needs to be taken into consideration by national and community institutions when
planning for a future earthquakes since it has implication not only to the preparedness process
and to the distribution of resources prior to the event, but also to the response and
rehabilitation phases which requires special attention to the needs of vulnerable groups; such
as increased demand for medical aid by the elderly, the chronically ill, the physically disabled
and by children (Leitmann, 2007). Accurate analysis of vulnerable populations can also be
leveraged for the purpose of pre-disaster risk assessment, for prevention and mitigation of
severe outcomes and for development and implementation of appropriate preparedness and
effective response (UNISDR 2015).
The casualty estimation process in the HAZUS-MH model did not previously take into
account human-related risk factors which may compromise the evaluations’ accuracy (Shapira
et al, 2015). These findings were also reflected in the validation process that was performed to
the casualty matrix currently used in HAZUS-MH, by local experts and revealed that the
majority of experts believed that the current rates might underestimate the casualty numbers
expected from an earthquake event in Israel. The arguments raised by the experts to justify
that position were, among others, that the population in Israel lacks the experience of dealing
with an earthquake event and that the general preparedness level of the population to this
scenario are low; evidence of this were found in a recent study that reported relatively low
awareness among Israeli population regarding earthquake’s potential risks and also of low
rates of compliance with recommended preparedness measures (Shenhar et al. 2015). The
77
results of the casualty estimation simulations for the city of Tiberias demonstrate the
differences in the expected number of casualties for three earthquake scenarios when utilizing
the original casualty rates used in HAZUS and for an alternative matrix. A substantial increase
in the number of casualties was observed mainly for the scenarios that caused greater damage
to structures (i.e. number of completely damaged structures; scenarios b and c). When
comparing the two matrices, a substantial increase in the number of casualties was attributed
mainly to moderate and severe injuries. The finding regarding the amount of damaged
structures as highly influential in terms of predicting casualty numbers is expected. Both
because this is a well-documented fact in the literature (Jaiswal et al. 2010; Peek-Asa et al.
1998; Noji & Kelen 1990; Spence et al. 2003; Wald et al. 2011; Porter et al. 2008; Jaiswal et
al. 2009) and since the current HAZUS social loss model is engineering-based and does not
take into account other human related factors when estimating casualties (Shapira et al, 2015).
However, the finding that suggest that the extant of moderate and severe casualties might
increase (in some simulations almost doubled) should a strong earthquake strike our region,
has an important clinical significance in terms of preparedness efforts. These casualties
require greater degree of medical care and the time of their evacuation to a hospital is critical
for their chances of survival and for avoiding further complications that might results in
various disabilities (Bayard 2010); the health system institutions (whether in the pre-hospital
or hospitalization settings) could benefit greatly from these projections in terms of adjusting
current emergency plans for this scenario (e.g. personnel augmentation, expansion of hospital
surge capacity, stockpile of medical supplies etc.) to increase their preparedness. The results
of the casualty estimation are presented in absolute numbers accompanied by a ‘range of
uncertainty’. This is due to fact that a high degrees of uncertainty is involved in
loss-estimation processes and models, which is a well-documented matter in the literature
(Kircher et al. 2006; Eguchi & Seligson, 2008; Remo & Pinter 2012). However, the relatively
fixed relationship between the number of casualties and the state of damaged buildings
78
assumed by most loss-estimation models and the fact that these models do not account for
factors such as occupants’ characteristics and behavior during the event or the availability and
response capabilities of medical care organizations and institutions may contribute to an even
higher degrees of uncertainty compared with other loss-estimation outputs (i.e. physical
damage to buildings and infrastructures and economic loss) (Spence and So, 2011; Shapira et
al, 2015). This could lead to inappropriate preparedness and may ultimately result in a
shortage or a waste of supplies such as drugs or other medical equipment. In the recent 2015
Nepal earthquake (Mw=7.8) the rapid assessment model (PAGER) estimated the death toll
from the event to be between 1400 and up to 10000 fatalities (OCHA, 2015), the wide range
of estimates hindered the efforts of providing the apt international aid; and indeed, five days
after the occurrence of the quake, the Nepali disaster management head published a message
urging the world not to “dump” inappropriate aid that creates logistic bottlenecks (ReliefWeb,
2015).
Limitations
The fact that the state of Israel has no recent experience in dealing with major earthquakes in
its region affects the ability to practically validate our estimates. Loss estimation models
depend greatly on accurate data catalogues but also on calibration according to past events
consequences which are absent in our region. Therefore, these results should be interpreted
cautiously.
Conclusions
Although that the results show that at severity 1 and severity 4 injuries levels there are no
significant differences between the simulation of the injured based on the FEMA matrix and
that based on the “new” matrix, the present study suggests to conceder applying the new
casualties' matrix (Kaplan and Yilmaz, 2007) in Israel since that for an earthquake of large
magnitude the injured number of in severity 3 level expected be larger than that of FEMA
matrix.
79
An alternative model is offered in attempt to improve casualty projection accuracy. In
addition, the population of the city of Tiberias was assessed in terms of vulnerability to
earthquakes and areas in high risk were identified. National and community organization and
other institutions may consider using this information to allocate resources more adequately to
enhance preparedness and mitigate future losses in case a large earthquake strikes our region.
To explore the role of natural hazards in the perspective of building long-term environmental
performance, as well as the environmental value of hazard mitigation, we have proposed an
innovative Life Cycle Assessment (LCA) framework that can incorporate building damage
due to hazards and converting this data into quantifiable environmental metrics (Wei et al.
2014a; b; c; d). Also, by incorporating buildings’ environmental impacts attributable to
hazards as derived from the LCA framework, we arrive at a Benefit-Cost Analysis (BCA) to
justify the environmental and economic desirability of hazard mitigation actions (Shohet et al.
2015; Wei et al. 2015a; b; c). Based on the result of the proposed report, we have also
introduced an innovative Public-Private Partnership framework for property owners, insurers
and governments to facilitate decisions related to hazard insurance and structural retrofit of
vulnerable buildings (Yao et al. 2015). All related published journal and conference papers
can be found in the Appendix.
80
Appendix
Journal Papers
1. Wei, H.H., Skibniewski, M., Shohet, I., Yao, X.J. (2015a), “Life Cycle Environmental
Performance of Natural Hazard Mitigation for Buildings,” ASCE Journal of Performance
of Constructed Facilities.10.1061/(ASCE)CF.1943-5509.0000803 (Accepted).
2. Wei, H.H., Shohet, I., Skibniewski, M., Shapira, S., Yao, X.J. (2015b), “Assessing the
Life-cycle Sustainability Costs and Benefits of Hazard Mitigation Designs for Reinforced
Concrete Buildings,” ASCE Journal of Architectural Engineering (Accepted).
3. Yao, X.J., Wei, H.H., Shohet, I., Skibniewski, M. (2015), “Public-Private Partnership for
Hazard Mitigation Involving Retrofit and Insurance,” Technological and Economic
Development of Economy (Accepted)
4. Wei, H.H., Shohet, I. M., Skibniewski, M. J., Levy, R., Shapira, S., Aharonson-Daniel, L.,
Levi, T., Salamon, A., and Levi, O. (2014a), “Economic Feasibility Analysis of
Pre-earthquake Strengthening of Buildings in a Moderate Seismicity/High Vulnerability
Area, ” Procedia Economics and Finance, 18, 143-150.
5. Wei, H. H., Skibniewski, M. J., Shohet, I. M., Shapira, S., Aharonson-Daniel, L., Levi, T.,
Salamon, A., Levy, R., and Levi, O. (2014b), “Benefit-Cost Analysis of the Seismic Risk
Mitigation for a Region with Moderate Seismicity: The Case of Tiberias, Israel, ”
Procedia Engineering, 85, 536-542.
Peer-Reviewed Conference Papers
1. Shohet, I., Wei, H.H., Skibniewski, M., Shapira, S., Levy, R., Levi, T., Salamon, A.,
Zohar, M. (2015), “Application of semi-empirical model in estimation of economic losses
in natural hazards,” Proceedings, Creative Construction Conference, Krakow, Poland,
June 21-24.
81
2. Wei, H.H., Skibniewski, M., Yao, X.J., Shohet, I. (2015c), “A New Approach for
Building Renovation in Mitigating Natural Hazard Risk through Public-Private
Partnership,” Proceedings, 2nd International Conference on Sustainable Urbanization,
Hong Kong, China, January 7-9.
3. Wei, H.H., Skibniewski, M., Yao, X.J., Shohet, I. (2014c), “Decision Models for
Sustainable Post-Disaster Reconstruction Strategies,” Proceedings, 3rd International
Conference on Urban Disaster Reduction, Boulder, Colorado, September 28-October 1.
4. Wei, H.H., Skibniewski, M., Shohet, I.M., Shapira, S. (2014d), “Catastrophe Risk
Assessment and Management,” Proceedings, Project Management Symposium, College
Park, Maryland, June 9-10.
82
References
Aleskerov, F., Say, A. I., Toker, A., Akin, H. L. and Altay, G. (2005), A cluster-based decision
support system for estimating earthquake damage and casualties. Disasters, 29: 255–276.
doi: 10.1111/j.0361-3666.2005.00290.x
Avirav, V. (2011), Veronic: An Earthquake Loss Estimation Program , Lod, Israel: The
Geophysical Institute of Israel.
Bayard, D. (2010), Haiti Earthquake Relief, Phase Two — Long-Term Needs and Local
Resources. The New England Journal of Medicine, 360(20): 1858-1861.
Coburn, A. and Spence, R. (2002), Earthquake Protection. Second ed. Chichester, England:
Wiley.
DGPC and IGN (2002), SES 2002: Technical Manual, s.l.: Directorate General of Civil
Protection & The National Geographic Institute.
Eguchi, R. T. and Seligson, H. A. (2008), Loss Estimation Models and Metrics. In: A.
Bostrom, S. French & S. Gottlieb, eds. Risk Assessment, Modeling and Decision Support.
s.l.:Springer Berlin Heidelberg, pp. 135-170.
Erdik, M., Aydinoglu, N., Fahjan, Y., Sesetyan, K., Demircioglu, M., Siyahi, B., Durukal, E.,
et al (2003). Earthquake risk assessment for Istanbul metropolitan area. Eearthq Eng Eng
Vib, 2(1): 1-23.
Kircher, C. A., Whitman, R. V. and Holmes, W. T., (2006b), Hazus-MH earthquake-loss
estimation methods. Nat Hazards Rev, 7(2): 45-59.
Leitmann, J. (2007), Cities and calamities: learning from post-disaster response in Indonesia.
J Urban Health, 84(1):144-153.
OCHA. (2015), Nepal: Estimated Population Exposed to Earthquake shaking (as of 25 April
2015). [Online]. Available at:
http://reliefweb.int/map/nepal/nepal-estimated-population-exposed-earthquake-shaking-25-
april-2015. [Accessed 28 April 2015].
83
OECD. (2014), Education at a glance 2014: OECD indicators, s.l.: OECD Publishing.
ReliefWeb. (2015), Nepal urges better international aid coordination. [Online]
Available at:http://reliefweb.int/report/nepal/nepal-urges-better-international-aid-coordination
[Accessed 29 April 2015].
Remo, J. W. and Pinter, N. (2012), Hazus-MH earthquake modeling in the central USA. Nat
Hazards, 63: 1055-1081.
Shapira, S., Aharonson-Daniel, L., Shohet, I.M., Peek-Asa, C., Bar-Dayan, Y. (2015),
Integrating Epidemiological and Engineering approaches in the Assessment of Human
Casualties in Earthquakes. Nat Hazards, DOI 10.1007/s11069-015-1780-0.
Shenhar, G., Radomislensky, I., Rozenfeld, M. and Peleg, K. (2015), The Impact of a
National Earthquake Campaign on Public Preparedness: 2011 Campaign in Israel as a Case
Study. Disaster Med Public Health Prep, 9(2): 138-44.
Spence, R. & So, E. (2011), Human casualties in earthquakes: modelling and mitigation.
Proceedings of the Ninth Pacific Conference on Earthquake Engineering Building an
Earthquake-Resilient Society. Auckland, New Zealand, paper number 224.
UNISDR. (2015), Sendai framework for disaster risk reduction 2015-2030, Sendai, Miyagi,
Japan: United Nations International Strategy for Disaster Reduction.
Yankelevsky, D., Schwarz, S., Leibovitch, E. and Ofir, Y. (2011), Basis for Preparation of
Databases Related to Existing Buildings in Israel. The Ministry of National Infrastructures
and the Technion Research and Development Foundation, Haifa, Israel.
Yeh, C. H. (2003), Taiwan Earthquake Loss Estimation System – TELES. National
Earthquake Engineering Research Center, Taipei, Taiwan.
Zuccaro, G. and Cacace, F. (2011), Seismic Casualty Evaluation: The Italian Model, an
Application to the L’Aquila 2009 Event. In: R. Spence, E. So and C. Scawthorn, eds.
Human Casualties in Earthquakes: Progress in Modelling and Mitigation. Springer, pp.
171-184.
84
תקציר
שיכוך התוצאות של רעידות אדמה פוטנציאליות מחייב הערכה של הנפגעים בגוף ובנפש כתוצאה מרעידות
פוטנציאליות. סקירה ספרותית מקיפה של ההשלכות של אירועי רעידות אדמה מרובי נפגעים, הראתה כי
הפרמטרים הבאים הם המשמעותיים בהערכת היקף הנפגעים בגוף ובנפש:
ם סייסמיים: שברים גיאולוגיים פעילים או חשודים כפעילים, התנזלות קרקע, גלישות קרקע . סיכוני1
והגברות של תנדות הקרקע;
. פגיעות של בניינים לסיכונים סיסמיים, באמצעות סימולציה של אירועים סייסמיים, הכוללת בסיסי 2
נתונים של מבנים, ונתונים גיאולוגיים וסייסמיים;
אקונומיים באזור הנחקר בכדי להעריך את האוכלוסיה החשופה לריוע -סוציו -נאים ה. הערכה של הת3
בתבניות זמן שונות;
ישראל מצויה באזור החשוף לאירועי רעידת אדמה. מחקרים שנערכו בעשורים האחרונים לא לקחו
רות זמן שונות בחשבון את סוגי המבנים הספציפיים באזור, את פרופיל נוכחות האוכלוסיה בבניינים ובמסג
אמפירית להערכת -)יום, לילה ותנועה לעבודה(. המחקר הנוכחי כולל פיתו ויישום של גישה אנליטית
הנזקים בגוף ובנפש כתוצאה מרעידת אדמה. העיר טבריה וסביבתה נבחרו כמקרה בוחן לפיתוח המודל,
רעידות אדמה עם תקופות תרחישים שונים. של 12שישמש בהמשך כמודל להערכת נזקים במדינה. נבחנו
חזרה שונות. ניתוח התרחישים מאפשר לצוות המחקר להעריך את מעטפת הסיכונים הנגרמת כתוצאה
מאירועירעידת אדמה ולפתח אסטרטגיות מוכנות של המבנים והאוכלוסיה לאירועים אלה.
המחקר השיג את אבני הדרך הבאות:
אקונומיים, גיאולוגיים ומבניים;-רמים סוציו. סקר ספרות מקיף של גורמיי רפואת חירום, גו1
.HAZUS. מיון של סוגי המבנים הקיימים לפי תקנים ישראליים בזמן הבנייה ולפי סיווג מתודולוגיית 2
אקונומיים;-. מיפוי האוכלוסיה בתבניות זמן שונות ובהתייחס למאפיינים דמוגםריים וסוציו3
להערכת הנזקים והתואם את חתך המבנים בישראל;אנליטי -. פיתוח והתאמה של מודל אמפירי4
. יישום תהליך דלפי להשגת קונסנסוס בין מומחים לגבי מטריצת הנפגעים במצבי נזק שונים של המבנים. 5
ללא נזק ועד התמוטוות כוללת של המבנה; -הוגדרו חמישה מצבי נזק
(;VULNERABILITYעות המבנים )באמצעות הערכת סיכונים מבניים ופגי -. הערכת הנזקים לבניינים 6
. הערכת הנפגעים באמצעות התאמת מטריצת הנפגעים בתהליך דלפי;7
. תיקוף הממצאים באמצעות השוואת הממצאים לממצאי מחקרים אחרים ונזקים ברעידות מקבילו 8
.1999 -טורקיה ב -כדוגמת רעידת האדמה באיזמיט
עצימות לא גבוהה של אירועים סייסמיים, הפגיעות המחקר מלמד כי למרות שישראל נמצאת באזור בעל
של המבנים והאוכלוסיה לאירוע רעידת אדמה גבוהים וכי יש לכין תכנית היערכות מפורטת ללא דיחוי.
2
אמפירי להערכת הקורבנות בגוף ובנפש בעיר גדולה -מודל אנליטי
טבריה כמקרה בוחן -בישראל כתוצאה מרעידת אדמה
חוקרים ראשיים:
4 , צפריר לוי2,3 , לימור אהרנסון דניאל1 יגאל שוחט )מרכז המחקר(
חוקרים:
2 דיין-וירון בר 2,3 יני, ברוריה עד1 , אורן וילנאי1 , דוד אורנאי4 , עמוס סלמון1 רוברט לוי
עוזרי מחקר:
Hsi-Hsien Wei 1,51,4 ואוהד לוי 1 , סתיו שפירא
.גוריון בנגב-המחלקה להנדסת בניין, אוניברסיטת בן 1 , ולטה למדעי הבריאותהפקהמחלקה לרפואת חרום, בית ספר למקצועות הבריאות ע"ש רקאנטי, 2
.גוריון בנגב-אוניברסיטת בן .גוריון בנגב-בן , אוניברסיטתלחקר המוכנות והמענה למצבי חירום ואסון PREPARED מרכז 3 .המכון הגיאולוגי 4 .וניברסיטת מרילנד, קולג' פרק, ארה"בא, ה אזרחית וסביבתיתהמחלקה להנדס 5
3-9618מחקר מענק
GSI/21/2016 דו"ח מס' 2016 יולי, ירושלים
משרד התשתיות הלאומיות האנרגיה והמים
המכון הגיאולוגי
גוריון-ןאוניברסיטת ב
בנגב