BUSHFIRE RISK ASSESSMENT AT THE URBAN-BUSH INTERFACE (UBI) IN SYDNEY, AUSTRALIA: AN INTEGRATED MODELLING APPROACH
Daminda Thushara Solangaarachchi BSc (Hons), MSc (GIS & RS)
A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy
School of Physical, Environmental & Mathematical Sciences The University of New South Wales Australian Defence Force Academy Canberra, ACT, 2600, Australia 31 August 2012 ii
ORIGINALITY STATEMENT
‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis.
I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’
Signed ……………………………………………......
Daminda Thushara Solangaarachchi
31 August 2012
iii
iv
ABSTRACT
Bushfire Risk Assessment at the Urban-Bush Interface (UBI) in Sydney, Australia: An Integrated Modelling Approach
Bushfires are one of the major threats to the environment and human systems in Australia. The recent 2009 Black Saturday bushfires in Victoria claimed more than 2,029 homes and 173 lives, and demonstrated that fire management authorities need to rethink their current risk and emergency management approaches. Rapid population growth and land use change at the urban-bush interface combined with favourable weather conditions for bushfires are causing a rapid increase in vulnerability of communities exposed to bushfires. Identifying vulnerability and risk before an event occurs are essential steps towards efficient and effective risk management. Global initiatives such as the United Nations International Decade for Natural Disaster Reduction (IDNDR) and Hyogo Framework for Action (HFA) have highlighted the importance of research for formulating the overall value of disaster risk reduction through national and local risk assessments. In order to achieve that, it is necessary to ‘measure’ the existing level of risk and the potential future risk that may be encountered in future bushfire events. In Australia, various institutions and agencies have developed a variety of bushfire risk assessment models. However, many of these models focus primarily on the hazard component of risk, which is mainly based on physical factors such as weather, fuel, and topography. A risk assessment model that integrates both the hazard component and the elements of vulnerability such as social vulnerability, physical vulnerability, and emergency response and coping capacity is yet to be developed. Risk assessments often use objective, quantifiable approaches. However, assessing the objective level of risk itself is not enough for efficient risk management decision-making. Understanding subjective judgements of residents living at urban-bush interface, the factors affecting their decisions and their perceptions of bushfire risk and attitudes towards current bushfire management strategies is also an important step towards effective bushfire risk management. Despite the considerable effort that has been directed towards encouraging bushfire preparedness in Australia, research on public perceptions of bushfire management strategies to reduce bushfire risk is relatively rare.
v
This thesis develops a multifaceted understanding of vulnerability and risk based on a holistic approach to risk. In this research, the hazard component is recognised as the product of the probability of occurrence and the severity of an event. Vulnerability is shown to arise from the inherent socioeconomic conditions of households, the exposure and physical succesptibility of locations and a community’s capactity to respond and cope with hazard events. Risk is identified as a function of hazard and vulnerability. To understand these different dimensions, a mixed methods approach was utilised in this thesis. A quantitative method was developed for a multidisciplinary evaluation of risk that assesses its different components individually and then combines them algorithmically. A GIS-based, Fuzzy Multi Criteria Evaluation (FMCE) method was utilized to integrate the components of risk. Such techniques also enable appropriate means of quantification and visualization of complex data in map form. Qualitative methods were primarily used to investigate subjective questions such as perceptions, household and community level preparedness activities. Household surveys and semi- structured interviews with local residents, community fire volunteers, local council members and others who participated in responses to the fires were conducted to capture such information. Exploratory data analysis was performed to understand these subjective judgements and the results were presented in graphical format. This thesis demonstrates the fundamental importance of understanding the multi- dimensional characteristics of risk in managing bushfire risk at the urban bush interface. The results revealed the spatial variation of composite risk as well as the elements of risk; hazard and vulnerability. It identified important physical and socioeconomic dimensions of vulnerability and the response and coping capacities of the communities. These variations help to prioritise different disk reduction initiatives in different areas. It also found different perceptions and attitudes of residents towards bushfire management activities. This information could help to further modify the risk reduction measures in order to address specific household and community level issues. The overall results of this thesis will provide a framework to strengthen the risk reduction measures that engage in anticipating future disaster risk, reducing existing exposure, hazard, or vulnerability, and improving community capacities to cope with hazard events.
vi
PUBLICATIONS RELATED TO THIS THESIS
Journal Article Solangaarachchi D., Griffin A.L., Doherty M.D., Social Vulnerability in the Context of Bushfire Risk at the Urban Bush Interface in Sydney: A Case Study of the Blue Mountains and Ku-ring-gai Local Council Areas, Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards, DOI: 10.1007/s11069-012-0334-y
vii
viii
ACKNOWLEDGEMENTS
First and foremost I want to thank my supervisor, Dr. Amy L. Griffin for her valuable guidance and encouragement throughout the research process. She provided constructive comments and helped me to navigate obstacles and recognise the opportunities during the process. She has always being patient and willing to listen. I am fortunate to have had the opportunity to study under her guidance. I also wish to express profound respect to my co-supervisor Michael D. Doherty for his constructive suggestions and guidance in improving this research.
This work would not have been possible without the interest of the members of the public in Ku-ring-gai and the Blue Mountains who received and responded to my survey. Most people were happy to participate in the survey, and even some people who didn’t participate found it useful. I would like to thank them for assisting me in gaining an in-depth knowledge of my study area.
The support I got from the local councils and NSW Fire and Rescue (NSWFR) was very helpful. I would like to express my thanks to Greg Buckley (NSWFR), Dr. Jenny Scott and Jennie Cramp (Ku-ring-gai Council), and Peter Belshaw (Blue Mountains City Council) for helping me to complete this research. The key data and information used for this study was provided by Local councils, Land and Property Management Authority, NSW Rural Fire Services and NSW Fire and Rescue. Without these data, this study would not have been possible. Their provision of data is greatly appreciated.
This research was financially supported by School of Physical, Environmental and Mathematical Sciences at the UNSW Canberra and, I extend my thanks and appreciation to the school for its support. My appreciation also extends to the PEMS administration team for the many ways they helped throughout my research. I am indebted to Ms Julie Kesby for her excellent help with regards to referencing, thesis formatting, and gathering resources for my study.
Special thanks to my dear colleagues in the Geographers’ room, especially Solomon, Dustin and Vijai. So many good ideas were born with what started as random ix conversations in the room. I will always remember the time we spent discussing, arguing, and laughing.
I am deeply grateful to my parents for teaching me the value of education, and who sacrificed the best times of their lives for my education. I would not have come this far without their support. My brothers, thank you for being my best friends during my journey. Your love and support has given me the strength I needed to walk along this path. My heartfelt thanks also go to my father-in-law who have always supported me and encouraged me to pursue my goals wherever they took me.
A final word of thank to my wife, Beyandi, for love, encouragement, support and good food. Thank you for being understanding during those late nights and early mornings. Without your continuous support, love and patience I would not have made it.
x
LIST OF ABBREVIATIONS
ABS Australian Bureau of Statistics AHP Analytical Hierarchy Process CCD Census Collection Districts CFU Community Fire Unit DEM Digital Elevation Model GIS Geographic Information Systems IDNDR International Decade for Natural Disaster Risk Reduction IDW Inverse Distance Weight IPCC Intergovernmental Panel on Climate Change KDE Kernel Density Estimation LGA Local Government Area LPMA The Land and Property Management Authority MCE Multi Criteria Evaluation NSW RFS New South Wales Rural Fire Services NSWFR New South Wales Fire and Rescue OWA Ordered Weighted Average PCA Principle Component Analysis SOVI Social Vulnerability Index SWS Static Water Supply UBI Urban Bush Interface UNISDR International Strategy for Disaster Risk Reduction WLC Weighted Linear Combination
xi
xii
TABLE OF CONTENTS
ORIGINALITY STATEMENT ...... iii
ABSTRACT ...... v
PUBLICATIONS RELATED TO THIS THESIS ...... vii
ACKNOWLEDGEMENTS ...... ix
LIST OF ABBREVIATIONS ...... xi
TABLE OF CONTENTS ...... xiii
LIST OF FIGURES ...... xix
LIST OF TABLES ...... xxiii
CHAPTER ONE: INTRODUCTION ...... 1
1.1 Background ...... 1
1.2 Disasters, Risks, Vulnerability ...... 2
1.3 Bushfires in Australia ...... 3
1.4 Risk Assessment ...... 6
1.5 Motivation for this Research ...... 7
1.6 Objectives ...... 10
1.7 Research Benefits ...... 11
1.8 Structure of the Thesis ...... 11
CHAPTER TWO: LITERATURE REVIEW ...... 15
2.1 Introduction ...... 15
2.2 Natural Hazards and Disasters ...... 16
2.3 Vulnerability ...... 18 2.3.1 Vulnerability as exposure to the hazard ...... 21 2.3.2 Vulnerability as (in)capability to cope ...... 21 2.3.3 Vulnerability as exposure and response ...... 22
xiii
2.4 Conceptual Frameworks for Understanding Vulnerability ...... 23
2.5 Risk ...... 26
2.6 Conceptual Frameworks for Risk ...... 28
2.7 Disaster Risk Management ...... 37
2.8 Disaster Risk Management Frameworks ...... 40 2.8.1 Australian/New Zealand Risk Management Framework ...... 43
2.9 Bushfire Risk Management Framework ...... 45
2.10 Summary ...... 49
CHAPTER THREE: HAZARD RISK AND VULNERABILITY ASSESSMENT 51
3.1 Introduction ...... 51
3.2 Risk Assessment ...... 51
3.3 Elements of Risk ...... 55
3.4 Hazard Assessment ...... 56
3.5 Vulnerability Assessment ...... 56
3.6 Elements of Vulnerability ...... 58
3.7 Integrated Risk Assessment ...... 61
3.8 GIS and Risk Assessment ...... 63
3.9 Scale of Risk Assessment ...... 65
3.10 Uncertainty in Risk Assessment ...... 66
3.11 Risk Assessment and Risk Perception ...... 66
3.12 Study Area ...... 67 3.12.1 Blue Mountains LGA ...... 69 3.12.2 Ku-ring-gai LGA...... 71
3.13 Summary ...... 72
CHAPTER FOUR: BUSHFIRE HAZARD ASSESSMENT ...... 75
4.1 Overview ...... 75
4.2 Introduction ...... 75
xiv
4.3 Bushfire Hazard ...... 76
4.4 Bushfire Hazard vs. Bushfire Risk ...... 78
4.5 Bushfire Hazard Assessment...... 79
4.6 The Conceptual Framework for Bushfire Hazard Assessment ...... 81 4.6.1 Ignition Probability ...... 82 4.6.2 Fire Severity ...... 82
4.7 GIS and Spatial Modelling in Risk Assessment ...... 83 4.7.1 Spatial Multi Criteria Evaluation Model ...... 86
4.8 Materials and Methods ...... 92 4.8.1 Ignition Probability ...... 93 4.8.2 Fire Severity ...... 95 4.8.3 Exploratory Data Analysis ...... 96 4.8.4 Recurrence Interval ...... 96
4.9 Results ...... 97 4.9.1 Fire History Database ...... 97 4.9.2 Exploratory Data Analysis ...... 100 4.9.2.1 Fire History ...... 100 4.9.2.2 Recurrence Interval ...... 102 4.9.3 Bushfire Hazard in Ku-ring-gai ...... 104 4.9.4 Bushfire Hazard in the Blue Mountains ...... 107
4.10 Discussion ...... 111
4.11 Conclusions ...... 111
4.12 Summary ...... 112
CHAPTER FIVE: SOCIAL VULNERABILITY IN THE CONTEXT OF BUSHFIRE RISK AT THE URBAN BUSH INTERFACE IN SYDNEY ...... 113
5.1 Overview ...... 113
5.2 Introduction ...... 113
5.3 Social Vulnerability ...... 115 5.3.1 Social Vulnerability Assessment ...... 116 5.3.2 Social Vulnerability Indicators ...... 119
5.4 Methodology ...... 120
5.5 Results ...... 124 5.5.1 Dimensions of Social Vulnerability in Ku-ring-gai ...... 125 5.5.2 Dimensions of Social Vulnerability in the Blue Mountains ...... 127 5.5.3 Social Vulnerability Index and Mapping ...... 130
xv
5.6 Discussion ...... 137
5.7 Conclusions ...... 140
5.8 Summary ...... 141
CHAPTER SIX: BUSHFIRE RISK ASSESSMENT: AN INTEGRATED MODELLING APPROACH ...... 143
6.1 Overview ...... 143
6.2 Introduction ...... 143
6.3 Bushfire Risk and Conceptual Frameworks...... 144
6.4 Integrated Approach ...... 147
6.5 Bushfire Risk Assessment - MCE Model ...... 148 6.5.1 Standardisation Weighting and Identifying Evaluation Rules ...... 150
6.6 Vulnerability Assessment ...... 152 6.6.1 Exposure and Physical Susceptibility ...... 152 6.6.2 Social and Economic Fragilities ...... 166 6.6.3 Response and Coping Capacities ...... 166 6.6.4 Integrated Vulnerability ...... 177
6.7 Integrated Risk ...... 182
6.8 Conclusions ...... 186
6.9 Summary ...... 188
CHAPTER SEVEN: UNDERSTANDING COMMUNITIES AT THE URBAN BUSH INTERFACE FOR BUSHFIRE PREPARATION, RESPONSE AND RECOVERY ...... 189
7.1 Introduction ...... 189
7.2 Treating Bushfire Risk at the UBI ...... 191
7.3 Social Construction of Bushfire Risk...... 192
7.4 What Makes Communities More Vulnerable and Less Resilient? ...... 194
7.5 Social Factors Shape the Level of Vulnerability and Resilience to Bushfires ...... 197 7.5.1 Perception of Bushfire Risk ...... 199 7.5.2 Risk Knowledge, Awareness ...... 200 7.5.3 Risk Communication...... 202 7.5.4 Experience and Local Environment ...... 204 7.5.5 Community Strength ...... 206
xvi
7.5.6 Available Resources ...... 208 7.5.7 Institutional Arrangements ...... 210 7.5.8 Shared Responsibility...... 211
7.6 Community Based Bushfire Management in the Study Area ...... 216
7.7 Household Level Activities in the Study Area ...... 218
7.8 Role of the Fire Management Authorities in the Study Area ...... 219
7.9 Methodology ...... 220 7.9.1 Household Survey ...... 221 7.9.2 Interviews with Community Members and Key Personnel ...... 222 7.9.3 Data Analysis ...... 224
7.10 Results ...... 226 7.10.1 Household Survey ...... 226 7.10.1.1 Demographics of Respondents ...... 226 7.10.1.2 Knowledge and Bushfire Awareness ...... 236 7.10.1.3 Household Level of Preparedness ...... 245 7.10.1.4 Household Level Preparedness and Related Issues ...... 253 7.10.1.5 Community Level Preparedness ...... 255 7.10.1.6 Community Preparedness and Related Issues ...... 264 7.10.1.7 Shared Responsibilities ...... 265 7.10.1.8 Response and Recovery Actions ...... 269 7.10.2 Household Interviews and Focus Group Discussions ...... 274
7.11 Discussion and Limitations ...... 282
7.12 Conclusions ...... 283
CHAPTER EIGHT: CONCLUSIONS OF THE RESEARCH ...... 287
8.1 Introduction ...... 287
8.2 Summary of Results ...... 287
8.3 Future Research ...... 292
8.4 Use of this Study ...... 294
REFERENCES ...... 297
APPENDIX I: PARTICIPANT INFORMATION STATEMENT ...... 329
APPENDIX II: HOUSEHOLD SURVEY QUESTIONNAIRE ...... 331
APPENDIX III: PAIRWISE COMPARISON TOOL ...... 342
xvii
xviii
LIST OF FIGURES
Figure 1: Concept of risk ...... 30
Figure 2: The risk triangle ...... 30
Figure 3: Conceptual framework of earthquake disaster risk ...... 32
Figure 4: Conceptual framework of risk ...... 33
Figure 5: Conceptual framework for risk that takes a holistic approach to disaster risk assessment and management ...... 34
Figure 6: Disaster impact framework ...... 36
Figure 7: UNISDR framework for disaster risk reduction ...... 41
Figure 8: The BBC conceptual framework...... 42
Figure 9: AS/NZS Risk management process ...... 44
Figure 10: ‘Cities Project’ interpretation of the risk management process ...... 47
Figure 11: Interpretation of the bushfire risk assessment within the risk management process ...... 48
Figure 12: Risk assessment process...... 53
Figure 13: Elements of vulnerability and risk in an integrated risk assessment...... 62
Figure 14: Urban Bush Interface ...... 69
Figure 15: Map showing the study areas and the Greater Sydney region...... 71
Figure 16: Conceptual framework for bushfire hazard assessment...... 82
Figure 17: Spatial distribution of ignition points in Ku-ring-gai...... 98
Figure 18: Spatial distribution of ignition points in the Blue Mountains...... 99
Figure 19: Number of ignitions in Ku-ring-gai (1980-2007)...... 100
Figure 20: Number of ignitions in the Blue Mountains (1980-2006)...... 101
Figure 21: Annual area burnt in Ku-ring-gai (1986-2007)...... 101
Figure 22: Annual area burnt in the Blue Mountains (1986-2006)...... 102
Figure 23: Recurrence interval and the annual burnt area in Ku-ring-gai...... 103 xix
Figure 24: Recurrence interval and the annual burnt area in the Blue Mountains...... 103
Figure 25: Input maps for the bushfire hazard map (Ku-ring-gai)...... 105
Figure 26: Bushfire hazard map for the Ku-ring-gai area...... 106
Figure 27: Hazard zonation map for the Ku-ring-gai area...... 107
Figure 28: Input maps for the bushfire hazard map (Blue Mountains)...... 108
Figure 29: Bushfire hazard map of the Blue Mountains area...... 109
Figure 30: Bushfire hazard zonation map of the Blue Mountains area...... 110
Figure 31: Social vulnerability by CCD at the UBI based on SoVI...... 131
Figure 32: Maps showing distribution of individual component scores by CCD in the Ku-ring-gai LGA (components 1-3)...... 133
Figure 33: Maps showing distribution of individual component Scores by CCD in the Ku-ring-gai LGA (components 4-6)...... 134
Figure 34: Maps showing distribution of individual component scores by CCD in the Blue Mountains LGA (components 1-3)...... 135
Figure 35: Maps showing distribution of individual component scores by CCD in the Blue Mountains LGA (components 4-6)...... 136
Figure 36: Integrated bushfire risk assessment model...... 148
Figure 37: MCE model for bushfire risk assessment. Adapted from...... 149
Figure 38: Different sigmoid fuzzy membership functions ...... 151
Figure 39: Fuzzy criterion maps of the indicators of exposure and physical susceptibility in Ku-ring-gai...... 159
Figure 40: Fuzzy criterion maps of the indicators of exposure and physical susceptibility in the Blue Mountains...... 161
Figure 41: Exposure and physical susceptibility maps - Ku-ring-gai...... 164
Figure 42: Exposure and physical susceptibility map - Blue Mountains...... 165
Figure 43: Exposure and physical susceptibility zonation map - Blue Mountains...... 165
Figure 44: Fuzzy criterion maps of the indicators of emergency response and coping capacity in Ku-ring-gai...... 171
Figure 45: Fuzzy criterion maps of the indicators of emergency response and coping capacity in the Blue Mountains...... 173
xx
Figure 46: Emergency response and coping capacity maps - Ku-ring-gai...... 176
Figure 47: Emergency response and coping capacity maps - Blue Mountains...... 177
Figure 48: Integrated vulnerability map - Ku-ring-gai...... 179
Figure 49: Integrated vulnerability zonation map - Ku-ring-gai...... 180
Figure 50: Integrated vulnerability map - Blue Mountains...... 181
Figure 51: Integrated vulnerability zonation map - Blue Mountains...... 182
Figure 52: Integrated risk map - Ku-ring-gai...... 183
Figure 53: Integrated risk zonation map - Ku-ring-gai...... 184
Figure 54: Integrated risk map - Blue Mountains...... 185
Figure 55: Integrated risk zonation map - Blue Mountains...... 185
Figure 56: Hazard and risk relationship (Paton & Johnston 2006)...... 195
Figure 57: Framework of community wildfire preparedness ...... 212
Figure 58: Social cognitive preparation model ...... 213
Figure 59: Bushfire preparedness model...... 215
Figure 60: Interviewing community members and key personnel...... 223
Figure 61: Gender demographics comparison with ABS data...... 229
Figure 62: Age demographics comparison with ABS data...... 230
Figure 63: Level of household income/week among respondents, in comparison with ABS data...... 231
Figure 64: Household type comparison with ABS data...... 232
Figure 65: Home ownership comparison with ABS data...... 233
Figure 66: Employment status comparison...... 234
Figure 67: Level of education of adults...... 235
Figure 68: Number of people at home during day and nighttime...... 236
Figure 69: Years of residence in their current suburb...... 237
Figure 70: Reason for living in Ku-ring-gai or the Blue Mountains...... 238
Figure 71: Previous bushfire experience...... 240 xxi
Figure 72: Sources of information...... 242
Figure 73: Awareness of fire ban rules and regulations...... 244
Figure 74: Perceived level of bushfire risk...... 245
Figure 75: Knowledge of the month when the bushfire season starts...... 246
Figure 76: Reasons for not conducting bushfire preparedness activities...... 248
Figure 77: Information that encourages bushfire preparedness activities...... 249
Figure 78: Bushfire preparedness - household level activities...... 250
Figure 79: Discussion of preparedness issues with fire management authorities...... 251
Figure 80: Perceived levels of preparedness...... 253
Figure 81: “I feel like I belong to the local community” ...... 255
Figure 82: Source of social connection...... 256
Figure 83: Community help each other in an emergency ...... 257
Figure 84: Community discussions about bushfire preparedness activities...... 258
Figure 85: Participation in CFU/fireguard programmes...... 260
Figure 86: Participating in community level activities...... 261
Figure 87: Best practice for bushfire preparedness and mitigation involvement...... 262
Figure 88: Access to resources...... 264
Figure 89: Perceptions of responsibility for community protection and bushfire management activities...... 267
Figure 90: Community protection and bushfire management activities...... 268
Figure 91: Immediate contact point...... 270
Figure 92: My family would evacuate in an event of life threatening bushfire...... 271
Figure 93: When is the right time to evacuate? ...... 272
Figure 94: Point of evacuation...... 273
xxii
LIST OF TABLES
Table 1: Differences between risk assessment and risk perception ...... 67
Table 2: Social vulnerability indicators and variables used in the PCA ...... 122
Table 3: Social Vulnerability Index comparison ...... 124
Table 4: Social vulnerability and its dimensions – Ku-ring-gai LGA ...... 125
Table 5: Social vulnerability and its dimensions – Blue Mountains LGA ...... 128
Table 6: Indicators of exposure and physical susceptibility ...... 154
Table 7: Pairwise comparison matrix of indicators of exposure and physical susceptibility ...... 156
Table 8: Indicators of exposure and physical Susceptibility - Fuzzy membership functions, evaluation criteria and weights ...... 157
Table 9: Vegetation communities in Ku-ring-gai and the Blue Mountains ...... 158
Table 10: Indicators of response and coping capacity ...... 167
Table 11: Pairwise comparison matrix of indicators of emergency response and coping capacity ...... 169
Table 12: Indicators of exposure and physical susceptibility - Fuzzy membership functions, evaluation criteria and weights ...... 169
Table 13: Pairwise comparison matrix of vulnerability factors ...... 178
Table 14: Factors of vulnerability - Fuzzy membership functions, evaluation criteria and weights ...... 178
Table 15: Variables and derived composite variables ...... 225
Table 16: Demographic profile of the respondents ...... 228
Table 17: Communication of total fire ban information ...... 243
Table 18: When do you normally start to prepare for bushfires? ...... 247
Table 19: Use of bushfire assessment tools ...... 252
Table 20: Community protection and bushfire management activities ...... 259
Table 21: Areas of community preparedness that respondents believe need further development ...... 263 xxiii
Table 22: Reason to not to evacuate ...... 274
xxiv
Chapter One: Introduction
1.1 Background
Every year natural disasters threaten the sustainability of human systems, disrupting their use of resources and threatening human lives. The Great East Japan earthquake occurred on 11 March 2011, 130 km off Japan’s eastern coast, causing a tsunami that, together with the earthquake, may have killed more than 20,000 people (UNISDR 2011). In 2010, western Russia experienced its hottest summer in 100 years. Around 800,000 hectares in western Russia were affected by severe wildfires between July and September. This resulted in the deaths of more than 50 civilians and fire fighters. Some 2,000 houses burnt down, and more than 30 villages were completely destroyed (UNISDR 2011). The earthquake that struck Haiti on 12 January 2010 translated into a massive disaster which caused 222,517 fatalities (UNISDR 2011).
In 2008, numerous major disasters threatened human development gains across the world. In May, tropical cyclone Nargis caused an estimated 140,000 deaths in Myanmar and China’s most powerful earthquake since 1976 affected Sichuan and parts of Chongqing, Gansu, Hubei, Shaanxi and Yunnan provinces, killing at least 87,556 people, injuring more than 365,000 and affecting more than 60 million people (UNISDR 2009a). In August 2008, the Kosi River in Bihar, India, broke through an embankment and changed its course 120 km eastwards, rendering useless more than 300 km of flood defences that had been built to protect towns and villages. Flowing into supposedly floodsafe areas, the flood affected 3.3 million people in 1,598 villages located in 15 districts. It was characterized as the worst flood in the area for 50 years (UNISDR 2009a).
In light of escalating disaster losses, the international community has prioritized promoting safer communities while campaigning for effective disaster risk reduction strategies. Global initiatives such as the United Nations’ International Decade for Natural Disaster Reduction (IDNDR) in 1999 and the Hyogo Framework for Action (HFA) 2005-2015, which is an outcome strategy from the 2005 World Conference, have
1 highlighted the importance of the research mandate for formulating the overall value of disaster risk reduction (UNISDR 2005; United Nations 1999). The latest initiative, the Hyogo Framework for Action (HFA), identified five key areas that promote disaster risk reduction. Those are governance, risk assessment, knowledge and education, risk management and vulnerability reduction, and disaster preparedness and response. Among other priorities, the HFA defines the identification, assessment and monitoring of disaster risks as one of the most important steps in disaster risk reduction.
“The starting point for reducing disaster risk and for promoting a culture of disaster resilience lies in the knowledge of the hazards and the physical, social, economic and environmental vulnerabilities to disasters that most societies face, and of the ways in which hazards and vulnerabilities are changing in the short and long term, followed by action taken on the basis of that knowledge” (UNISDR 2005, p. 7).
1.2 Disasters, Risks, Vulnerability
In addition to these international initiatives, the scientific community has also put much effort into increasing its understanding of the concepts of vulnerability and risk in order to develop planning strategies and assessment tools to reduce potential losses. Disasters are often viewed as complex interactions between a hazard and a vulnerable human population, which often result from political, economic, and development failures (Oliver-Smith 1996; Smith & Petley 2009; Villagrán de León 2008). A common understanding of disasters is hampered by the various definitions of hazard, vulnerability and risk, derived from concepts and theories from various schools of thought. Risk assessment depends on understanding these key, underlying terms. Moreover, understanding risk factors and communicating risk to decision makers and the general public still remain significant challenges (Cardona et al. 2012). Disaster risk is derived from the combination of physical hazards and the vulnerabilities of exposed elements (Cardona et al. 2012). Therefore, the disaster risk management community often emphasises that the association between hazard and vulnerability determines both probabilities and consequences. This association often depends on the physical, social and geographic conditions that prevail in the area of interest (Lavell et al. 2012).
2
Hazard is defined as the possible future occurrence of a natural or human induced physical event that may cause impacts on exposed populations (Cardona et al. 2012). Often a hazard is considered to be an external element of risk. The probability of occurrence is an estimate of how often a hazard event occurs. A review of historic events and favourable environmental conditions for the event assists with this determination.
Vulnerability is broadly defined as the potential for loss, and is an essential concept in natural hazards research. Vulnerability is generally perceived as a predisposition of societies to be affected by hazards, including the incapacity to cope with an event that results in adverse effects on those exposed (Cardona et al. 2012; Villagrán de León 2006). It arises from the present conditions of communities that may be exposed in the future. In the field of natural hazards, linkages between unsafe conditions and different dimensions of vulnerability have been investigated using integrated approaches. Integrated vulnerability assessments go beyond traditional vulnerability modelling to provide a wider and more comprehensive explanation of vulnerability, one which differentiates exposure and physical susceptibility, social and economic fragilities, and lack of ability to cope and recover as different dimensions of vulnerability (Cardona & Barbat 2000; Cardona & Hurtado 2000; Cardona et al. 2012).
1.3 Bushfires in Australia
Bushfires are one of the major threats to the environment and human systems in Australia. The recent 2009 Black Saturday bushfires in Victoria claimed more than 2029 homes and 173 lives, and demonstrated that fire management authorities need to rethink their current bushfire risk and emergency management approaches. In Australia the majority of bushfire impacts occur at the urban-bush interface (UBI) (Buxton et al. 2011; Dixon 2005). 0 , 31&,2#0$ !# ! , #"#$',#" 1b ,7 0# 5#1203!230#10#1'"#,2' *Q',"3120' *Q0#!0# 2'-, *-0 %0'!3*230 * 0#*-! 2#" "( !#,2-0 +-,%!-+ 312' *# 31&* ,"$3#*1c -220#**TRRWQ.TSSRTClimate change scenarios predict that south-eastern Australia will experience increased temperatures, decreased precipitation, and high winds, which will result in increased bushfire intensity, longer fire weather seasons, and increased frequency of extreme or
3 catastrophic fire risk days (Blue Mountains Bush Fire Coordinating Committee 2008; Hennessy et al. 2005). With population growth and urbanization in areas of high fire risk along with insufficient bushfire management activities, it becomes clear that high consequence events are likely to become more frequent and more intense (Pitman et al. 2007).
According to Collins (2005), fires in the UBI have risen to prominence for three reasons; the occurrence of destructive fires in the area; rapid growth of the population in areas that are biophysically hazardous; and inadequate mitigation efforts of federal, state and local governments and private land management. The urban interface is a complicated environment for fire agencies to successfully operate within. Challenges include the sheer number of properties at risk, the dynamic nature of fires, a lack of community awareness and education, large-scale self-evacuation, narrow streets, failure of essential services such as electricity and gas, and lack of interagency communication.
In traditional bushfire models, fires are started by lightning strikes in rural locations. The development of large scale fires often results from adverse weather conditions and large volumes of combustible vegetation, causing the fires to make runs that threaten urban communities in their path (Bradstock et al. 1998). These urban interface fires develop quickly, and have the potential to overrun local fire fighting resources (Lowe 2008). In these scenarios, planning and developing mitigation and response strategies prior to any bushfire event play an important role in managing hazard impacts.
Fires at the UBI pose the greatest challenges facing Australian fire management agencies today (Preston et al. 2009). To utilize scarce response resources most efficiently, it is important to have prior knowledge about the level of vulnerability and risk that are posed to communities at the UBI. Such information can also help to identify the most vulnerable populations, and to minimize potential losses from future bushfire events by undertaking planning and preparation activities. Losses can be minimized if bushfire management activities are designed to target the most vulnerable people.
4
The UBI in the Sydney metropolitan area is one of the more bushfire-prone densely populated areas in Australia. The level of bushfire risk is being increased by an ever growing population and the concentration of high value properties and industrial and commercial assets (Chen 2005). The potential for future increases in bushfire risk is of particular concern in the area with projections of climate change that suggest the region is likely to experience more fire-weather days (Preston et al. 2009). The 1994 bushfires covered 75 percent of all bushland in the Sydney metropolitan area, caused significant economic losses, disrupted essential services, and caused both injury and death (Dixon 2005). The Black Christmas fires in 2001 destroyed 121 homes, and the 2002 fires destroyed 10 homes (EMA Disaster Database 2012). All of these events caused significant economic loss. Bushfires pose significant but largely unmeasured risks to people and their property (Bradstock et al. 2008). Bushfire risk is not simply determined by the severity of fire event. It is rather a product of the nature of fires and the vulnerability of the communities that are exposed to fires (Whittaker 2008). Therefore deeper analysis of this interaction is required for effective fire management.
In Australia, disaster and emergency management agencies have proposed a stronger focus on anticipation, mitigation, and recovery and resilience in order to achieve safer, more sustainable communities (Gabriel 2009). Such disaster management activities should be driven by better knowledge of vulnerability and risk assessments. Identifying levels of hazard, vulnerability and risk as well as the factors that influence them and their interactions facilitates the development of effective disaster risk reduction strategies that target the needs of specific groups. Thus, these activities are more effective at the local level because the impact of disasters is a product of interactions between local scale conditions such as weather, vegetation and fuel loads, topography, social characteristics, level of preparedness and emergency response and coping capacities. Furthermore, risk mitigation and emergency response measures such as planning and preparedness at the local level are often driven by collective actions of residents and local councils (Preston et al. 2009).
Despite the many unforgettable bushfire events that have occurred in Australia, and their significant impacts, the lack of a risk assessment framework has been a primary
5 limitation to quantitative risk assessment (Middelmann 2007). This hinders the use of comprehensive bushfire risk assessments in the planning process. Thus there is limited awareness of the geography of bushfire risk, implications of climate change for future bushfire hazards, changing patterns of social vulnerability, or how to develop appropriate adaptive responses (Preston et al. 2009). The Victorian Bushfires Royal Commission (2009) also highlighted the importance for effective emergency response of having information on vulnerable populations. Such information can guide emergency managers to recognise the specific needs of vulnerable people such as those who might need evacuation assistance or separate consideration, particularly on high fire risk days (Victorian Bushfires Royal Commission 2009).
1.4 Risk Assessment
Within the hazard research community, risk assessment can be defined differently in various contexts. However, it primarily concerns the degree to which a population, the built environment, and socioeconomic activities are susceptible to damage from a hazard event and includes physical aspects of hazard events such as location, magnitude, frequency and process (Chen et al. 2003). Although different disciplines conceptualise risk in different ways, the literature suggests that there is a common emphasis on the interaction between the hazard agent and a vulnerable community. In most of the research, the equation risk = hazard x vulnerability has been used to elaborate the relationship between these three concepts. As hazards and vulnerability are spatially distributed, risk is inherently a spatial phenomenon, and risk assessment should address both the degree of risk and its spatial variations.
Risk assessments often use an objective quantifiable approach. However, assessing objective levels of risk alone is not sufficient for efficient risk management decision- making. Quantitative risk assessments need to be complemented with qualitative approaches (Cardona et al. 2012). The public often views and evaluates risk differently from researchers and experts. Understanding how the public constructs their perceptions of risk can greatly improve risk communication and direct risk reduction strategies most appropriately (Cottrell et al. 2009). This emphasizes that taking only a theoretical approach to analysing bushfire risk is not sufficient to make decisions; the possible mix
6 of underestimating risk and overconfidence in facing a risk may increase the vulnerability of the community. Therefore it is important to identify factors that influence the decisions of residents living at the UBI about bushfire management. Understanding subjective judgements of residents living at UBI, the factors affecting their decisions and their perception of bushfire risk, and attitudes towards current bushfire management strategies are also important steps towards effective bushfire risk management.
Despite the fact that considerable effort has been directed towards encouraging bushfire preparedness in Australia, research on public perceptions of bushfire management strategies to reduce bushfire risk is uncommon. Available studies focus on a particular region or bushfire management strategy (Bushnell et al. 2007; Cottrell et al. 2008; Lowe et al. 2008). Social, cultural and institutional processes influence household and community level risk management and preparedness activities (Bushnell et al. 2007; Miceli et al. 2008; Paton et al. 2000). Therefore people from one social group may not perceive risk in the same way as others, and even within the same community, people perceive risk differently (Paton et al. 2000). An understanding of community perceptions and attitudes will support the development of improved risk communication and risk reduction strategies (Cottrell et al. 2008; Paveglio et al. 2009).
1.5 Motivation for this Research
Local governments are one of the key governance scales for bushfire risk management. In the context of bushfires, most bushfire preparedness activities are conducted at the local level. While a range of institutional arrangements and policy measures for mitigation of bushfire risk have been implemented within Australian local governments, the exploration of bushfire risk at the local level is not well addressed (Preston et al. 2009). Nevertheless, local governments need information on bushfire risk in order to understand the implications of bushfire risk management in the context of local conditions as well as guidance on how such information can be incorporated into local level planning, bushfire risk and emergency management, and resource management, in order to facilitate the implementation of relevant policies and measures.
7
It is difficult to implement preparedness and response activities across the UBI in an equitably distributed manner because there are never enough resources to cover all areas. To manage bushfires, fire management agencies have to work in the areas that are considered to be at highest risk of damage from bushfires. Easy identification of areas where high-risk populations live and spatial comparisons to prioritize high-risk areas across the UBI allows agencies to funnel resources to those areas. If hazard, and the dimensions of vulnerability can be incorporated into a risk assessment model that is capable of determining risk, the effectiveness of emergency management may be increased. Such an integrated model will answer questions like: Where is the hazard high? Where are the most vulnerable people living? Why are they vulnerable? Where is risk high? Where would potential losses be high? Furthermore, preparedness activities can be designed to address specific issues (i.e., reduce social vulnerability, reduce physical vulnerability or increase community capacity). This contributes to risk reduction and ultimately reduces the possibility of future disasters.
At present, at the local government level, bushfire management and planning rely heavily on ‘bushfire prone area’ maps that are developed using vegetation buffering (NSW Rural Fire Service 2006). These ‘bushfire prone’ maps only provide information about proximity to flammable vegetation. Therefore, to understand the level of risk at the UBI, a more holistic approach is needed. The assessment of bushfire risk at the local level is complex and challenging as the different elements of risk need to be assessed individually and then integrated through some kind of algorithm to produce a holistic picture (Villagrán de León 2006). Thus far, at the local level there is no consensus on how to address risk holistically while measuring the elements of risk.
To address this issue, a variety of bushfire risk assessment models have been developed by various institutions and agencies. Many of them have only focused on the hazard component of risk, and are therefore based mainly on physical factors such as weather, fuel, and topography. Existing wildfire risk models only show the area of assets or value potentially impacted by fires. Such approaches do not provide information on the pre- existing condition of vulnerability. Therefore, tools for assessing bushfire risk that integrate hazard and pre-existing conditions of vulnerability for a particular geographic area need to be developed. To prepare an effective disaster mitigation plan, planners 8 need tools that simultaneously address all of the dimensions of vulnerability. To address this need, planners and government agencies need to develop simplified models to understand communities and to assist communities with the process of risk and vulnerability assessment.
This thesis develops a multifaceted understanding of bushfire vulnerability and risk at the UBI based on a holistic approach to risk. Based on the developed risk framework, a spatial bushfire risk assessment is performed. In this research, hazard is recognised as the probability of occurrence and the severity of the event. Vulnerability is shown to arise from the inherent socioeconomic conditions of households, exposure and physical susceptibility of locations, and the level of response and coping capacities within communities. Risk is identified as a function of hazard and vulnerability.
A methodology for bushfire risk assessment is presented for the Local Government Area (LGA) level, using two LGAs. A GIS-based, Fuzzy Multi Criteria Evaluation (FMCE) method was utilized to assess and integrate the individual elements of risk. Such techniques also provide a means of quantification and visualization of complex data in map form. Specific conditions in the LGA, data availability and the bushfire management setting are taken into account.
This research also highlights the importance of understanding different perceptions and attitudes of residents towards bushfire management activities. Such information helps to further modify risk reduction measures in order to address specific household and community level issues. Qualitative methods were primarily used to investigate questions such as subjective perceptions and household and community level preparedness activities. Household surveys and semi-structured interviews with local residents, community fire volunteers, local council members and others who participated in responses to fires were conducted to capture such information. Exploratory data analysis was performed to understand these subjective judgements. The overall results of this thesis provide an advanced framework to strengthen risk reduction measures that engage in anticipating future disaster risk, reducing existing exposure, hazard, or vulnerability, and improving community capacities to live with hazards. 9
1.6 Objectives
This research is devoted to developing a conceptual model of risk that provides a holistic perspective on bushfire risk assessment. It incorporates hazard; social vulnerability; exposure and physical susceptibility; and response and coping capacity at the UBI as well as information on a variety of social and cognitive processes that affect judgements, perceptions, and behaviour with respect to bushfire management and preparedness activities at the local level. The key objectives of the research are to:
1. Develop and implement an integrated bushfire risk assessment framework that would assist existing bushfire risk management processes.
2. Develop a GIS-based integrated bushfire risk assessment model to diagnose the bushfire risk at the Urban Bush Interface (UBI).
3. Understand household and community characteristics that influence subjective decisions about bushfire management activities at the local level.
To accomplish these objectives, the following tasks were carried out:
Review and analyse existing definitions, frameworks and models of hazard, vulnerability, risk and risk management to understand the most suitable conceptual and analytical framework for bushfire risk assessment in the context of bushfire risk management. Bushfire hazard assessment and mapping based on bushfire hazard occurrence probabilities and severity in the study area using the available bushfire history data. Develop a social vulnerability index for bushfires at the census collection district level using available census data. Assessment of vulnerability to bushfires based on physical exposure and susceptibility, social vulnerability, and response, recovery, and coping capacities at the UBI using a GIS- based Fuzzy Multi Criteria Evaluation model. Assessment of bushfire risk at the UBI based on the Hazard-Risk model.
10
A household level assessment was performed using a mixed methods approach in order to identify factors affecting community and household level decision making processes that influence local level bushfire management activities at the UBI.
1.7 Research Benefits
The findings of this research contribute to bushfire risk mitigation and reduction at the UBI in the Sydney metropolitan area. Integrated bushfire risk assessment will help to differentiate hazard, vulnerability and risk. Major contributions of this research are: 1. The development of a conceptual framework for bushfire risk assessment in the context of bushfire risk management. 2. The development of a bushfire hazard assessment model using fire history data. 3. The development of a social vulnerability index to understand the socioeconomic conditions at the UBI using census data. 4. The development of a GIS-based fuzzy multicriteria evaluation model to analyse bushfire risk at the local government level using available data sources. 5. Contribution to local level bushfire risk management processes by understanding spatial variations in bushfire hazard, risk and vulnerability across the study area. 6. The development of a household and community bushfire preparedness model. 7. Contribute to household and community level bushfire management processes by describing the different perceptions, attitudes and behaviours of residents living at the UBI.
1.8 Structure of the Thesis
This thesis is comprised of four parts: Part I (Chapters 1 to 3) lays the foundation that motivates this research into integrated bushfire risk assessment. The theoretical aspects, and conceptual and assessment frameworks that relate to the study are discussed. In Part II (Chapters 4 to 6), bushfire risk at the UBI is assessed using an integrated approach. The elements of risk are investigated and an integrated model is presented. Part III (Chapter 7) discusses household and community level bushfire management processes; the factors affecting the level of household and community level bushfire preparedness
11 are also assessed. Part IV (Chapter 8) concludes the overall research findings and highlights the importance of those findings. Furthermore, it describes a scope for future work related to the integrated bushfire risk assessment process for effective bushfire management.
Chapter 2 presents a literature review of definitions, conceptual frameworks, and their applications in disaster risk management. This is important because the lack of a common understanding of these basic concepts is a barrier to developing an integrated risk assessment framework. Different conceptual frameworks for vulnerability, risk and risk management are discussed, and a risk assessment framework for bushfire risk in the context of bushfire risk management in Australia is presented.
Chapter 3 provides an overview of hazard, vulnerability and risk analysis. It explains different assessment approaches and presents the analytical framework for the study based on the holistic risk assessment framework discussed in Chapter 2. It also provides an overview of the study areas.
In this research, bushfire hazard is considered to be an element of bushfire risk. Bushfire hazard assessment is presented in Chapter 4. It describes the methods used to analyse available data and generate fire hazard maps using information on ignition probability and fire severity. The result of the bushfire hazard assessment is used as an input for the overall risk assessment in Chapter 6.
Chapter 5 discusses the socioeconomic dimension of vulnerability. Social vulnerability at the UBI is modelled using a Social Vulnerability Index (SoVI). The results, including the development of the SoVI, data and indicator selection and factor identification, are presented.
Chapter 6 explores the integrated risk assessment process. It describes the analysis of overall vulnerability, including exposure and physical susceptibility, social vulnerability, and response and coping capacities. It then combines the hazard and vulnerability assessment results to develop an integrated bushfire risk map.
12
Chapter 7 investigates household and community level bushfire management processes. It presents a bushfire preparedness model and based on the household survey conducted, the different perceptions, attitudes and behaviours of residents living at the UBI are discussed.
Chapter 8 discusses the overall findings, limitations and challenges of this research. It further describes how useful these findings are for supporting effective bushfire management. It concludes by suggesting policy implications to minimise potential bushfire impacts and future work that is needed to improve the integrated bushfire risk assessment model developed in this thesis.
13
14
Chapter Two: Literature Review
2.1 Introduction
The number of disasters, number of people affected and the estimated economic losses generated by disasters have increased over the past decades (Hagenlocher 2012). Every recent disaster has illustrated the process of translating the impact through the level of exposure and vulnerability (UNISDR 2011). Therefore, the importance of policies, programmes and mechanisms to reduce exposure and vulnerability and to promote more resilient societies is widely recognised at all levels (Hagenlocher 2012; UNISDR 2005). Thus identification, assessment and monitoring of disaster risk have become key priorities among researchers and decision-makers. Contemporary research on natural hazards and disasters is multidisciplinary in nature. Natural scientists study the nature of the extreme events involved in hazards whereas social scientists study the human dimensions of both impacts and responses.
Geographers often try to provide a perspective that integrates physical and human dimensions of hazards while placing emphasis on the spatiotemporal distributions of hazards, their impacts and vulnerability (Alexander 1993). According to Cutter et al. (2000), hazards research in geography began when Harlan Barrows employed a human ecology approach in 1923 to find out how people and society adjust to extreme environmental events, such as floods. Later, Gilbert White (1936, 1945) argued against the conception of hazard as an isolated geophysical event and incorporated a social perspective on natural hazards to develop a human ecological framework that operates at the interface of both natural and human systems (Smith 2001). This broadened the field of natural hazards from a narrow focus on technological and engineering solutions to human adjustments (Burton et al. 1978). Since then, social scientists effectively began their own research on the interaction between natural hazards, people and society.
During the 1970s, hazards research split into three distinct perspectives (Smith & Petley 2009). Scientists who believed hazards are geophysical events extended their work on environmental control to improve predictions about extreme natural events and the
15 construction of physical works designed to resist them. Geographers continued to work on a hazard-based approach predicated on the unifying concept of human ecology and the notion of mitigating losses by adding various human adjustments to the existing use of physical control structures. Sociologists, on the other hand, adopted a disaster-based view with an emphasis on understanding the role of human behaviour at times of community crisis and the need to improve preparedness for such mass emergencies.
There has been considerable evolution of theories and concepts in hazards research in the decades since the 1970s. At present, hazards research considers more than just the hazards themselves. It considers hazards within their particular context such as geography of the event, physical properties of the hazard and the social, economic, political, spatial, temporal and organizational structure of the system where the hazard takes place (Cutter et al. 2000).
This chapter explores some of the key concepts in hazards research; hazard, vulnerability and risk. It reviews definitions and conceptual frameworks for vulnerability and risk in order to understand their multidisciplinary nature. It further discusses the risk management frameworks that help to minimise levels of exposure and risk. Finally, it highlights the importance of an integrated risk framework and develops a bushfire risk assessment framework for bushfire risk management based on a holistic approach to risk.
2.2 Natural Hazards and Disasters
The basic term ‘natural hazard’ implies a potentially damaging physical event, phenomenon or human activity that may cause a loss of life or injury, property damage, social and economic disruption, and/or environmental degradation (UNISDR 2004). “A hazard may be regarded as the pre-disaster situation, in which some risk of disaster exists, principally because the human population has placed itself in a situation of vulnerability” (Alexander 1993, p. 7). Different researchers have employed different approaches to define hazards and disasters and their relationship to many disciplines. The term hazard sometimes creates confusion and it has been used imprecisely and with different implicit meanings, but in addition, the term has evolved with understanding of
16 the components that interact to comprise hazardousness (Tobin & Montz 1997). Alexander (1993, p. 4) defines hazard in four ways based on previous literature: “A naturally occurring or man-made geologic condition or phenomenon that presents a risk or is a potential danger to life or property (American Geological Institute 1984); An interaction of people and nature governed by the co-existent state of adjustment of the human use system and the state of nature in the natural event system (White 1973); Those elements in the physical environment which are harmful to man and caused by forces extraneous to him (Burton and Kates 1964); The probability of occurrence within a specific period of time and within a given area of a potentially damaging phenomenon (UNDRO 1982)”.
From these four perspectives on hazards, it is clear that a hazard represents the potential for impact on human beings and their environment (Alexander 1993; Tobin & Montz 1997).
Burton et al. (1978) further elaborated upon and modified the concept of interactions between humans and extreme natural events. They argued that natural systems are neither benevolent nor maliciously motivated towards humans: they are neutral, in the sense that they neither prescribe nor set powerful constraints on what can be done with them. It is people who transform the environment into resources and hazards, by using natural features for economic, social, and aesthetic purposes. Tobin and Montz (1997) also employed a similar concept. They stated that a natural hazard represents the potential interaction between humans and extreme natural events and represents the potential or likelihood of an event. In these views, hazards are seen as a result arising from the interaction of social and natural systems. If the result of these interactions causes disruptions in the human environment, it will create natural hazards (Burton et al. 1993).
A natural disaster, in contrast to a hazard, can be defined as a rapid, instantaneous or profound impact of the natural environment upon the socio-economic system
17
(Alexander 1993). UNISDR (2009b) defined a disaster as “a serious disruption of the functioning of a society, causing widespread human, material, or environmental losses which exceed the ability of affected society to cope using only its own resources.” These disruptions in human society are mainly due to the interaction between a hazardous physical event and vulnerable social conditions (Lavell et al. 2012). Disasters are often classified sudden or slow onset (according to their speed), or natural or man-made (according to their cause). A disaster might also be defined as an extremely hazardous event that differs significantly from the norm and that disrupts the workings of society (Alexander 1993; Tobin & Montz 1997). However, there is no definite threshold that determines exactly what size of natural event can be used to categorically define an event as a disaster (Alexander 1993; Tobin & Montz 1997).
The natural and the human are inextricably bound together in almost all disaster situations (Wisner et al. 2004). Therefore, if there is no significant interaction between an extreme natural event and a human system, there will be no disaster (Alcántara- Ayala 2002; Alexander 1993; Burton et al. 1993). Villagrán (2006) elaborated upon this idea by noting that a disaster is preceded by at least two predispositions: the possibility that the triggering event takes place, usually called a hazard at this potential state; and a pre-existing vulnerability, the pre-disposition of people, processes, infrastructure, services, organisations, or systems to be affected, damaged, or destroyed by the event.
2.3 Vulnerability
Various definitions and terms exist for the term “vulnerability” in the hazard and disaster management literature. In general, vulnerability is identified as the propensity for or predisposition to be adversely affected by an event. The concept of vulnerability is used frequently in many diverse research and policy communities such as sociology, geography, engineering, natural sciences and anthropology. Recently its use has become more prominent in areas such as natural hazards, including climate change (Cutter et al. 2000; Simpson & Human 2008), poverty and human security (Chaudhuri et al. 2002; O’Brien & Leichenko 2007), environmental change and sustainability (Barnett et al. 2008; Metzger & Schröter 2006; Sands & Podmore 2000), and food security (Burg 2008). As a result, the concept of vulnerability has been employed within multiple
18 contexts, with multiple dimensions and at multiple geographical and temporal scales (Hufschmidt 2011). These different views on vulnerability have emerged as a consequence of the conceptual needs required to address particular aspects of the potential impacts of disasters (Villagrán de León 2006). Yet the various definitions of and meanings ascribed to the concept of vulnerability have hampered a common understanding of how to measure vulnerability.
The concept of vulnerability achieved prominence in the 1960s, when it was viewed as a measure of the propensity for the damage a structure could face from a hazard of a given intensity. This conceptualization of vulnerability developed out of the engineering perspective and it was predominantly based on structural and physical aspects of the environment (UNDRO 1980). In the late 1970s the concept of vulnerability acquired a hazard-centric perspective through the work of White (1973) and Burton et al. (1978). They stated that vulnerability is not just an attribute of physical structures but of social groups and therefore vulnerability is socially produced or at least influenced. This concept was used to understand the extent to which people experience a particular hazard differently: as a result of their likelihood of exposure (to the hazard) and their capacity to withstand the hazard, which is often in turn influenced by their socio- economic circumstances (Cannon 1994; Schneiderbauer & Ehrlich 2004). Since then human aspects of vulnerability and its role in natural disasters have gained wide attention from scholars around the world and as a result, various definitions, concept and frameworks have emerged over the decades.
Timmerman (1981) conceptualised vulnerability as the degree to which a system reacts adversely to the occurrence of a hazardous event. This adverse reaction includes response, absorption and recovery from the event. According to this conceptualization, exposure to a hazardous event and the reaction of society to it can be seen as the key elements of vulnerability. Chambers (1989) described these elements as the internal and external sides of vulnerability. The external side of vulnerability is the set of shocks and stresses to which an individual or household is subjected, and the internal side is defencelessness, meaning a lack of ability to cope without experiencing a damaging loss.
19
While the theoretical basis of vulnerability progressed over the last two decades, discrepancies in the meaning of vulnerability have also arisen in different research and practical settings. Cutter (1996, p. 532) noted: “Despite more than a decade’s worth of collective research experience about the concept, vulnerability still means different things to different people.” Although there is a variety of different academic meanings for the term vulnerability, it also has a commonplace meaning: being prone to or susceptible to damage or injury (Wisner et al. 2004). Vulnerability has also been broadly defined as the potential for loss of property or life from environmental hazards (Cutter et al. 2000). However, this general definition of vulnerability does not specify the type of loss or the characteristics of the individuals, groups, or societies experiencing the loss. Cardona (2004) identified vulnerability as an internal risk factor that is mathematically expressed as the potential that the exposed system may be affected by the hazard phenomenon.
The UNISDR (2004) definition for vulnerability describes it as the conditions determined by physical, social, economic and environmental factors or processes that increase the susceptibility of a community to the impact of hazards. The physical aspect of vulnerability refers to susceptibilities of location and the built environment, and can be represented through factors such as population density, remoteness of a settlement, and the construction materials and design used for critical infrastructure and housing. Social factors include levels of health and well-being of individuals, gender ratios of the population, levels of literacy and education, the existence of peace and security, access to human rights, social equity, traditional values, beliefs, and organisational systems. Economic factors include poverty and the levels of individual, community, and national economic reserves, levels of debt, degrees of access to credits, loans, and insurance, and economic diversity. Environmental factors include natural resource depletion and degradation.
From the brief survey above, the literature clearly shows that there are many possible ways to categorize the various connotations and concepts related to vulnerability, based on how vulnerability has been conceived. The most common themes are vulnerability as exposure, vulnerability as (inadequate) coping capacity and vulnerability as the
20 intersection of exposure and (inadequate) coping capacity (Cutter 1996; Gall 2007; Villagrán de León 2006; Whittaker 2008).
2.3.1 Vulnerability as exposure to the hazard
In this perspective, vulnerability is referred to as direct exposure to a potential hazard; how the system deals with the hazard, its characteristics and its impacts. It mainly describes vulnerability as a physical property and is also viewed as the degree of loss resulting from the occurrence of a natural event (Buckle et al. 2000). Alexander (1993) stated that vulnerability is a function of the costs and benefits of inhabiting areas at risk of natural disasters. He emphasized that human occupancy of and humans’ activities in hazardous zones are the central domain of vulnerability. Exposure depends on proximity to the natural hazard, the rapidity of its onset, its duration, spatial impacts or areal extent, and the probability with which a hazard of a specific magnitude and frequency occurs (Cutter 1996). Cutter used the term biophysical vulnerability to express the idea of vulnerability as exposure. Although this view emphasises that vulnerability is only determined by the exposure to an external event, others argue that the characteristics of the human system and the relationship between human society and the external event also play a critical role (Dolan & Walker 2006).
2.3.2 Vulnerability as (in)capability to cope
This perspective views vulnerability as socially constructed and it is conceptualized as the pre-disaster state of the social system, which is determined by social conditions and historical circumstances. It is further described as (inadequate) coping responses, including a community’s lack of social resilience and resistance to hazards (Cutter 1996; Villagrán de León 2006). Wisner et al. (2004) explained this idea as the characteristics of a person or group and their situation that influence their capacity to anticipate, cope with, resist and recover from the impact of a natural hazard. They further described vulnerability as a result of social processes and structures that constrain access to resources, resulting in an inability to cope with hazard impacts. Differences in societal conditions also determine the level of exposure and thus the potential damage among different social groups (Cutter et al. 2003; Wu et al. 2002). As an example, low-income people might move into a flood prone area because of the
21 affordability of properties. Living in a flood zone might increase both their level of exposure and potential losses. In this perspective, researchers have emphasised the role of the socioeconomic position of individuals and communities within the system, which includes characteristics such as income, gender, race and age, irrespective of the physical properties of hazard exposure.
2.3.3 Vulnerability as exposure and response
In this view, both exposure and the underlying social conditions of a population that make it incapable of coping with hazard events are considered to be determinants of vulnerability. Therefore this perspective defines vulnerability as a relationship between physical events and the social characteristics of populations affected by these events, and integrates both the physical events and underlying causal characteristics of vulnerability (Dolan & Walker 2006). Along these lines, Watts & Bohle (1993) defined vulnerability in terms of exposure, capacity and potentiality. They emphasised that “the most vulnerable individuals, groups, classes and regions are those most exposed to perturbations, who possess the most limited coping capability, who suffer the most from crisis impact and who are endowed with the most circumscribed capacity for recovery” (Watts & Bohle 1993, p. 45). Bohle et al. (1994) elaborated upon this definition by suggesting that vulnerability is an aggregate measure of human welfare that integrates environmental, social, economic, and political exposure to a range of potential harmful perturbations. Villagrán (2006) further modified this definition by defining vulnerability as a multi-layered and multi-dimensional social space defined by the political, economic, and institutional capabilities of people in specific places and times.
An earlier expression of this idea can be found in Cutter et al.’s (2000) contention that the degree to which populations are vulnerable to hazards is not solely dependent upon the physical nature of the hazard and that social factors play a significant role. In their ‘hazards of place’ model, the hazard potential interacts with the underlying social fabric of the place to create social vulnerability. Social fabric includes socio-demographic characteristics as well as perceptions of and experience with risk and hazards. The geographic filter includes the site and situation of the place and its proximity to hazard sources and events, and interacts with hazard potential to produce the place’s
22 biophysical vulnerability. The social and biophysical vulnerability elements are mutually related and produce the overall vulnerability of the place.
Vulnerability in the context of climate change has been defined as “the degree to which a system is susceptible to, or unable to cope with, adverse effects of climate change, including climate variability and extremes. Vulnerability is a function of the character, magnitude, and rate of climate variation to which a system is exposed, its sensitivity, and its adaptive capacity” (McCarthy 2001, p. 995). In this definition, vulnerability is seen as an integrated measure of the expected magnitude of adverse effects to a system caused by a given level of certain external dimensions (exposure) and internal dimensions (sensitivity and adaptive capacity).
2.4 Conceptual Frameworks for Understanding Vulnerability
Over the years, different conceptual frameworks for understanding vulnerability have been developed to explain vulnerability in the context of natural hazards. Widely accepted vulnerability frameworks include the Pressure and Release (PAR) and Access models (Blaikie et al. 1994; Wisner et al. 2004), the Vulnerability in Sustainability Science model (Turner et al. 2003), the Vulnerability of Place model (Cutter 1996; Cutter et al. 2000), the IPCC model (Lavell et al. 2012; McCarthy 2001), the holistic perspective on vulnerability (Cardona & Barbat 2000; Cardona & Hurtado 2000), and other integrated models (Dolan & Walker 2006; Füssel 2007; UNISDR 2004).
The PAR model discusses the root causes of vulnerability (unsafe conditions) and various components and processes that create dynamic pressures that result in unsafe conditions. The PAR model of vulnerability recognises three elements that cause the progression of vulnerability; root causes, dynamic pressures and unsafe conditions. It starts with root causes, which include a lack of resources and power along with ideologies of political and economic systems. These root causes generate dynamic pressures such as a lack of skills or local institutions, and macro forces such as population change and urbanization. Dynamic pressures, in turn, drive a system towards unsafe conditions. Unsafe conditions are seen as a combination of factors involving the physical environment, the local economy, social relations, and public actions and
23 institutions. However, within the PAR model, this progression of vulnerability is taken as separate to and independent of hazard characteristics.
In their framework for vulnerability in sustainability science, Turner et al. (2003) viewed vulnerability as a component of the human environmental system. This system operates at multiple spatial, functional and temporal scales. They identified exposure, sensitivity and resilience as key elements in their vulnerability framework. Exposure is characterised by the frequency, magnitude and the duration of the hazard. Sensitivity is determined by both human conditions (population and economic structures) and environmental conditions (biophysical endowments). Resilience is characterised by coping responses, impact responses and adjustments, and adaptation responses. Sensitivity and resilience jointly determine the consequences of a hazard. In this framework, characteristics of the hazard are considered to be an element (exposure) that determines the level of vulnerability.
In the ‘vulnerability of place’ model, the interaction between biophysical vulnerability and social vulnerability creates place vulnerability (Cutter 1996). It also assumes vulnerability is a pre-existing condition and distinctive to particular places and times. This model recognises that hazard potential is lessened or amplified by the social fabric of society (social vulnerability) as well as place-based characteristics such as proximity to the hazard source (biophysical vulnerability). According to this framework hazard potential is filtered through both of these elements and the interaction and intersection of social and physical elements of vulnerability results in place vulnerability.
Vulnerability frameworks in the context of global climate change address three key factors; exposure, sensitivity and adaptive capacity (McCarthy 2001; Preston et al. 2009). In these frameworks exposure is seen as an external stressor, sensitivity as the responsiveness of a system to the external stressor, and adaptive capacity is viewed as the inherent property that defines capacity to deal with exposure (Dolan & Walker 2006; Preston et al. 2009). This type of vulnerability framework is only suitable for slow onset disasters such as drought and climate change because adaptation is considered to be a longer term and more-sustained adjustment (Lavell et al. 2012). In the context of a sudden onset disaster like bushfires, instead of adaptive capacities, 24 emergency response capabilities play an important role in minimising impacts. Although climate change frameworks include exposure within the notion of vulnerability, it might be problematic when interpreting the concept of risk in relation to vulnerability and the hazard. This is because the term ‘hazard’ is widely used in risk research to explain the specific hazard conditions while ‘exposure’ is used to represent hazard conditions in the context of climate change (Hufschmidt 2011). This needs to be considered in analyses of risk because the hazard component may therefore be double- weighted.
The holistic perspective on vulnerability provides a more comprehensive explanation of vulnerability. It presents vulnerability as a dynamic system that is characterised by physical exposure and susceptibility, socioeconomic fragility and a lack of resilience to cope and recover (Birkmann 2006b; Cardona & Barbat 2000; Cardona & Hurtado 2000). These factors determine both direct and indirect impacts of hazard events. Integrated frameworks often consider the inherent properties of physical and social environments as an interrelated and interdependent human-environment system (Dolan & Walker 2006). Within risk-hazard models, vulnerability is often seen as a separate element. Vulnerability is characterised by factors such as physical infrastructure, population, and social and political systems (Davidson 1997). The objective of integrated vulnerability frameworks is to include all relevant dimensions of vulnerability, i.e. economic, social, physical as well as some aspects of ecological vulnerability (Fuchs et al. 2011). In most of the integrated frameworks, vulnerability is characterised by physical exposure, fragility of the socio-economic system and lack of resilience to cope and recover (Dolan & Walker 2006). Other authors view the level of vulnerability as determined by the level of exposure and susceptibility (Birkmann 2006a).
Many of the above mentioned frameworks have highlighted vulnerability as the main driver that determines the existing and future level of risk (PAR, holistic perspective, IPCC-2012, sustainability science model and some integrated models). Therefore, risk is defined as a function of hazard and vulnerability (Birkmann 2006b).
25
Although a number of different social and environmental characteristics have been addressed in vulnerability frameworks, they do not represent the actual interaction between hazard and vulnerability. Often vulnerability frameworks have not been able to suggest specific vulnerability reduction measures such as hazard mitigation and emergency response. The lack of clarity about the relationship between hazard and vulnerability within these frameworks makes it difficult to understand the concept of disaster risk. With recent developments in natural hazard research, disasters are often viewed as a result of complex interactions between hazardous physical events and a vulnerable society (Lavell et al. 2012). The outcome of a disaster is characterised by the hazard phenomenon, the degree of vulnerability of exposed elements and their capacities to withstand disaster events (Granger 2003; Lavell et al. 2012). This interaction between hazardous conditions and the vulnerable elements needs to be understood in order to identify specific risk management strategies, either through hazard reduction or vulnerability reduction measures. Therefore, this study focuses on the concept of risk. Hazard is considered to be an external component of risk while vulnerability is considered an internal component of risk (Cardona 2004).
2.5 Risk
The concept of risk has gained favour in disaster management over the past two decades as it signifies the possibility of adverse effects in the future (Lavell et al. 2012). Risk is driven by the interaction of social environmental processes, from the combination of physical hazards and the vulnerabilities of exposed elements (Lavell et al. 2012). The term “risk” is multidisciplinary in nature and is used in natural hazards research with different definitions in various situations (Thywissen 2006). Nevertheless, risk seems to have two broad foci: event occurrence on one hand and the consequences of the event on the other hand (Bachmann 2001). Risk is often defined as probabilistic in nature (Bachmann & Allgower 2001; Blanchi et al. 2002; Brooks 2003), relating either to; the probability of occurrence of a hazard that acts to trigger a disaster or series of events with an undesirable outcome, or the probability of the outcome (loss) of an event or combining the probability of the hazard event with a consideration of the outcome of the hazard.
26
This concept of probability is mainly driven by the likelihood of hazard occurrence and the degree of loss. Concepts of risk that emphasise the hazard component are not rare in the hazard literature. Therefore the risk of these events is characterized by the magnitude of the hazard (including size and spread), the frequency and duration, and the history of hazard occurrence (Alwang et al. 2001; Schneiderbauer & Ehrlich 2004). The consequences are mainly measured in terms of causalities and economic loss and are mainly used in the insurance sector to estimate potential losses from future disasters.
In the context of natural disasters, the combination of probability of occurrence and consequences provides a quantitative representation of a qualitatively defined hazard (Lavell et al. 2012). Consequences are largely influenced by the nature of hazard and the vulnerability of the exposed elements. Hazard and vulnerability can each contribute to the probability of occurrence. Wisner et al. (2004) defined risk as a complex combination of vulnerability and hazard. The relationship between hazard and vulnerability that explains root causes, dynamic pressure and unsafe conditions shows the consequences and likelihood. UNISDR (2004, p. 16) provided a more comprehensive definition of risk as the “probability of harmful consequences, or expected loss of lives, people injured, property, livelihoods, economic activity disrupted (or environment damaged) resulting from interactions between natural or human induced hazards and vulnerable conditions”. These proposed concepts of risk lead to a conventional equation of risk given below:
Risk = Hazard x Vulnerability Equation (1)
However, the conventional equation has been further modified to elaborate upon interactions between hazard and vulnerability. Most efforts have added new elements while the main concept remains same. Crichton (1999), for example, proposed risk as a function of three elements, hazard, vulnerability, and exposure.
Risk = Hazard*Exposure*Vulnerability Equation (2)
Granger (1999) substituted the term ‘elements at risk’ for exposure. Risk = Hazard*Elements at Risk*Vulnerability Equation (3) 27
Both attempts have tried to identify the exposure in terms of lives or properties at risk in a given area as a separate component of vulnerability.
Recent publications define risk by incorporating terms such as coping capacity, emergency response and recovery capacity, and deficiencies in preparedness (Lavell et al. 2012). Coping capacities refer to the means by which people or organisations use available resources and capacities to face adverse consequences related to a disaster. In general, such capacities involve management of resources before, during, and after the disaster. Therefore it is suggested that there is an inverse relationship between risk and coping capacities (Villagrán de León 2006).
5- dz0') -$'$/4 Risk = *+$)"+$/$ . Equation (4)
However, the relationship between risk and capacities has also been described as a linear relationship. In such conceptualizations, coping capacities are assigned a negative value because of their negative relationship with the level of risk (Hahn 2003, p. 115).
Risk = Hazard + Exposure + Vulnerability - Coping Capacities Equation (5)
All of these efforts represent risk as a probability of harmful consequences, or expected losses resulting from interactions between natural or human-induced hazards and vulnerable conditions. However, there is no consensus on how to measure any of the components thus far. The assessment of risk itself is highly context-specific and rather complex, as the components need to be assessed individually and then combined based on the specific risk model used for the study.
2.6 Conceptual Frameworks for Risk
In addition to the numerous mathematical expressions of risk, various conceptual frameworks for risk are also available in the literature. Many of the conceptual approaches for risk had their origin in studies of technological hazards and some of them were adapted for natural disaster risk (Carreño et al. 2007). After Gilbert White (1964) initiated the concept of risk in natural hazards, sociologists, geographers and
28 engineers devoted their efforts to explaining how the term risk related to their disciplines. Since then sociologists began to focus on social responses to disasters, civil engineers’ focus was on physical and technical aspects of disasters and geographers’ focus was on natural hazard risk with an emphasis on a socio-ecological perspective (Carreño et al. 2007). Kates (1971) provided an example model of risk while considering the interactions between nature, humans and technology. When Whitman (1973) started damage assessment, methodologies were developed for physical risk assessment (Carreño et al. 2007).
Later, various analytical concepts and models of risk were developed to systematise its study. In recent decades, a more integrated vision of risk that combines both hazard and vulnerability has evolved. During the 1990’s, stimulated by the International Decade for Natural Disaster Reduction (IDNDR), many studies that dealt with disaster risk emerged around the world. The topic gained wide attention and it is being increasingly recognized as an important step in disaster risk reduction. It also highlighted that the terms hazard, vulnerability and risk have had different meanings and implications from both methodological and practical angles (Cardona, 2004).
The risk analysis guidelines developed by GTZ (2004) (Figure 1), identify hazard and vulnerability as the essential elements of risk. According to this framework, risk only arises when hazard and vulnerability coexist. This framework implements the mathematical model presented in equation 1. However, it does not elaborate on components of vulnerability, and exposure is considered to be a component of hazard, while self-protection capabilities are a component of vulnerability. The relationship of these factors and their contributions toward disaster risk reduction are addressed. Therefore applying this model in to this framework may be highly context specific. For an example, this model may be more suitable for individual level assessment rather than community level assessment.
29
Figure 1: Concept of risk (GTZ 2004, p. 17).
Several authors have portrayed risk as a triangle. This concept was initially developed by Crichton (1999). According to him, the area inside the triangle represents the risk and the sides of the triangle represent the independent factors that contribute to risk (Figure 2). Those independent factors include exposure, vulnerability and hazard.
Figure 2: The risk triangle (Crichton 1999, p. 130), Source: Middelmann (2007).
Birkmann (2006b) has discussed the triangular risk framework that was developed by Villagrán De León in 2004. Villagrán’s triangle of risk also consists of three components: vulnerability, hazard and deficiencies in preparedness. In this framework, exposure is not directly mentioned. However, he views exposure as a component of hazard. In risk triangles, each component of risk contributes in equal proportions, and changing any one of these three factors may change the level of overall risk.
30
These triangular frameworks are implementations of equation 2. Although some elements such as exposure and deficiencies in preparedness are seen as separate elements of risk, the relationship between these elements and the level of vulnerability and risk is not clear, nor does the model clarify the relationship between these elements and levels of vulnerability and risk. Therefore, how these elements contribute to practical implications is difficult to ascertain from these triangular models.
Davidson (1997) (Figure 3) introduced two new components into the traditional risk framework: external context and emergency response and recovery capabilities. His framework is directed to sudden onset disasters like earthquakes. In this framework hazard represents the geophysical phenomenon that serves as an initiating event (e.g., an earthquake). Vulnerability is conceptualised as the potential impact on physical structures and populations. Exposure describes the size of the area affected; it includes a list of everything that is subjected to effects of the hazard, including physical infrastructure, population, economy and social and political systems. External context is included to describe how damage to a certain place (e.g., a city) affects those outside that place (e.g., those in other cities). Emergency recovery and response capability describes how effectively and efficiently a city can respond to and recover from short and long-term impacts through formally organized activities that are performed after, before or during the disaster.
31