2.2. | IL MARCHIO, IL LOGOTIPO: LE DECLINAZIONI
School of Civil, Environmental and Land Management Engineering MSc Civil Engineering for Risk Mitigation
Post-disaster loss accounting vs disaster forensic: insights from the November 2013 flood in the Umbria region
Supervisors: Scira Menoni and Daniela Molinari
MSc graduation thesis by: Maria Camila Rodriguez Parra Matricola: 815983
December 2015 2 Acknowledgements
This thesis is dedicated to my dad that has unconditionally supported me on each of my dreams, making them a reality, even now from far away. Together with mi mom they made me the person I am today. IwouldliketothankmyfamilyandfriendsinColombiaandItalyforencour- aging me to keep going, specially in rough times. This work would not have been possible with out all the great people that sewed a net not to let me down. I would like to thank Jose for being my life companion through adventures and misadventures, but moreover, the person capable of holding me together when it seemed impossible to me. Many thanks to all the professors in Politecnico di Milano that shaped my learning process but specially to Prof. Scira Menoni and Prof. Daniela Moli- nari that as my thesis supervisors, guided this thesis through the path it took with valuable advises from their expertise. I am grateful to the people in Umbria region for all their help and in general to all the people that made this thesis possible.
3 4 Contents
Aknowledgements 3
Abstract 11
Abstract italiano 13
Introduction 15
1Framework 17 1.1 Flooddamage...... 17 1.2 Collection and recording of damage data ...... 19 1.2.1 Loss accounting ...... 20 1.2.2 Disaster forensics ...... 21 1.2.3 Risk modelling ...... 22 1.2.4 Compensation ...... 22 1.2.5 Overview of the applications ...... 23 1.3 Disaster forensics: the FORIN project ...... 26 1.3.1 Usesofforensicinvestigations ...... 27 1.3.2 Methods for forensic investigations ...... 28
2 Practices on damage data collection and recording 31 2.1 In Europe ...... 32 2.2 In Italy ...... 36 2.3 Challenges of these practices ...... 40
3 The November 2013 flood in Umbria 43 3.1 TheUmbriaregionriskmanagement ...... 44 3.2 Transition tests ...... 47 3.3 Results 2013 event ...... 47 3.3.1 Physic scenario ...... 48 3.3.2 Data processing ...... 49 3.3.3 Results...... 62
5 4 Discussion of the results 81 4.1 Data quality ...... 81 4.2 Challenges for forensic use of data ...... 86
5 Conclusions 91 5.1 Lessons learned from this project ...... 91 5.2 Future recommendations ...... 93
A General maps 97
B Point maps 111
6 List of Figures
1.1 Working scale and scope for loss accounting. Adapted from [De Groeve et al., 2014] ...... 21 1.2 Working scale and scope for disaster forensics. Adapted from [De Groeve et al., 2014] ...... 22 1.3 Working scale and scope for risk modelling. Adapted from [De Groeve et al., 2014] ...... 23 1.4 Working scale and scope for compensation ...... 24 1.5 Role of loss data in national processes of risk management [De Groeve et al., 2014] ...... 26 1.6 Target levels of the questions in the FORIN framework [Inte- grated Research on Disaster Risk, 2011] ...... 29
3.1 Administrative limits in the Tiber basin. [Mondo del Gusto, 2009] 44 3.2 Scheme of the cyclic conceptualization Poli-RISPOSTA project . 46 3.3 Flooding levels reached during 2013 event. [Centro Funzionale RegioneUmbria,2013] ...... 49 3.4 A↵ected municipalities during the November 2013 - February 2014 event. [Protezione Civile Regione Umbria, 2014] ...... 50 3.5 Timeline of source tables and strikes of the flood event...... 52 3.6 Scheme of the correlation process between the three source tables 56 3.7 Change of the financial request over time ...... 64 3.8 Change of the financial request and prioritization over time . . . 66 3.9 Change of the financial request and responsible stakeholder over time ...... 66 3.10 Change of the financial request and sector over time ...... 67 3.11 Frequency analysis for emergency management sector: type of activity ...... 68 3.12 Frequency analysis for public area sector: type of area and type of direct damage ...... 69 3.13 Frequency analysis for public good sector: type of infrastructure and type of direct damage ...... 70
7 3.14 Frequency analysis for road sector: type of infrastructure and typeofdirectdamage...... 71 3.15 Frequency analysis for public hydrogeological sector: type of direct damage ...... 72 3.16 Frequency analysis for water/sewage system sector: type of in- frastructure, cause of damage and type of direct damage . . . . 73 3.17 General map for road sector 4 months after the event ...... 75 3.18 General map for the road sector 1 year 3 months after the event 76 3.19 Point map for road sector and road network of Umbria . . . . . 78 3.20 Point map for road sector and blocked streets in the road net- workofUmbria ...... 79
4.1 Conectivity between types of uncertainty [De Groeve et al., 2014] 82 4.2 Example of measurement uncertainty obtaining coordinates of an intervention ...... 83
A.1 General map for emergency management sector 4 months after theevent...... 98 A.2 General map for public area sector 4 months after the event . . 99 A.3 General map for public area sector 1 year 3 months after the event...... 100 A.4 General map for public good sector 4 months after the event . . 101 A.5 General map for public good sector 1 year 3 months after the event...... 102 A.6 General map for road sector 4 months after the event ...... 103 A.7 General map for road sector 1 year 3 months after the event . . 104 A.8 General map for hydrogeological protective measures sector 4 months after the event ...... 105 A.9 General map for hydrogeological protective measures sector 1 year 3 months after the event ...... 106 A.10 General map for water/sewage system sector 4 months after the event...... 107 A.11 General map for water/sewage system sector 1 year 3 months after the event ...... 108 A.12 General map for electric sector 4 months after the event . . . . . 109 A.13 General map for electric sector 1 year 3 months after the event . 110
B.1 Pointmapforpubicareasector ...... 112 B.2 Pointmapforpublicgoodsector ...... 113 B.3 Point map for road sector and road network of Umbria . . . . . 114 B.4 Point map for road sector and blocked streets in the road net- workofUmbria ...... 115
8 List of Tables
1.1 Examples of types of damage. Adapted from [Messner et al., 2007] ...... 19 1.2 Application areas of disaster damage/loss data. Adapted from [De Groeve et al., 2014] ...... 25
2.1 Analysisofnationaldriversforlossdata ...... 33 2.2 Analysis methodology of loss data collection ...... 34 2.3 Analysis methodology of loss data recording ...... 35 2.4 AnalysisforAVIItaliandatabase ...... 37
3.1 List of responsible stakeholders for each sector. Adapter from [Politecnico di Milano & Umbria Region, 2015] ...... 51 3.2 Structure of source tables: Preliminary table, March 2014; First emergency table, July 2014; Final table, February 2015 . . . . . 53 3.3 Raw data resume of financial resources required by stakehold- ers: Preliminary table, March 2014; First emergency table, July 2014;Finaltable,February2015...... 55 3.4 Summary of the source tables ...... 55 3.5 Required information for a disaster forensic analysis for emer- gency management sector ...... 58 3.6 Required information for a disaster forensic analysis for public areas and public goods sectors ...... 59 3.7 Required information for a disaster forensic analysis for roads and hydrogeological protective measures sectos ...... 60 3.8 Required information for a disaster forensic analysis for wa- ter/sewage system and electric sectors ...... 61 3.9 List of output of the disaster forensic analysis for all the sectors 63
4.1 Simplified example of an intervention request for Gubbio mu- nicipality from partial source table ...... 84 4.2 Simplified example of an intervention request for Perugia province for hydraulic protection from partial source table . . . 85
9 4.3 Simplified example of an intervention request for Gualdo Cat- taneo municipality from final source table ...... 85 4.4 Simplified example of an intervention request for Campello sul Clitunno municipality from partial and final source tables . . . . 86 4.5 Comparison between priority letters definition according to gen- eral Law and specific-event Order ...... 88
10 Abstract
With increasing disasters provoking increasing damage along the world, dam- age data collection has become a raising concern as the basis to determine the e↵ectivity of risk management strategies. However present methodologies for damage data collection and recording after a disaster, as well as the uses given to these data, di↵er along countries, creating inconsistencies. In such a context, this thesis focuses on the feasibility of using damage data obtained with current methodologies for collection and recording for a specific use that is disaster forensic investigation. Aliteraturereviewwasperformedtodenethedi↵erentrequirementsofdamage data for each of their four applications: loss accounting, risk modelling, disaster forensic analysis and compensation. Special attention was given to disaster forensic analysis given its utility to supply a post-event comprehensive overview of the disaster. Through literature review it was also possible to describe the current practices for damage data collection and recording in Europe and specifically in Italy. This let to the identification of key challenges for the use of damage data for a di↵erent application than its original purpose. The November 2013-February 2014 flood in the Umbria region was chosen as the case study to perform a disaster forensic analysis with data collected with current practices in Italy, which are compensation oriented. As a result, it was possible to identify overlapping characteristics of data for compensation and disaster forensics applications. Moreover, a second set of specific challenges for the use of data was set, like: dealing with uncertainty leading to low quality of data in a smarter way, improving standards, having a more coherent way to account for the change of data on time, decreasing the time needed for data elaboration. Future developments on this field are encouraged to address these challenges. This will also set a step forward towards compatible methods for the collection of loss data worldwide that eventually can lead to European or global structured databases.
11 12 Abstract Italiano
Con il continuo aumentare dei disastri naturali e dei danni conseguenti in tutto il pianeta, la raccolta dei dati di danno `ediventata una delle principali priorit`a, come base per determinare l’e cacia delle strategie di gestione del rischio. Ciononostante, le attuali metodologie di raccolta e archiviazione dei dati di danno, cos`ıcome gli usi che si danno a questi dati, divergono da paese a paese, creando incompatibilit`a. In questo contesto, questa tesi si concentra sulla possibilit`adi usare i dati di danno ottenuti attraverso le attuali metodologie di raccolta e archiviazione per il particolare uso di analisi dell’evento calamitoso di tipo forense. In primo luogo, `estata fatta un’analisi della letteratura per identificare i req- uisiti dei dati di danno per ognuno dei quattro usi possibili: contabilit`adelle perdite, modellazione del rischio, analisi forense e risarcimento. Particolare attenzione `estata rivolta alle analisi di tipo forense, data la loro utilit`anel fornire un quadro comprensivo dell’evento. Attraverso l’analisi della letter- atura `estato anche possibile descrivere le attuali pratiche per la raccolta e archiviazione dei danni in Europa, e specicatamente in Italia. Questo ha per- messo di identicare alcune criticit`aper l’uso di dati di danno per uno scopo diverso da quello originale. L’alluvione del Novembre 2013-Febraio 2014 in Umbria `estata scelta come caso studio per un’analisi di tipo forense dell’evento con i dati raccolti con le pratiche attuali in Italia, finalizzate al risarcimento. Come risultato, `estato possibile identicare le caratteristiche comuni dei dati per gli scopi di risarci- mento e analisi forense. Quindi, ulteriori criticit`aper l’uso dei dati sono state identificate, tra cui: la gestione dell’incertezza legata alla qualit`adel dato, la gestione coerente dell’evoluzione temporale del dato, la riduzione dei tempi necessari all’elaborazione del dato. I futuri sforzi di ricerca dovrebbero con- centrarsi sul superamento di tali criticit`a.Questo permetterebbe fare un passo avanti verso metodi compatibili per la raccolta dei dati di danno a livello Eu- ropeo e globale, archiviati in database compatibili.
13 14 Introduction
The floods at the european level are the most frequent disasters and also the ones causing more losses with e52 billion between 1998 and 2009 [European Environmental Agency, 2011]. Moreover it is suspected that climate change will increase the occurrence of floods, making them also more severe (more frequent, more magnitude) [Messner et al., 2007]. New strategies leading to diminish the losses of floods have become an increasing concern worldwide; but the question is: what losses do these strategies want to diminish and how. Disaster forensic analysis is a tool for the reconstruction of a disastrous event and its causes; it can help understand what went wrong during the event, leav- ing room for improvements in future events. This thesis project focuses on the feasibility of using data obtained with the current methodology for collection and recording of loss data for a disaster forensic analysis. The damage data collection and recording after a disaster di↵ers along Europe, as well as the purposes for which the recorded data is used for. The November 2013-February 2014 flood event in Umbria region is the case study of the project, where a disaster forensic analysis will be performed with the data obtained from the current practices. This will permit to establish existing limitations and pro- pose challenges to be faced in the future with improvements on methodologies for damage data collection and recoding. The work starts with the definition of the framework in chapter 1. Basic key definitions regarding flood damage will be provided as well as the four applications that can be given to loss data. This chapter will also introduce the disaster forensic as one of the applications and will provide a wider description of the methods to perform it. Chapter 2 will give an overview of the current practices for damage data collec- tion and recording in Europe and more specifically in Italy. This will provide the context on which the case study is set on. The final part of this chapter will outline the identified challenges of using these practices. The case study will be presented in chapter 3, starting form the description of
15 the research context on which it is developed. After that, the physic scenario of the event is described and the data processing steps done to the raw data to perform a disaster forensic investigation will be outline. The chapter ends with the presentation of the results obtained from the described processing steps. The discussion of the results of the case study will be presented in chapter 4. The data quality and the challenges of using the data collected from current practices for a disaster forensic investigation will be presented. The final chapter 5 will present the conclusion of the thesis through the pre- sentation of the lessons learned from it. Moreover it will give some suggestions for future development of this field, basically based on the identified challenges to be faced. It can be seen as a the guiding path for future improvements of the practices for damage data collection and recording.
16 Chapter 1
Framework
This chapter is focuses on setting the reader into the framework of the project by establishing some key definitions and tools that will guide the whole un- derstanding of the subsequent reading. First, a key definition of flood damage will be given, as well as concepts of types of damage and losses that will be used along the whole document. Second of all, the main reasons for collecting and recoding damage data are described based on recent guidelines set at the European level [De Groeve et al., 2013], giving the identified applications. Fi- nally in the last section, a expanded definition of disaster forensic is given as well as the various uses that can be given to this investigations and some of the suggested methods to perform it based on the results of the FORIN project [Integrated Research on Disaster Risk, 2011]. This final section is important because it will define the required characteristics to perform a disaster foren- sic investigation like the one that will be done for the case study in chapter 3.
1.1 Flood damage
Disaster according to UNISDR [UNISDR, 2009] is a disruption of a commu- nity or a society involving widespread human, material, economic or environ- ment losses and impacts, which exceeds the ability of the a↵ected community or society to cope with using its own resources. From this, it is possible to classify a flood as a potential disaster, where a flood is a condition of inun- dation of a normally dry area [FEMA, 2014]; the causes of a flood can be: heavy rains causing overflow of a water source, snow melt, dam break among others. Floods, when in contact with a vulnerable and exposed community, can cause the disruption of the community or society. Moreover, floods are
17 often accompanied by other kinds of hazards that are also triggered by the same or similar causes: landslides, erosion, debris flows; this group of hazards are called hydrogeological hazards and are the one to be treated in the case-study.
According to the Economic Commission for Latin America and the Caribbean [ECLAC, 2003], disaster damage is defined as the total or partial destruc- tion of physical assets existing in the a↵ected area, measured in physical units, occurred after a disaster. And according to the Nuclear Regulatory Commis- sion [NRC, 1999] the disaster loss is the market-based negative economic impact caused by a disaster; in this project the disaster loss consists of the transformation of the disaster damage (both direct and indirect as it will be shown below) into the equivalent economic units representing the replacement or compensation cost; this process is also known as monetarization. For exam- ple, the destruction of a 100 m2 of floor in a house after a flood is the disaster damage; if every m2 of that floor costs e20 to be replaced, then the disaster loss would be e2000.
The process of monetarizing the disaster damage is not always straightfor- ward, take for instance the example of transforming the human damage into an economic value. Two types of disaster damage can be identified, tangi- ble damage when it can be monetarized and intangible damage when it cannot.
Another classification of disaster damage/loss distinguishes between direct damage and indirect damage. ECLAC [ECLAC, 2003] defined the direct damage of a disaster as the physical damage to capital assets due to direct physical contact with the hazard, in this case the flood water (or the land- slide). While according to Benson & Clay [2000] the indirect damage is the damage to the flow to goods and services due to a dysfunctionality of services, infrastructure, businesses. Note also that when the direct or indirect damage can be monetarized (meaning it is tangible) it can be translated into direct loss or indirect loss respectively. Some examples on the types of damage are given in table 1.1.
The estimation of indirect damage is normally more complicated because it requires a more detailed and a long term evaluation, take of instance the case of e↵ects to the economic productivity of a business, where the e↵ects will only be evident some months after the event (and there are cases where they are evident some years after the event). The accounting of intangible damage also represents a challenge, take for example the di↵erent categories that can derive for a↵ected people: fatalities, injured, evacuated, isolated [De Groeve et al., 2014]; another example is the damage to cultural heritage sites or goods or the
18 Table 1.1: Examples of types of damage. Adapted from [Messner et al., 2007] Tangible Intangible Physical damage to buildings Illnesses Physical damage to stock Effects to the environment Direct Damage to agricultural products Damage to cultural heritage … … Emergency management cost Trauma Indirect Reduction of business production Loss confidence in authorities … … damage to memorabilia1.Thisiswhymostofthedamageassessmentsdonot include these e↵ects in the accounting and many databases only include the direct tangible losses and a small portion of the direct intangible damage (in terms of a↵ected people).
1.2 Collection and recording of damage data
The European Union (EU) represents an environment to discuss create and further individually apply the legislative basis and guides in various fields. The risk management is an important field frequently debated on this environment and a particular interest has been given to the need of an European level loss data recording system; this need stands, among others on the concern of EU on transboundary2 risk management, and to check the e ciency of Disaster Risk Reduction (DRR)3 policies [De Groeve et al., 2013]. This is why nowadays many of the legislative tools in Europe, within the EU disaster prevention framework [Council of the European Union, 2009], require or addresses loss data recording, some of them are: the Flood Directive [European Parliament, 2007], the European union solidarity fund (EUSF 2014), the union civil pro- tection legislation (2013), the ISPIRE directive (2007), the green paper on insurance of natural and manmade disasters (2012). However, all these tools lack in requiring a specific standard for the loss data
1Memorabilia is the losses of high a↵ective-value objects that cannot be replaced like family photographs or pets 2Transboundary refers to things that are not contained within administrative boundaries. Many things can be transboundary: e↵ects, management, resources, etc. 3Disaster Risk Reduction is the concept and practice of reducing disaster risks through systematic e↵orts to analyse and reduce the causal factors of disasters, further information can be found in [UNISDR, nd]
19 collection, recording, analysis and uses. This is why the European Commis- sion set up a specific working group on the topic; the results are collected in three series of reports on disaster loss data recording [De Groeve et al., 2013, De Groeve et al., 2014, Corbane et al., 2015] which constitute the latest e↵ort at European level of approaching the need of loss data recording and o↵er some guidelines and recommendations for standardization that permits the required compatibility for sharing purposes. These recommendations and guidelines are further intended to permit the future creation of an European level disaster loss database through the aggregation of compatible and consis- tent national databases.
According to the reports, the uses that can be given to loss and damage data are numerous; from checking the e↵ectiveness of a determined mitigation mea- sure during a disaster, to understanding the global e↵ects of climate change. But all of them are, in some way, directed to risk mitigation strategies; this means that the loss data are the base to define the success of mitigation strate- gies after an event, as well as to design new mitigation strategies for future events. From the various uses that can be given this data, there are some clear applications that group these uses. Initially the first report of the JRC [De Groeve et al., 2013] identified three applications to give to damage data (loss accounting, risk modelling and disaster forensic), but the second report [De Groeve et al., 2014] included a fourth application that was identified form the study of the current practices in Europe (Compensation). In the follow- ing subsections the four applications are described, together with the required characteristics of the data for each application based on the review of the mentioned reports. Some of the basic requirements are the scale, meaning the detail of recording, and the scope referring to the geographic coverage respectively.
1.2.1 Loss accounting
Loss accounting is the principal motivation for recording data. This application uses the data to document the trends of the losses, to measure the e ciency of DRR policies or for decision making on balancing prevention budget with loss compensation funds. Figure 1.1 shows in red the working area to perform a loss accounting in terms of scale (vertical axis)-scope (horizontal axis) of data. The process for the collection and recording basically intends the collection at small levels (small scale) and the aggregation of the data to reach larger levels (larger scope) where the data can be analyzed. Compatibility is required for the aggregation process.
20 Take for instance the documentation of a regional trend (scope) of disaster losses due to flooding. This can be accomplished for example with options 1 or 2 illustrated in the figure 1.1: by collecting data (scale) at the regional level, or at the asset level that can later on be aggregated until the regional level is reached respectively.
ACCOUNTING
Global
Nation
Region 1 Scale
Munic
Asset 2
Asset Munic Region Nation Global Scope
Figure 1.1: Working scale and scope for loss accounting. Adapted from [De Groeve et al., 2014]
1.2.2 Disaster forensics
The objective of disaster forensics is to identify the loss causes during a disaster by understanding the unfolding of the event form the recorded damage data. The analysis measures the contribution of exposure, vulnerability, resilience, mitigation and response to later estimate what aspects need to be strength- ened. This kind of analysis permits to evaluate the e↵ectiveness of specific disaster prevention measures. Figure 1.2 shows in green the working area to perform a disaster forensic investigation in terms of scale-scope of data. More detailed information is required to perform disaster forensic investigations and this is why the maximum scale is the regional level. For a regional disaster figure 1.2 shows two ways to perform a disaster forensic investigation: with information gathered at the regional level (number 1), or with information gathered at asset level (number 2).
21 FORENSIC
Global
Nation
Region 1 Scale
Munic
Asset 2
Asset Munic Region Nation Global Scope
Figure 1.2: Working scale and scope for disaster forensics. Adapted from [De Groeve et al., 2014]
1.2.3 Risk modelling
Risk modelling is the last application identified in the first report of the JRC [De Groeve et al., 2013]. This application uses the loss data for the calibration and validation of risk models, on any of its components (hazard, vulnerability, exposure, damage). Correctly calibrated and validated models are used to es- timate future disasters and are also useful for the decision making process for the preparedness of future events. Figure 1.3 shows in purple the working area to perform risk modelling in terms of scale-scope of data. A clear example is shown in figure 1.3 with number 1 and refers to the use of damage data at building leve and the height reached by the flood water to construct and validate depth-damage curves (also known as vulnerability curves). A second example is shown in the same figure with the number 2, referring to the con- struction of national hazard, vulnerability and exposure maps using regional data.
1.2.4 Compensation
The second JRC report [De Groeve et al., 2014] identified compensation as an additional application from the study of the current practices in Europe. The compensation process is part of the recovery and can occur at di↵erent levels: national, European or international. It is guided by solidarity mechanism
22 RISK MODELLING
Global
Nation
Region 2 Scale
Munic
Asset 1
Asset Munic Region Nation Global Scope
Figure 1.3: Working scale and scope for risk modelling. Adapted from [De Groeve et al., 2014] or insurance markets and requires a detail loss assessment. One example of compensation is the European Union Solidarity Fund, established in 2014. Figure 1.4 shows in blue the working area scale-scope to perform compensation. The number 1 in this figure represents the example of a regional event that collects data at the asset level and aggregates it until the regional level to request the compensation of losses.
1.2.5 Overview of the applications
Table 1.2 shows additional features for the applications, permitting a critical comparison to understand the overlapping aspects as well as the di↵erences between the applications. This table evaluates 6 features for each of the ap- plications: Driver, referring to the reasons to record loss data in a certain ap- plication; the relevant legislation agreements that require loss data recording; the loss period evaluated for each of the applications; the interested stake- holders on recording loss data; required scale level to collect the data for each of the applications; and the required characteristics to be recorded. From this table it is possible to see that some of the features are overlapping for the applications, for example the direct monetary losses are recorded for all the applications; this overlapping can be a starting point towards compatible national loss databases towards the objectives of the JRC reports.
23 COMPENSATION
Global
Nation
Region Scale
Munic
Asset 1
Asset Munic Region Nation Global Scope
Figure 1.4: Working scale and scope for compensation
Such level of compatibility will required that the recorded data has to serve to each one of the applications the states are currently using them for (compen- sation, loss accounting, disaster forensics and risk modelling); it means that if the loss data of a Member State is currently being used for compensation, in the future a shift of its methodologies will be required so that the recorded data serves also for loss accounting, forensics and risk modelling, since other Member states are using the data for those purposes. Aconsistentstructureddatabasethatcanbecompatiblewiththedatabases of other Member States (or other extra-european states) can be then used for several purposes and permit a flow of information like the one illustrated in figure 1.5, where the database is both fed and used by three di↵erent actors during the processes of risk management: scientists, using the data for example for the assessment of hazard and risk, as well as feeding by supporting minimum requirements of the data model; Decision makers, using the data for instance to check the objectives and the e ciency of investments; and practitioners, using the data for example for the definition of the plans and feeding it with the loss collection. Other actions performed by the three actors can be found on the mentioned figure. Since the ideal collection and recording of damage data has small scale but large scopes, it puts more pressure on smaller territorial levels to gather the information while the larger levels are benefitting the most form the use of this information. Therefore, one of the most important recommendations for the design of the databases is to engage the local actors to establish a loss database
24 Table 1.2: Application areas of disaster damage/loss data. Adapted from [De Groeve et al., 2014] Use Compensation Accounting Forensics Risk modelling Fair and efficient Avoiding sovereign Evaluate prevention Accurate risk assessment solidarity mechanism- insolvency measures and protection based on local loss insurance market exceedance curves
Driver Balance prevention and Improve prevention Develop economic loss compensation budgets policy models to estimate indirect losses National legislation on National legislation on EU Council conclusions EU Flood directive compensation of victims disaster prevention and on risk management and government aid risk assessment capability
Insurance policy HFA-2 Union Civil Protection HFA-2 Mechanism and agreements and
Relevant legislation legislation Relevant EU solidarity fund EU Council Conclusions on a Community framework on disaster prevention Event-based Cover future losses Event-based Use archive to estimate
Loss future losses period Monitor trends in losses MSs with public MSs with high annual MSs Emergency MSs potentially affected compensation scheme average losses-high Management authority by climate change maximum probable loss
European Union
Insurance industry Financial system Regional and local Scientific community emergency management United natons authorities Insurance industry Interested stakeholders Interested Civil society EU Member States and Institutions Asset-based National-regional Event-based Asset-based (sampling) aggregates scale Required Required Direct monetary losses Direct monetary losses Direct and indirect Direct and indirect monetary losses monetary losses
Human losses Human losses
Uncertainty Dynamics of impacts Specific asset-related (population movements, information (number of evacuations), response floors, water height, (decisions, actions) and level of damage, etc.) hazard (evolution) Narrative What needs to be recorded be to needs What Human losses Uncertainty HFA-2: Hyogo Framework for Action beyond 2015; MS: Member State
25 INTRODUCTION
What needs to Direct monetary Direct monetary Direct (and indirect) Direct and indirect be recorded losses losses monetary losses economic losses
Human losses Human losses Human losses
Uncertainty Dynamics of impacts Specific asset-related (population information (number movements, of floors, water height, evacuations), response local soil type, level of (decisions, actions) and damage, etc.) hazard (evolution) Narrative
Uncertainty
1.2 NATIONAL LEVEL: PROCESS VIEW
The applications of loss data are interwoven and stakeholders typically are involved in multiple strands at once. At national level and seen as a process, loss data recording can be described as in Figure 1.
Disaster loss data collection involves a number of stakeholders, such as decision makers, scientists and practitioners with each their responsibility and function.
Figure 1. Conceptual information flow for the implementation of crisis management plans and the role Figure 1.5: Role of lossof disaster data loss in data national within this process processes of risk management [De Groeve et al., 2014] Disaster risk awareness, unless made obvious by a disastrous event, is often brought to attention as an issue for the safety of societies by academia/and or practitioners (S1 in the figure 1). Policy thatmakers they may can act use on the on scientific their risk evidence management by establishing routines, a national andrisk assessment, that can and later there on be aggregatedfrom defining at disaster national risk reduction and international objectives through level the for allocation the robust of resou strategicrces and drafting purposes [Delegislation Groeve (P1). et al., It is then 2013]. taken up by the mandated bodies that draft the implementation plan (T1) and execute the plans (T2). The appropriateness of risk reduction and prevention measures is evaluated over time (P2, e.g. through peer review processes or internal review processes) and 1.3 Disaster forensics: the FORIN project 19
The Integrated Research on Disaster Risk IRDR within the FORIN project developed between 2010 and 2011 a series of disaster forensic investigation case studies. All the case studies are available to public in the webpage bank of the IRDR. As a result of these case studies, an approach of a methodology to perform disaster forensic investigations was developed; the methodology was written in 2011 in the report of the FORIN approach [Integrated Research on Disaster Risk, 2011]. This section is intended to provide a wider defini- tion of disaster forensics, its uses and the possible methods to perform this investigations based on the mentioned report. Most of the post disaster investigations focus their attention on the physical aspect of the events (magnitude, frequency, distribution, causal mechanism), some of them also investigate the response of the emergency, but few of them deeply investigate the whole decision making process and the policy justifying it. These aspects are the link of the physical aspect of the event and the response given to the emergency [Integrated Research on Disaster Risk, 2011]. The decision making process, together with the socio cultural aspects of the community are the key aspects shaping the resilience and vulnerability of a
26 community; therefore it is important to pay more attention on such aspects during the aftermath of a disaster. This is the principal motivation to perform structured disaster forensic investigations.
1.3.1 Uses of forensic investigations
The uses of disaster forensic investigations are multiple and can vary from lo- cal scopes like determine the e ciency of a specific mitigation measure during a disaster, to global scopes like the identification of the main weaknesses of policies driving the unequally distribution of disasters occurrence and losses between developed and developing countries. In the FORIN approach [Inte- grated Research on Disaster Risk, 2011], the identification of such uses required first the establishment of the research elements and the final identification of the five categories of objectives that summarize the uses that can be given to forensic investigations. The four key research elements for the FORIN project are the following [In- tegrated Research on Disaster Risk, 2011]: 1) to identify the circumstances, causes and consequences of loss, as well as the conditions that have prevent loss; 2)toidentifyandtestaseriesofhypothesisofdamagecasualty;3)to identify key factors in the expanding number of losses in disasters and how they enter into risk and disaster; 4)toinvestigatetheuseofexistingscientific knowledge in risk assessment and management. The identification of the elements lead to the establishment of the objectives of the FORIN project [Integrated Research on Disaster Risk, 2011], grouped into five categories: Policy objectives to experiment with multi-disciplinary and multi- stakeholder inputs; to encourage participation of decision makers; to guide policy as well as public and private investments on risk reduction. Management objectives to link the research findings with policy improve- ments through the establishment of a case-study bank for the e↵ective communication of causes of disasters. Scientific research objectives to advance methodological diversity; to test existing theories and concepts; to build interdisciplinary capacity of policy-oriented research. Development objectives to promote learning culture; to guide recovery and reconstruction e↵orts; to communicate key messages required for paradigm change; to advance understanding of causal factors of disasters
27 like: justify that solutions are locally e↵ective, that e↵ects of disasters are impediments to development and also that some development initiatives can be causal factors of disasters. DRR objectives to help and support DRR and Hyogo Framework for Action implementations with the case-studies, giving priority to the reduction of human losses and shifting paradigm of responsibilities (from nature to social sectors of the society) To sum up, the disaster forensic investigations permit to improve policies, to improve the management of information; to improve the scientific research methods; to understand and improve the dynamics of development; to improve the implementation of Disaster Risk Reduction within the Hyogo Framework. All these uses, as it was mentioned, vary form a local scope to the global scope, but all look towards the reduction of disaster losses.
1.3.2 Methods for forensic investigations
The FORIN approach [Integrated Research on Disaster Risk, 2011] identifies six levels of driving factors for disaster risk reduction. These levels constitute the framework of the FORIN approach and can be seen in figure 1.6: level 1, governance; level 2, risk assessment, including the causal agents, social sys- tems and infrastructure; level 3, understanding & awareness; level 4, outcomes and impacts, including the distinction between sector, spatial and susceptive population; level 5, risk reduction; and level 6, resilience enhancement. All of them evaluated during the antecedent conditions, the emergency response and the long-term recovery. Forensic investigations should therefore understand the role of each of the levels on the disaster and it is also suggested to address all the levels during the investigation. Once the framework of the FORIN investigations has been understood, the report suggests four di↵erent methods to guide the path of a forensic investi- gation. The choice of a method depends on the context of the disaster, as well as the objectives of the investigation: Critical case analysis Seek to identify the root causes of the disaster events with a multi-disciplinary approach that integrates social, environmental and technical assessments. Meta-analysis Systematic reviews of the available literature carried out to identify and assess consistent findings across diverse studies for causal linkages as well as the e↵ectiveness of interventions.
28 5. The FORIN framework
A conceptual framework for the FORIN investigations is shown diagrammatically in Figure 1. This is derived in part from the Hyogo Framework for Action, which guides the work of governments and international organizations under the United Nations International Strategy for Disaster Reduction (UNISDR).