AFTER 10 YEARS OF CAPRA

Eduardo REINOSO1, Mario ORDAZ2, Omar-Dario CARDONA3, Gabriel A. BERNAL4, Marcial CONTRERAS5

ABSTRACT

The open source software CAPRA (Comprehensive Approach to Probabilistic Risk Assessment) has been used for more than 10 years. It is an on-going initiative that has been developed in different phases with the financial support, in the beginning, of the World Bank, the Inter-American Development Bank and the UNISDR. The goal, 10 years ago, was to improve the understanding of disaster risk due to natural hazard events, such as earthquakes, tsunami, hurricanes and floods, among others, and to generate incentives to develop planning solutions and results to reduce and mitigate potential damages. The original version of CAPRA is a platform that provides, for different type of users, tools, information and data to evaluate risk. CAPRA applications include a set of different software modules for different types of hazards, a standard format for exposure of different components of infrastructure, a vulnerability module with a library of vulnerability curves, and an exposure, hazard and risk mapping geographic information system. Probabilistic techniques of CAPRA employ historic and stochastic approaches to simulate hazard intensities and frequencies across a country’s or any territory. This hazard information can then be combined with the data on exposure and vulnerability, and spatially analyzed to estimate the resulting potential damage. Results are expressed in risk metrics such as the exceedance probability curve, expected annual loss and probable maximum loss for any given return period, useful for multi-hazard risk assessment. The platform’s architecture has been developed to be modular and open, enabling the possibility of harnessing various inputs and contributions. This approach enables CAPRA to become a living instrument. CAPRA’s innovation extends beyond the risk-modeling platform; a community of disaster risk users is now growing in the countries. Training and workshops have been developed in many countries in the last decade. Using CAPRA, it has been possible to design risk transfer instruments, the evaluation of probabilistic cost-benefit ratio, providing a tool for decision makers to analyze the benefits of risk mitigation strategies, such as building retrofitting. This model has been useful for land use planning, evaluation of loss scenarios for emergency response, early warning systems, on-line loss assessment systems, and for the holistic evaluation of disaster risk based on indicators, facilitating the integrated risk management by the different stakeholders involved in risk reduction. CAPRA has been used in Central and South America and in some countries of Africa, Europe and Asia, for the evaluation of the country’s risk profile for most countries of the Americas and risk metrics for 200+ countries in the framework of the United Nations Global Assessment Report GAR, among other projects. This paper illustrates how CAPRA has been used in many places for hazard and risk assessments, including seismic risk.

Keywords: Probabilistic seismic risk assessment; Building damage; Risk reduction; Loss scenarios; CAPRA.

1. INTRODUCTION

The loss of life and economic impact of natural hazards are most severe if they have been not properly evaluated and considered as determinants of development planning. At present, in many countries, it is necessary to help decision-makers understand risk, estimate probable losses, and reconstruction costs, so they can assess disaster risk reduction strategies. Acknowledging this need, in January 2008 the World Bank through the Global Facility for Disaster Reduction and Recovery (GFDRR) launched the CAPRA initiative (originally as Central America Probabilistic Risk Assessment) to evaluate hazard and risk in any scale or resolution (Cardona et al. 2012; Marulanda et al. 2013). It was initially aimed to

1 Professor, Instituto de Ingeniería, UNAM, Mexico City, Mexico, [email protected] 2 Professor, Instituto de Ingeniería, UNAM, Mexico City, Mexico, [email protected] 3 Associate Professor, Universidad Nacional de Colombia, Manizales, Colombia, [email protected] 4 Technical Manager, Ingeniar: Risk Intelligence Ltda., Bogotá D.C., Colombia, [email protected] 5 Research associate, ERN, Evaluación de Riesgos Naturales, Mexico City, Mexico, [email protected]

establish a common methodology and set of tools to assess disaster risk in the region and help develop risk reduction strategies. It started with pilot studies for and Nicaragua. CAPRA has also received support from the Central American Coordination Center for Disaster Prevention (CEPREDENAC in Spanish) and the United Nations Office for Disaster Risk Reduction (UNISDR), SFLAC6 and DFAT-AG7. In the years that followed, the Inter-American Development Bank (IDB) joined the effort through the Multidonor Disaster Prevention Trust Fund, expanding CAPRA to other Central and South American countries. Ten years later, the program success has turned CAPRA8 into a world class risk assessment platform, that has allowed several countries in the Americas, Africa, Europe and Asia to evaluate sovereign risk for earthquake, tsunami, tropical cyclone (wind and storm-surge), excess of rain and flooding, landslides and volcanic eruptions; with the financial support of the UNISDR (GAR and DEVCO projects) and the World Bank. Since 2017, the University of Los Andes, in Bogota, was selected to manage the eCAPRA website (Uniandes, 2017) and keep open and freely available the early version of CAPRA (WB, 2016a, 2016b). The new developments, as result of the evolution and current applications of the platform worldwide, are supported by ERN Ingenieros Consultores, S.C., INGENIAR: Risk Intelligence and CIMNE9; the main joint venture members of the ERN-AL (Evaluación de Riesgos Naturales - America Latina) consortium that developed the CAPRA methodologies and tools.

2. OVERVIEW OF HAZARD AND RISK ASSESSMENT TOOLS

There are sundry hazard modeling programs. GFDRR conducted a review and ranking of open source codes (GFDRR, 2014a) finding only a few combine exposure and vulnerability to assess multi-hazard risk as CAPRA does. The SHARE project (GRCG, 2010) developed a common methodology and tools to evaluate earthquake hazard in Europe. With a similar scope, CRISIS, the hazard module of CAPRA (R-CRISIS, 2017) and the Open Quake program, of the Global Earthquake Model, aim to further the development of modules for earthquake hazard and risk analysis. HAZard U.S. (HAZUS) (FEMA, 2011) has been a multi-hazard risk assessment software focused on North American typical construction types, that has been successfully used in the U.S. and Puerto Rico. InaSAFE is free and open source software that produces realistic natural hazard impact scenarios; it works as a plugin for the free GIS editor QuantumGIS and was developed with the support of the GFDRR in the framework of the Australian-Indonesia Facility for Disaster Reduction and by the Government of Indonesia. Other platforms with regional specific vulnerability curves are Risk-EU, GA and PCRAFI. Risk-EU provides capacity curves for several construction classes for European countries. Geoscience Australia (EQRM, TCRM, TsuDAT, ANUGA) developed vulnerability functions for the Asia-Pacific region and they were used with CAPRA for the Global Assessment Report on Disaster Risk Reduction (GAR) of the UNISDR (2015a) in 2015. The Pacific Catastrophe Risk Assessment & Financing Initiative provides risk-related geospatial data sets based on AIR Worldwide proprietary software. Lastly, the Global Risk Model of UNISDR, based on CAPRA, has been used for the GAR 2011, 2013, 2015 and recently for the GAR Atlas 2017. It combines key elements from regional-scale studies and other platforms (vulnerability functions) to procure comparable measures of risk between countries over time for six perils.

On a different-scope, sophisticated risk models have also been developed for the Caribbean Catastrophe Risk Insurance Facility (CCRIF, 2017) also supported by the GFDRR. CCRIF’s risk assessment framework is specifically aimed to model risks for a community of countries subscribing natural hazard insurance. In 2007 the facility started by pooling risk among 16 countries and it has announced plans to provide similar coverage to Central American governments. CCRIF provides rapid liquidity when a policy is triggered in the wake of an earthquake, heavy rainfall, or a hurricane. The 2016-2017 models update offers a still more sophisticated risk assessment framework. For more details on CCRIF, see ERN-RED (2015).

6 Spanish Fund for Latin America and the Caribbean. 7 Department of Foreign Affairs and Trade of the Australian Government. 8 Some reports redefine CAPRA as Comprehensive Approach to Probabilistic Risk Assessment to reflect the broader scope of the platform. 9 International Center for Numerical Methods in Engineering (CIMNE in Catalonian).

2.1 FONDEN software in Mexico

With collaboration between officials from public infrastructure and buildings in México, UNAM and the firm ERN, it was possible to compute losses and loss scenarios for Mexico’s government decision- makers. This collaboration has included engineers, researchers, architects, physicians and administrative people, among others. This effort has meant the development of databases and software for financial risk assessment in Mexico. On one hand, all the information has been gathered to create a database of almost all public infrastructure and buildings. What the government has now is the location (in terms of GIS systems), the number of stories, structural and contents value, structural type, year of construction (therefore, design building code) among others. It also has detailed information of inspections and reports by local maintenance personal with information such as recent damage and behavior due to natural hazards.

The R-FONDEN software has been developed as part of full probabilistic loss estimation software for earthquake, tsunami and hurricane risks that assess losses for structures, infrastructure, contents and non-structural components, among others (ERN, 2008, 2009, 2011, 2013). These models, developed at first for the insurance sector, have been very useful to draw maps of hazard and risk to point out those structures at the highest risk and, therefore, those that should be attended and retrofitted first. They include (Table 1) the earthquake hazard in terms of spectral acceleration, the tsunami hazard in terms of flooding areas and the hurricane hazard in terms of three internal hazards: wind, flood, and storm surge (to assess the hazard due to low frequent events and for non-hurricane areas, other non-hurricane wind, flood, and storm-surge hazards has been included in the so-called hurricane model). The models are capable of computing, formally, as the CAPRA platform, the average annual loss, the probable maximum loss for any return period, deterministic losses for a given event, among others. Table 1. Perils and hazards included in the R-FONDEN software and their intensity parameter.

Earthquake Volcanic eruption Hazard Intensity parameter Hazard Intensity parameter Strong ground motion Response spectra Ash fall Ash thickness Tsunami Water speed and depth Lava flow Lava footprint Landslide Lateral soil (Newmark) displacement Lahar Lahar footprint Surface fault Displacement Pyroclastic Pyroclastic footprint Liquefaction Ground deformation

Hurricane Heavy storms Hazard Intensity parameter Hazard Intensity parameter Wind Peak gust wind speed Landslide Destructiveness Storm surge Extent, depth Wind Peak gust wind speed Flood Extent, depth Flood Extent, depth Hail Kinetic energy

Tornado Drought Hazard Intensity parameter Hazard Intensity parameter Wind Peak gust wind speed Humidity Humidity

The information of almost all assets within Mexico is already in a GIS format and with minimum information to assess losses using R-FONDEN. This effort took almost five years and, as expected, it is still in progress since it is difficult to gather all desired information. The estimated exposure value of all assets is around 1,500 billion USD. What the exposure database contains is the follows: • Standard buildings: dwellings, hospitals, schools and offices. • Public buildings: Malls, hotels, touristic places and churches. • Cultural heritage buildings such as museums and archeological sites. • Lifelines: roads, bridges, railways, water supply systems, electricity and telecommunication networks. • Places with hazardous materials such as gas stations. • Hydraulic infrastructure: dams, channels and water distribution networks. This model informed the design and placement of the Government of Mexico’s risk transfer instruments, including the US$315 million catastrophic bond MultiCat Mexico 2012 as well as an indemnity-based excess-of-loss insurance program of US$425 million (GFDRR, 2013).

3. METHODOLOGICAL APPROACH

The CAPRA platform is based, overall, on a probabilistic risk assessment methodology that evaluates expected losses on the exposure inventories of buildings and assets within the study area. First, hazard information obtained from historical data and published investigations is used to build probabilistic hazard models. Then, intensity measures -seismic, tsunami, strong winds, storm-surge, flood, volcanic products- are related to economic loss through physical vulnerability functions. The main output risk metrics are the loss exceedance curve, the average annual loss, the probable maximum loss and pure risk premium (Grossi and Kunreuther, 2005; Cardona et al., 2008; ERN-AL 2010a, 2011a; Cardona et al., 2012; Niño et al. ,2015; Jaimes and Niño, 2017) from individual or simultaneous perils (Ordaz, 2015): • The loss exceedance curve (LEC) represents the annual frequency with which determined economic loss will be exceeded. A hybrid loss exceedance curve has been developed as well, integrating the costs of strong large-return-period events and small frequent events (ERN-AL, 2011g). • The average annual loss (AAL) is the mathematical expected annual loss. That is the sum for all stochastic events of the product of the frequency of occurrence and expected loss for each event. • Probable maximum loss represents the value of the worst-case scenario for a specific return period. • Pure risk premium corresponds to the value of the AAL divided by the replacement value of the asset. It indicates the cost that must be paid annually in order to cover expected losses in the future. The platform includes software tools grouped in modules. The hazard modules allow to model hazard frequency and intensity for earthquakes, tsunami, tropical cyclones, floods, and volcanic hazards. The exposure model uses information to depict the inventory of assets, including tools for the localization, classification, qualification and valuation of infrastructure potentially exposed to the hazard being considered. The physical vulnerability module is used for the development of physical vulnerability functions specific for each hazard and class of asset considered (structural classification, building size and occupancy type), and assign vulnerability functions to the exposed elements. Finally, the loss module is used to calculate the potential for losses for user-defined return periods or specific scenarios (WB, 2016a). The spatial variation of results may be presented using the integrated geographical information system CAPRA-GIS (ERN-AL, 2010d). CAPRA is also a useful tool for risk assessment taking into account climate change and variability (IDB, 2016e).

4. OUTCOME OF THE INITIATIVE

The CAPRA initiative has advocated proactive data sharing and risk identification activities throughout the [Central American] region (GFDRR, 2008). Some outstanding outcomes are the seismic risk assessment of schools in the Andean Region and in Central America (ERN-AL, 2010e). Technical Assistance Projects (TAPs) of the GFDRR, have targeted risk assessment using the CAPRA platform in several countries. These projects included workshops and capacity building on risk management, reduction and transfer. The CAPRA initiative has also obtained the holistic evaluation of risk for different countries and cities. This comprehensive approach is made based on social fragility and lack of resilience to cope with catastrophe risk. Total risk has been obtained by affecting the physical risk by an impact factor representative of socioeconomic and governance conditions (Marulanda et al., 2018).

The IDB has supported National Profile of Catastrophic Risk studies and hazard and risk specific reports in the most member states using CAPRA as the risk assessment platform. The Risk Profile is an important technical input to identify actions and investments with regards to disaster risk management as well as emergency response action plans. Table 2 presents a description of some of the studies and their outcomes while Figure 1 shows a timeline and a map summary of most studies. With only ten years after the first studies publishing, it is difficult to assess the impact in loss reduction. However, it is clear the studies have provided an important step for most countries, where applications including cost-benefit analyses of risk mitigation plans, land-use planning, capacity building programs, analysis of premiums and access to a Cat-DDO (Catastrophe Deferred Drawdown Option) for some countries have been done.

Table 2. Example of studies and reports using CAPRA.

Country Risk Analysis and Application Argentina Application and updating of the risk and risk management index system, conducted by ERN (2009) for the IADB. Flood risk at Santa Fe City, national earthquake hazard and risk and earthquake risk at El Gran Mendoza. Hazard evaluation: forest fire and volcanic activity, conducted by Ingeniar & CIMNE (IDB, 2016a). Bangladesh Impact of spatiotemporal distribution of population on earthquake loss. Case study for Sylhet City (Ara, 2014) Barbados Disaster risk management index (ERN-AL, 2008-2009), conducted by ERN. Update of Disaster Risk and Risk Management Indicators to 2014, conducted by Ingeniar & CIMNE. Belize Disaster risk assessment for Belize City (ERN-AL, 2011f). Capacity building, conducted by Ingeniar & CIMNE. Bhutan Earthquake risk assessment at Thimphu Thromde, to contribute towards the development of a holistic strategy to mitigate human and physical damage (ERN, 2013). Training on probabilistic risk assessment tools (WB, 2014), conducted by ERN. Bolivia Application and updating of the risk and risk management index system, conducted by ERN (2009) for the IADB. Flooding and landslides (IDB, 2016b), conducted by ITEC. Brazil Flooding at Santa Catarina, Pernambuco and Alagoas. Flooding and landslides at Rio de Janeiro by ERN (BM, 2008, 2010a, 2010b, 2011). Drought in northeast region (Bernal et al., 2017). Chile Disaster risk management index, conducted by ERN (2009) for the IADB. Earthquake and tsunami at Antofagasta and Atacama by ERN (ERN-AL, 2011h). National seismic hazard and risk and flood risk in Carahue, conducted by Ingeniar & CIMNE (IDB, 2016c). Probabilistic risk estimation for selected IIRSA infrastructure (ports and airports) for earthquake and tsunami, with support of the IDB, conducted by ERN (2015). China Seismic risk physical and human loss estimation on Sichuan Province (Yi et al., 2015) Colombia Application and updating of the risk and risk management index system, conducted by ERN (2009) for the IADB. Earthquake hazard and risk assessment for Pereira, Medellin, Bogota, Manizales and at national level. Volcanic risk for Pasto municipality conducted by Ingeniar & CIMNE (ERN-AL, 2011b, 2011g, 2012; Salgado-Galvez et al. 2010, 2012, 2016, 2017). Flooding risk at La Mojana (Cardona et al., 2017e) Costa Rica Risk scenario modeling for water and sanitation systems, and flooding risk due to hurricanes at Orosí district and river basin. Earthquake risk at Cantón de San José. Tsunami at and Puerto with application to land-use planning. Landslides at Cinchona, Orosí, Río Savegre basin. Volcanic risk at San José (ERN-AL Result 5, 2011c; ERN, 2014), conducted by ERN. Dominican Disaster risk management index (ERN-AL, 2008-2009) and national risk evaluation for earthquakes and Republic tropical cyclones, conducted by ERN for the IADB. Earthquake in Santo Domingo and Santiago de Los Caballeros in 2012 and a compendium of hazard and risk maps, conducted by Ingeniar & CIMNE (CIMNE et al., 2012) Ecuador Disaster risk management index (ERN-AL, 2008-2009) conducted by ERN for the IADB. Seismic microzonation of Quito and risk evaluation of water lifelines for EPMAPS, conducted by Ingeniar & CIMNE. El Salvador Earthquake and hurricane at national scale. Analysis of premiums by socioeconomic level. Case study for San Salvador metropolitan area portfolio of education, health and public buildings, conducted by ERN on behalf of the ERN-AL consortium. Flooding at national scale. Based on these studies, the national government is designing a seismic vulnerability reduction program with an initial focus on the education sector. Cost-benefit analysis of risk mitigation at schools (ERN-AL Result 2, 2010b, 2011d; IDB, 2016d; GFDRR, 2012). Drought risk (Cardona et al., 2017c). Guatemala Risk analysis at Ciudad de Guatemala and others. Cost-benefit analysis of risk mitigation at schools (ERN-AL Result 3) conducted by ERN. Risk assessment for tropical cyclones considering the climate change, conducted by Ingeniar & CIMNE (CIMNE et al, 2013). Drought risk (Cardona et al., 2017c). Guyana Flood risk assessment at national level and for Georgetown, conducted by Ingeniar and CIMNE on behalf the ERN-AL consortium. Honduras Risk at Tegucigalpa, Puerto Cortes and El Progreso. Cost-benefit analysis of risk mitigation in schools, and insurance scheme for Tegucigalpa (ERN-AL Result 4) conducted by ERN. Drought risk (Cardona et al., 2017c). India Multi-hazard risk modeling and assessment in coastal India, conducted by ERN (WB, 2015). Italy Loss of life and economic impact of Amatrice’s earthquake (Salgado-Gálvez et al., 2016)

Table 2 (cont.). Selection of studies using CAPRA. Country Risk Analysis and Application Jamaica Application and updating of the risk and risk management index system, conducted by ERN (2009) for the IADB. Earthquake and hurricane. Analysis of premiums by layers of loss by Ingeniar & CIMNE (ERN-AL, 2008-2009; IDB, 2014a). Mexico Earthquake risk at national level, conducted by ERN (ERN-AL, 2011g; ERN, 2008-2009) for UNISDR, disaster risk management index, conducted by ERN (2009) for the IADB. Loss estimation for health, education, telecommunications, transport and low-income housing at national level due to earthquakes and hurricanes (Reinoso et al. 2010). Cost-benefit analysis of public school seismic mitigation options (Jaimes and Niño, 2017). Nepal Earthquake risk at national level (ERN-AL; 2011g) for UNISDR. Comparison of fatalities and direct economic losses for the Gorkha M7.8 earthquake (Salgado-Gálvez, 2015). Upper Koshi river dike failure probabilistic flood risk (Oliver et al., 2018). Table 2 (cont.). Selection of studies using CAPRA. Nicaragua Earthquake and Volcanic risk at Managua. Tsunami at San Juan del Sur and Miramar. Hurricane at Bluefields and Corinto and San Dionisio. Landslides at San Dionisio. Outcomes include land-use planning and insurance design for Managua (ERN-AL Result 7; 2009). Pakistan Earthquake at Muzaffarabad, Pakistan. (Jean Ingenieros, 2014), PML and Risk Premium to contribute towards the development of a catastrophe risk financing strategy (ERN, 2013). National fiscal risk assessment with emphasis on the provinces of Punjab and Sindh, conducted by ERN (GFDRR, 2014b). Panama Application and updating of the risk and risk management index system, conducted by ERN (2009) for the IADB. Earthquake at city of David, with portfolio analysis of housing, education and health facilities. Outcomes include a plan of vulnerability reduction actions and capacity building programs, conducted by Ingeniar & CIMNE (ERN-AL, 2011e; IG-UP, 2012; WB, 2013). Paraguay Disaster risk and climate change vulnerability assessment, considering flood by overflow of the Paraguay River and small rivers inside at Asuncion city. Conducted by ERN for the IDB (2014). Peru Disaster risk management index, conducted by ERN (2009) for the IADB. National earthquake hazard maps and for Cusco, Lima and Arequipa. Earthquake risk evaluation at schools and hospital in Lima. Flooding at three distinct characteristic river basins. (ERN-AL, 2010c, 2013; PUCP, 2013; IDB, 2014b, 2015a). Risk assessment of lifelines of SEDAPAL, conducted by Ingeniar & CIMNE. Probabilistic risk estimation for selected IIRSA infrastructure (ports and airports) for earthquake and tsunami, with support of the IDB, conducted by ERN (2015). The Technical assistance to develop a disaster risk financing strategy and action plan, conducted by ERN Philippines (GFDRR, 2014). South- Training on Seychelles, Mauritius and Madagascar on hazard and risk modules of the CAPRA platform West related to earthquakes and tropical cyclones for the Indian Ocean Commission, conducted by Ingeniar. Indian Resulting in Seychelles being the first African country to access a Cat-DDO (GFDRR, 2014). Mauritius, Ocean Madagascar, Seychelles, Union des Comoros and Zanzibar. AAL, PML, and probabilistic cost benefit Region analysis. (UNISDR, 2015c, 2015d, 2015e, 2015f, 2015g, 2015h). Spain Earthquake at Barcelona. Disaster risk assessment at country level for all perils, conducted by CIMNE. (Marulanda et al., 2013; UNISDR, 2015a). Sub- Earthquake risk assessment for Ethiopia, Kenia, Uganda, Cabo Verde, Malawi, and Mozambique Saharan (GFDRR), conducted by ERN & RED (2017). The risk profiles provide loss metrics related to the countries affected population, damage to physical assets and impact on GDP. Trinidad & Disaster risk management index, national and Port of Spain multi-hazard risk assessment (ERN-AL, Tobago 2008-2009) conducted by ERN for the IADB, and National multi-hazard risk assessment, conducted by Ingeniar & CIMNE (2013). Uruguay Drought risk at national level and floods in the Branco River (Cardona et al., 2017b). USA Preliminary loss assessment for hurricane Irma and Maria using GAR 15 (Cardona et al. 2017d). Venezuela Earthquake risk at national scale and flood risk in Caracas, conducted by Ingeniar & CIMNE (IDB, 2015b).

Lastly, CAPRA have been used for the Global Risk Model in 2013, 2015, including the Catastrophe Risk Profiles for 216 countries (Cardona et al. 2015; UNISDR, 2015b,) and in the framework of the UN Atlas-GAR: Unveiling Disaster Risk (Cardona et al. 2017; UNISDR, 2017; Marulanda et al. 2018).

CAPRA

Earthquake

Volcanic eruption

Land slide

Tsunami

Hurricane 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Flooding Visit the full-size map at www.ern.com.mx/web/capra_en.html

Figure 1. Timeline and map showing the sites of some studies using CAPRA. The full-size map can be visited at www.ern.com.mx/web/capra_en.html

5. EXAMPLE OF APPLICATION FOR EARTHQUAKE RISK

Several evaluations of earthquake risk at city level, building by building, have been made using CAPRA

(MarulandaERN el at. 2013; Salgado-Galvez et al. 2016, 2017). There are outstanding examples such as Barcelona, Bogota, Medellin, Manizales, Chacao (Caracas), Ciudad David, Mendoza, Managua, Santo 4 Inventario de Elementos Expuestos Domingo, Santiago de los Caballeros, San José de Costa Rica, among others. Figure 2 shows a view of the4.1 urban Levantamiento zones dein la San Información José City, Costa Rica, land use density for commerce, industry, government and Dado que no fue posible conseguir la información catastral oficial de la ciudad, se procedió housinga hacer un buildings, levantamiento empleando the str lauctural herramienta type web de and Zonif icaciónthe replacement Urbana de value grouped by district. The inventory was CAPRA (disponible en www.ecapra.org/zonhu.php). Dicha herramienta permiteERN identificar, 4. Inventario de elementos expuestos alsosobre classified imágenes satelitales for de number Google Maps, of zonas stories de exposición and homogénea, population es decir, density for daytime and nighttime scenarios. This waszonas the en dondefirst pued caseen identificarse study condicionesat city delevel uso, niveles made de ocupación,using costoCAPRA y in 2011. densidades de construcción similares. Cada zona es luego calificada en términos de porcentajes identificados de tipos constructivos, con relación a lo observado durante el levantamiento. El mapa de zonas homogéneas de la ciudad se presenta en la Figura 4-1. (a) (b)

Figura 4-4 ERN 4. Inventario de elementos expuestosDistribución de sistemas estructurales asignados a los predios

(c) (d)

Figura 4-1 Mapa de zonas urbanas homogéneas de San José

Estas zonas homogéneas fueron luego discretizadas, para simular los predios de la ciudad. El proceso de discretización consiste en ubicar aleatoriamente puntos dentro de cada zona homogénea, asignando a cada punto un costo y ocupación consistente con los valores identificados en la zona, y un tipo constructivo en función de los porcentajes previamente definidos. El número total de predios ubicados por zona es consistente con la densidad de construcciones identificada en el levantamiento.

4-1

Figure 2. Exposed assets ofFigura San 4- 4José City. Urban zones (a), land-use zoningFigura ( b4-)5, structural system (c) and Distribución de sistemas estructurales asignados a los predios Distribución del valor físico de predios replacement value in 2011 USD (d). Adapted from ERN-AL (2011c).

4-4

Figura 4-5 Distribución del valor físico de predios

4-4

The vulnerability functions for all classes of assets were proposed and calibrated against observations and numerical models. Each vulnerability function is input to CAPRA as a set of values defining the expected loss as a percentage of the replacement value and a measure of dispersion. The function usually takes the form of a cumulative log-normal distribution.

The seismic hazard for the region was evaluated using CRISIS 2007, the probabilistic seismic hazard analysis (PSHA) program included as the earthquake module for CAPRA, the seismic sources are discretized in finite regions as those shown in Figure 3 for interplate, intraplate and cortical earthquake

sources in Central America. 1. Amenaza sísmica 1. Amenaza sísmica 1. Amenaza sísmica

Figure 3. Seismic sources for Central America represented on CRISIS 2007 (PSHA included in CAPRA).

Figura 1-3 Interplate sources (a), intraplate sources (b) and corticalFigura 1-5 sources (c). Adapted fromFuentes de tipo ERN cortical. Proyecto-AL RESIS (2011c). II Figura 1-4 Fuentes de tipo intraplaca. Proyecto RESIS II Fuentes de tipo interfase. Proyecto RESIS II

1.3.4 Modelo de atenuación de movimiento fuerte

The risk assessment methodology forComo earthquakemodelo de atenuación de ondas (Avelar,sísmicas se adoptó el propuesto 2007; por Climent Huerta et al. et al., 2007; Ordaz et al., 1999; ERN1994, el cual fue generado con sismos de la región y otras regiones5. Resultados tectónicamente de la evaluación Reinoso et al., 2006) will output thesi milares.expected La función de atenuación annual se basa en la siguienteloss ecuación or general. pure ERN risk premium PP m for hazard 5m. Resultados as, de la evaluación

ln A c c M c ln R c R c S 1 2 3 4 5 (Ec. 1) 8,000.0 PPm = l E(Pl) PAl 7,000.0 (1)

Σ Millones 6,000.0 ERN América Latina ERN América Latina 1-6 5,000.0 TR1000 PML(17.7%) 1-8 ERN América Latina 1-7 TR 500 PML(13.8%) where E(Pl) is the expected value of intensity4,000.0 for event l and PAl is the annual probability of occurrence 3,000.0 TR 250 PML(10.2%) Pérdida, [USD] Pérdida, of event l. The loss exceedance probability2,000.0 TR v100m PML(6.4%)(p) for hazard m and a loss equal or greater than p is, 1,000.0 0.0 vm(p) = Σl Pr(Pl>p) PAl 0 500 1,000 1,500 2,000 2,500 3,000 (2) Periodo de retorno, [años]

Figura 5-8 where sum in both (1) and (2) implies the numericalPérdida máxima integration probable from all events belonging to source m. 5.2.2 Mapas de riesgo por distritos Figure 4 presents some basic resultsA continuación for se muestran San los Jose mapas de( ERNriesgo probabilístico-AL, 2011c)generados por wheremedio de la it was shown higher losses ERN 5. Resultadosevaluación de larealizada evaluación. Los resultados han sido agrupados a nivel de Distrito, lo que permite are expected at the housing portfverlosolio de manera, p robable más globalizada. losses Los resultados are se between presentan en términos 6% deand pérdidas 10% of theFigura total 5-3 replacement anuales esperadas, PAE. Pérdida física agrupada por distritos value, while the worst-case scenario accumulates up to 30% loss.

Figura 5-3 Figura 5-9 Figura 5-4 Figure 4. PérdidaResults física agrupadafor San por distritosJosé City groupedPérdida by district. anual esperada Physical agrupada por distritosloss ( left) andNúmero expected de víctimas annualagrupadas por loss distritos. (center Escenario) Día in 2011 millions of USD. Life losses in number of habitants for a daytime scenario (right). Adapted from ERN-AL

(2011c) conducted5-6 by ERN. 5-3 6. SOME UPDATES AND NEW DEVELOPMENTS

In 2014, RMS and ERN partnered to implement the analysis of CAPRA models on the RMS(one) platform, additionally, thanks to the support of the World Bank and the UNISDR, governments have now free access to the RMS(one) platform (RMS, 2014).

Figura 5-4 Número de víctimas agrupadas por distritos. Escenario Día

5-3

As abovementioned, at present the University of Los Andes is the current manager of the eCAPRA website, where the public can download the main modules of the early version of CAPRA platform. The vulnerability and GIS modules of this original version have received minor updates. The flood kernel is now the HEC-RAS (USACE, 2017) software (closed source), and the seismic hazard module still uses the 2007 version of CRISIS10 because most researchers consider it is good enough for this application.

CAPRA outputs have been successfully used for disaster risk management and to advice the design of risk transfer strategies, but the placement of financial protection has not always been successful because additional efforts beyond the scope of the original CAPRA initiative are needed, while the projects must be developed in the right context reaching to key decision-makers. An example of successful outcomes is the CCRIF, an initiative aimed to provide insurance to a pool of countries, that is currently considered one of the best models and tools (closed source) for the Caribbean. Interestingly, CAPRA and CCRIF started at about the same time, both with support from GFDRR. Also, some of the risk management specialists for the latest update of CCRIF participated in the development of CAPRA nearly a decade before. It appears the development of open and free software tools is effective for the advancement of said tools and for capacity building only while the incentives and support are available. But not for the development of highly sophisticated methods and software that usually require a higher number of development iterations and considerable resources. The same trend has been observed in the software industry many times. Open-source projects like CAPRA serve as a trigger for awareness, capacity building and development of novel platforms.

As result of the natural advancement of methodologies, two new generation sets of tools compatible with the original CAPRA platform are now available, both from original members of the CAPRA initiative. The R-CAPRA system developed by ERN with special focus on financial protection placement as in R-FONDEN and the latest update of CCRIF. For more details see www.ern.com.mx/web/capra_en.html. And the CAPRA software suite by Bernal and Cardona (2018) in these proceedings, also used in IDIGER (2016) for Bogotá.

7. CONCLUSIONS

The CAPRA platform has become an unavoidable reference for risk assessment practitioners worldwide. An innovative developing scheme based on opening the source code and datasets spurred a field of study not attended before in Central America, and shortly after in the rest of world. In recent times, CAPRA continues to be used for basic risk assessments. However, the most advanced updates and innovation of CAPRA have been part of efforts, made by the original developers in the framework of different new projects and achievements, showing the need to either update or move on to a different platform with a broader scope, like those involved in financial protection placing, territorial ordering, and climate change. Projects, like the GAR-Atlas and CCRIF, have received indirect benefits from research conducted during the CAPRA program development, while stakeholders and local governments continue to resort to the CAPRA platform for cost effective basic studies to help develop land-use planning, and to assess disaster risk reduction initiatives.

8. REFERENCES

Ara, Sharmin (2014). Impact of temporal population distribution on earthquake loss estimation: A case study on Sylhet, Bangladesh. Int. J. Disaster Risk Sci 5:296-312 DOI 10.1007/s13753-014-0033-2 Banco Mundial (2008). Avaliação de Perdas e Danos: Inundações Bruscas em Santa Catarina. In Portuguese. Banco Mundial (2010a). Avaliação de Perdas e Danos: Inundações Bruscas em Alagoas. In Portuguese. Banco Mundial (2010b). Avaliação de Perdas e Danos: Inundações Bruscas em Pernambuco. In Portuguese. Banco Mundial (2011). Avaliação de Perdas e Danos: Inundações e Deslizamentos na Região Serrana do Rio de Janeiro. In Portuguese.

10 The latest version of CRISIS is from 2017. It has been renamed as R-CRISIS (www.r-crisis.com).

Bernal GA, Cardona OD. (2018) Next generation CAPRA software. Proceedings of 16ECEE, Thessaloniki, Greece. Bernal GA, Escovar MA., Zuloaga D, Cardona OD. (2017). Agricultural Drought Risk Assessment in Northern Brazil: An Innovative Fully Probabilistic Approach. In V. Marchezini, B. Wiesner, S. Saito, & L. Londe (Eds.), Reduction of Vulnerability to Disasters: from Knowledge to Action (pp. 331–356). Cardona OD, Ordaz MG, Yamín LE, Marulanda MC, Barbat AH (2008). Earthquake Loss Assessment for Integrated Disaster Risk Management. Journal of Earthquake Engineering, 12 (S2): 48-59, Taylor & Francis, Philadelphia, PA. Cardona OD, Ordaz MG., Reinoso E, Yamín LE, Barbat AH (2012). CAPRA- Comprehensive Approach to Probabilistic Risk Assessment: International Initiative for Risk Management Effectiveness. Proceedings of 15WCEE, Lisbon. Cardona OD, Bernal GA, Ordaz MG, Salgado-Gálvez MA, Singh SK, Mora MG, Villegas CP (2015). Update on the Probabilistic Modelling of Natural Risks at Global Level: Global Risk Model - Global Earthquake and Tropical Cyclone Hazard Assessment. Disaster Risk Assessment at Country Level for Earthquakes, Tropical Cyclones (Wind and Storm Surge), Floods, Tsunami and Volcanic Eruptions. Background paper for GAR15. Barcelona/Bogotá. Cardona OD, Bernal GA, Marulanda MC, Marulanda PM (2017). World at Risk: Revealing Latent Disasters, INGENIAR Risk Intelligence, www.ingeniar-risk.com, Bogotá. Cardona OD, Bernal G, Escovar MA, Villegas C, Brenes A, Velázquez C (2017b). Perfil de riesgo por sequía e inundación de Uruguay – Análisis retrospectivo de consecuencias y evaluación probabilista de la amenaza. Preparado para el Banco Interamericano de Desarrollo BID. Consorcio INGENIAR – CIMNE. Bogotá. In Spanish. Cardona OD, Bernal G, Escovar MA, Brenes A, Velázquez, C (2017c). Perfil de riesgo por sequía de El Salvador, Guatemala y Honduras– Análisis retrospectivo de consecuencias y evaluación probabilista de la amenaza. Preparado para el Banco Interamericano de Desarrollo BID. Consorcio INGENIAR – CIMNE. Bogotá. In Spanish. Cardona OD, Bernal GA., Villegas CP (2017d) Loss assessment for hurricane Irma. Ingeniar & CIMNE Cardona OD, et al. (2017). Modelación probabilista de inundaciones en La Mojana. Preparado para el Fondo Adaptación. Bogotá. In Spanish CCRIF (The Caribbean Catastrophe Risk Insurance Facility), (2017). ERN-RED consortium. Recovered from the CCRIF website at http://www.ccrif.org/content/aboutus/ccrif-team in November, 2017. CIMNE, ITEC and Ingeniar Ltda. (2012). Amenazas y riesgos naturales en la República Dominicana. Compendio de Mapas. In Spanish. CIMNE, ITEC, Ingeniar Ltda. and EAI (2013). Estimación de la amenaza y el riesgo probabilista por huracán en Guatemala, incorporando el impacto asociado al cambio climático. Informe final de consultoría. Colombia. In Spanish. ERN (Evaluación de Riesgos Naturales), (2008). Integración, análisis y medición de riesgo, para establecer los mecanismos financieros para el FONDEN. Project for the National Autonomous University of Mexico. In Spanish. ERN (Evaluación de Riesgos Naturales), (2009). Integración, análisis y medición de riesgo de sismo, inundación y ciclón tropical en México para establecer los mecanismos financieros eficientes de protección al patrimonio del fideicomiso FONDEN del Banco Nacional de Obras y Servicios Públicos (BANOBRAS). Project for AGROASEMEX ERN (Evaluación de Riesgos Naturales), (2008-2009). Aplicación y actualización de los sistemas de índices de riesgo y manejo de riesgo para Argentina, Barbados, Bolivia, Chile, Colombia, Ecuador, Jamaica, México, Panamá, República Dominicana, Trinidad y Tobago. Project for the Inter-American Development Bank. ERN (Evaluación de Riesgos Naturales), (2011). Sistema R-FONDEN para la estimación de pérdidas por amenazas naturales del fondo nacional de desastres naturales de México. Project for the Mexican Ministry of Finance (SHCP). ERN (Evaluación de Riesgos Naturales), (2013). Sistema de cuantificación de pérdidas, control de recursos y análisis de riesgo para el FONDEN. Project for the Mexican Ministry of Finance (SHCP) and FONDEN. ERN (Evaluación de Riesgos Naturales), (2014). Valoración de la vulnerabilidad física y riesgo ante sismos en distritos clave del Cantón de San José, de la Provincia de San José, Costa Rica. Technical report to CNE (Comisión Nacional de Prevención de Riesgos y Atención de Emergencias). Mexico City. In Spanish. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2009). Nicaragua. Riesgo por tsunami en San Juan del Sur. Technical Report s for task T2. In Spanish. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2010a). Application of the System of Indicators of Disaster Risk and Disaster Risk Management 2010. Summary Report. Inter-American Development Bank.

ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2010b). El Salvador. Perfil de riesgo catastrófico, mapas de riesgo y análisis de concentración del riesgo. Technic al Report ERN-CAPRA-T3.3 In Spanish. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2010c). Mapa de peligro sísmico nacional del Perú y a nivel local en Cusco, Lima y Arequipa. Technical Assistance Project report. Perú. In Spanish ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2010d). CAPRA-GIS v2.0. Program for the probabilistic risk assessment. Software. Available at: www.ecapra.org Accessed May 15th, 2013. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2010e). Seismic Risk Assessment of Schools in the Andean Region and Central America. Exposure Estimation and Seismic Risk Modeling. Report to the International Labor Office. April 2010. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2011a). Probabilistic Modelling of Natural Risks at the Global Level: The Hybrid Loss Exceedance Curve. Development of Methodology and Implementation of Case Studies Phase 1A: Colombia, México, Nepal. Informative document prepared for the Global Assessment Report on Disaster Risk Reduction GAR 2011. Geneva, Switzerland: EIRD/ONU. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2011b). Evaluación probabilista de riesgo sísmico para el municipio de Pereira. Technical Assistance Project report. Colombia. In Spanish. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2011c). Modelación probabilista de escenarios de riesgo para algunos sistemas de agua y saneamiento en Costa Rica. Technical Assistance Project report. In Spanish. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2011d). Modelación probabilista de escenarios de riesgo sísmico para el área metropolitana de San Salvador. Incluye análisis de los portafolios de educación, salud y gobierno. Technical Assistance Project report. El Salvador. In Spanish. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2011e). Modelación probabilista de escenarios de riesgo sísmico para la Ciudad de David (Panamá), incluye análisis de los portafolios de vivienda, educación y salud. Technical Assistance Project report. Panamá. In Spanish. ERN-AL (Evaluación de Riesgos Naturales - América Latina), (2011f). Disaster Risk Assessment for Belize City. Technical Report Subtask 4.2B. ERN-AL (Evaluación de Riesgos Naturales - América Latina), (2011g). Probabilistic Modelling of Disaster Risk at Global Level: The Hybrid Loss Exceedance Curve – Development of a Methodology and Implementation of Case studies, Phase 1A: Colombia, Mexico, Nepal. Report for the GAR 2011, Bogotá, D.C. ERN-AL (Evaluación de Riesgos Naturales – América Latina), (2011h). Modelación probabilista de escenarios de riesgo sísmico y tsunami para las regiones de Antofagasta y Atacama (Chile). Technical Assistance Project report. Chile. In Spanish. ERN-AL (Evaluación de Riesgos Naturales - América Latina), (2012). Evaluación probabilista del riesgo volcánico para el municipio de Pasto. Technical Assistance Project report. Colombia. In Spanish. ERN-AL (Evaluación de Riesgos Naturales - América Latina), (2013). Evaluación probabilista del riesgo sísmico de escuelas y hospitales en Lima. Technical Assistance Project report. Perú. In Spanish. ERN-RED, (2015). ERN-RED appointed as Risk Manager Specialists of CCRIF SPC. Press release. Available at https://www.ern.com.mx/boletines/CCRIF_SPC_ERN_&_RED.pdf FEMA (Federal Emergency Management Agency), (2011). HAZUS-MH MR3. Multi-hazard Loss Estimation Methodology. Hurricane Model. FEMA, 2011. GRCG (German Research Center for Geosciences), (2010) Seismic Hazard Harmonization in Europe SHARE. Development of a common methodology and tools to evaluate earthquake hazard in Europe. Theme 6: Environment. German Research Center for Geosciences. GFDRR (Global Facility for Disaster Reduction and Recovery), (2008). Annual report 2008. GFDRR (Global Facility for Disaster Reduction and Recovery), (2012). Annual report 2012. GFDRR (Global Facility for Disaster Reduction and Recovery), (2013). Quantify Contingent Liabilities Associated with Natural Disasters: Towards effective post-disaster public financial management. GFDRR Document 97977 GFDRR (Global Facility for Disaster Reduction and Recovery), (2014a). Understanding Risk: Review of Open Source and Open Access Software Packages Available to Quantify Risk from Natural Hazards. Report number 92165. The World Bank. Washington. GFDRR (Global Facility for Disaster Reduction and Recovery), (2014b). Annual report 2014.

IDB (Inter-American Development Bank), (2014a). Disaster Risk Profile for Jamaica / Inter-American Development Bank (Technical note IDB-TN-635). IDB (2014b). Perfil de Riesgo de Desastres para Perú / Banco Interamericano de Desarrollo (Technical note IDB-TN- 634). In Spanish. IDB (2015a). Perfil de Riesgo por Inundaciones en Perú: Informe Nacional / Banco Interamericano de Desarrollo (Technical note IDB-TN-844). In Spanish. IDB (2015b). Perfil de Riesgo de Desastres para Venezuela / Banco Interamericano de Desarrollo (Technical note IDB- TN-831). In Spanish. IDB (2016a). Perfil de Riesgo de Desastres: Informe Nacional para Argentina / Banco Interamericano de Desarrollo (Technical note IDB-TN-1082). In Spanish. IDB (2016b). Perfil de Riesgo de Desastres para Bolivia: Informe Nacional / Banco Interamericano de Desarrollo (Technical note IDB-TN-1100). In Spanish. IDB (2016c). Perfil de Riesgo de Desastres: Informe Nacional para Chile / Banco Interamericano de Desarrollo (Technical note IDB-TN-1079). In Spanish. IDB (2016d). Perfil de Riesgo de Desastre por Inundaciones para El Salvador: Informe Nacional / Banco Interamericano de Desarrollo (Technical note IDB-TN-877). In Spanish. IDB (2016e). Climate Change Data and Risk Assessment Methodologies for the Caribbean / Inter-American Development Bank: Environmental Safeguards Unit (Technical note IDB-TN-633). IDIGER (Instituto Distrital de Gestión de Riesgos y Cambio Climático), (2016). Sistema de Modelación de Amenazas y Riesgos de Bogotá (Hazard and Risk Modelin System of Bogotá). Proyect Report December 2015- October 2016. DOI: 10.13140/RG.2.2.31828.60800. In Spanish. IG-UP. (2012). Proyecto CAPRA: Modelación probabilista de riesgo sísmico para la Ciudad de David (Panamá), incluye análisis de los portafolios de vivienda, educación y salud. Instituto de Geociencias, Universidad de Panamá. Jaimes, M., Niño, M. (2017). Cost-benefit analysis to assess seismic mitigation options in Mexican public school buildings. Bulletin of Earthquake Engineering. 15:3919–3942 DOI 10.1007/s10518-017-0119-5 Jean Ingenieros (2014). Probabilistic earthquake loss for Pakistan. Technical report to The World Bank. Mexico City. Marulanda MC, Carreño ML, Cardona OD, Ordaz MG, Barbat AH (2013). Probabilistic earthquake risk assessment using CAPRA: Application to the city of Barcelona, Spain. Natural Hazards 69(1): 59–84. Marulanda MC, Cardona OD, Marulanda P, Carreño ML, Barbat AH (2018). Evaluating seismic risk from a holistic perspective to improve resilience: The UN evaluation at global level. Proceedings of the 16th European Conference on Earthquake Engineering, Thessaloníki, Greece. Niño M, Jaimes M, Reinoso E (2015). A risk index due to natural hazards based on the expected annual loss. Natural Hazards, Springer Ed. DOI 10.1007/s11069-015-1837-0, Online ISSN: 1573-0840 Ordaz MO, Aguilar A, Arboleda J (2007). CRISIS2007. Program for computing seismic hazard. Universidad Nacional Autónoma de México. Mexico. Ordaz MO (2015). A simple probabilistic model to combine losses arising from the simultaneous occurrence of several hazards. Natural Hazards 76:389-396. DOI 10.1007/s11069-014-1495-7 PUCP (Pontificia Universidad Católica del Perú). (2013). Evaluación Probabilista del riesgo sísmico de escuelas y hospitales de la ciudad de Lima. Componente 2: Evaluación probabilista del riesgo sísmico de locales escolares en la ciudad de Lima. Project Report. Perú. In Spanish. Reinoso-Angulo E, Jaimes-Téllez MA, Ordaz-Schroeder M, Niño-Lázaro MA, (2010). Pérdidas en la infraestructura en México ante sismos y huracanes. Revista Digital Universitaria. Universidad Nacional Autónoma de México. 11(1) ISSN: 1067-6079 In Spanish. RMS (2014). RMS Partners with the United Nations and The World Bank to Advance Global Catastrophe Resilience. Press Release. Newark California, July 1, 2014. Salgado-Gálvez MA, Bernal GA, Yamín LE, Cardona OD (2010). Evaluación de la amenaza sísmica de Colombia. Actualización y uso en las nuevas normas colombianas de diseño sismo resistente NSR-10. Revista de Ingeniería Universidad de Los Andes 32, 28-37. Salgado-Gálvez MA (2015). Comparison of fatalities and direct economic losses for the Nepal (Gorkha) M7.8 earthquake using the GAR15 Global Exposure Dataset and the CAPRA risk assessment tool. Technical Note.

Barcelona, Spain. Salgado-Gálvez MA, Bernal GA, Cardona OD (2016). Evaluación probabilista de la amenaza sísmica de Colombia con fines de actualización de la Norma Colombiana de Diseño de Puentes CCP-14, Revista Internacional de Métodos Numéricos para Cálculo y Diseño en Ingeniería 32, 230-239. In Spanish. Salgado-Gálvez MA, Ordaz MG, Cardona OD (2016). Preliminary estimation of Amatrice’s Earthquake (August 24th 2016) direct economic losses and fatalities using GAR15 seismic hazard and risk datasets and tools. Technical Note. Geneva, Switzerland. Salgado-Gálvez MA, Zuloaga D, Velásquez CA, Carreño ML, Cardona OD, Barbat AH (2016). Urban seismic risk index for Medellín, Colombia, based on probabilistic loss and casualties’ estimations, Natural Hazards. 80(3):1995-2021. Salgado-Gálvez MA, Bernal GA, Zuloaga D, Marulanda MC, Cardona OD, Henao S (2017). Probabilistic Seismic Risk Assessment in Manizales, Colombia: Quantifying Losses for Insurance Purposes. International Journal of Disaster Risk Science 8(3): 296-307. Oliver J, Qin XS, Larsen O, Meadows M, Fielding M (2018). Probabilistic flood risk analysis considering morphological dynamics and dike failure. Natural Hazards. 91: 287-307. DOI: 10.1007/s11069-017-3126-6 Uniandes (Universidad de los Andes in Bogotá, Colombia), (2017). CAPRA, www.ecapra.org UNISDR (United Nations Office for Disaster Risk Reduction), (2015a). Update on the probabilistic modelling of natural risks at global level: Global risk model. Background paper for the 2015 GAR on Disaster Risk Reduction. UNISDR (United Nations Office for Disaster Risk Reduction), (2015b). World summarized catastrophe risk profiles. Background paper prepared for the 2015 Global Assessment Report on Disaster Risk Reduction. UNISDR (United Nations Office for Disaster Risk Reduction), (2015c). Working Papers on Public Investment Planning and Financing Strategy for Disaster Risk Reduction: Review of Mauritius, 2015, UNISDR. Geneva. UNISDR (United Nations Office for Disaster Risk Reduction), (2015d). Working Papers on Public Investment Planning and Financing Strategy for Disaster Risk Reduction: Review of Madagascar, 2015, UNISDR. Geneva. UNISDR (United Nations Office for Disaster Risk Reduction), (2015e). Working Papers on Public Investment Planning and Financing Strategy for Disaster Risk Reduction: Review of Seychelles, 2015, UNISDR. Geneva. UNISDR (United Nations Office for Disaster Risk Reduction), (2015f). Working Papers on Public Investment Planning and Financing Strategy for Disaster Risk Reduction: Review of Union des Comores, 2015, UNISDR. Geneva. UNISDR (United Nations Office for Disaster Risk Reduction), (2015g). Working Papers on Public Investment Planning and Financing Strategy for Disaster Risk Reduction: Review of Zanzibar, 2015, UNISDR. Geneva. UNISDR (United Nations Office for Disaster Risk Reduction), (2015h). Working Papers on Public Investment Planning and Financing Strategy for Disaster Risk Reduction: Review of South-West Indian Ocean Region, 2015, UNISDR. Geneva. UNISDR (United Nations Office for Disaster Risk Reduction), (2017). GAR Atlas: Unveiling Global Disaster Risk, Recovered from http://www.unisdr.org/files/53086_garatlaslr2.pdf on November 207. Geneva. WB (World Bank), (2012). Latin America: Putting disaster preparedness on the radar screen. World Bank News, Washington DC, October 9. WB (World Bank), (2013). Panama: plan, prepare, mitigate – key actions for disaster prevention. Recovered from http://blogs.worldbank.org/latinamerica/trade/ppps/education/voices/governance/psd/voices/comment/reply/742 on November 2017. WB (World Bank), (2014). Seismic risk assessment in Thimphu Bhutan. Technical paper, GFDRR. WB (World Bank), (2015). India - Second Phase of the National Cyclone Risk Mitigation Project. Washington, D.C. WB (World Bank), (2016a). EOI for CAPRA management, deadline: Aug 29 / Understanding Risk. Recovered from https://understandrisk.org/opportunity/eoi-for-capra-management-deadline-aug-29/ on November, 2017. WB (World Bank), (2016b). Decision on the call for EOI-CAPRA / Understanding Risk. Recovered from the World Bank website https://understandrisk.org/decision-on-the-call-for-eoi-capra/ on November, 2017. Yi W, Shen H, Cheng C (2015). Multi-criteria simulation analysis of seismic risk based on CAPRA platform. Zhongnan Daxue Xuebao (Ziran Kexue Ban) Journal of Central South University (Science and Technology). 46. 603-609. 10.11817/j.issn.1672-7207.2015.02.031.