92165 GFDRR UNDERSTANDING RISK Review of Open Source and Open Access Software Packages Available to Quantify Risk from Natural Hazards GLOBAL FACILITY FOR DISASTER REDUCTION AND RECOVERY © 2014 International Bank for Reconstruction and Development / International Development Association or The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with Rights and Permissions external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily The material in this work is subject to copyright. Because reflect the views of The World Bank, its Board of Executive The World Bank encourages dissemination of its Directors, or the governments they represent. knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution The World Bank does not guarantee the accuracy of to this work is given. the data included in this work. The boundaries, colors, denominations, and other information shown on any map Any queries on rights and licenses, including subsidiary in this work do not imply any judgment on the part of The rights, should be addressed to the Office of the Publisher, World Bank concerning the legal status of any territory or The World Bank, 1818 H Street NW, Washington, DC 20433, the endorsement or acceptance of such boundaries. USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ACKNOWLEDGMENTS This report was prepared based on analysis and model The team gratefully acknowledges the contributions of testing undertaken by James Daniell, Center for Disaster Marc Forni (World Bank Group) and Anne Himmelfarb in the Management and Risk Reduction Technology at Karlsruhe development of this report. Institute of Technology. Substantial contributions to the The team greatly appreciates the support and guidance of analytical design and reporting were made by Alanna Francis Ghesquiere, Zoubida Allaoua, Rachel Kyte, James Simpson, Rick Murnane, Annegien Tijssen, Ariel Nunez, and Close, and Ede Jorge Ijjasz-Vasquez. Vivien Deparday (Global Facility for Disaster Reduction and Recovery); Rashmin Gunasekera, Abigail Baca, and Oscar Ishizawa (World Bank Group); and Andreas Schäfer (Center for Disaster Management and Risk Reduction Technology at Karlsruhe Institute of Technology). 01 02 ABBREVIATIONS AAL annual average loss IRD Institut de recherche pour le AIFDR Australia-Indonesia Facility for Disaster développement Reduction IST-SUPSI Instituto scienze della Terra–Scuola ANU Australian National University universitaria professionale della API application programming interfaces Svizzera italiana BCR benefit-cost ratio ITB Bandung Institute of Technology CEDIM Center for Disaster Management and KIT Karlsruhe Institute of Technology Risk Reduction Technology LNEC Laboratório Nacional de Engenharia DEM digital elevation model Civil DRM disaster risk management MDR mean damage ratio EPA Environmental Protection Agency METU Middle East Technical University EU European Union MEOW maximum envelope of water FEM finite element method MMI Modified Mercalli Intensity FEMA Federal Emergency Management MOM maximum of maximums Agency NCEP National Centers for Environmental GA Geoscience Australia Prediction GEM Global Earthquake Model NCREE National Center for Research on GFDRR Global Facility for Disaster Reduction Earthquake Engineering and Recovery NOAA National Oceanic and Atmospheric GIS geographic information system Administration GM ground motion NTU National Technical University (Athens) GMPE ground motion prediction equation OGC Open Geospatial Consortium GUI graphical user interface OS operating system IAHR International Association for Hydro- OSSN Italian National Seismic Survey Environment Engineering and Research PGA peak ground acceleration PHIVOLCS Philippine Institute of Volcanology and Seismology PML probable maximum loss PRACE Partnership for Advanced Computing in Europe PSHA probabilistic seismic hazard analysis Sa spectral acceleration SKM Sinclair Knight Merz UNEP United Nations Environment Programme USGS U.S. Geological Survey USU Utah State University EXECUTIVE SUMMARY The World Bank’s disaster risk management (DRM) regarding some flood models. In addition, leaders activities utilize a range of open access and open in developing risk models in the public sector, source computational modelling tools to quantify such as Geoscience Australia (EQRM, TCRM, the risk posed by natural hazards. An important TsuDAT, ANUGA) and CAPRA (ERN-Flood, goal of these activities is to build capacity among Hurricane, CRISIS2007), are launching and/or national and local governments and international helping many other initiatives. As we achieve greater development professionals working in disaster risk interoperability between modelling tools, we will management. A key decision in these activities is also achieve a future in which open source and open the choice of modelling tool that will be used to access modelling tools are connected and adapted address the hazard, exposure, and/or risk question to unified multi-risk model platforms and highly under consideration. This document presents an customized solutions. objective analysis of freely available hazard and risk modelling software in order to facilitate selection of appropriate tools for various DRM activities. There have been previous evaluations of freely available modelling tools across various natural hazards, but this is the first multi-hazard systematic review using a set of consistent criteria. The analysis covers hazard risk models for cyclone (wind), storm surge and tsunami, earthquake, and flood. Over 80 open access software packages—excluding commercial software packages—were considered in the evaluation. A preliminary analysis was used to determine whether the models were currently supported and if they were open access. Based on the results, a subset of 31 models was selected for more detailed analysis; these included 8 earthquake models, 4 cyclone models, 11 flood models, and 8 storm surge/tsunami models. The detailed analysis evaluated the models on the basis of over 100 criteria and provided a synopsis of available open access natural hazard risk modelling tools. The quality and availability of open access/open source software has grown significantly over the past few years. For example, private entities such as Deltares now have an open source policy 03 04 Introduction: Purpose and Use of This Document The demand for risk assessment data and modelling face (for example, complicated installation or poor tools in the disaster risk management (DRM) documentation), though as modelling tools are community is high, and many open access software updated, they may address the challenges identified packages for natural hazards have been created over here. the last few decades. However, it is often difficult This review provides initial guidance to users on to assess the advantages and disadvantages of the the appropriateness of the various modelling tools different tools, given the wide range of contexts, for specific purposes, and offers an introduction purposes, and users’ technical expertise. to the connectivity that is possible between This review is a technical document intended for a models. We emphasize here, however, that the technical audience. It aims to highlight modelling final decision about which tool to use must also tools’ strengths (for example, sophisticated be based on downloading and testing of a variety graphical user interfaces [GUIs], straightforward of possible tools. Finally, the document highlights installation, frequent updates, and capacity for where collaborative efforts between modelling tool customization). It also highlights some of the developers could substantially improve our current challenges that a user of a modelling tool might understanding of risk. 1.0 – Background This systematic assessment of software packages a certain intensity of hazard at a location and is that simulate natural hazards and quantify risk was usually determined by an historical or user-defined motivated by the interest of the World Bank and the scenario, probabilistic hazard assessment, or Global Facility for Disaster Reduction and Recovery other methods. Some hazard modules can include (GFDRR) in supporting DRM efforts. They envisaged secondary perils (such as soil liquefaction or fires an objective assessment of the functionality, quality, caused by earthquakes, or storm surge associated and usability of risk model software packages that with a tropical cyclone or extra-tropical cyclone). would help potential users identify the optimal Vulnerability accounts for the susceptibility model(s) for addressing the hazard and risk to damage of the assets exposed to the forces question(s) of concern. They also saw the review as generated by the hazard. Fragility and vulnerability a way to address issues related to the development functions estimate the damage ratio and consequent and use of open data and open source models, which mean loss respectively, and/or the social cost (e.g., they strongly support. number of injured, homeless, and dead) generated by a hazard, given the specified exposure. This assessment focused on open access and open source software packages only; thus no commercial The wide range of available loss estimation packages models were considered. The 82 software packages means there are multiple ways to simulate each examined in this report simulate a number of component. Users may choose from software different perils. In addition to models designed for a packages that are proprietary, open access, or open single peril, the assessment also evaluated multi-risk source, and that have varying degrees of complexity software packages. The perils modelled with the and usability. However, other considerations may packages were divided into the following four groups also be important for users choosing a software on the basis of peril characteristics: (a) earthquakes package. For example, the most appropriate model and their secondary effects such as liquefaction, may vary by region and hazard, because data fire, landslides, etc.; (b) inland flooding; (c) winds availability and specificity also vary. Alternatively, from either tropical or extra-tropical storms; and a user may wish to modify the software to generate (d) coastal flooding due to tsunamis and/or storm more loss outputs or derive a particular type of loss surge. Each peril in the multi-risk software packages metric. Also, the technical skills of users can vary was assessed separately (though such packages may greatly. Thus a simple model driven by a GUI may be particularly valuable, given their ability to solve be appropriate in some cases, but a more complex, multiple problems). command-line controlled model may provide additional flexibility for an advanced modeller. In Risk assessments are produced in order to estimate other ways, the knowledge of the user may control possible economic, infrastructure, and social the suitability of the model; an advanced user may impacts arising from a particular hazard or multiple desire control over definition of technical content, hazards. Three components are usually considered whereas an entry-level user may want a hardwired when assessing risk and probable loss: exposure, quick version. Thus a range of criteria should be hazard, and vulnerability. Exposure represents the considered when selecting a software package for a stock of property and infrastructure exposed to a risk assessment. hazard, and it can include socioeconomic factors. Hazard is defined as the probability of experiencing 05 06 For this assessment, the criteria were grouped software package is directly downloadable, but that into 11 modules. A summary of the modules, and the source code is not editable or viewable. There the number of criteria and descriptors associated are also different forms of open access, including with each, is provided in table 1-1. Details on the full open access, partial open access (i.e., certain criteria and descriptors for each module are given components are open), and partial source code. in appendix C. The assessment methodology and criteria follow the OPAL methodology as developed “Open” also refers to the fact that payment is not by Daniell (2009) in order to rank and evaluate required for using a software package, though software using a scorecard/multi-criteria decision for the technical and research community, open analysis approach. The criteria and descriptors source provides a transparent, user-community- used for the assessment were modified in response driven solution beyond the free price. In addition to discussions with experts from GFDRR and the to making access to the software code available free World Bank, who suggested adding useful criteria of charge, open source software packages have the such as number of users in the community for a following advantages: particular software package, as well as the package’s • The models and calculations are transparent (not particular GNU license. a black box), so the science and assumptions It is important to note at this stage the difference behind the models can be checked and sensitivity between open source and open access and to analyses undertaken. articulate some of the key reasons why open source • The software allows users to detect and correct provides a more transparent framework than open errors and to make direct improvements. access software packages. Open source, as the name suggests, refers to a software package model whose • Collaborative development of some packages source code (programming language) is available for means that many experts work on improving, and access and viewing. There are various formats for solving problems within, the same models. open source, with some software code being directly editable, some viewable but not directly editable, • All their processes can be easily replicated and and some requiring registration before viewing. checked, which is critical for validation of the Open access, on the other hand, means that the outputs. Table 01—1  Module Abbreviation Criteria Descriptors Description of Software accessibility SA 20 81 Modules and GUI GU 3 7 Number of Criteria Software details SW 22 65 Used for Ranking Technology TE 5 15 Software Packages Exposure EX 17 53 Vulnerability VL 18 56 Note: See appendix C for details on the descriptors. Risk RK 14 46 Post-event PS 9 25 Forecasting FC 3 9 Output OU 6 17 Hazard–Flood HF 29 81 Hazard–Hurricane/Wind HS 21 59 Hazard–Earthquake HE 27 81 Hazard–Wave HT 18 57 2.0 – METHODOLOGY An initial analysis of 82 open access software The detailed assessment of the software packages packages was used to select a subset for detailed involved the following steps: assessment. The initial analysis looked for packages 1. The packages were installed and tested using that met three criteria: they had to be (a) open their accompanying tutorials, along with various access or open source; (b) active (currently data sets and examples, in order to create supported); and (c) available. To be considered open outputs. The advantages and disadvantages of access, the software package had to be downloadable these software packages were compiled using and testable; this limitation reduced the number a set of 117 criteria under 10 key classification of packages to 60. Twenty-two packages provided modules common to all hazard groups and an manuals, papers, and/or methodologies but did not additional 18 to 29 criteria associated with each provide access to the software itself. If a package specific hazard module (table 1-1). was inactive or unavailable, then it could not be tested in a rigorous manner. The initial analysis 2. The written descriptors for each module were identified a total of 48 software packages for the converted to a numeric system using a point four hazard groups that met these simple criteria. score having between two and five levels for each An additional subjective ranking based on output, criterion. An example of the descriptors and the hazard, risk, and available user support further numerical values assigned to word descriptions reduced the number of modelling tools for detailed is given in table 2-1. It should be noted that some testing to 31. A description of the initial analyses is of the criteria are subjective, although every provided in appendix D. effort has been made to define objective criteria. The relative importance of different factors can All software was tested on a machine with a 2.5GHz be adjusted by assigning weights to the various Intel Core i5 with 4GB of RAM, running Windows 7 criteria. This allows users to rank the software Supplementary in the Windows operating system (OS), and Ubuntu packages according to what is important for Spreadsheet 12.04 under Linux. To verify that software packages them. A supplementary spreadsheet to be available https://www.gfdrr.org/ advertised as compatible with Mac OS actually ran online shows how each software package RASoftwareReview under Mac OS, we relied on user reports. performed against each criterion. Criterion code Point score Word criterion Descriptor no. Table 02–1 SA-002 4 Fast, easy download without registration`` SA-0024 Sample Criterion for SA-002 3 Fast, easy download with registration SA-0023 Assessing Software SA-002 2 Moderate download with registration SA-0022 Packages (with SA-002 1 Slow download with registration SA-0021 five descriptors) SA-002 0 Registration did not work SA-0020 Note: The criterion shown in the table concerns software accessibility. 07 08 The criteria associated with the 11 modules listed in as the software is well designed and extensible, table 1-1 are discussed below, with an emphasis on and the developer(s) can be contacted via email, which criteria should be considered most desirable. website wiki, and/or in open discussion, then any coding language can be used. Contact details 1. Software accessibility. The 20 criteria were available for all software packages analyzed. /// /// associated with the software accessibility module cover a variety of software-related issues, such Additionally, the code should have a version as licensing, availability, ease of use, and the user code, a bugtracker, and some indication of release community. Reflecting the criteria for the initial dates, as well as patches and a free non-login analysis, all the packages analyzed in detail are access virtual community to aid knowledge open access, and most are open source (note exchange. The coding and software should also earlier definitions). be user-oriented, with separate documentation Open source can mean many things. The diversity available for those wishing to modify or of open source license structures reflects extend the tools and leverage any available the variety of interpretations of open source. APIs (application programming interfaces), Examples of open source software licenses and with tutorials, sample data, and expected include GNU GPL (or just GPL), Apache, and results available for training and testing model Creative Commons. Each of these has different installation. reuse options, and usually software must be released with the same license if source code is The software package should include all required reused. software and ideally be open source if coding changes are required; otherwise open access is Many factors influence the software package’s sufficient. An example of open-closed software ease of use, including the operating system, the is freely available Hazus-MH, which requires language used for coding, the sophistication a commercial geographic information system of the developer and user communities, (GIS) package, ArcGIS, to run the model. Another and the inclusion of additional libraries and important software accessibility component is documentation that facilitate the use of the related to data access. If the software package software package. Linux is currently the most requires particular data to run the model, then common operating system for supercomputers; these data should be available to all users, however, most basic users have no experience preferably as open data—i.e., as a library of of Linux. This review therefore considered only generalized vulnerability functions. software packages that can run on Windows operating systems. 2. Graphical user interface. The GUI module is /// /// an extremely important factor in the assessment In terms of coding languages, an open source because it determines the usability of a software coding language such as Fortran or Python is package. Few users have the technical skills that broadly considered best practice. Java is also a allow them to execute models using command very good language that allows for fast analysis lines alone. For non-experts, grappling with risk styles. Other common languages include C++ and assessment concepts is usually quite difficult; Octave, both of which are open source. As long attempting to come to grips with what is being modelled using a new software package makes 4. Technology. The code should be written to /// /// things even harder. Thus, simple software that maximize accuracy with minimal computational allows a user to point, click, and then understand effort. Hence, the software package review takes is best for a non-expert. Two examples of various hardware and software requirements into software packages that achieve this level of account. All the software packages can be run simplicity are InaSAFE and TCRM. Both have on a standard PC (2.5GHz with 4GB of RAM and help commands and simple descriptions that a 500GB hard drive); however, computationally expensive algorithms and GIS-based systems may the user can easily understand. MAEviz/mHARP require more computing power. Ideally, users is another package that provides an easy-to-use determine whether the software algorithms are GUI as well as hazard, exposure, and vulnerability reasonable for their computational purposes. analysis, so that users have more control over The actual physical computation is generally their analysis. not computationally demanding, but where memory is insufficient, the large volumes of 3. Software details. The software details module /// /// data (exposure or hazard event sets) can cause captures a variety of factors that influence the problems. For deterministic use in post-disaster usability of the software. A wiki-type system studies, all of the software packages can be run for updating code and leaving ideas, as well as in a reasonable time (assuming the region is not a direct system for handling bugs, is desirable. extremely large and the data are available). Rapid In cases where software has been derived from response data can be problematic, however, if other models or other software, this relationship data sets are not publicly available for reuse. should be clear. The package should be integrated In contrast, computing power plays a much with the Internet and allow updating with recent more central role in stochastic or probabilistic open edits to the code. The ability to update modelling—i.e., in the simulation of 10,000+ the codes is important for facilitating adaption years of hazard events analyzed against exposure to current conditions, for ensuring that the data sets of varying sizes. software is not a black box, and for enhancing the community’s ability to debug code. GIS analysis software licensing can be prohibitively expensive for many users; ideally, Most software can be run on a standard PC and therefore, open source models should be able to run without the need for a separate GIS has been optimized by the software developers to license and platform, and ideally all programs execute efficiently; however, it is useful to have should be freeware. It is also ideal for the GIS the option to change bin sizes and the capacity to to be integrated within the software, and for fully optimize the code for the user’s particular the data output to be in OGC (Open Geospatial computer. This could mean allowing for parallel Consortium) standardized format, or for the computing, or limiting the backup systems in software to be a plugin for GIS. QGIS (www.qgis. place. Learning and tutorials are extremely org) provides a very robust freely open source important, and contribute to the software’s software package or plugin. This approach also potential ease of use. A full run-through with allows technological updates to be directly a test data set showing all the features of the applied to the software package as it evolves, software is very useful, as is the ability to produce rather than relying on the GIS package version graphics and user-defined plots. (as is the case, for example, with Hazus-MH). 09 10 5. Exposure. A critical factor for any risk /// /// 6. Vulnerability. One of the fundamental factors /// /// assessment is exposure data. Thus for the influencing a risk assessment is vulnerability software packages, the critical factors are of the exposed assets. The availability of data the tools for managing exposure data. These for input, calibration, and validation governs include the ability to handle and utilize common the quality of the vulnerability module, and construction and occupancy information, and ideally the software package should be able to the ability to handle site-specific as well as handle all types of vulnerability functions. The aggregate data—particularly given that some software package should use empirical methods packages were developed with a regional focus. (historical trending of data) or analytical To aid in the collection of exposure data, pictures methods (mathematical or mechanical approach) and examples (such as the World Housing to calculate vulnerability (the susceptibility to Encyclopedia) should be used to illustrate the damage or loss). The vulnerability functions various exposure classification criteria. should be computationally simple to allow for rapid response as well as consistent with Depending upon the vulnerability functions, observations of historical damage. InaSAFE exposure information can be restricted to is an example of a software package that structural features, or it can extended to provides excellent documentation with very nonstructural features such as building contents good explanations of the algorithms used, and and to infrastructure such as lifelines and that offers a transparent framework for the emergency response facilities. Most software determination of damage from hazard. packages classify the various exposure elements using construction and occupancy information The number of damage states included in an associated with location information. It may analysis often determines the level at which the also be possible to account for temporal analysis is useful. With too few damage states, the changes in various elements. Needless to say, analysis cannot be used for detailed loss analysis. this information should be compatible with the With numerous damage states, a function can vulnerability functions provided by the software be characterized as continuous. Open source package. Flexible, open source software allows software allows the addition, alteration, or advanced users to provide additional building improved resolution of damage distributions as and classification types, temporal variability data become available. in population and demographics, new risk indicators, and supplemental socioeconomic The software package’s vulnerability functions parameters once relevant checks have been made must be consistent with the spatial resolution to the applicability of the vulnerability, hazard, of the exposure. Some software packages have and loss modules. site-specific vulnerability functions that are developed for modelling buildings (in some cases Awareness of any restrictions the software only residential buildings). Other packages have package might impose on the spatial resolution only generalized vulnerability functions that are of the analysis is critical. Ideally, the exposure, designed for aggregate data. Both site-specific vulnerability, hazard, and socioeconomic and aggregate vulnerability functions might damage and loss calculations can be completed accept additional exposure types such as bridges, at multiple levels, including global, continental, roads, lifelines (utility systems), and critical national, regional, city, and district/suburban. emergency service and response facilities. Many software packages simulate only physical Particularly when using aggregate exposure and vulnerability, but functionality to include aggregate vulnerability functions, it is common socioeconomic vulnerability is very useful. to use the mean damage ratio or coarse data sets InaSAFE, for example, allows for the calculation (such as gross national product or gross domestic of impact and needs, including gender and product) as a basis for loss. In addition, using a age disaggregation; RiskScape provides model for land-use planning and/or cost-benefit socioeconomic disruption modelling. analysis may be relevant (such as in Kalypso for flood decisions, or MAEviz for earthquake), 7. Hazards. Whether packages include single- but this capability is highly dependent on the /// /// peril capabilities or multi-peril capabilities resolution of the model. (with multiple primary perils and/or secondary perils associated with the primary hazard), the 9. Post-event scenarios. Speed and simplicity of /// /// quality of the hazard module is a fundamental use are critical assets for post-event scenarios, consideration for evaluating software packages. which require quick access to information. Thus This assessment considered four groups of the different software packages were rated on hazards: inland flood, earthquake, coastal their ability to generate products that would flooding from storm surge and tsunamis, and complement post-event response, recovery, and wind hazard from tropical and extra-tropical reconstruction efforts. Among these products storms. Many of the criteria listed in appendix are maps, which after a disaster facilitate C are hazard dependent; however, there are collaboration among the users of the software. many similarities among the hazards and some GIS capabilities can also be important for post- criteria are applicable to all. The hazard module event analysis; but even where this capability is is used to rank the perils on the basis of their not built to an optimal level, a GIS-compatible representation of the primary hazard, the output can often provide post-event viewing availability of secondary perils (liquefaction, speed. Many software packages for earthquake landslide, tsunami, and fire for earthquake; and for flood, as well as the four multi-risk landslide, soil erosion/land degradation, packages, provide GIS output compatibility for sedimentation, salinity, and fire for floods), quick viewing. The ability to analyze consecutive spatial resolution, and the availability of historical events (the historical event catalogue is just as events such as earthquake aftershocks or relevant for floods versus earthquakes as is the combined flood peaks was also considered in this spatial resolution). module. 8. Risk components. Risk can be quantified in a /// /// 10. Forecasting. Because some emergency /// /// variety of ways. Losses may be calculated via a preparedness measures are taken in response damage-loss conversion that—because loss data to forecasts, a module that accounts for several are absent—is often the least researched part of criteria related to forecasting was included. the entire process. The economic losses generally Forecasting depends on the speed of analysis in account for direct loss; estimates of indirect much the same way that post-event analysis does. loss are less common. Most software packages In cases where a disaster resembles an historical simply use the mean damage ratio (repair or modelled event scenario, forecasting also to replacement cost) and variability from a depends on the ability to quickly apply a database vulnerability function to derive an economic loss. of tested scenarios. 11 12 11. Outputs. Model results are the most important /// /// for probabilistic analysis. Both sensitivity output of the risk analysis and were therefore analysis and the ability to view uncertainties in one of the modules included in the assessment. deterministic scenario runs are also important. Software should make it easy to view outputs in The uncertainty parameters in the criteria GIS (such as OGC standard formats) in terms of should be weighted by users, given that they will hazard, exposure, and vulnerability. A one-page ultimately know what the desired uses of the summary is also important, since it allows the software outputs are. In addition, it should be key losses and products of the software to be possible to rerun the analysis through either a simply and quickly viewed. Loss statistics should saved file or automatic repeat function. Finally, be available in a way that enables analysis and model outputs such as a benefit-cost metric comparison of statistics and that makes possible (offered by OPENRISK, a package not reviewed production of loss exceedance curves, event loss directly in this study), are also valuable. tables, and return period analysis. Ideally, the models should take into account all uncertainties 3.0 - Results A one-page summary of each of the 31 software packages assessed in detail is presented here, ordered by peril. Earthquake Loss Flood Loss Estimation Estimation 16 CAPRA Earthquake 28 BASEMENT 17 EQRM 29 CAPRA-Flood Model 18 Hazus-MH Earthquake Module 30 Delft-3D-FLOW 19 InaSafe -Earthquake 31 Hazus-MH Flood Module 20 MAEviz/mHARP 32 HEC-RAS/HEC-HMS/HEC-FDA/HEC-FIA HEC Suite 21 OpenQuake 33 InaSAFE Flood 22 RiskScape-Earthquake 34 Kalypso 23 SELENA 35 NoFDP IDSS 36 RiskScape-Flood 37 Sobek Suite 1D/2D with HIS-SSM 38 TELEMAC-MASCARET Wind/Hurricane/ Storm Estimation Tsunami/Storm 24 CAPRA-Hurricane Surge /Wave Loss 25 Hazus-MH Hurricane Model Estimation 26 RiskScape-Storm/Wind 39 CAPRA-Tsunami and Storm Surge 27 TCRM 40 Delft-3D-WAVE (SWAN) 41 InsaSAFE-Tsunami 42 OsGEO Tsunami (R.Tsunami) 43 RiskScape-Tsunami 44 SLOSH 45 TOMAWAC Wave 46 TsuDAT using ANUGA 13 14 CAPRA EARTHQUAKE Software Name Peril License Current Version Open Source Operating Systems CAPRA Earthquake Apache 2.0, CC3BY V2.0.0 Yes Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help CRISIS-2007 (Hazard), Vulnerabilidad (Vuln.), CAPRA-GIS, WWJ Visual Basic .NET N Y Y Mapviewer Goal and Summary of the Software The software calculates deterministic and event set probabilistic risk for buildings using continuous fragility functions. CRISIS2007 is the hazard module that can create an event set using 3D source geometries of a particular annual frequency. Different ground motion (GM) parameters can be used, but spectral acceleration (Sa) is the most common. Once there is output, the vulnerability module (ERN-Vulnerabilidad) allows computation of fragility functions that are essentially user driven. Uncertainty is user driven with a simple variance. CAPRA-GIS is used for the quantification of the loss based on the input hazard set, for a particular exposure. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Spectral ordinates are used in terms of peak ground acceleration (PGA) and Sa. Calculations use ground motion prediction equations (GMPEs) *.ame (main), *.txt, *.atn *.fvu, *.dat/*.xml *.shp and source-site distance, source geometry, and seismicity resulting from the CRISIS analysis. Description of Software Risk Outputs Human losses can be calculated directly from a vulnerability function. In addition, economic losses are output in a *.res format file. For the list of earthquakes, the exposure value as well as EP (expected loss), VarP (variance of the loss), and the beta distribution (a, b) are output. Annual average loss (AAL) over a set of buildings or one building, probable maximum loss (PML), and exceedance curves for loss are output. 3.1 – Earthquake Loss Estimation Losses are displayed per building in the CAPRA-GIS window, providing an easy view of loss. Liquefaction analysis can also be undertaken and a map produced (for landslide and tsunami also). *CAPRA does not have a formal manual currently and instead uses support and tutorials. Advantages and Disadvantages ✓ Hazard is not hardwired, and can be input from any other program as long as the file is in the right format. ✓ The log files are very good, and the individual file production means the rerun capabilities are very good. ✓ The process of going through the hazard, vulnerability, and exposure, and then building the functions, helps the basic user to understand the problem. Variability and uncertainty are handled well. ✓ Inbuilt GIS related directly to the loss calculations is very useful; GIS is modular and extendable. ✕ The fatality functions and economic functions lack a lot of diversity, with only a direct relationship available. ✕ The damage distribution is not calculated directly and only an MDR (mean damage ratio) is available. ✕ CRISIS2007 has a strange way of assigning fault regimes to faults: assignment can be done only through the GMPE, not directly. ✕ No formal manual is provided, and with its mix of Spanish and English, the entire GUI is quite difficult to maneuver. Recommended Improvements for Greater Utility The software would benefit from the input of fragility/casualty/economic functions from other projects. It could benefit from synergy with EQRM or MAEviz to add more functionality. EQRM 3.1 – Earthquake Loss Estimation So tware Name Peril License Current Version Open Source Operating Systems EQRM Earthquake GPL 2.2svn1183 Yes Windows, Mac, Linux Pre erred Speci ic In ormation Coding Language So tware Modules Manual GUI Help Python, Matlab, C EQRM Python Y N Y Goal and Summary o the So tware Earthquake Risk Model (EQRM) is a model or regional earthquake risk assessment that has been developed by Geoscience Australia (GA) or application to Australian cities. The model is utilized in the orm o a Python or Matlab-based program ounded on the Hazus model. It has been adapted to Australian conditions with the building/bridge types and other changes, especially the geological conditions within the hazard section. It also includes a regional seismicity model, attenuation model, regolith site response model, elements at risk (social demographics, building inventory), vulnerability o those elements at risk (building vulnerability model [capacity]), casualty model, injury model, and economic loss model. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Spectral ordinates are used in terms o a continuous Sa spectrum. Calculations use GMPEs and source-site distance, source geometry and *.xml, *.csv *.csv *.csv, *.par seismicity, and return period. MMI (Modi ied Mercalli Intensity scale) is also possible. Description o So tware Risk Outputs Structural, nonstructural, and contents damage based on lognormal ragility curves with standard deviation calculated in terms o spectral displacement (per the Hazus methodology) via per ormance point are used to create damage state probability. Total economic loss split into components (structural, contents, etc.) is calculated by usage, damage state, loor area, and cost per m2. Fatalities and injuries are also calculated. Both Australian values and those or Hazus are inbuilt. Outputs include aggregated annualized loss, annual loss, risk exceedance curves (PML), exceedance curves, and disaggregated losses by a number o options, such as distance, magnitude, Screenshot o the disaggregated loss or Newcastle (Robinson, Ful ord, construction type, and spatial unit in *.csv orm. Various plots are and Dhu 2006). available. Advantages and Disadvantages The so tware o ers a large number o visualization options or hazard (uni orm hazard spectra, hazard exceedance, and probabilistic seismic hazard analysis [PSHA]) and or risk (AAL, PML, disaggregation), including or a large number o building typologies. This so tware was the irst to calculate event-based PSHA with this level o detail and analysis; it still leads or physical risk output options in terms o annualized losses and risks. It is completely open source and extendable, which allows or easy modi ication o parts o the code. Integration with GIS is lacking; this could be changed in uture editions. There is no GUI, which makes it di icult or basic users. The so tware simply needs to be combined with MAEviz! Recommended Improvements or Greater Utility Socioeconomic indicators should be added, and there should be greater depth in GUI and GIS (as in MAEviz). EQRM would also combine well with CAPRA, given its event-based nature, despite the di erence in vulnerability. EQRM needs a GUI or non-experienced users in order to become mainstream and could combine well with other so tware rom GA (TCRM, TsuDAT) as well as lood so tware rom Deltares, although rewriting would be necessary. 15 16 Hazus-MH Earthquake Module Software Name Peril License Current Version Open Source Operating Systems Hazus-MH Earthquake Single User I V2.1 (MR5) No Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help VB6, C++ Hazus-MH, ArcGIS Desktop, AEBM, CDMS, SQL Server Y Y Y Goal and Summary of the Software The software calculates earthquake damage to infrastructure and populations over a census tract, county, or state in the United States. The hazard is based on an input of a set of earthquakes, or a scenario quake using NGA (next generation attenuation) relations. The vulnerability method is based on the capacity spectrum method—finding the performance point between demand and capacity. This allows for the calculation of losses to buildings, infrastructure, and lifelines, as well as social loss. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Spectral ordinates are used in terms of PGA and Sa (0.1, 0.3, 1.0, 3.0 Inbuilt Inbuilt *.csv, *.xls, *.res, *.dbf sec). Calculations use GMPEs as well as source-site distance and source geometry. Description of Software Risk Outputs Losses are based on buildings, essential facilities, transportation, and utilities. Damage states are calculated based on a lognormal pdf versus hazard metric. The output is in the form of an ArcGIS display of hazard and the relative losses to buildings and population. Shelter, deaths, injuries, and other social loss functions are calculated through calibration with historical losses and building damage. In addition, direct and indirect economic losses are taken into account 3.1 – Earthquake Loss Estimation with downtime and business interruption functions. These are calibrated for U.S. conditions. Many other earthquake loss estimation procedures have been based on this Hazus method. Screenshot of Hazus Earthquake damage states. Advantages and Disadvantages ✓ The software includes a detailed technical and user manual with full disclosure as to all data and assumptions related to fragility, hazard, and loss functions. Benefit-cost ratio (BCR) calculators and mitigation aspects are also part of the decision module. ✓ The software includes many groupings of buildings and loss estimates based on U.S. building typologies and expert judgment. ✓ GUI and system of analysis for earthquakes work well and even allow ShakeMap input. ✕ The package is heavily calibrated to U.S. conditions. Many loss functions have never been fully calibrated, given the lack of large loss events in the U.S. since 1994 (Northridge). ✕ Although free, the package cannot operate without commercial software (ArcGIS). A problem arose because .NET had not been installed, creating a conflict with the SQL server. Recommended Improvements for Greater Utility Hazus has already been adapted and has influenced EQRM, SELENA, MAEviz, etc. In terms of its functionality, it could become more global (adapting fragility functions to locations outside the United States) and open source (allowing changes in code, and changing GIS). As it develops, it should be monitored and its functions included in other software packages. The fire-following-earthquake, liquefaction, and input-output models can be applied to most other software packages. InaSAFE-Earthquake 3.1 – Earthquake Loss Estimation Software Name Peril License Current Version Open Source Operating Systems InaSAFE Earthquake GPLv3 V1.0.1 YES Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Python InaSAFE, InaSAFE QGIS Plugin Y Y Y Goal and Summary of the Software This software is a link between the science community and those in the planning and policy community seeking to understand an earthquake scenario. Created as a project of the Australia-Indonesia Facility for Disaster Reduction (AIFDR), World Bank, and GFDRR, it is a plugin that takes exposure inputs (population, buildings) and hazard inputs (MMI, intensity raster over the scenario area) from any model, then uses simple vulnerability functions to calculate an output through a simple-to-use GUI in a QGIS plugin form. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics MMI is used via the input hazard file. Hazard is calculated outside the *.shp n.a. *.shpf program and is then switched inside the program to create the impact functions. Description of Software Risk Outputs InaSAFE is currently in production; it currently outputs building loss (as a function of MMI) and damage state, and can also calculate economic losses based on floor area and contents/building value. It calculates fatalities using a PAGER function or an ITB (Bandung Institute of Technology) function. Displaced people are also calculated via population density. It has a great tool for measuring shelter needs, even calculating the amount of rice, drinking water, family kits, and toilets needed. The losses, shown in the form of a GIS file within QGIS, are easy to view. The sidebar also provides an easy-to-view loss summary. Screenshot of the Padang 2009 earthquake (inasafe.org 2013). Advantages and Disadvantages ✓ This is a wonderful tool that allows the “plug and play” addition of hazard and exposure layers into the system. ✓ It explains concepts clearly so that novice users are able to understand them. ✓ It is supported by a very good developer community and a very good manual system. ✓ It is easy to adjust functions and to read the code. ✕ It uses an ITB fatality function based on four earthquakes in which the MMI was only simulated and did not match the actual event (though the beauty of InaSAFE is that it can be changed and is extendable). ✕ It is still in the test phase, and still needs additional functionality. Recommended Improvements for Greater Utility This software would fit well both with ShakeMap utilities for rapid loss and with detailed scenario hazard estimates. It would benefit from the following: synergy with MAEviz to explore possible end modules to be coded; synergy with an additional hazard module (possibly MAEviz or EQRM); and some form of additional socioeconomic analysis with respect to indicators. 17 18 MAEviz/mHARP Software Name Peril License Current Version Open Source Operating Systems MAEviz Earthquake Single User V3.1.1 Build12 YES Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Java using Eclipse RCP Many risk modules—NCSA GIS, MAEviz, EQvis+ Y Y Y Goal and Summary of the Software Another Hazus-based application, MAEviz (Mid-America Earthquakes Visualization) was developed to perform seismic risk assessment in the middle U.S. states. At first glance, it seems specialized; however, its huge potential can be seen in the flowchart of analysis procedures (48 and counting) and its complete Hazus system, including detailed algorithms. The visually driven system uses a combination of Sakai (an open source web portal), NEESgrid (a framework of tools to allow researchers to collaborate), and SAM (Scientific Annotation Middleware) in order to allow users to add their own hazard data. It is easily extendable; the European Union (EU) project SYNER-G, for example, has added a large fragility function manager to it, in addition to other tools. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Spectral ordinates are used in terms of PGA and Sa. Calculations use .txt, .csv .xml *.shp GMPEs and source-site distance, source geometry, and seismicity. Description of Software Risk Outputs Damage estimates include options for multiple mitigation strategies, testing of scientific and engineering principles, and estimation of the earthquake hazard impact on lifelines and social or economic systems (based on Hazus and extra analysis). The outputs are all types of economic losses (direct, indirect, downtime, business interruption), social losses (social vulnerability, fatalities, injuries, homeless), and management options. Many modules (50+) have been produced for applications (like disruption analysis etc.). 3.1 – Earthquake Loss Estimation Simple reports and data views are given. The software creates all scenario outputs (disaggregated and not). An overview of the MAEviz options (McLaren et al. 2008). Advantages and Disadvantages ✓ The software is completely open source and features inbuilt GIS; it is well formatted with the GIS user interfaces. ✓ It is easily the best software for scenario risk assessment and decision support (mitigation, benefit-cost). ✓ It has an outstanding array of modules that provide end analysis such as shelter needs or business interruption. ✓ There is a developer community, and the function codes are easy to read and improve. ✓ Basic users find it easy to use; it offers a large array of infrastructure types for analysis. ✓ Combining detailed hazard, detailed vulnerability, and management and risk modelling, the software is easily extendable. ✕ It is currently tuned only for deterministic analysis. Recommended Improvements for Greater Utility mHARP will give this fantastic software an additional use. It should be integrated with Deltares or other risk software, given the common structure. It has already been integrated in HAZturk and SYNER-G. A combination with EQRM for probabilistic modelling would be useful. An InaSAFE-style command system could simplify the software even further for the most basic users, but it is currently fairly user-friendly. OpenQuake 3.1 – Earthquake Loss Estimation Software Name Peril License Current Version Open Source Operating Systems OpenQuake Earthquake Affero GPL (Apr 2013) V0.4.6 YES Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Python Separated Modules for Risk and Hazard Y N Y Goal and Summary of the Software Currently in the production phase, the software is being designed to calculate earthquake risk transparently for any location on the globe at various levels (country, regional, local). The release date was moved from 2013 to 2014. In the demonstration of the software, hazard can be calculated using multiple PSHA methods (classical and event based) as well as deterministic scenarios analyzed via the NRML XML files (Natural Markup Language). Vulnerability is then input in the form of fragility functions via xml to create a variety of risk outputs. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Software uses PGA, spectral acceleration at various periods via a wide range of source geometries with a large number of GMPEs. It uses *.xml .xml *.xml various site effects methods, including Vs30. It will include many recurrence relation methods for probabilistic analysis, and will also include MMI. Description of Software Risk Outputs Risk outputs will include losses for nonstructural, structural, contents, and occupancy for residential buildings. The software creates loss curves, aggregated loss curves, loss maps (currently output in xml), fractional loss ratios, benefit-cost ratios, damage distribution and various aggregated loss ratios, and event loss tables for a particular taxonomy, or scenario losses and damage. Handling of fatalities and social losses is still being developed but will likely use PAGER-type loss functions associated with structural loss, or the empirical functions. A view of the NRML .xml for vulnerability functions. *v1 was originally due for release in April 2013; however, it was not available at the time of publication. The software should be retested once a full version is released. Advantages and Disadvantages ✓ Software includes a wide range of hazard and risk analysis tools, with a very good hazard section allowing for all types of sources, as well as fault and strain rate analysis. ✓ The source code and test case are promising for Messina 1908. ✓ It currently offers the most in-depth probabilistic analysis of any of the reviewed software packages for earthquake, accounting for classical PSHA as well as event-based PSHA. ✓ It provides a consensus from some earthquake experts globally through a stakeholder process for some parts of the software. ✕ It looks only at residential buildings. ✕ It is not known whether all data are freely available and open, or whether the all components of the software are open source. ✕ No GUI is currently available; the installation procedure through OATS had many difficulties, and the software will likely be difficult for non- engineers to use. Recommended Improvements for Greater Utility For the software to be useful externally, a stand-alone GUI with data is required. The installation procedure needs to be improved (note that software is still in the test, preproduction phase). Before the software’s release, it is hard to propose possible synergies, but given that the Python-coded EQRM joins well with the Global Earthquake Model (GEM), there is a natural synergy between OpenQuake and these two software packages. friendly. 19 20 RiskScape-Earthquake Software Name Peril License Current Version Open Source Operating Systems RiskScape Earthquake Licensed (2-month) V0.2.82 No Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Java RiskScape, Asset, Aggregation, Hazard, Fragility, Mitigation Builder Y Y Y Goal and Summary of the Software The software creates deterministic and (in the future) probabilistic direct and indirect socioeconomic loss estimates for earthquakes for all types of assets, networks, and population, and includes all secondary hazards. It uses a wide array of builders, combining assets (input of buildings, infrastructure, etc.), aggregation (method to combine assets), hazard (defining the hazard model to be used), fragility (to create fragility curves), and mitigation (to perform analysis of changes in infrastructure quality). File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Ground motion is measured in MMI or other metrics and is calculated *.rksh *.rksh *.rksa using intensity prediction equations and source-site distance, source geometry (point, fault, or historical earthquake), and soil effects. Description of Software Risk Outputs An empirical method is used to create loss using MDRs for different infrastructure types from the intensity with a continuous function to 1.0. Casualties are calculated in much the same way from a mean collapse rate. Damage states, monetary losses via replacement, contents, nonstructural costs, a number of socioeconomic interruption parameters, fatalities, injuries, homeless, and uninhabitable buildings are all calculated over the *.shp aggregation selected. 3.1 – Earthquake Loss Estimation *.kml outputs are in Google Earth in 3D. *.shp outputs can be viewed in any GIS program. *.pdf summary shows the key aspects and maps desired from the starting panel. Screenshot of the latest RiskScape v. 0.2.82 front page. *.xls gives an Excel readout of the various losses. RiskScape was very easy to run, with all models tested inside a few minute. The user interface is great. Advantages and Disadvantages ✓ Graphical user interface and tool builders are very easy to understand and a pleasure to use. ✓ Output of the analysis can be done in many forms (pdf, xls, in software, kml, shp). ✓ Historical earthquakes and the various builders are very easy to use. This is one of the most user-friendly packages! ✕ There is not a transparent explanation of how the different functions are calculated. ✕ The software is not itself open source, and the code is unavailable. ✕ The software itself is quite simplistic in terms of the features to calculate—e.g., simple curves. Recommended Improvements for Greater Utility The integration of an EQRM-style analysis would be useful. Essentially, the RiskScape model is dominated by the hazard layer, which is limited to MMI in New Zealand. The probabilistic model will provide an additional boost to the software. Other software packages should understand the benefits of working with RiskScape, given the large effort that has gone into making the software user-friendly. The software should be open source. SELENA 3.1 – Earthquake Loss Estimation So tware Name Peril License Current Version Open Source Operating Systems SELENA Earthquake GPLv2 V6.0 Yes Windows, Mac, Linux Pre erred Speci ic In ormation Coding Language So tware Modules Manual GUI Help Octave, C SELENA v6.0 Y Y Y Goal and Summary o the So tware SELENA (Seismic Loss EstimatioN using a logic tree Approach) has been produced by NORSAR with support rom the International Centre or Geohazards, Norway, and essentially uses the Hazus damage probability methodology in a stand-alone Octave ormat, which has been calibrated to Oslo conditions. SELENA uses a logic tree approach based on the weighting o the input parameters in order to consider epistemic uncertainty. Hazard analysis is probabilistic, real-time, or deterministic; the capacity spectrum method is used to ind the per ormance point and damage. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics So tware uses PGA and Sa (0.3, 1.0s). It also uses NEHRP/EC8 soil classes and spectral shapes rom IBC, EC8, or IS1893. Real-time data *.txt *.txt *.txt can also be input. Many GMPEs are used. Sources are modelled as inite aults. Description o So tware Risk Outputs The so tware produces text iles o ive damage states in terms o built area, number o buildings, and probability per building type. The cost o repair is simply calculated by repair/replacement per m2, giving a total direct economic loss value on a geounit basis. Uninhabitable units and debris computation are also included on a geounit level. Social losses are calculated in terms o deaths and injuries (three levels), based on occupancy ratio, the structural damage, and a user input casualty rate. Occupancy patterns are taken into account using the Hazus methodology (night, day, to/ rom work), and the text iles give the results or these three times. Plotting can occur in RiSe (a Google Earth display acility as part o SELENA) as an aggregated or disaggregated ile with 16 percent, 84 RISE output rom SELENA and the GUI inter ace. percent, and logic tree options. Advantages and Disadvantages The so tware has an easy-to-use GUI, and the Hazus methodology is easy to use once the text iles are prepared. It allows all types o disaggregation and logic trees in order to calculate loss. It uses Octave, an easy programming language, and also has a nice viewer (RiSE). Outputs are quite di icult to manipulate compared to other so tware packages. The number o text iles that need to be input makes it complicated to run without errors; headers need to be in the right spot, and the use o many geounits and building types and occupancy contributes to the di iculty. Recommended Improvements or Greater Utility The original text ile system should be re ormatted, and the GUI inter ace should be integrated with this system. Currently, there is no intensity measurement, though one would be a use ul addition or low-moderate seismicity countries with a lack o ground motion records. The logic tree component lends itsel to being adapted into other so tware packages. 21 22 CAPRA-Hurricane Software Name Peril License Current Version Open Source Operating Systems CAPRA-Hurricane Storm/Wind Apache2.0, CC3BY V1.0.0.0 Yes Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help CAPRA-Hurricane, CAPRA-RainNH, CAPRA-Flood - CAPRAVuln, CAPRA- Visual Basic .NET N* Y Y GIS Goal and Summary of the Software The software calculates deterministic and event set probabilistic risk for buildings using continuous fragility functions. CAPRA-Hurricane is the hazard module that can create a set of events for hurricane paths. CAPRA NHRain and CAPRA-Flood are combined for creating water column depth. Once there is output, the vulnerability module (CAPRA-Vulnerabilidad) allows computation of fragility functions that are essentially user driven and based on flood height and wind speed. Uncertainty is user driven with a simple variance. CAPRA-GIS is used for the quantification of the loss. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Metrics include maximum wind speed (3 second gust) influenced by *.pcf,*.atl (path) - *.shp, *.grn, topography and roughness; flood depth via spatial distribution of *.fvu, *.dat *.ame *.dat (topo) rainfall; and storm surge flood depth via bathymetry and hurricane path. 3.2 – Wind/Hurricane/Storm Loss Estimation Description of Software Risk Outputs Human losses can be calculated directly from a vulnerability function. In addition, economic losses are output in a *.res format file. From a hurricane event list, the exposure value as well as EP (expected loss), VarP (variance of the loss), and the beta distribution (a, b) are output. AAL over a set of buildings or one building, PML, and exceedance curves for loss are output. Losses are displayed per building in the CAPRA-GIS window, providing an easy view of loss. Landslides can also be calculated in relation to losses, as rainfall often causes problems. Screenshot of the Nicaragua hurricane example. Advantages and Disadvantages ✓ Hazard is not hardwired, and could be input from any other program as long as the file is in the right format. ✓ The log files are very good, and the individual file production means the rerun capabilities are very good. ✓ The CAPRA-Hurricane package works well, CAPRA-Vuln and CAPRA-GIS combine well to allow users to understand the loss. ✓ Inbuilt GIS related directly to the loss calculations is very useful. GIS is modular and extendable. ✕ The fatality functions and economic functions lack diversity, with only a direct relationship available, the damage distribution is not calculated directly, and only an MDR is available. ✕ No formal manual is provided, and with its mix of Spanish and English, the entire GUI is quite difficult to maneuver – with novice users, it will is difficult due to lack of help options. Recommended Improvements for Greater Utility The wind speed measurement should be calculated using pressure and other parameters available. Software could learn from TCRM about the hazard module. Some Hazus functions should be applied, and hazard file converters should be more functional. The methodology would work well in a MAEviz environment.. Hazus-MH Hurricane Model 3.2 – Wind/Hurricane/Storm Loss Estimation Software Name Peril License Current Version Open Source Operating Systems Hazus-MH Hurricane Single User © V2.1 (MR5) No Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help VB6, C++ Hazus-MH, ArcGIS Y Y Y Goal and Summary of the Software The software calculates hurricane damage to infrastructure and populations over a U.S. census tract, county, or state. The exposure module has around 5,000 unique building types, based on roof types, etc. There are many functions for each building type. The hazard module is based on user-defined or historical hurricanes, which can be converted for probabilistic analyses or analyzed individually. The vulnerability method is based on the peak gust speed versus a damage probability using empirical curves. This approach allows for the calculation of losses to buildings and essential facilities (other types not supported as yet). File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Pressure and hurricane category versus return period is calculated for *.csv, *.xls, *.res, *.dbf, landfall versus non-landfall. Peak gust wind speed is generally the Inbuilt Inbuilt *.grd hazard parameter influenced by roughness, land cover. Rainfall is also used. Description of Software Risk Outputs Losses are based on buildings and essential facilities (loss of use and damage state). Software calculates the damage states based on a lognormal pdf versus the hazard metric. Debris is calculated along with tree blowdown. The output is in the form of an ArcGIS display of hazard and the relative losses to buildings and population. Tables and reports are also created. Shelter, temporary housing, and displaced households are calculated through calibration with historical losses and building damage. In addition, direct (structural, nonstructural, contents, inventory) and indirect economic losses are taken into account with downtime and business interruption functions. These are calibrated for U.S. conditions. Screenshot of Hazus Hurricane in action. Advantages and Disadvantages ✓ There is a detailed technical and user manual with full disclosure concerning data and assumptions for fragility, hazard, and loss functions. BCR calculators and mitigation aspects are also part of the decision module. ✓ The package worked well combined with earthquake and flood. ✓ It includes many groupings of buildings and loss estimates that are based on U.S. building typologies and expert judgment. ✓ The GUI and system of analysis for hurricanes works well, allowing inclusion of new storm tracks. ✕ The software is heavily calibrated to U.S. conditions and difficult to apply to other locations. ✕ Although free, the package cannot operate without commercial software (ArcGIS). ✕ No fatality modelling is currently undertaken. Recommended Improvements for Greater Utility Hazus is not currently open source, and there are no hurricane loss analyses apart from the CAPRA and RiskScape software packages. The addition of transportation and power outages would be useful. Hazus could be integrated into the MAEviz methodology. 23 24 RiskScape-Storm/Wind Software Name Peril License Current Version Open Source Operating Systems RiskScape Wind Licensed (2-month) V0.2.82 No Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Java RiskScape, Asset, Aggregation, Hazard, Mitigation Builder Y Y Y Goal and Summary of the Software The software creates deterministic and (in the future) probabilistic direct and indirect socioeconomic loss estimates for wind gusts for all types of assets, networks, and population, and includes all secondary hazards. It uses a wide array of builders, combining assets (input of buildings, infrastructure, etc.), aggregation (method to combine assets), hazard (defining the hazard model to be used), fragility (to create fragility curves), and mitigation (to perform analysis of changes in infrastructure quality). For wind, a few test cases have been undertaken using the BLASIUS, RAMS, and GERRIS modelling of wind fields for locations in New Zealand. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Terrain and roughness are influencing factors. Gust wind velocity, *.rksh *.rksf *.rksa however, is the only modelled parameter (in m/s) solved via flow equations (i.e., Navier-Stokes) 3.2 – Wind/Hurricane/Storm Loss Estimation Description of Software Risk Outputs An empirical method is used to create loss using MDRs for different infrastructure types from the intensity with a continuous function to 1.0. Casualties are calculated in much the same way from a mean collapse rate. Damage states, monetary losses via replacement, contents, nonstructural costs, a number of socioeconomic interruption parameters, fatalities, injuries, homeless, and uninhabitable buildings are all calculated over the *.shp aggregation selected. Currently not all functions are available for wind. *.kml outputs are in Google Earth in 3D. *.shp outputs can be viewed in any GIS program. *.pdf summary shows the key aspects and maps desired from the starting panel. *.xls gives an Excel readout of the various losses. Screenshot of Hazus Hurricane in action. Advantages and Disadvantages ✓ The GUI and tool builders are very easy to understand. ✓ Output of the analysis can be done in many forms (pdf, xls, in software, kml, shp). ✓ The various builders are very easy to use. ✕ There is not a transparent explanation of how wind speed is calculated (individual model equations are available for the three methods so far used in manuals, but not for the application in RiskScape). ✕ The software is not open source as such, and the code is unavailable online. ✕ The software itself is quite simplistic in terms of the features to calculate—e.g., simple curves. Recommended Improvements for Greater Utility The integration of a TCRM-style analysis would be useful. Essentially, the model is dominated by the hazard layer, and the vulnerability functions are not transparent. Much more study related to wind storm modelling is needed. Currently, RiskScape cannot really be integrated with other software packages. TCRM Software Name Peril License Current Version Open Source Operating Systems 3.2 – Wind/Hurricane/Storm Loss Estimation TCRM Tropical Cyclone GPLv3, CC3.0BY V1.0.2 Yes Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Python and C TCRM Y Y Y Goal and Summary of the Software TCRM (Tropical Cyclone Risk Model) is a stochastic event simulator for tropical cyclone hazard. It uses wind field modelling distributions, which can be mathematically and statistically derived from a variety of methods, to simulate a user-input number of years of activity and create either a probabilistic view of wind speeds or just a single scenario. Having so many distributions available allows the user to see the sensitivities associated with stochastic modelling. Although created in Australia, this software could potentially be used anywhere. Lists of historical events can be input from the IBTrACS data set using tropical cyclone serial numbers and locations, and then defining radius of maximum wind, mean sea level pressure, and distribution to fit the wind field profile and boundary layer method. Different return periods are then input for the wind hazard calculation File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Cyclonic wind speed (maximum gust in m/s) is created for each return period at each grid point, including confidence intervals. Wind field, *.nc, *.txt, *.csv None None pressure, location, beta parameter, and bearing are used in terms of pdf. Description of Software Risk Outputs The hazard outputs are in the form of tracks, which can then be used in risk analysis Example of Port Headland—Tutorial (TCRM 2011). Advantages and Disadvantages ✓ The software package is easy to use—and widely used—for creating tropical cyclone tracks with many different wind field profiles and distributions. ✓ The data set supplied has all the data needed to run the model and is provided freely. ✓ The GUI is fantastic—the format is easy to use and help files are included. ✕ The advconfigeditor.exe should be renamed in order to prevent confusion. ✕ As with EQRM, installing MinGW, SciPy, and NumPy can be problematic depending on the version of Windows being used. There are fewer problems with Linux. Recommended Improvements for Greater Utility This software would work well with nearly every model, given the diversity and ease of coding. It currently addresses only hazard, so it cannot be compared with risk models. 25 26 BASEMENT Software Name Peril License Current Version Open Source Operating Systems BASEMENT Flood Single user V2.1.1 No Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Visual Basic .NET BASEMENT, BASEchain, BASEplane, BASEMesh Y Y Y Goal and Summary of the Software Basement is a numerical simulation software for computation of environmental flow and natural hazard events. It was programmed by ETH Zurich to determine the impact of river corrections. The whole program is developed for hydraulic purposes to analyze river flow and flood potentials. It offers many modelling possibilities including sediment transport, erosion and both steady and unsteady modes. It also includes many algorithms and calculation methods (optimisation techniques). Basechain is a 1D numerical tool for river reach simulation. BASEplane is a 2D numerical tool for river reach simulation. Both of them are primarily hazard based. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Flow depth and flow velocity can be calculated, as well as 1D and 2D *.bmc, ascii, *.txt n.a. *.shp options related to steady and unsteady flow limited by friction, sediment transport, and topography. Description of Software Risk Outputs There are essentially no risk outputs, as there are no modules for vulnerability or exposure analysis included. The hazard model is the output of the software and produces a detailed inundation map. 1D Saint-Venant, 2D shallow water, and 3D Navier-Stokes equations are solved in all domains, as well as the output of depth and velocity. The software also explores geography of channels. Some nice visualizations are possible through the BASEviz module. 3.3 – Flood Loss Estimation Dam break scenario in BASEMENT. Advantages and Disadvantages ✓ The manual explains the process well and also explains the parallel performance available in the code. ✓ There are good topographic input and grid components, as well as a large range of numerical models. ✓ The hazard model is sound and provides nice solutions to channel transport and flow. ✕ The software needs a lot of time to run (slow calculation speed). ✕ It is not easy to use and takes longer to learn than other flood software. ✕ There is no vulnerability or risk module. Recommended Improvements for Greater Utility This software would benefit from an associated risk module. It is not currently open source, meaning that changing many of the functions is not possible. The software itself provides a good format, being coded with links to Python; thus it could be used in some existing pursuits such as InaSAFE. CAPRA-Flood Model Software Name Peril License Current Version Open Source Operating Systems 3.3 – Flood Loss Estimation CAPRA-Flood Flood Apache2.0, CC3BY V2.1 Yes Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Visual Basic .NET CAPRA-Flood, HEC-RAS v2.1 linked., CAPRA-Vuln., CAPRA-GIS N* Y Y Goal and Summary of the Software CAPRA-Flood is a simple tool for determining flood hazard. It uses rainfall data produced by the module, CAPRA-Lluvia, and hydraulic calculations, which are performed by a link to HEC-RAS. It takes multiple factors into account, such as mean velocity, depth, and rainfall, and also includes uncertainties. Its use of the HEC-RAS calculation engine makes it a fast and rather simple tool for evaluating risk and return periods based on precipitation data. For basin analysis, it calculates rainfall data based on geographical data (provided by simple shp files) and PADF (precipitation, area, duration and frequency) curves, which should be provided as txt files. A short manual is also included, which explains the theoretical background and the functionality of the program. Once there is output, the vulnerability module (CAPRA-Vulnerabilidad) allows computation of modifiable inundation depth fragility functions. CAPRA-GIS is used for the quantification of the loss based on the input hazard set for a particular exposure. The software is in both Spanish and English. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Rainfall intensity (mm/24h), storms/simulations, and PADF (precipitation, area, duration, frequency) curves are produced for *.ame(main), *.txt, *.grn *.fvu, *.dat *.shp rainfall. For floods, rainfall runoff is calculated to give a flood depth. Unit velocity can also be calculated, and various hydrograph methods can be used. Description of Software Risk Outputs Human losses can be calculated directly from a vulnerability function. In addition, economic losses are output in a *.res format file. For flood events, the exposure value, as well as EP (expected loss), VarP (variance of the loss) and the beta distribution (a, b), are outputs from the input list. AAL over a set of buildings or one building, PML, and exceedance curves for loss are output. Losses are displayed per building in the CAPRA-GIS window, providing an easy view of loss. Screenshot of an inundation scenario in CAPRA-Flood, Advantages and Disadvantages ✓ Hazard is not hardwired, and could be input from any other program as long as the file is in the right format. ✓ The log files are very good, and the individual file production means the rerun capabilities are very good. ✓ The process of going through the hazard, vulnerability, and exposure, and then building the functions, helps the basic user to understand the problem. Variability and uncertainty are handled well. ✓ Inbuilt GIS related directly to the loss calculations is very useful; GIS is modular and extendable. ✕ The fatality functions and economic functions lack diversity, with only a direct relationship available. ✕ The damage distribution is not calculated directly and is based only on inundation depth, not flow. ✕ No formal manual is provided, and with its mix of Spanish and English, the entire GUI is quite difficult to maneuver. Recommended Improvements for Greater Utility Capra-Flood could be combined with Kalypso or Deltares for flood hazard; specifically, some end functions from these software could be added to CAPRA-Flood. 27 28 Delft-3D-FLOW Software Name Peril License Current Version Open Source Operating Systems Delft-3D-FLOW Flood GPLv3 V4.01.00rc3 Yes Windows Preferred Specific Information Coding Language Software Modules Manual GUI Help C++, DeltaShell Delft-3D-FLOW Y Y Y Goal and Summary of the Software Delft-3D-Flow calculates solutions for unsteady flow using meteorological and tidal forcing upon a curvilinear (or rectangular) grid. The software can therefore model all tidal flow regimes, tsunamis, and river flow simulations. The software is compatible with all other modules in Delft3D. It allows for 3D flow modeling within and around engineering structures, making it a useful tool for locations or projects that may consider flood control structures. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics On a curvilinear or rectangular grid, the flow, direction, etc. can be modelled on a number of points. Cross sections as well as atmospheric, Ascii - .mdf n.a. Too many (river details) tidal, harmonic, wind, and temperature conditions are also looked at, including all stresses and conservation. Description of Software Risk Outputs The inundation depth, flow, and other hydrodynamic characteristics, as well as turbulence quantities, are output to a file after processing. All forms of 2D/3D equations are solved from Navier-Stokes incompressible free surface flow. Boundary conditions and other details such as wind, control structures, and cross sections are also output to files. This approach makes rerun very easy. Coupled with HIS-SSM, there are certain opportunities and possibilities to change the calculated hazard values into risk values. These changes have not been implemented in open source as yet. The software could potentially use the flood layers calculated as part of a 3D-FLOW scenario. 3.3 – Flood Loss Estimation Just one application of Flow (geography.exeter.ac.uk 2013). *HIS-SSM is currently not combined in the software, but can be requested free of charge. Advantages and Disadvantages ✓ The software is well suited to modelling 3D flow; it takes all boundary phenomena and details into account. ✓ It offers many application possibilities (rivers, storm, tide, tsunami) and is coupled with all other Deltares software. ✓ Large-scale analysis is possible (limited only by computing power). ✓ The user can choose from many inbuilt functions for each part of the calculation—e.g., fitting coefficients. ✕ For risk assessment, the computation is quite difficult to use, with many inputs needed (3D modelling). ✕ To review results, the software needs external tools such as Delft3D-WAVE. Recommended Improvements for Greater Utility This software package is state-of-the-art. Further development of risk products on the end of the software is potentially a collaboration point. Combining with MAEviz, TCRM, or CAPRA would also be a natural progression to improve the software’s risk output capability. Hazus-MH Flood Module Software Name Peril License Current Version Open Source Operating Systems 3.3 – Flood Loss Estimation Hazus-MH Flood Module Flood Single User © V2.1 (MR5) No Windows Preferred Specific Information Coding Language Software Modules Manual GUI Help VB6, C++ Hazus-MH, ArcGIS Y Y Y Goal and Summary of the Software The software calculates flood damage to infrastructure and populations over a U.S. census tract, county, or state. The hazard is based on an input of a set of floods, or a scenario flood using depth-discharge frequency. A combination of hydraulic and hydrologic modelling is added to a digital elevation model (DEM). Hydrologic modelling is done via stream gauge data and a regression equation for discharge frequency for each reach. Hydraulic modelling is done in 2D, using cross sections, Manning’s n, and all flow regimes. The estimates are confined to within the floodplain. The relationship to inundation depth is generally used within the flood vulnerability module. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Inundation depth is used in the analysis. A combination of hydraulics *.csv, *.xls, *.res, *.dbf, Inbuilt or .flt Inbuilt and hydrology is used in combination with DEM and roughness data, as *.grd well as various regressions upon return periods. Description of Software Risk Outputs Losses are based on buildings, essential facilities, transportation, and utilities. The software calculates the damage based on occupancy and count. A lognormal pdf versus inundation depth is used. The output is in the form of an ArcGIS display of hazard and the relative losses to infrastructure and population. Social losses are calculated via a simple function, whereas shelter needs are analyzed in depth. In addition, direct (cost of repair, income loss, and agricultural damage) and indirect economic losses are taken into account with downtime and business interruption functions. These are calibrated for U.S. conditions. The Hazus front screens Advantages and Disadvantages ✓ The software includes a detailed technical and user manual with full disclosure as to all data and assumptions used with respect to fragility, hazard, and loss functions. BCR calculators and mitigation aspects are also part of the decision module. ✓ The software includes many groupings of buildings and loss estimates based on U.S. building typologies and expert judgment. ✓ The GUI and system of analysis for floods are good. The analysis is very sound, using 1D/2D and hydraulic and hydrological modelling. ✕ The software is heavily calibrated to U.S. conditions; the depth-damage functions are from the U.S. Army Corps of Engineers and are applicable to U.S. building types. ✕ Although free, the package cannot operate without commercial software (ArcGIS). Recommended Improvements for Greater Utility Hazus is not itself open source, but it can be considered a learning tool for other open source software packages. It provides a useful estimate of flood risk for anywhere in the United States. It allows for transference (or testing of the transference) of flood loss curves which could be used for other software packages like CAPRA or in other locations (as long as the US damage functions and assumptions hold). 29 30 HEC-RAS/HEC-HMS/HEC-FDA/HEC-FIA–HEC Suite So tware Name Peril License Current Version Open Source Operating Systems HEC Flood User License V4.10 No Windows Pre erred Speci ic In ormation Coding Language So tware Modules Manual GUI Help HEC-2 - Fortran HEC-RAS, HEC-HMS, RASMapper, HEC-1 links, etc., HEC-FIA Y Y Y Goal and Summary o the So tware The HEC Suite allows the modelling o lood risk in its entirety— rom hazard through to loss and then decision support in an easy-to-use Windows environment. It was designed by the U.S. Army Corps o Engineers or U.S. lood modelling. HEC-RAS undertakes 1D low measurements to model hydraulic low o rivers. It is probably the best known lood hazard so tware in the world. It can also model lood structures and sediment transport. HEC-HMS is the rain all-runo model; it includes urban and rural overland modelling, watershed runo , and other water supply processes. These are combined with other tools or lood requency statistics calculation (HEC-SSP) and geospatial modelling (HEC-GeoRas). The risk components are then modelled in HEC-FDA and HEC-FIA. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Inundation depth and low are modelled within the HEC Suite, including *g01, *dss HEC-FDA ormats *.shp, *ti , *cad *MrSID the temporal aspects through hydrographs. Description o So tware Risk Outputs HEC-FIA uses hydraulic inputs in the orm o depth grids and duration grids or hydrograph data to calculate risks or single events using an entire hydrograph. Economic losses are then calculated using location data input rom Hazus, parcel data, or existing point data. Li e loss calculations can be undertaken in a module called Li eSim, which is the most advanced atality system among so tware packages that use indicator systems to model potential atalities. Agricultural losses can be calculated using crop and harvest details. HEC-FDA is a probabilistic or deterministic methodology using the peak o the hydrographs. It looks at contents loss analysis, too. Annualized loss, annual damage, annual exceedance probabilities, and other conditional data are calculated, and uncertainty analysis is carried out as well. Risk analysis and visualization are also created. Floodway determination: Perspective plot in HEC-RAS. 3.3 – Flood Loss Estimation Advantages and Disadvantages HEC Suite provides an impressive array o so tware, given the ull lood risk approach. There are around 15 tools that make up the suite, all o them reeware, and they have been developed over many years. Li e loss and economic losses are calculated easily rom the outputs o FDA and FIA. Tutorials with reely available data are o ered, and there are very good manuals and data sets. The GUIs are very easy to use, and or irst-time users make a lot o sense. The major drawback is the absence o source code, meaning that the so tware is not open source Recommended Improvements or Greater Utility I the source code could be obtained, then this so tware would be the best or lood. HEC-RAS will also change to 2D modelling soon. For now, the best that can be done is to look at and learn rom the methodology, and then apply it to a so tware such as Kalypso or Deltares Sobek. InaSAFE-Flood Software Name Peril License Current Version Open Source Operating Systems 3.3 – Flood Loss Estimation InaSAFE Flood GPLv3 V1.0.1 Yes Windows, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Python InaSAFE, InaSAFE QGIS Plugin Y Y Y Goal and Summary of the Software This software links the science community to those in the planning and policy community seeking to understand a flood impact scenario. Created as a project of the AIFDR, World Bank, and GFDRR, it is a plugin that takes exposure inputs (population, buildings from OpenStreetMap, or other shapefiles) and hazard inputs (inundation depth raster over the scenario area, from any open software), and then uses simple vulnerability functions to calculate an output through a simple-to-use GUI in a QGIS plugin form. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Inundation depth is used via the input hazard file. Hazard is calculated *.shp n.a. *.shp outside the program. Description of Software Risk Outputs InaSAFE is currently in production; however, it currently outputs building loss (as a function of inundation depth) in terms of a 1 or 0 function. This could be adapted by the user. Displaced people are also calculated using population density and the buildings impacted by the flood raster. The software looks not only at the number of displaced, but also at the resources needed to support them. InaSAFE has a great tool for measuring various shelter needs, even calculating the amount of rice, drinking water, family kits, and toilets needed. The losses are shown in the form of a GIS file within QGIS that is easy to view. Screenshot of a Jakarta flood example. Advantages and Disadvantages ✓ This is a wonderful tool that allows the “plug and play” addition of hazard and exposure layers to the system. ✓ It explains concepts clearly so that novice users are able to understand them. ✓ The developer community makes it possible for researchers to adapt functions and easily contribute. ✓ OpenStreetMap and QGIS connectivity are very easy to use. ✕ The software is still in the test phase, and needs additional functionality. ✕ There is no numerical model for flood or water depth inside the software, though this is not necessarily a negative. Recommended Improvements for Greater Utility This software would fit well with many decision support modules of Kalypso and SOBEK for rapid loss, as well as any detailed scenario hazard estimates. It would benefit from some synergy with MAEviz to explore possible end modules to be coded. The software would also benefit from synergy with an additional hazard module, possibly a simple form such as NoFDP IDSS. Some form of additional socioeconomic analysis with respect to indicators would also improve it. 31 32 Kalypso Software Name Peril License Current Version Open Source Operating Systems Kalypso Flood LGPLv2 (very good) V12.11.1 Yes Windows Preferred Specific Information Coding Language Software Modules Manual GUI Help Java Kalypso Hydrology 1D/2D, WSPM, Risk, Flood, BASE, Evacuation Y Y Y Goal and Summary of the Software Kalypso is a multi-module program for calculations about hydrology (rainfall-runoff), water-level analysis, 1D/2D analysis, flood calculation, and risk determination. All these modules are linked together. The main focus is on deterministic hydrological calculations of river basins and floods. The user can start with hydrological analysis, which takes into account elements such as precipitation, rainfall, etc., and then continue to the risk calculation. It was originally built for German river locations. The risk calculation uses land-use data based on user input. The resolution of the calculations is quite nice, while the complexity of the modules decreases moving from the hazard to risk, meaning that the hydrological part is quite complex, while the modules for flood and risk are rather simple. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Metrics include flow, inundation depth, and volume of water. 1D/2D *.1d,*.2d .xml / functions *.shp, *.asc rainfall modelling can take snow into account. Roughness is taken into account via land use. Description of Software Risk Outputs The risk outputs include entire hazard inundation maps for various return periods (between 2 and 100 years). Land use and other parameters can be input in order to give the risk. In Kalypso Risk, the parameters are combined as a damage function in an open format, using a function versus inundation depth or duration/frequency. This can have the form “x-” or a straight value, and it is then multiplied by the economic value to give loss. This design allows the software to be easily manipulated, and fatalities could technically be calculated by manipulating the software. Kalypso Evacuation models an entire flooding scenario evacuation, including bus route changes and many other features. 3.3 – Flood Loss Estimation Results for risk zones and damage potential. Advantages and Disadvantages ✓ Multiple modules are included within the analysis, which allows for all-in-one loss analysis, and the system can easily be updated. ✓ The wiki-style system and manual for each component make the software easy to use. The results also have multiple export options. ✓ The GUI and GIS (through QGIS) are state-of-the-art and are easy to use, well-developed, and up-to-date. ✓ There are multiple options for optimization in terms of short-time and long-time modelling. ✕ As with most flood analyses, the risk and flood modules are too simple. The vulnerability analysis is essentially constructed with the hazard parameter overlaid on exposure. ✕ The software focuses mostly on German conditions, and the manuals are mostly in German. Recommended Improvements for Greater Utility TThis software should be combined with MAEviz, given the modular structure. It would benefit from the many modules on the end of it, as well as common language types in the structure. The shelter and evacuation modules are well suited for combining. The addition of fragility functions would help Kalypso. InaSAFE could learn from the simplicity of Kalypso’s modules. NoFDP IDSS Software Name Peril License Current Version Open Source Operating Systems 3.3 – Flood Loss Estimation NoFDP IDSS Flood GPLv3, LGPL (source) V1.0 Yes Windows Preferred Specific Information Coding Language Software Modules Manual GUI Help Java NoFDP IDSS Y Y Y Goal and Summary of the Software Produced as part of the INTERREG III project, NoFDP IDSS is an easy-to-use 1D program for developing, optimizing, and analyzing solutions for risk reduction. It contains the Deltares hydrological calculation tool SOBEK. IDSS is completely open source, as is SOBEK, so the two can now be combined. IDSS covers multiple modules for the import of geo-data and includes a simple GIS engine to edit the data. It also contains a link to ISAR, a calculation concept based on vegetation factors and renaturation. It is the precursor to Kalypso and has many links back to SOBEK. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics 1D hydrodynamic modelling is undertaken using the modelling of *.shp, *.asc, *ZML/XML, ASCII, *.csv n.a. SOBEK. The flow, inundation depth, duration, and frequency are tif modelled using input roughness, land use/land cover data, etc. Description of Software Risk Outputs The program calculates the flooded areas due to different flood scenarios (return periods of 50 years, 100 years, etc.) and uses land-use information by CORINE to calculate economic damage (much as the Kalypso Risk module does). Multiple building and single building analysis are not covered within the calculation, but this information can be obtained by more detailed land-use data. A variety of flood structures can also be modelled. The focus is on determining and calculating risk reduction variants and comparing them to each other by economic social parameter. This decision-support aspect is the most interesting part of NoFDP IDSS. The program also contains a number of collaboration tools, such as automatic report generator (based on Open Office), a screenshot database, and a link to Google Earth. NoFDP IDSS scenario screen. Advantages and Disadvantages ✓ The decision module is very easy to use for basic users, and allows for multivariate analysis. ✓ There is a risk reduction focus as well as collaboration tools. ✓ The open source outputs are useful, and the Google Earth screens are easy to understand and use. ✕ 1D analysis undertaken with SOBEK has not been updated since the end of INTERREG III in 2006. ✕ There is little information about the steps within the risk analysis. Recommended Improvements for Greater Utility Not all options that are in NoFDP IDSS have yet been added to Kalypso, though they will probably be added in the future. The software would be improved with a powerful model such as SOBEK 1D/2D. 33 34 RiskScape-Flood So tware Name Peril License Current Version Open Source Operating Systems RiskScape Flood Licensed (2-month) V0.2.82 No Windows, Mac, Linux Pre erred Speci ic In ormation Coding Language So tware Modules Manual GUI Help Java RiskScape, Asset, Aggregation, Hazard, Mitigation Builder Y Y Y Goal and Summary o the So tware The so tware creates deterministic and (in the uture) probabilistic direct and indirect socioeconomic loss estimates or loods (riverine and coastal) or all types o assets, networks, and population, and includes all secondary hazards. It uses a wide array o builders, combining assets (input o buildings, in rastructure, etc.), aggregation (method to combine assets), hazard (de ining the hazard model to be used), ragility (to create ragility curves), and mitigation (to per orm analysis o changes in in rastructure quality). File Types Used: *.rskm = aggregation Hazard Vulnerability Exposure Key Hazard Metrics Metrics include inundation velocity, inundation depth, and ponding, as *.rksh *.rks *.rksa well as inundation duration on each level. Description o So tware Risk Outputs An empirical method is used to create loss utilizing MDRs or di erent in rastructure types rom the intensity with a continuous unction to 1.0. Casualties are calculated in much the same way rom a mean collapse rate. Damage states, monetary losses via replacement, contents, nonstructural costs, a number o socioeconomic interruption parameters, atalities, injuries, homeless, and uninhabitable buildings are all calculated over the *.shp aggregation selected. *.kml outputs are in Google Earth in 3D. *.shp outputs can be viewed in any GIS program. *.pd summary shows the key aspects and maps desired rom the starting panel. *.xls gives an Excel readout o the various losses. NoFDP IDSS scenario screen. 3.3 – Flood Loss Estimation Advantages and Disadvantages The GUI and tool builders are very easy to understand and a pleasure to use. Output o the analysis can be done in many orms (pd , xls, in so tware, kml, shp.) Historical loods examined within the so tware and the various builders are very easy to use. This allows or a ast view o the possible analysis that can be undertaken. There is not a transparent explanation o how the di erent unctions are calculated. The so tware is not itsel open source, and the code is unavailable. The so tware itsel is quite simplistic in terms o the eatures to calculate—e.g., simple curves. Recommended Improvements or Greater Utility The way RiskScape portrays the loss in ormation with outputs and shows the need or detailed DEMs is impressive and could be combined with a probabilistic engine or a Deltares-type so tware. Sobek Suite 1D/2D with HIS-SSM So tware Name Peril License Current Version Open Source Operating Systems 3.3 – Flood Loss Estimation Sobek 1D/2D Flood GPLv3 V2.13.002 Yes Windows Pre erred Speci ic In ormation Coding Language So tware Modules Manual GUI Help SOBEK1D (Pipe, Overland), SOBEK2D (Rural, Urban, River), HIS-SSM (Risk C++, DeltaShell Y Y Y outputs) Goal and Summary o the So tware SOBEK undertakes all types o 1D/2D hydrodynamic modelling by solving low equations on both 1D network systems and 2D horizontal grids. This approach can be used or river lood orecasting and modelling, overland looding, drainage system modelling, and engineering structure testing (dam breaks, breaches, rural and urban looding). It is computationally e icient and allows or all low regimes. SOBEK 1D can be run through two modules—pipes and examining overland low. It can also be coupled with D-Rain all Runo Open Water, which creates event sets or separate events. SOBEK 2D gives three options or downloading, including rural, urban, and river modelling, which consists o an overland low module in 2D and a rain all-runo model. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Flow and inundation depth are the key metrics. They can also be plotted as a time series. The modelling includes wetting and drying Many n.a. Many processes and physical phenomena, and looks at all low types, roughness, and mass conservation. Description o So tware Risk Outputs The main output is the inundation depth at di erent time periods on the spatial scale de ined, along with the low velocities. In both 1D and 2D, the hydrograph methods are changeable, and the complete set o Saint-Venant equations is solved. The rain all-runo model can be either distributed or lumped; it allows analysis o many catchments and input o historical rain all or production o rain all patterns. Note, however, that there are essentially only hazard outputs. They can be viewed in a GIS plat orm along with all 1D network and 2D grid structures. HIS-SSM could be connected to Sobek in order to calculate casualties and damage associated with representative scenarios o the outputs rom the SOBEK suite. Screenshots o SOBEK in action. *HIS-SSM is currently not combined in the so tware, but can be included ree o charge upon request. Advantages and Disadvantages The so tware provides all solutions or 1D and 2D lood modelling, with pipes, overland, rural, urban, and river modules. It is simple to install and very easy to use. The end products are easily viewable. It is compatible with OpenMI (along with other Del t products), which allows or user adaption. The programming is designed to be very ast, with e icient computation structures. Unless the so tware is combined with HIS-SSM, only hazard outputs are produced. Not currently ully open source at time o review.* Recommended Improvements or Greater Utility This so tware o ers a antastic basis or lood hazard modelling, and the so tware structure works very well. SOBEK (and all the Deltares products) would combine well with CAPRA, TSUDAT, InaSAFE, or MAEviz. * Sobek 1D/2D is currently being released under https://publicwiki.deltares.nl/display/nghs/Development. The 1D components are open source, and there is currently a partial source code available; there will be components released in 2014 and a ull release in 2015. 35 36 TELEMAC-MASCARET Software Name Peril License Current Version Open Source Operating Systems TELEMAC Flood GPLv3 and LGPL V6.2 Yes Windows, Unix Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Fortran, Python, Perl TELEMAC2D-3D, MASCARET, ARTEMIS, SISYPHE, SEDI-3D, STBTEL Y Y Y Goal and Summary of the Software The software is a set of mathematical solvers for various equations related to free surface flow. Many of these can be used for floods and waves. It has been very well tested, with over 200 applications worldwide to date. A finite element method (FEM) grid is set up and discretized into triangles, with the numerical modelling then being undertaken to solve 1D/2D/3D equations. The various module parts are ARTEMIS (wave modelling in harbors), MASCARET (1D surface), TELEMAC-2D (2D Saint-Venant), TELEMAC-3D (3D Navier-Stokes), SISYPHE (2D sediment transport), and SEDI-3D (3D sediment transport). STBTEL is used to build the grid interface. The GUI is provided through BlueKenue or a software called FUDAA, which allows the grid to be generated. POSTEL-3D also allows creation of 2D sections from 3D simulations. It consists of a series of Fortran subroutines and functions that are easily adapted. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Amplitude (inundation depth) and flow are calculated through the Many n.a. ASCII, SELAFIN various equations. Description of Software Risk Outputs Some risk outputs of TELEMAC include the inundation modelling based on dam breaks, embankment failures, and other structural breaks. Risk as such is not calculated; however, a few applications have used TELEMAC-MASCARET for risk production. Though often overlooked, TELEMAC is a very powerful solver suite; in the 1D it can solve all flow regimes (subcritical, supercritical), as well as steady and unsteady flows. TELEMAC-2D has been well set up for supercomputing and is currently part of the PRACE (Partnership for Advanced Computing in Europe) project that models flooding of the Rhine. There is the potential to include an open source risk output package on 3.3 – Flood Loss Estimation the end of each of these modules. Malpasset dam break (opentelemac.org 2013). Advantages and Disadvantages ✓ Validation is a key parameter checking the output; the software has been ratified by the International Association for Hydro-Environment Engineering and Research (IAHR), and there is documentation of these validation cases. ✓ The software provides all the tools necessary for wave and inundation modelling via solving equations. ✓ The code is in Fortran and easily downloadable for use in any application. ✓ There is quite a large user community for these tools. ✕ The software is not for basic users, and using it for applications requires considerable manipulation. ✕ There is no risk calculation within the software. Recommended Improvements for Greater Utility This software should be examined in greater depth; however, it definitely offers a very useful set of libraries and tools to integrate into flood and wave modelling. It has natural synergies with any of the risk modules, including HEC-FIA, Kalypso, Sobek HIS-SSM, CAPRA, or NoFDP IDSS. CAPRA-Tsunami and Storm Surge Software Name Peril License Current Version Open Source Operating Systems 3.4 – Tsunami/Storm Surge/Wave Loss Estimation CAPRA Tsunami/Surge Apache2.0, CC3BY V2.0.0 Yes Windows,Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help CRISIS-2007 (Hazard), CAPRA-Hurricane, Vulnerabilidad (Vulnerability), Visual Basic .NET N* Y Y CAPRA-GIS, WWJ MarbleBlue Goal and Summary of the Software The software calculates deterministic and event set probabilistic risk for tsunamis, using continuous fragility functions. CRISIS2007 is the hazard module that can create an event set using source geometries of a particular annual frequency for tsunami generation. Inundation depths are then produced from topography and bathymetry. Once output, the vulnerability module, CAPRA-Vulnerabilidad, allows computation of fragility functions, which are essentially user driven. Uncertainty is user driven with a simple variance. CAPRA-GIS is used for quantifying the loss based on the input hazard set for a particular exposure, and then Map Viewer CAPRA-WWJ, a visualization tool using the NASA WorldWind Java SDK engine, is used. Storm surge is handled as part of CAPRA-Hurricane. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Inundation depth is the modelled parameter, using either storm surge *.ame(main), *.tsu, *.fvu, *.dat *.shp, *.grn, *.dat from hurricane tracks or tsunami (wave height). Both need the *.pcf bathymetry and topography data. Description of Software Risk Outputs Human losses can be calculated directly from a vulnerability function. In addition, economic losses are output in a *.res format file. For the list of tsunamis and/or storm surge events, the exposure value as well as EP (expected loss), VarP (variance of the loss), and the beta distribution (a, b) are output. AAL over a set of buildings or one building, PML, and exceedance curves for loss are output. Losses are displayed per building in the CAPRA-GIS window, providing an easy view of loss. Screenshot of the San Juan tsunami. Advantages and Disadvantages ✓ Hazard is not hardwired, and could be input from any other program as long as the file is in the right format. ✓ The software provides both tsunami and storm surge options. ✓ The log files are very good, and the individual file production means the rerun capabilities are very good. ✓ The tutorials help a basic user to understand the problem. ✓ Inbuilt GIS related directly to the loss calculations is very useful; GIS is modular and extendable. ✕ The fatality functions and economic functions lack diversity, with only a direct relationship available. ✕ The damage distribution is not calculated directly and is based only on inundation depth, not velocity; the model is a simple one. It is unclear where roughness is used in terms of assumptions. ✕ No formal manual is provided, and with its mix of Spanish and English, the GUI is quite difficult to maneuver. Recommended Improvements for Greater Utility CAPRA would benefit from a more formal tsunami methodology, such as that of TsuDAT. An inbuilt historical tsunami database would also help users to create the event set. There should be additional calibration of loss functions, though there are fewer functions available in other software tools. 37 38 Delft-3D-WAVE (SWAN) Software Name Peril License Current Version Open Source Operating Systems Delft-3D Wave GPLv3 4.01.00rc.03 (40.91) Yes Windows,Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Fortran77/90, C/C++, DeltaShell Delft3D Y Y Y Goal and Summary of the Software Delft-3D-Wave is a GUI-based application using the SWAN model and the HISWA model, both of which are strong calculation engines for hydraulic flow and waves. The software can look at wind-generated waves in coastal waters, examining all depths of water, and can model wave generation, propagation, and breaking problems of short-crested random waves. It needs a lot of advanced-user input, but is one of the most powerful numerical tools for calculating large-scale wave propagation and storm surges. It is usable for estuaries, tidal inlets, lakes, barrier islands, channels, and coastal regions. It is coupled to Delft3D-FLOW. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics 3.4 – Tsunami/Storm Surge/Wave Loss Estimation On a curvilinear grid, wave height, direction, etc. can be modelled on a .dep, .grd, .enc, .wnd, Ascii - .mdw n.a. number of points. All types of roughness, topography, bathymetry, .bnd, .pol, .loc, .obs, .pol wind, etc. are taken into account. Description of Software Risk Outputs This software offers only a hazard output; however, it creates the most detailed 3D wave models available. A binary file is output with the data in time series format as a 2D or 3D map. The distance, depth, mean wave period/direction, directional spreading, dissipation rate, mean wave length, and current velocity are modelled on the grid and wind components. Also modelled are frequencies, densities, and spectral nautical directions in the spectra files. There is no attached risk module. A 2D plot (Delft3D-Wave 2013). Advantages and Disadvantages ✓ The software is well developed for modelling 3D waves. ✓ It is very fast, using as it does the set of Fortran90 codes based on the third generation SWAN model—though it also allows use of the second generation model. ✓ There are many application possibilities (storm, tide, tsunami). ✓ Large-scale analysis is possible (limited only by computing power). ✕ For risk assessment, the computation is quite difficult to carry out, with many inputs needed (3D modelling). ✕ It needs external tools to review results. Recommended Improvements for Greater Utility Delft3D-Wave using SWAN achieves what it sets out to achieve. It could be combined as an option for detailed scenario modelling in complex regions and is a very useful tool combined with Delft3D-FLOW. InaSAFE-Tsunami Software Name Peril License Current Version Open Source Operating Systems 3.4 – Tsunami/Storm Surge/Wave Loss Estimation InaSAFE Tsunami GPLv3 V1.0.1 Yes Windows,Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Python(+libraries), QGIS InaSAFE, InaSAFE QGIS Plugin Y Y Y Goal and Summary of the Software This software is a link between the science community and those in the planning and policy community seeking to understand a tsunami impact scenario. Created as a project of the AIFDR, the World Bank, and GFDRR, it is a plugin that takes exposure inputs (population, buildings) and hazard inputs (tsunami inundation depth) from any Openstreetmap data or software, then uses simple vulnerability functions to calculate an output through a simple-to-use GUI in a QGIS plugin form. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Tsunami inundation depth is used via the input hazard file. Hazard is *.shp n.a. *.shp calculated outside the program and is then switched inside the program to create categories of inundated (1) or not inundated (0). Description of Software Risk Outputs InaSAFE is currently in production; however, it currently outputs buildings that are inundated as a result of a tsunami wave as well as damage state, and it can also calculate economic losses based on floor area and contents/building value. Displaced people are also calculated via the population density. It has a great tool for measuring various shelter needs, as in the other InaSAFE modules. The losses are shown in the form of a GIS file within QGIS and are easy to view. Startup screen including the exposure data. Advantages and Disadvantages ✓ This is a wonderful tool that allows the “plug and play” addition of hazard and exposure layers to the system. ✓ It explains concepts clearly so that novice users are able to understand them. ✓ Its functions and coding are easy for inexperienced users to understand. ✓ There is a good developer community that promotes interaction. ✕ The software is still in the test phase, and needs additional functionality. Recommended Improvements for Greater Utility This software would fit well with many functions for rapid loss and any detailed scenario hazard estimates. It would benefit from some synergy with MAEviz to explore possible end modules to be coded, and could use run-up heights from TsuDAT or CAPRA. The software would also benefit from synergy with an additional vulnerability function builder for various tsunami impacts globally. Some form of additional socioeconomic analysis with respect to indicators would make the software more useful for decision makers. 39 40 OsGEO Tsunami (R.Tsunami) So tware Name Peril License Current Version Open Source Operating Systems OsGEO Tsunami Tsunami GPL No vers. Yes Windows,Mac, Linux Pre erred Speci ic In ormation Coding Language So tware Modules Manual GUI Help R, GRASS-GIS GRASS-GIS, R.Hazard.Tsunami Y Y Y Goal and Summary o the So tware R.tsunami is a plugin or the open-source GIS tool GRASS-GIS. Within GRASS-GIS, it is a numerical model or calculating inundation depth, storm surges, and run-up or tsunami waves. It has been tested in two areas, the Ligurian coast and the Virgin Islands. R.tsunami calculates the hazard o tsunami events, but does not address exposure or vulnerability; there ore urther plugins or GRASS-GIS are needed. Users can drag in exposure data and then use vulnerability unctions based on U.S. Army Corps o Engineers unctions or lood heights with building stocks o the respective countries. Exposed stock is simply intersected with the hazard layer to create statistics. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics .shp 3.4 – Tsunami/Storm Surge/Wave Loss Estimation Inundation depth and run-up are calculated using the plugin based on Evaluated directly (poly or point) Multiple location, bathymetry, topography, and roughness. Description o So tware Risk Outputs The so tware needs a lot o data, including • digital terrain model raster map • roughness raster map • height o the generated wave in the source point • depth o the sea in the source point • sea depth near the coast • output looded area map name • output lood height map name • coordinates o a point on the sea It is tailored to EU conditions, using CORINE land use or roughness calculations. Equations o economic loss or di erent return periods are characterized or a high event, medium event, and low event (as de ined by the user) or residential and agricultural buildings per m2. Startup screen including the exposure data. Advantages and Disadvantages This is a straight orward plugin or GRASS GIS that is easily understandable and open source. It attempts to calculate the economic impact o tsunamis and characterizes damage states. It uses roughness based on CORINE land cover, which provides a good basis or the rest o Europe. The damage-stage relations will usually be di erent or tsunami/storm surge than or lood—but in the absence o in ormation they are perhaps reasonable. Much data needs to be input. Not many case studies have been tested around the world. No manuals or tutorials have been provided. Recommended Improvements or Greater Utility The GRASS-GIS plugin is very simple but could provide a use ul additional testbed or applying InaSAFE and or converting the unctions rom R to Python. RiskScape-Tsunami Software Name Peril License Current Version Open Source Operating Systems 3.4 – Tsunami/Storm Surge/Wave Loss Estimation RiskScape Tsunami Licensed (2-month) V0.2.82 No Windows ,Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Java RiskScape, Asset, Aggregation, Hazard, Mitigation Builder Y Y Y Goal and Summary of the Software The software creates deterministic and (in the future) probabilistic direct and indirect socioeconomic loss estimates for tsunamis for all types of assets, networks, and population, and includes all secondary hazards. It uses a wide array of builders, combining assets (input of buildings, infrastructure, etc.), aggregation (method to combine assets), hazard (defining the hazard model to be used), fragility (to create fragility curve), and mitigation (to perform analysis of changes in infrastructure quality). The RiCOM model, which solves a Reynolds-averaged 3D wave equation, has been used so far in test cases. High-resolution topography and bathymetry are also used within the analysis. File Types Used: *.rskm = aggregation Hazard Vulnerability Exposure Key Hazard Metrics Tsunamis are measured in terms of the same parameters as flood. *.rksh *.rksf *.rksa Inundation depth (m), duration (hrs), velocity (m/s), and ponding are included. Description of Software Risk Outputs An empirical method is used to create loss, using MDRs for different infrastructure types from the intensity with a continuous function to 1.0 using an empirical fragility function. RiskScape-Tsunami currently allows everything except human susceptibility to be calculated. Damage states, monetary losses via replacement, contents, nonstructural costs, a number of socioeconomic interruption parameters, fatalities, injuries, homeless, and uninhabitable buildings are all calculated over the *.shp aggregation selected. This is not available for all options in RiskScape. *.kml outputs are in Google Earth in 3D. *.shp outputs can be viewed in any GIS program. *.pdf summary shows the key aspects and maps desired from the starting panel. *.xls, *.csv give an Excel readout of the various losses. Screenshot of the RiskScape loss output for Hawke’s Bay tsunami. Advantages and Disadvantages ✓ GUI and tool builders are very easy to understand and a pleasure to use. ✓ Output of the analysis can be done in many forms (pdf, xls, in software, kml, shp). ✓ Mitigation factors, tide levels, and predefined tsunamis can be calculated easily to see the process of tsunami risk assessment. ✕ There is not a transparent explanation of how the different functions are calculated. ✕ The software is not open source, and the code is unavailable. ✕ The software itself is quite simplistic in terms of the features to calculate—e.g., simple curves. Recommended Improvements for Greater Utility There is a natural link between the Delft3D models, TELEMAC3D models, and the high-resolution modelling attempted by RiskScape. These models within RiskScape attempt to calculate very in-depth hazard parameters, but there is no one model for hazard in RiskScape. TsuDAT and RiskScape also should collaborate, given the proximity of the development teams and given that Riskscape currently lacks probabilistic hazard calculations. 41 42 SLOSH So tware Name Peril License Current Version Open Source Operating Systems SLOSH Storm Surge Single User V1.65i Yes Windows, Linux Pre erred Speci ic In ormation Coding Language So tware Modules Manual GUI Help Python SLOSH Y Y Y Goal and Summary o the So tware SLOSH is a simple hurricane storm surge calculator (using a 1D wave model). It uses an historical database and prede ined modelling grids or common geographic basins, including coastal areas o North America (the Atlantic), parts o the Caribbean, and some Paci ic coasts. SLOSH can also be run in a mode that works out di erent scenarios, changing land all directions, Sa ir-Simpson categories, orward speeds, and sea levels. It then creates a maximum envelope o water (MEOW). File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Surge heights are calculated based on a simple 1D wave equation 3.4 – Tsunami/Storm Surge/Wave Loss Estimation Many n.a. n.a. solver combined with data on wind, topography, bathymetry, pressure, and orward direction. Description o So tware Risk Outputs SLOSH is a hazard model that creates inundation depths over the coastal areas. It does, however, have two use ul eatures that are essentially risk indicators: a MEOW and a MOM (maximum o the maximums), which is the maximum o all the singular MEOWs or a particular basin. Many basins have been modelled, and SLOSH includes ive MOMs (Cat 1 MOM, Cat 2 MOM, etc.). These risk indicators are very use ul or planning. Another program, PHISH/PSURGE, can be used or probabilistic analysis o storm surge heights. 2009 New Orleans hurricane model.. Advantages and Disadvantages The latest version dates to December 2012 and is still based on the engine o the 2003 version—which is simple to use and transparent. The so tware has a great historical database that is integrated into the so tware. It is very well suited or basic users: easy to install, easy to use, and very ast. It has only a ew output/export options into .rex, .pcx, and .txt and these are not very use ul. The so tware is technologically outdated, and it will be no longer developed. Recommended Improvements or Greater Utility This so tware has some great eatures, but was last updated 18 months ago and is no longer being developed. Some o the tools would be use ul inputs into CAPRA and into the planning components o other so tware where the maximum characterization makes sense. TOMAWAC Wave Software Name Peril License Current Version Open Source Operating Systems 3.4 – Tsunami/Storm Surge/Wave Loss Estimation TOMAWAC Wave GPLv3 V2.0.0 Yes Windows, Mac, Linux Preferred Specific Information Coding Language Software Modules Manual GUI Help Visual Basic .NET TOMAWAC, TELEMAC 2D, TELEMAC 3D Y Y Y Goal and Summary of the Software The software is part of the set of mathematical solvers for various equations related to free surface flow, a few of which can be used for floods and waves. In this case, the same FEM grid is set up and discretized into triangles, with the numerical modelling then being undertaken to solve 1D/2D/3D equations. The various module parts are TELEMAC-2D (2D Saint-Venant), for tsunamis with wavelengths in the shallow water domain (wavelength > 20 times the depth); TELEMAC-3D (3D Navier-Stokes), for shorter wavelengths; and TOMAWAC, for modelling wave propagation in coastal areas. TOMAWAC uses a solution of spectroangular density of wave action (F = E/pg). STBTEL is also used to build the grid interface. POSTEL-3D also allows creation of 2D sections from 3D simulations. It consists of a series of Fortran subroutines and functions that is easily adapted and can be viewed in various programs such as BlueKenue. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics Amplitude in terms of peak wave height is output. In addition, Many n.a. ASCII, SELAFIN frequency and direction for mean and peak are output. Wave-induced currents are also calculated at each point of the mesh. Description of Software Risk Outputs This set of solvers unfortunately does not follow through to risk; however, the hazard calculation could be integrated directly into risk assessments. The currents, as well as frequency and duration outputs, are useful. Tomawac output of waves. Advantages and Disadvantages ✓ Validation is a key parameter checking the output. The software model has been ratified by IAHR, and there is documentation of these validation cases. It provides another useful solver. ✓ The software provides all the tools necessary for wave modelling via solving equations. ✓ The code is in Fortran and easily downloadable for use in any application. ✓ There is quite a large user community. ✕ This is not for basic users, and much manipulation is required in order to use the software for applications. ✕ There is no risk calculation within the software. Recommended Improvements for Greater Utility The tool needs to be examined in greater depth; however, it definitely offers a very useful set of libraries and tools to integrate into wave modelling. It has natural synergies with any of the risk modules for tsunami, such as Hazus and CAPRA. 43 44 TsuDAT using ANUGA Software Name Peril License Current Version Open Source Operating Systems TsuDAT Tsunami GPL, CC by AU V2.0 Yes Linux, Windows (Client) Preferred Specific Information Coding Language Software Modules Manual GUI Help Python TsuDAT, GeoNode, TsuDAT Client Y* Y Y Goal and Summary of the Software TsuDAT is a calculation software for tsunami inundation, based on the Probabilistic Tsunami Hazard Assessment by Geoscience Australia, which allows many individual scenarios to be created from 76,000 different modelled synthetic tsunamis. The hazard is calculated at a point of 100m depth offshore around Australia. The hazard can then be disaggregated for an event with a defined height and return period, giving a table of results. The scenario is then run using the hydrodynamic model, ANUGA, in order to solve the shallow water wave equations over the DEM topography and bathymetry data. Offshore time series and ANUGA scripts can be edited, meaning that any inundation code can be used. TsuDAT has a variable mesh resolution and friction that can be drawn as polygons by the user. File Types Used Hazard Vulnerability Exposure Key Hazard Metrics 3.4 – Tsunami/Storm Surge/Wave Loss Estimation Inundation depth, via wave height, is modelled through a finite volume *.csv into py n.a. *.shp, *grd method mesh, solving a 2D shallow water wave equation with respect to momentum, roughness, boundary, and forcing conditions. Description of Software Risk Outputs The software presents an inundation map of the location that the user uploads as a DEM. For an engineer or a decision-maker, it is a very good tool for creating a map. TsuDAT Client also allows the user to analyze the given scenario through the NCI (a cloud computing system in Canberra) and then receive the results back. Currently, no risk model is attached, though a couple of test cases have been undertaken internally; on the other hand, the software is extendable and the inundation map could be uploaded to InaSAFE or analyzed within another environment of flood inundation modelling, potentially NoFDP IDSS or Kalypso. TsuDAT picture of Batemans Bay inundation (OpenGeo 2013). It should be noted that testing TsuDAT fully was not possible, as it was being updated during this study. Advantages and Disadvantages ✓ The software’s tsunami hazard methodology is the best of all wave software packages reviewed. ✓ The methodology for disaggregating a suitable scenario is better than that of any other package. ✓ The ability to “plug and play” the code with any hydrodynamic model is useful. ✓ The use of GeoNode and fully open source systems is useful. ✕ The code can be looked at locally, but running TsuDAT requires a connection to the server at the NCI. Thus if Internet connections are not working, analysis cannot take place. Recommended Improvements for Greater Utility TsuDAT should be promoted as much as possible and seems the best candidate to be combined with other software. The system for scenario selection and disaggregation lends itself to a MAEviz system. It should be implemented in InaSAFE to allow risk outputs as well as in other modules for tsunami globally. 4.0 – Discussion and Conclusion The information presented here is intended to guide There is a potential for valuable synergy between Supplementary users in selecting suitable software packages. Users existing software packages. A number of open source Spreadsheet can draw on additional information (included in the software packages could be combined to generate https://www.gfdrr.org/ supplementary spreadsheet to be available online) to a multi-risk model with multiple views of a hazard. RASoftwareReview further align software selection with their specific An example is a hypothetical “super package” requirements. But it is important for users to that combined MAEviz, EQRM, Kalypso, Deltares, actually test software packages in order to make TCRM, and TsuDAT (but the options are limitless). informed decisions about which is best for their One goal of this review has been to provide a purposes. platform for dialogue between all open source and The findings of this review—some perhaps open access software package developers and users. surprising—are as follows: Ideally, it will also inspire collaboration between developers, who have thus far done a great job with 1. Open source software has a reputation for poor open access and open source packages. GUIs and difficult installation. But as the quality of open source software continues to improve, Typically, risk model development and software this reputation is proving unfounded. Many of the development are two distinct activities. But this software packages (Deltares tools and RiskScape, review suggests that the ideal situation is one in for example) were first developed for commercial which scientists and engineers develop the risk purposes and have advanced GUIs, and many are model—inputs, outputs, calibration—and software simple to install with a few clicks. developers work closely with this team to build efficient and user-friendly tools that are easily 2. Deterministic analysis is the most common extended and adapted to suit a wide range of function among the software packages reviewed. applications. However, probabilistic or stochastic event set modelling can be undertaken in many software packages, including OpenQuake, CAPRA, and EQRM. 3. Many software tools could be improved by enabling user-defined exposure and vulnerability. Without this capability, many tools can be used only regionally. 4. It is becoming increasingly easy to use multiple packages for a single region and/or hazard as a means of characterizing the uncertainty in the risk, or checking for the sensitivities in the analysis. 45 46 References Daniell, J. E. 2009. “Open Source Procedure for Assessment of Loss Using Global Earthquake Modelling (OPAL Project).” CEDIM Earthquake Loss Estimation Series, Research Report No. 09-01, Karlsruhe, Germany. Lucca, S., and L. Valentini. 2008. “Tsunami Risk Exercise: Geographic Information Systems (GIS).” GRASS-GIS Course Notes, Politecnico Milano, Milan, Italy. McLaren, T. 2008. “MAEVIZ.” Presentation at EclipseCON, Santa Clara, CA, March 17–20. Robinson, D., G. Fulford, and T. Dhu. 2006. “EQRM: Geoscience Australia’s Earthquake Risk Model: Technical Manual: Version 3.0.1.” Record 2005/001. Geoscience Australia, Canberra. Appendix A. References and Other Information for Software Packages Initially Assessed NAME OF SOFTWARE DEVELOPING INSTITUTION REGION DEVELOPED DOWNLOAD LOCATION / LITERATURE PACKAGE ADCIRC University of North Carolina U.S. www.adcirc.org Australian National ANUGA Australia http://sourceforge.net/projects/anuga/. University (ANU) and GA ANUGA and TsuDAT ANU and GA Australia, Indonesia https://github.com/AIFDR/tsudat2. Institut de recherche pour le ATHYS France http://www.athys-soft.org/v1/Index.html. développement (IRD) http://www.basement.ethz.ch/services/download/ BASEMENT ETH Zurich (ETH-Z) Switzerland box_feeder/BASEMENT_v2.2.1.zip. World Bank, United Nations http://www.ecapra.org/sites/default/files/ CAPRA (ERN-Flood, ERN- Office for Disaster Risk Central America softwares/Setup_ERN-Inundaci%C3%B3n%20 Lluvia) Reduction, GFDRR v2.1_100812.rar. CAPRA (ERN-Hurricane) Central America http://www.ecapra.org/ERN-Hurricane. CAPRA (Earthquake, ERN- Surge, ERN-Tsunami via Central America http://www.ecapra.org/crisis-2007. CRISIS2007) Federal Emergency CATS Management Agency (FEMA), U.S. https://www.saic.com/products/security/cats/. ESRI Champ2.0 FEMA U.S. http://www.fema.gov/. ComMIT (MITgcm) MIT U.S. http://nctr.pmel.noaa.gov/ComMIT/. http://isec.nacse.org/models/coulwave_download. COULWAVE Cornell University, Texas A&M U.S. php. Tyagunov, S. 2006. “CREST Software.” Karlsruhe, Center for Disaster Germany; Tyagunov, S., G. Grünthal, R. Wahlström, Management and Risk CREST Germany L. Stempniewski, and J. Zschau. 2006. “Seismic Reduction Technology Risk mapping for Germany.” Natural Hazards Earth (CEDIM) Systems Science 6: 573–86. DBELA EUCENTRE Italy https://github.com/VSilva/DBELA. http://oss.deltares.nl/web/delft3d/source-code;jses Delft-3D-FLOW, -WAVE Delft (Deltares) Worldwide sionid=3A3C93DEEAE13D06A0FCAE8A7D39E5D6. U.S. Geological Survey DR3M U.S. http://water.usgs.gov/software/DR3M/. (USGS) http://www.koeri.boun.edu.tr/depremmuh/eski/ ELER JRA-3, NORSAR, Imperial Europe ELER/eler_dvd.rar. Eguchi, R. T., Goltz, J. D. , Seligson, H. A. Flores, P. J. N., Blais, C. , Heaton, T. H. and Bortugno, E. 1997. “Real-Time Loss Estimation as an Emergency EPEDAT EQE International California, U.S. Response Decision Support System: The Early Post- Earthquake Damage Assessment Tool (EPEDAT).” 47 Earthquake Spectra 13 (4): 815–32. 48 NAME OF SOFTWARE DEVELOPING INSTITUTION REGION DEVELOPED DOWNLOAD LOCATION / LITERATURE PACKAGE Robinson, D., Fulford, G. and Dhu, T. 2006. “EQRM: EQRM Matlab GA Australia Geoscience Australia’s Earthquake Risk Model.” Record 2005/01, Geoscience Australia, Canberra. EQRM Python GA Australia http://sourceforge.net/projects/eqrm/files/. Markus, M., Fiedrich F., Leebmann, J., Schweier, C. & Karlsruhe Institute of Steinle, E. 2004. “Concept for an Integrated Disaster EQSIM Germany Technology (KIT) Management Tool.” Proceedings of the 13th World Conference on Earthquake Engineering, Vancouver. Frolova, N., Larionov V. & Bonnin, J. [2006]. “Multi- Hazard Risk Assessment at Different Levels Extreme. Situations Extremum Russia with Extremum System Application”, The Paper Research. Center. Ltd. presented at Third International Conference on Early Warning, Bonn, Germany. http://water.usgs.gov/nrp/gwsoftware/gsflow/ GSFLOW USGS USAU.S. gsflow.html. National Center for Research Yeh, C.H., Loh, C.H. & Tsai, K.C. [2006]. “Overview of HAZ-Taiwan (TELES) on Earthquake Engineering Taiwan Taiwan Earthquake Loss Estimation System”, Natural (NCREE) Hazards, Vol. 37, No. 1-2, pp. 23–37. HAZUS Flood, Earthquake, FEMA U.S. http://www.fema.gov/hazus. Hurricane http://www.hec.usace.army.mil/software/hec-ras/ HEC-RAS USA Army Corps of Engineers U.S. downloads.aspx. Instituto scienze della Terra–Scuola universitaria https://svn.osgeo.org/grass/grass-addons/grass6/ HydroFOSS Worldwide professionale della Svizzera HydroFOSS/. italiana (IST-SUPSI) National Technical University Hydrognomon Greece http://hydrognomon.org/download.html. (NTU) Athens InaSAFE Earthquake, Flood, AIFDR Indonesia https://github.com/AIFDR/inasafe. Tsunami InLET ImageCat, Inc. California, U.S. http://rescue-ibm.calit2.uci.edu:8888/inlet/inlet.php. Hamburg University of Kalypso Technology and Bjoernsen Germany http://sourceforge.net/projects/kalypso/. Consulting Engineers Sousa, M. L., Campos Costa, A., Carvalho, A. & Coelho, E. [2004]. “An Automatic Seismic Scenario Laboratório Nacional de Loss Methodology Integrated on a Geographic LESSLOSS/LNECLOSS Engenharia Civil (LNEC), Lisbon Information System.”, Proceedings of the 13th World Consortium Conference on Earthquake Engineering, Vancouver., Canada, Paper No. 2526. http://earthquake.ncsa.illinois.edu:8080/ MAEViz University of. Illinois U.S. release/3.1.1/maeviz-3.1.1.setup.exe. mHARP (MAEviz and Eqvis University of. Illinois, and U.S. http://mharp.ncsa.illinois.edu/?page_id=48. combined) collaborators NAME OF SOFTWARE DEVELOPING INSTITUTION REGION DEVELOPED DOWNLOAD LOCATION / LITERATURE PACKAGE http://siteresources.worldbank.org/EXTDISMGMT/ MIRISK GFDRR Grant Japan Resources/MIRISkprofScawthorn.pdf. noFDP IDSS with SOBEK- Darmstadt University. / Germany as part of http://nofdpidss.sourceforge.net/. River (1D Hydraulic Model) Collaborators INTERREG III OpenQuake GEM Italy https://github.com/gem/oq-engine/. http://www.risk-agora.org/index.php/files/ OpenRISK SpaRisk LLC Worldwide category/4-openrisk-software. OsGEO Tsunami (R.tsunami) IST-SUPSI Italy http://svn.osgeo.org/grass/grass-addons/. http://www.risk-agora.org/index.php/files/ OSRE Kyoto University Japan category/1-osre. PAGER USGS U.S. earthquake.usgs.gov/pager. PIHM and GIS and QGIS Penn State U.S. http://www.pihm.psu.edu/pihm_downloads.html. plug in QLARM-WAPMERR WAPMERR/Max Wyss Switzerland www.wapmerr.org. http://www.fema.gov/library/viewRecord. Quick20 FEMA U.S. do?id=2316. http://grasswiki.osgeo.org/wiki/AddOns/ R.Hazard.Flood M. D. Leo, M.D. Italy GRASS_6#r.hazard.flood. United Nations Environment http://worldbank.mrooms.net/course/view. RADIUS India Programme (UNEP) php?id=483&pageid=3053. REDARS MCEER, ImageCat, Inc. California, U.S. ftp://mceer.buffalo.edu/Stoyle/redars/. REDAS PHIVOLCS Philippines www.phivolcs.dost.gov.ph/images/IEC/redas.pdf. Centre for Ecology & ReFH Hydrology (CEH), and United Kingdom http://www.ceh.ac.uk/Feh2/FEHSoftware.html. Wallingford Solutions Sinha, Ravi, Aditya K. S. P. & Gupta A. 2008. “GIS- Based Urban Seismic Risk Assessment Using RISK. Risk.IITB IIT Bombay India IITB.” Journal of Earthquake Technology 45 (3–4): 41–63. National Institute of Water RiskSCAPE Earthquake, and Atmospheric Research New Zealand http://www.riskscape.org.nz/. Flood, Tsunami, Wind (NIWA), and GNS Science Sinclair Knight Merz (SKM), http://www.eng.monash.edu.au/civil/research/ RORB Australia Monash University centres/water/rorb/. http://www.fema.gov/library/viewRecord. Runup v2 FEMA U.S. do?id=3392. Fast based SELENA – Zschau, J., Gasparini, P., Papadopoulos, G., & SAFER Consortium . [2007] “Status of the SAFER Project (Seismic eArly warning SAFER Multiple EU partners Europe For EuRope).”, Paper presented at American Geophysical Union (AGU) Fall Meeting, San Francisco. 49 50 NAME OF SOFTWARE DEVELOPING INSTITUTION REGION DEVELOPED DOWNLOAD LOCATION / LITERATURE PACKAGE Anagnostopoulos, S., Providakis, C., Salvaneschi, P., Athanasopoulos, G. & Bonacina, G. (2008). University of Patras, and “SEISMOCARE: An Efficient GIS Tool for Scenario- SEISMOCARE Greece Consortium type Investigations of Seismic Risk of Existing Cities.” Soil Dynamics and Earthquake Engineering, 28 (2), 73–84. SEISVARA NORSAR India http://eqrisk.info/seis.php. SELENA NORSAR Norway http://sourceforge.net/projects/selena/files/. Roca, A., Goula, X., Susagna, T., Chàvez J., Gonzàlez, M. J. & Reinoso, E. [2006]. ‘‘A Simplified Method for Director General for Dir. Of SES2002 and ESCENARIS Spain Vulnerability Assessment of Dwelling Buildings and Civil Protection Estimation of Damage Scenarios in Spain’’, Bulletin of Earthquake Engineering, Vol. 4, No. 2, pp. 141–158. Di Pasquale, G., Ferlito, R., Orsini, G., Papa, F., Pizza, A. G., Van Dyck, J. & Veneziano D. [2004]. “Seismic Italian National Seismic Scenario Tools for Emergency Planning and SIGE Italy Survey (OSSN) Management.”, Paper presented at the 29th XXIX General Assembly of the European Seismological Commission, Potsdam, Germany. National Oceanic and SLOSH Atmospheric Administration U.S. http://slosh.nws.noaa.gov/sloshPriv. (NOAA) http://www.deltares.nl/en/software/108282/sobek- Sobek Suite – 1D, 2D Deltares Netherlands suite. Borzi, B., Crowley, H. & Pinho, R. [2008]. “Simplified Pushover-based Earthquake Loss Assessment (SP- SPBELA EUCENTRE Italy BELA) Method for Masonry Buildings.”, International Journal of Architectural Heritage, Vol. 2 (No. 4) pp. 353-376. Kandilli Observatory and Earthquake Research Institute. 2002. “Earthquake Risk Assessment for Middle East Technical Istanbul Metropolitan Area.”, Kandilli Observatory StrucLoss/KOERILoss Turkey University (METU) and Earthquake Research Institute, Istanbul, available from URL: http://www.koeri.boun.edu.tr/ depremmuh/EXEC_ENG.pdf. SWAN Delft (Deltares) Netherlands http://swanmodel.sourceforge.net/. Environmental Protection http://www.epa.gov/nrmrl/wswrd/wq/models/ SWMM U.S. Agency (EPA) swmm/. http://www.vce.at/syner-g. Contact EU project SYNER-G (Eqvis) Consortium Europe consortium partners for Eqvis and OOFIMS (available on website). http://hydrology.usu.edu/taudem/taudem5.0/ TauDEM Utah State University (USU) U.S. downloads.html. TCRM Geoscience Australia Australia http://code.google.com/p/tcrm/. NAME OF SOFTWARE DEVELOPING INSTITUTION REGION DEVELOPED DOWNLOAD LOCATION / LITERATURE PACKAGE Collaboration between TELEMAC-MASCARET France http://www.opentelemac.org/index.php/download. Germany, UK, France TOMAWAC and WAVE – Collaboration between France http://www.opentelemac.org/index.php/download. TELEMAC Germany, UK, France USGS Water Resources Applications Software USGS U.S. http://water.usgs.gov/software/. (NSSv6, OWLS, PKFQ) WaSIM ETH-Z Switzerland http://www.wasim.ch/products/model_r9-2-0.htm. National Centers for http://polar.ncep.noaa.gov/waves/wavewatch/ Wave Watch 3 Environmental Prediction U.S. wavewatch.shtml. (NCEP) 51 52 Appendix B. Additional Software Packages and Links Table B-1 lists—and includes links to—software packages that were not reviewed in their entirety but will be in the future. Some were discovered too late in the review process to be included; some solve only a very simple set of equations and may not be extensive in their application to risk assessment; others were known to be only hazard engines, and were not reviewed if enough risk software packages existed in their group (e.g., earthquake). The applicable packages will be reviewed over the coming months and years, and a dynamic list for the worldwide user community will be provided in order to keep the list of potential open source risk assessment software up-to-date. Table B—1  Software SOFTWARE PACKAGE LINK Packages to Be A Java/C routine: for distributed rainfall-runoff https://code.google.com/p/jgrass/. Reviewed in the Atmospheric models from the MITGCM (which http://mitgcm.org/. Future are also to be used in the COMMIT software) Clawpack/Diego Melgar/Geoclaw https://github.com/dmelgarm https://github.com/clawpack/geoclaw. DHSVM: Distributed hydrologic model http://www.hydro.washington.edu/Lettenmaier/Models/DHSVM/ documentation.shtml. ECOMSED: 3D hydrodynamic and sediment http://www.hydroqual.com/ehst_ecomsed.html. transport computer code. FLDTA: 1D Flow http://csdms.colorado.edu/viewvc/midas/. FVCOM: A set of solvers http://fvcom.smast.umassd.edu/FVCOM. Hydrotrend http ://csdms.colorado.edu/viewvc/hydrotrend/. LBRM: Large Basin Runoff (Distributed) http://www.glerl.noaa.gov/wr/lbrmexamples.html. MoCaHAZ: Wiemer – ETH-Z : – seismic hazard http://www.seismo.ethz.ch/static/stat_2010_website/stat-website- assessment (Wiemer – ETH-Z) pre2010/www.earthquake.ethz.ch/research/Swiss_Hazard/downloads/ software_downloads.html; - contact authors. NSHMP: U.S. seismic hazard mapping program http://earthquake.usgs.gov/hazards/products/conterminous/2008/ software/. OHAZ: Probabilistic hazard for BSHAP (Balkans) www.wbbalkanmaps.org; - contact authors. OPENFOAM: A wide range of coastal solvers for http://www.openfoam.org/download/. fluid equations OpenSHA: earthquake hazard, global use, http://www.opensha.org. probabilistic Parflow: Watershed model http://inside.mines.edu/~rmaxwell/maxwell_software.shtml.b POM http://www.aos.princeton.edu/WWWPUBLIC/htdocs.pom/. PREVAH: Distributed rainfall-runoff (login) http://www.hydrologie.unibe.ch/PREVAH/. RORB v6.14 http://eng.monash.edu.au/civil/research/centres/water/rorb/. SEISRISKIII: A global earthquake hazard engine http://earthquake.usgs.gov/hazards/apps/seisrisk/seisrisk.zip. (probabilistic) Stvenant: 1D model from 1992 to 1994 http://csdms.colorado.edu/wiki/Model:STVENANT. STWAVE: Nearshore wind-wave growth http://chl.erdc.usace.army.mil/chl.aspx?p=s&a=SOFTWARE;9. SWASH: Simulating near-surface waves http://swash.sourceforge.net/. TOPKAPI: Advanced rainfall-runoff software http://www.progea.net/prodotti.php?c=Acquista&p=Acquista_TOPKAPI. Topoflow: A group of distributed hydrologic tools http://csdms.colorado.edu/viewvc/topoflow/. TSUNAMOS https://nees.org/warehouse/filebrowser/577, TUNAMI http://code.google.com/p/tunami/. WaSIM: Distributed rainfall-runoff model http://www.wasim.ch/en/products/model_r9-2-0.htm. Appendix C. Software Criterion Used in Evaluation Table C-1 provides a verbal description of the criteria for the modules used in the analysis of the 31 software packages. The criteria are not exhaustive and can be changed, added to, removed, and adjusted, to reflect the opinions of future users and software updates. The process and most of the criteria are derived from the OPAL procedure described in Daniell (2009), with additional criteria suggested during meetings with World Bank and GFDRR experts. CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE Table C—1  Description SOFTWARE ACCESSIBILITY and Score Levels for Internet based and normal software Module Criteria SA-001 Global availability 4 versions, current Not available just on Internet; SA-002 Ease of download 4 downloadable, fast, usable SA-003 Instructions for use Manual updated every few months 4 Clear methodology, results, and SA-004 Documentation 4 updating with new versions SA-005 Versions of software Windows, Mac, or Linux based 4 Email/phone contact, updated SA-006 Ability to contact developer 4 website, and open discussion Extensible, clear, public domain, all SA-007 Ease of coding 4 components Fortran, C++, web-based, Java, VB, SA-008 Coding language used XML, Excel, Python, Matlab, SOSEWIN, 4 self-organizing systems Which components of the system are SA-009 Number of open source components 4 open source and which closed source GNU GPL licensing type—e.g., GPLv3, SA-010 Licensing 4 GPLv2; reuse features No cost associated with GIS; no trial SA-011 GIS license 4 period; fully integrated into system SA-012 Other software needed Types of other software needed 4 Availability of mailing lists (user and SA-013 Mailing lists 4 developer) Number of people registered to the SA-014 Popularity mailing list, and extent to which the 4 systems used SA-015 Code versioning Latest version of the code 4 SA-016 Bugtracker Any form of bugtracker? 4 Frequency of new releases and How often has the software been SA-017 4 patches updated? SA-018 Data access Does the user have access to all data? 4 Setup of virtual communities such as 53 SA-019 Virtual Communities Matlab Central; user-added code, and 4 help. 54 CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE Possible for any user, anywhere, for personal use; checked for public use; end-user oriented, with separate SA-020 Extensible architecture 4 documentation available for those wishing to modify or extend the tools or leverage the APIs GRAPHICAL USER INTERFACE GU-001 GUI presence Does the software have a GUI? 4 GU-002 GUI quality Quality of the GUI (subjective rating) 10 GU-003 GUI help Help 4 SOFTWARE DETAILS Online updates available when online; SW-001 Integration with Internet 4 user community page also current Wiki-style interface allows users to SW-002 Open editing (Wiki) update code, leave ideas, and update 4 list of and/or fix bugs in software Testing for the various optimization/ SW-003 Computation speed 4 minimization of coding English used for the analysis, SW-004 Global language translatable, simple architecture; 4 globally available software language Functions are hardwired in many SW-005 Hardwiring 4 cases into the coding—SELENA, etc. Not a wiki update but a direct system SW-006 Adaptability (allow additions) to handle bugs; email-associated with 4 a user number Able to use new versions of flood SW-007 Recent developments 4 hazard, vulnerability modules, etc. Update is not via DVD; email user list SW-008 Update method (versions) 4 exists; updates downloadable Use of more than one method to SW-009 Hybrid methods adapt to different situations (multiple 4 level systems of all risk components) Collaboration, or previous models; SW-010 Reliance on past methods 4 inherent decision-making Relevance to current conditions and SW-011 Current application 4 population/settlement data Inclusion of development status and SW-012 Development status 4 updated list of bugs fixed SW-013 Updated recently How recently updated? 4 Level of hardware required, GHz, GB, SW-014 Computing power required 4 and method SW-015 Licensing Free access installation 4 Windows, Linux, Unix, Mac, Vista, SW-016 Global testing under all platforms Windows95, and various service 4 packs; no need to change hardware CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE Learning and tutorials provided for SW-017 Learning and tutorials 4 user awareness Software optimized with no slow parts SW-018 Optimization 4 and user warnings provided Integration with user-based data SW-019 Data plotting 4 plotting; many options Backup systems in place for loss of SW-020 Backup systems during analysis 4 data, more data sets, etc. User knowledge check, advanced and SW-021 Ease of use 4 normal settings Run-through explaining methodology, SW-022 Tutorial-based loss analysis 4 to the final results TECHNOLOGICAL ASPECTS GIS closed or open source (open TE-001 GIS licensing 4 source does not always use open GIS) Integrated GIS, or output in GIS TE-002 GIS production format; use of GIS for exposure 4 selection Technology updated; updates applied TE-003 Dynamic improvement 4 to software package % accuracy in terms of height, building type, GIS systems, and other TE-004 Remote sensing accuracy methods; what methods used; how 4 well have they been improved and checked? Updating of systems in place to apply TE-005 Rapid response/real data real data to software then rapid 4 response EXPOSURE COMPONENTS EX-001 Test locations used Location of test sites 4 Method of collection, collaboration, EX-002 Collection 4 reliance on technology, etc. % global population covered; EX-003 Global cover building type, cost, age of building, 4 demographics, etc. Is portfolio analysis allowed; can EX-004 Portfolio analysis 4 groups of buildings be input? What is the use of the elements and EX-005 Inventory elements 4 their importance? Damage modelling and risk analysis EX-006 Critical + lifelines 4 possible for most important elements EX-007 Format of collected material Format of the collected data 4 Is there an existing taxonomy to be EX-008 Ontology-taxonomy 4 adhered to? 55 56 CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE Temporal changes to the buildings recorded and satellite images; EX-009 Temporal data 4 retrofitting history, seismic code (if any) adhered to Spatial changes able to be recorded EX-010 Spatial data 4 and collected Accuracy of the exposure calculation EX-011 Location 4 at the location being tested Number of bins used and able to be EX-012 Accuracy of typology 4 used to maintain computational speed Not set and hardwired; new building EX-013 Allow addition of building types 4 types able to be calculated Accuracy of the population at any time EX-014 Population assessment 4 of day, demographics Exposure data used to identify risk EX-015 Risk indicators 4 indicators Ease of data access, time taken to EX-016 Data management 4 access, easy storage method Inclusion of costs or human EX-017 Socioeconomic 4 components RISK COMPONENTS Direct, Indirect, analysis, GNP basis, RK-001 Economics (level of layers) 4 repair, MDR Deaths, levels of injuries, homeless, RK-002 Social (level of layers) 4 shelter needs, etc. Complexity of calculations from RK-003 Complexity 4 damage-loss conversion RK-004 Output accuracy Calibration with real disasters 4 Extent to which dynamic changes RK-005 Dynamic vulnerability of vulnerable regions are taken into 4 account in equations Inclusion of age, demographics, RK-006 Social vulnerability community awareness programs 4 included in the analysis and output Age and use of the equations for RK-007 Age of equations social and economic costing (new 4 data, new accuracy) Extent to which uncertainty is RK-008 Uncertainty constrained; use of logic tree 4 approaches Visualization via GIS or graphs of RK-009 Visualization 4 comparative scenarios Logic tree approach employed or RK-010 Decision possibilities 4 expert opinion setup Consideration of single building and Single or portfolio—risk insurance/ RK-011 portfolio of buildings for insurance 4 financing purposes CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE Possibility of disaggregating economic RK-012 Disaggregation 4 data RK-013 Cost-benefit analysis Possibility of cost-benefit analysis 4 GIS layering and scenario overlaying RK-014 Use for land planning and zoning 4 to help governments VULNERABILITY COMPONENTS Use of empirical methods in one level VL-001 Empirical ability 4 or more Intensity/inundation depth or flow; VL-002 Fragility function parameters sediment transport/speed of wind/ 4 wave height Ease with which vulnerability damage VL-003 Rapid response is constrained exactly for rapid 4 response VL-004 Correlation with damage Damage state correlation 4 VL-005 Analytical method used Use of analytical methods 4 Influence on knowledge and coding VL-006 Complexity of algorithm 4 complexity Specific damage states used; extent VL-007 Damage states 4 to which they are well defined Extent to which uncertainty is well VL-008 Uncertainty 4 construed Level on which location data are VL-009 Location data 4 analyzed Universal application for vulnerability VL-010 Can this be applied globally? 4 functions Methods used to pick up vulnerability VL-011 Ability to detect indicators 4 indicators VL-012 Occupancy Use and occupancy type and rate 4 Basic structural features (materials used, irregularities in plan and VL-013 Structural elevation, building height and internal 4 characteristics, complex failure mechanisms) Age and number of age brackets, VL-014 Quality of stock correlation to disaster codes, 4 variability Probabilistic techniques and VL-015 Material variability/building differences distributions to quantify building 4 variability Presence of pictures and GIS of VL-016 Pictures/GIS vulnerable houses via screening 4 methods, etc.; panoramic/street view Additional non-structure-based Nonstructural, contents, business VL-017 4 damage elements interruption VL-018 Aggregation of data Data aggregated to create a result? 4 57 58 CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE POST-EVENT ANALYSIS Maps produced or not? Collaboration PS-001 Maps with disaster managers and event loss 4 analysts? Use of lifeline data and the forecast to PS-002 Direct management implement directly into management 4 after an event Technological tools used before and PS-003 Disaster management tools 4 after flood as part of the system Speed from source to an output loss PS-004 Speed of calculation 4 estimate Comparison with previous disaster PS-005 Accuracy 4 events in terms of loss Use of screening methods, before PS-006 Optimization of building choice and after; presence of disaster 4 management at most at-risk areas Ease with which information is PS-007 Functionality/communications transferred to relevant people (army, 4 government) Output of topography and GIS data for PS-008 Topography/GIS 4 use on the ground In tune with the hazards looked at; PS-009 Consecutive event modelling 4 warning system production FORECASTING Speed for a certain large-scale FC-001 Speed 4 calculation Large-scale accuracy based on FC-002 Accuracy 4 comparison with past floods Availability for use in post-disaster FC-003 Database of tested scenarios analysis (can apply quick previous 4 estimates) OUTPUTS OF THE SOFTWARE Format easily visible within the GIS OU-001 GIS 4 program in the package Similar to Onepager but with more OU-002 One-page summary features (optimization visualization, 4 uncertainties) Simplicity and ease of application OU-003 Ease of reading results/format for use in Open Office Math; results 4 already graphed Event loss tables, year probabilities, OU-004 Loss statistics annual rate of exceedance, etc.; 4 primary and secondary uncertainties Presence of view and analysis for OU-005 Statistics module 4 comparing statistics OU-006 Rerun module Ease of rerunning the analysis 4 CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE HAZARD – FLOOD Possibility of using station data in the HF-001 Station data 4 calculation HF-002 Numerical model Type of numerical modelling used 4 Types of dimensional wave equations HF-003 1D/2D/3D modelling 4 solved HF-004 Empirical data (raster measured data) Use of empirical data for analysis 4 HF-005 Theoretical data (modelled) Use of theoretical data for analysis 4 HF-006 Ability to update hydrographs Possibility of updating hydrographs 4 Ways in which data can be HF-007 Disaggregation 4 disaggregated Rainfall, riverine, and coastal hazards HF-008 Rainfall/riverine/coastal 4 accounted for HF-009 Link to storm surge Storm surge accounted for 4 Use of intensity-frequency-duration HF-010 Intensity-frequency-duration 4 curves Are zoning systems for flood inbuilt, HF-011 Zoning systems for flood inbuilt and can flood control structures be 4 accounted for? Possibility of producing stochastic HF-012 Stochastic catalog production 4 catalogs Specific optimization/minimization HF-013 Optimization techniques 4 techniques used Possibility of using historical flood HF-014 Historical flood event databases 4 event databases Digital elevation model use/ Approach to accounting for HF-015 4 topography use topography Calculation of hazard on floodplain HF-016 On floodplain vs. off floodplain hazard 4 only, or off floodplain as well Water speed (flow, velocity) taken into HF-017 Water speed analysis 4 account Possibility of accounting for sediment HF-018 Sediment transport 4 transport HF-019 Model resolution What is the model resolution? 4 HF-020 Rainfall model (resolution and type) Types of rainfall models used 4 Catchment (simple scaling and Types of catchment that can be HF-021 4 transposition OR rainfall-runoff) modelled Type of rainfall-runoff model: HF-022 Rainfall runoff (distributed or lumped) distributed (spatially variable) or 4 lumped (spatially non-variable) River reach models (channel-flow Type of river reach models that can HF-023 routing models or hydrodynamic river 4 be used models) Model’s use of single studies vs. HF-024 Probabilistic and/or deterministic 4 calculation of event return periods 59 60 CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE Possibility of undertaking entire HF-025 Location specificity process in any location vs. necessity of 4 undertaking process in fixed location Possibility of taking snow, etc. into HF-026 Snow and other forms of precipitation 4 account Possibility of undertaking flood control HF-027 Flood control input/secondary effects 4 input HF-028 Type of hazard files Types of hazard files used; ease of use 4 Possibility of accounting for climate HF-029 Temperature/climate change 4 change and temperature HAZARD – WIND / STORM / HURRICANE Possibility of using station data in the HS-001 Station data 4 model HS-002 Numerical model Type of numerical modelling used 4 Possibility of accounting for 1D, 2D, 3D HS-003 1D/2D/3D modelling 4 modelling Empirical data (raster measured data Possibility of using empirical data in HS-004 4 from meteorology centers) the modelling Possibility of producing synthetic/ HS-005 Synthetic theoretical data (modelled) 4 theoretical storms HS-006 Surge Consideration of surge 4 HS-007 Hail Possibility of modelling hail 4 Possibility of using tornados and HS-008 Tornado/lightning 4 lightning HS-009 Pressure Possibility of applying pressure 4 Peak wind speed or multi-wind-speed- Use peak wind speed vs. multiple wind HS-010 4 based speeds HS-011 Location and topographic data Specific location and topographic data 4 Model’s use of radius of maximum HS-012 Radius of maximum winds 4 winds HS-013 Modelling details Specific modelling details? 4 Possibility of modelling deterministic HS-014 Deterministic 4 storms Possibility of accounting for HS-015 Probabilistic 4 probabilistic modelling HS-016 Local effects or not Types of local effects accounted for 4 HS-017 Spatial resolution of the wind model Spatial resolution of the wind model 4 HS-018 Temporal resolution of the wind model Temporal resolution of the wind model 4 HS-019 Use of historical events Possibility of using historical events 4 HS-020 Use of severity Indices Types of severity indices used 4 HS-021 Format of hazard file Specific format of hazard file 4 HAZARD – EARTHQUAKE AND ASSOCIATED EFFECTS Use of spectrum-based technique vs. HE-001 Spectrum-based vs. intensity 4 intensity-based method vs. hybrid CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE Possibility of users defining HE-002 User-defined earthquakes and events 4 earthquakes or GM HE-003 Magnitude/location Type of magnitude and location used 4 Use of past equations or real-time HE-004 Observed GMs 4 GMs to produce GMs HE-005 Empirical GMs GMPEs and their global application 4 Theoretical modelling of GMs produced; level of theoretical HE-006 Theoretical GMs/basis 4 modelling of source, path, and site effects Possibility of user applying regional HE-007 Ability to update GMs 4 GMs and new GMs HE-008 Seismic source types available Types of seismic sources analyzed 4 Probabilistic techniques used; Probabilistic seismic hazard analysis HE-009 Poissonian or time dependent, 4 (reliance) multiple options Either observed or deterministic Deterministic seismic hazard analysis HE-010 techniques used; possible logic tree 4 (reliance) applied to parameters Possibility of disaggregating data into HE-011 Disaggregation components (magnitude-distance- 4 epsilon) Inclusion of uncertainty in hazard HE-012 Uncertainty in hazard parameters 4 parameters Spatial and temporal correlation HE-013 Spatial or temporal correlation 4 between GMs possible Application of stochastic catalogs for HE-014 Stochastic catalogs 4 completeness Demand data and other optimization/ HE-015 Optimization techniques 4 minimization Extent to which historical catalogs HE-016 Historical earthquake catalogs 4 are used Extensiveness of active fault HE-017 Active fault database database; fault history in terms of 4 earthquakes Possibility of using instrumental data HE-018 Instrumental earthquake catalog 4 and applications Extent of the site classification HE-019 Soil database/site class scheme used + conversion system 4 used to final ordinates Quality and accuracy, collection, code types and data, transfer and HE-020 Geodetic data standards 4 exporting, metadata, and accessibility for GPS and GIS Liquefaction taken into account, HE-021 Liquefaction 4 intensity based, detailed or not 61 62 CRITERION CODE MAIN CRITERION DESCRIPTION OF CRITERION TOTAL VALUE Fault rupture taken into account, GIS HE-022 Fault rupture 4 based, detailed or not Landslides and slope stability taken HE-023 Landslide/slope stability into account, empirical vs. analytical, 4 detailed or not Tsunamis, links to other software, HE-024 Tsunami (linking into other software) 4 basis, details, etc. HE-025 Fire Fire analysis 4 Other: aftershock + volcano + quake Aftershocks, volcanoes, quake lakes HE-026 4 lakes taken into account (basis and details) HE-027 Format of hazard file Ease of using format of hazard file 4 HAZARD – TSUNAMI / STORM SURGE / WAVE Intensity metric used for tsunami or HT-001 Wave height, energy 4 storm surge Possibility of users defining HT-002 User defined earthquakes and events 4 earthquakes or GM Storm model if model is only for storm HT-003 Storm model 4 surge HT-004 Fault model Type of block modelled 4 Model type (1D, 2D, or 3D), equation HT-005 Wave model 4 used Roughness coefficients used; basis for HT-006 Roughness coefficients 4 coefficients Probabilistic tsunami hazard analysis HT-007 Probabilistic techniques used 4 (reliance) Deterministic tsunami hazard analysis HT-008 Deterministic techniques used 4 (reliance) Possibility of disaggregating data into HT-009 Disaggregation components (magnitude-distance- 4 epsilon) Application of stochastic catalogs for HT-010 Stochastic catalogs 4 completeness? Historical tsunami or storm surge Extent to which historical tsunami HT-011 4 catalogs catalogs are used Resolution defined for inundated HT-012 Coastal resolution 4 coastline HT-013 Bathymetry Specific bathymetry data needed 4 Extensiveness of the active fault; HT-014 Active fault database 4 history on fault Possibility of using instrumental data HT-015 Instrumental earthquake catalog 4 and applications Quality and accuracy, collection, code types and data, transfer and HT-016 Geodetic data standards 4 exporting, metadata, and accessibility for GPS and GIS Other: aftershock + volcano + quake Aftershocks, volcanoes, quake lakes HT-017 4 lakes taken into account (basis and details) HT-018 Format of hazard file Computer format 4 Appendix D. Summary of Software Packages in Initial Assessment A list of all software packages used in the assessment and the results of the initial review are provided in table D-1. The table categorizes the packages by the four peril categories, indicates whether the software package was selected for more detailed analysis, and provides a quick synopsis of the package’s features. The name of each software package is hyperlinked to the software’s URL. (There are a number of other packages for flood and wave that were not included in this table; links for these are listed in appendix A.) The software packages selected by the initial review are listed in table D-2. Further analysis of the packages used subjective criteria that accounted for the packages’ outputs, the hazard and risk they addressed, and their Table D—1 accessibility. Table D-2 highlights in green the packages selected for detailed review. Summary by Hazard Although the inland flood model RORB is listed in Table D-1, it was not included in the assessment and should be Group of Software looked at in the future. The source code is not available; however, it is state of the art in terms of rainfall-runoff Packages Included in simulations used to calculate flood hydrographs from rainfall. It allows for distributed, nonlinear modelling and the Analysis can work in urban and rural catchments. Also, a new Australian rainfall and runoff model, AR&R, is coming out soon and should be examined in the future. TYPE OF SOFTWARE TYPE OF SOFTWARE DOCUMENTATION OPEN ACCESS OR SOFTWARE CODE DOWNLOADABLE CLOSED ACCESS PROGRAMMING INSTITUTION AVAILABLE? LANGUAGE LOCATION PACKAGE CHOSEN ACTIVE CAPRA Yes Earthquake YES Open YES YES YES Visual Basic.NET World Bank Central America CATS No Earthquake NO Closed YES YES NO ESRI ArcView FEMA, ESRI U.S. ArcGIS, ArcObject, CREST No Earthquake YES Closed NO YES YES CEDIM Germany VBA DBELA No Earthquake YES Open YES YES YES Matlab EUCENTRE Italy JRA-3, NORSAR, ELER Yes Earthquake YES Open YES YES NO Matlab Europe Imperial Windows-based, EPEDAT No Earthquake NO Closed Half YES NO EQE International California, U.S. Mapinfo EQRM No Earthquake NO Open YES YES YES Matlab GA Australia Matlab EQRM Yes Earthquake YES Open YES YES YES Python GA Australia Python EQSIM No Earthquake NO Closed NO YES NO C++, xmf KIT Germany Windows-based, Extreme Situations Extremum No Earthquake NO Closed YES YES NO Russia GIS Research Center Ltd. HAZ-Taiwan Microsoft Visual No Earthquake NO Closed YES YES NO NCREE Taiwan (TELES) C++ and MapInfo 63 64 TYPE OF SOFTWARE TYPE OF SOFTWARE DOCUMENTATION OPEN ACCESS OR SOFTWARE CODE DOWNLOADABLE CLOSED ACCESS PROGRAMMING INSTITUTION AVAILABLE? LANGUAGE LOCATION PACKAGE CHOSEN ACTIVE Hazus-MH Yes Earthquake YES Open YES YES NO VB6, C++, ArcGIS USGS U.S. InaSAFE Yes Earthquake YES Open YES YES YES Java, QGIS plugin AIFDR Indonesia InLET No Earthquake NO Closed YES YES NO js, Windows ImageCat, Inc. California, U.S. LESSLOSS/ No Earthquake NO Closed NO YES NO Fortran LNEC, Consortium Lisbon LNECLOSS EclipseRichClient, MAEviz Yes Earthquake YES Open YES YES YES Uni. Illinois U.S. Geotools EclipseRichClient, mHARP Yes Earthquake YES Open YES YES YES University of Illinois U.S. Geotools MIRISK No Earthquake YES Closed NO YES NO HTML, Java, PHP Kyoto University Japan OpenQuake Yes Earthquake YES Open YES YES NO Python, Java GEM Italy Object-oriented, OpenRISK No Earthquake YES Open YES YES YES SpaRisk LLC Worldwide Web, GUI Windows-based OSRE No Earthquake YES Open NO YES YES Kyoto University Japan GUI, Java PAGER No Earthquake NO Closed YES YES NO Matlab, unknown USGS U.S. QLARM- Internet-based, No Earthquake NO Closed YES YES NO ETHZ and WAPMERR Switzerland WAPMERR Java, PostgreSQL RADIUS No Earthquake YES Closed NO YES NO Excel UNEP India GUI Windows, REDARS No Earthquake YES Closed YES YES NO MCEER, ImageCatInc California, U.S. Basic GUI Windows, REDAS No Earthquake NO Closed YES YES NO PHIVOLCS Philippines Basic ArcGIS, ArcObject, Risk.IITB No Earthquake YES Closed YES YES YES IIT Bombay India VBA Java—GIS not RiskScape Yes Earthquake YES Open YES YES NO NIWA and GNS New Zealand needed SAFER No Earthquake NO Closed NO YES NO Same as SELENA Multiple EU Europe SEISMOCARE No Earthquake NO Closed NO YES NO GIS based, VBA Multiple EU Greece SEISVARA No Earthquake YES Open YES YES NO Excel based NORSAR India Matlab, C++ SELENA Yes Earthquake YES Open YES YES YES depending on NORSAR Norway version Visual Basic, dll SES2002 and Gen Dir. Of Civil No Earthquake NO Closed YES YES NO using MapObjects Spain ESCENARIS Protection 2.1 Visual Basic, dll SIGE No Earthquake NO Closed YES YES NO using MapObjects OSSN Italy 2.1 SPBELA No Earthquake NO Closed NO YES NO n/a EUCENTRE Italy TYPE OF SOFTWARE TYPE OF SOFTWARE DOCUMENTATION OPEN ACCESS OR SOFTWARE CODE DOWNLOADABLE CLOSED ACCESS PROGRAMMING INSTITUTION AVAILABLE? LANGUAGE LOCATION PACKAGE CHOSEN ACTIVE SPBELA No Earthquake NO Closed NO YES NO n/a EUCENTRE Italy SYNER-G No Earthquake YES Closed YES YES NO Matlab Consortium Europe (Eqvis) StrucLoss/ MapBasic and No Earthquake NO Closed YES YES NO METU Turkey KOERILoss MapInfo U.S. Army Corps of HEC-RAS Yes Flood YES Open YES YES NO Fortran originally U.S. Engineers ANUGA No Flood YES Open NO YES YES Python and C ANU and GA Australia ATHYS No Flood YES Open YES YES NO Fortran IRD France Python code BASEMENT Yes Flood YES Open YES YES NO in some parts, ETH-Z Switzerland unknown for some Windows- Champ2.0 No Flood YES Open YES YES NO interfaced Visual FEMA U.S. Basic DR3M No Flood YES Open NO YES YES Fortran 77 USGS U.S. Fortran 90 and C, GSFLOW No Flood YES Open YES YES YES USGS U.S. with GUI InaSAFE Yes Flood YES Open YES YES YES Python AIFDR Indonesia Hamburg University of Technology and Kalypso Yes Flood YES Open YES YES YES Java Germany Bjoernsen Consulting Engineers USGS (NSSv6, No Flood YES Open Some YES YES DOS-based USGS U.S. OWLS, PKFQ) DOS-based, Visual Quick20 No Flood YES Open YES YES NO FEMA U.S. Basic, Binary Spreadsheet–OS; YES and CEH and Wallingford ReFH No Flood Closed YES YES NO full software– United Kingdom No Solutions closed RiskScape Yes Flood YES Open YES YES NO Java GNS and NIWA New Zealand WaSIM No Flood YES Open YES YES NO C+ ETH-Z Switzerland HydroFOSS No Flood YES Open NO YES YES GrassGIS plugin IST-SUPSI Worldwide Linked to non-free Hydrognomon No Flood YES Open YES YES NO NTU Athens Greece libraries, unknown CAPRA (ERN- Flood, ERN- Yes Flood YES Open YES YES YES VB World Bank Central America LLuvia) R.Haz.Flood No Flood YES Open YES YES YES GrassGIS plugin, R M. D. Leo Italy 65 66 TYPE OF SOFTWARE TYPE OF SOFTWARE DOCUMENTATION OPEN ACCESS OR SOFTWARE CODE DOWNLOADABLE CLOSED ACCESS PROGRAMMING INSTITUTION AVAILABLE? LANGUAGE LOCATION PACKAGE CHOSEN ACTIVE TELEMAC- Collaboration between Yes Flood YES Open YES YES YES Fortran France MASCARET Germany, UK, France Hazus Flood Yes Flood YES Open YES YES NO VB6, C++, ArcGIS FEMA U.S. Monash University/ RORB Yes Flood YES Open YES YES NO Fortran Australia SKM C++, GIS and Sobek Suite Yes Flood YES Open YES YES YES other connecting Deltares Netherlands - 1D, 2D languages noFDP IDSS Germany with SOBEK- Yes Flood YES Open YES YES YES Eclipse, Java Darmstadt University as part of River (1D) INTERREG III TauDEM No Flood YES Open YES YES YES C++, VB USU U.S. C++, GIS and Delft-3D- Yes Flood YES Open YES YES YES other connecting Deltares Worldwide FLOW languages PIHM and GIS and No Flood YES Open YES YES YES C, C++ and QGIS Penn State U.S. QGIS plug in Fortran up to v4, Flood/ SWMM No YES Open YES YES YES now C in v5 with EPA U.S. stormwater rewrite ANUGA and Tsunami/ Australia, Yes YES Open YES YES YES Python ANU and GA TsuDAT storm surge Indonesia CAPRA (ERN-Surge, Tsunami/ ERN- Yes YES Open YES YES YES VB World Bank Central America storm surge Tsunami via CRISIS2007) Coastal surge/ RiskScape Yes YES Open YES YES YES Java GNS and NIWA New Zealand tsunami University of North ADCIRC No Storm surge NO Open YES YES NO Fortran 90 U.S. Carolina Cornell University, COULWAVE No Storm surge YES Open YES YES YES Fortran 90 U.S. Texas A&M Delft-3D- WAVE Yes Storm surge YES Open YES YES YES C++ Delft (Deltares) Netherlands (SWAN) InaSAFE Yes Tsunami YES Open YES YES YES Python AIFDR Indonesia OsGEO Tsunami Yes Tsunami YES Open YES YES YES GrassGIS plugin, R IST-SUPSI Italy (R.tsunami) TYPE OF SOFTWARE TYPE OF SOFTWARE DOCUMENTATION OPEN ACCESS OR SOFTWARE CODE DOWNLOADABLE CLOSED ACCESS PROGRAMMING INSTITUTION AVAILABLE? LANGUAGE LOCATION PACKAGE CHOSEN ACTIVE Tsunami/ Runup v2 No YES Open YES YES YES DOS-based FEMA U.S. storm surge Python and C vers SLOSH Yes Storm surge Yes Open YES YES YES NOAA U.S. 3.94 TOMAWAC Wave/storm Collaboration between and WAVE - Yes Yes Open YES YES YES Fortran France surge　 Germany, UK, France TELEMAC Wave Watch 3 No Storm surge YES Open YES YES YES Fortran NCEP U.S. SWAN Yes Storm surge YES Open YES YES YES Fortran Delft University Netherlands ComMIT No Tsunami NO Open YES YES No Java MIT U.S. (MITgcm) Hazus-MH Yes Hurricane YES Open YES YES No VB6, C++, ArcGIS FEMA U.S. RiskScape Yes Windstorm YES Open YES YES No Java GNS and NIWA New Zealand CAPRA Hurricane Yes YES Open YES YES YES VB World Bank Central America Hurricane rainfall Tropical Python and TCRM Yes cyclone YES Open YES YES YES GA Australia some C modelling 67 GFDRR 1818 H ST NW, WASHINGTON, DC 20433