Simplified Flood Risk Assessment Tool (FRAT)

GLOBAL RISK IDENTIFICATION PROGRAMME Better Risk Information for Sound Decision Making Simplified Flood Risk Assessment Tool (FRAT) Methodology and Tools GRIP UN-HABITAT IFRC Empowered lives. Resilient nations.

We encourage the free use of the contents of this report with appropriate and full citation. Acknowledgements: This Simplified Flood Risk Assessment Tool was developed by GRIP, UN- HABITAT, IFRC and the ProVention Consortium on behalf of the Global Emergency Shelter Cluster. The technical services of OYO International, Japan were contracted for the technical development of the tool. Peer review and comments from several experts and agencies during the development process are gratefully acknowledged. Disclaimer: GRIP is hosted by the United Nations Development Programme (UNDP). The information and opinions expressed in this publication do not necessarily reflect the views of GRIP, UN-HABITAT, IFRC, OYO and UNDP. This publication is available from: Global Risk Identification Programme (GRIP) Bureau for Crisis Prevention and Recovery (BCPR) United Nations Development Programme (UNDP) 11-13, chemin des Anémones, Châtelaine, CH-1219, Geneva, Switzerland http://www.gripweb.org

Better Risk Information for Sound Decision Making Simplified Flood Risk Assessment Tool (FRAT) September 2009 Empowered lives. Resilient nations.

Project Participants Coordination team Carlos Villacis (GRIP) Esteban León (UN-HABITAT) Anna Maria Selleri (IFRC) Ian O Donnell (ProVention Consortium) Technical implementation team Overall coordinator Mr. Osamu Nishii (Vice Director of OIC) Technical coordinator Mr. Fumio Kaneko (Chief Engineer of OIC) Flood analysis team Leader : Mr. Susumu Nakamura (Manager of Technical Division of OYO) Member: Dr. Mahbub Reza (Chief of Water Management Division of OIC) System team Leader : Mr. Jun Matsuo (Chief of Geophysical Division of OIC) Member: Mr. Shukyo Segawa (Chief of Earthquake Engineering Division of OIC) Member: Dr. Koichi Hasegawa (Researcher of Engineering Division of OIC) Software team Leader : Mr. Mitsuhiro Murasaki (Manager of IT Division of OYO) Member: Mr. Yoshihiko Noguchi (Researcher of IT Division of OYO) 4 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

Acronyms and Abbreviations DEM Digital elevation model DMP Disaster Management Programme FRAT Simplified Flood Risk Assessment Tool GIS Geographic information system GRIP Global Risk Identification Programme IFRC International Federation of Red Cross and Red Crescent Societies RADIUS Risk Assessment Tools for Diagnosis of Urban Areas against Seismic Disasters SRTM Shuttle Radar Topography Mission UNDP United Nations Development Programme UN-HABITAT United Nations Human Settlements Programme GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 5

List of Tables & Figures Tables Table 1: Table 2: Figures Figure 1: Figure 2: Figure 3: Figure 4: Figure 5: Figure 6: Figure 7: Figure 8: Figure 9: Figure 10: Figure 11: Figure 12: Figure 13: Figure 14: Figure 15: Figure 16: Figure 17: The 12 selected cases Example of the simulated results for inundation depths Main flow of FRAT Input 1: Target area Input 2: Topography data (DEM) An example of 50m grids in Japan (Usu-san Mt. sheet) Input 3: Precipitation data Monthly average precipitation for different climate classes Input 4: Hydrograph data Examples of typical hydrographs from around the world Hydrographs of different categories Analysis 1: Flood simulation Analysis 1: Flood simulation graphs Inundation map Result 1: Inundation maps Input 5: Demographic data Input 6: Economic indices (vulnerability) Analysis 2: Flood risk estimation Result 2: Risk (damage/loss) table 17 20 12 13 14 15 15 16 16 17 17 18 18 19 19 21 21 22 23 6 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

Table of Contents Project Participants Acronyms and Abbreviations List of Tables & Figures 1. Preface 2. Background 3. Overview of FRAT 3.1 Purpose of FRAT 3.2 Technical details 3.2.1 Target area 3.2.2 Size & number of grids 3.2.3 Size of map 3.2.4 Demographic data 3.3 FRAT outputs 3.4 Users of FRAT 4. FRAT Process 4.1 Input 1: Target area 4.2 Input 2: Topography data (DEM) 4.3 Input 3: Precipitation data 4.4 Input 4: Hydrograph data 4.5 Analysis 1: Flood simulation 4.6 Result 1: Inundation maps 4.7 Input 5: Demographic data 4.8 Input 6: Economic indices (vulnerability) 4.9 Analysis 2: Flood risk estimation 4.10 Result 2: Risk (damage/loss) table 5. Conclusions 4 5 6 8 9 10 10 10 10 10 10 10 11 11 12 12 14 15 16 18 19 21 21 22 23 24 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 7

1. Preface This Simplified Flood Risk Assessment Tool, hereinafter FRAT, was developed by GRIP, UN-HABITAT, IFRC and the ProVention Consortium under the Disaster Management Programme (DMP Nairobi/Geneva) and in the context of the Global Emergency Shelter Cluster. The technical development of the tool was contracted to OYO International, Japan. 8 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

2. Background Over the past years, climatic hazards, principally floods, have become more frequent and are associated with a significant portion of disaster-related loss and damage. Various measures exist for reducing disaster risk, which include structural measures such as for example flood levies, ocean wave barriers, earthquakeresistant construction, and evacuation shelters, and non-structural measures such as building codes, public awareness programmes etc. The extension of damage due to a flood disaster of a certain magnitude requires the urgent provision of shelter to protect the lives and well-being of the survivors and avoid a large number of casualties. Shelter provision is also the first step towards the recovery of the affected communities. Therefore, quick and efficient provision of shelter is a key factor not only after the disaster has happened, but it is before the disaster hits that emergency and recovery plans must be developed, reviewed and implemented, including shelter plans. The main objective of this initiative is to improve the capacity to predict and address future post-disaster shelter recovery needs as well as to manage information about on-going risks after a disaster and their implications for shelter planning. This improved capacity will ensure quick and efficient shelter provision after a disaster and will help to optimize resource utilization at both national and international levels. By applying and calibrating the methodology in several pilot countries using actual past events, the applicability and usefulness of the methodology will be demonstrated before its large-scale deployment and utilization. It is envisaged that in the future, the methodology will be applied in pilot locations with the main objective of gaining experience and testing the methodology for providing improved emergency shelter response. Usually, a chaotic situation at all levels that results from the physical, human, technical and economic impacts of the disaster, makes it difficult to provide quick and efficient shelter response. It makes sense therefore, to develop in advance strategies for shelter procurement, preparation and deployment. These strategies need to be based on realistic and adequate risk estimates of the potential damage that provide information not only on the magnitude of damage but also on its geographic distribution, characteristics, and causes. GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 9

3. Overview of FRAT 3.1 Purpose of FRAT FRAT is a simplified flood risk mapping tool that provides an estimation of the inundation area caused by riverine floods in plain settings. The publication describes every step of FRAT in detail and provides practical examples for each of the steps. A preliminary study on existing softwares was done and led to the compilation of several ones being available: HAZUS, ICHARM and RADIUS. The first one has the disadvantages of being only for the US, it needs a geographic information system (GIS) and internet. The second one is now being developed in Japan. RADIUS was then selected as a reference for FRAT development, as it does not need expertise, internet and GIS. FRAT is based on the Risk Assessment Tools for Diagnosis of Urban Areas against Seismic Disasters (RADIUS) and has been developed as a simplified flood risk mapping tool to easily provide an inundation depth estimation map for a flood anywhere in the world. FRAT is an easy-to-use tool, which enables anyone to create inundation estimate maps anywhere, anytime with a certain level of accuracy and among its applications is shelter planning. FRAT is an easy-to-use tool with the following advantages: - usable by non-experts; - no direct internet connection is needed; - based on non-proprietary softwares; - output will be used by GIS; - The methodology can be used by anyone, for instance using Microsoft Excel or GIS; - whether local data is available or not, the smallest amount of input or default input can be used; - user support materials are available such as trainings and guidelines, which include information such as applicability and limitations. 3.2 Technical details 3.2.1 Target area The target area should not be wider than 20km x 20km considering the width of the targeted rivers. On the contrary, when the smallest area is 1 km 2, it may be less accurate. 3.2.2 Size & number of grids The accuracy level of FRAT will correspond to the size of the grids for the analysis. Considering the condition above, the shelter size will be around 100 m 2 etc., the grid size for analysis will be less than 100m. Considering the calculation duration and input difficulties, less than 100,000 grids are the maximum. Calculation duration will be around several minutes. It may be very difficult to have users put in various types of information manually. 3.2.3 Size of map The resulting maps will be used by local governments or at public meetings. The scale of the maps will be larger than 1 to 100,000 and the paper size will be around A3 to A4. 3.2.4 Demographic data In order to estimate the flood risks, such as inundated buildings and affected population etc., demographic data are necessary and the amount of work associated with this task needs to be considered. FRAT will provide preliminary estimates of risk and damage. 10 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

3.3 FRAT outputs The final outputs of FRAT will be: - flood inundation map; - figures; - estimates of damage to buildings, agriculture and population; - FRAT manual. The results of FRAT will represent basic information for: - developing shelter plans; - emergency evacuation measures; - flood forecasting. 3.4 Users of FRAT - Local government officers of developing countries: - FRAT is expected to be a tutorial program; - estimating roughly the flood and its disaster at the site. - Users of FRAT will understand: - how to approach flood estimation; and - will know roughly how to consider shelter planning, based on location, construction of shelters, emergency evacuation measures and flood forecasting, provided by FRAT. GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 11

4. FRAT Process The FRAT process includes the following aspects: - Main flow of FRAT; - Input 1: Target area; - Input 2: Topography data (digital elevation model (DEM)); - Input 3: Precipitation data; - Input 4: Hydrograph data; - Analysis 1: Flood simulation; - Result 1: Inundation map; - Input 5: Demographic data; - Input 6: Economic index (vulnerability) data; - Analysis 2: Flood risk estimation; - Result 2: Risk (damage/loss) table. The main process is summarized in Figure 1. The process and each of its components is described below. 4.1 Input 1: Target area - Target size: - A local basin (drainage/watershed); - A basin like the Ganges would be too big (unrealistic); - Max 27km x 27km (fixed). - Grids size: - Around 90m (SRTM data); - # of Grids: 300 x 300 = 90,000; - (manual input is difficult). - Sheet size: - Scale 1:100,000 (27cm x 27cm): A3 size; - Scale 1:200,000 (13.5cm x 13.5cm): A4 size. For topographic data (altitude), Shuttle Radar Topography Mission (SRTM) data is available for all over the world with around 90m grids. Considering a total size around 20km x 20km, a grid size of 300 x 300 will be sufficient. For Figure 1: Main flow of FRAT Target area Demographic data Input 2 Input 3 DEM (topographic data) Precipitation data Input 6 Economical indices (vulnerability) Input 4 Hydrograph data Analysis 1 Flood simulation Result 1 Inundation maps Analysis 2 Flood risk estimation Result 2 Flood risks (tables) Application to shelter planning 12 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

user friendliness, size is fixed. Also, as below, for selection of the target area in the left side of the screen, the information has to be inputted. Figure 2: Input 1: Target area Index - Area name - user name - date - case # - etc. Area selection - Lat, long - Left-Top GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 13

4.2 Input 2: Topography data (DEM) Figure 3: Input 2: Topography data (DEM) - Worldwide data>default (stored) - SRTM (NASA) - 90m horizontal resolution - 1m vertical resolution - N50deg to S50deg HAZUS-MH-MR3 - provide DEM only for US - using GIS through internet - Manual input > option - Difficult - Need exact format - Vast amount of work - (grid by Grid) - No GIS usage Example of SRTM (90m grid) data Stream line/network, embankment data are difficult to collect. Since topographic data have a large size, the world should be divided into 10-20 regions with one DVD for each. For instance: America Asia Europe Africa Oceania North-East North-West Central South-North South-South East East-South South West Central Middle-East East Central West North South North-East North West Central South Australia 14 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

Figure 4: An example of 50m grids in Japan (Usu-san Mt. sheet) 4.3 Input 3: Precipitation data - Worldwide data are very limited - Exact hourly data is necessary - Though yearly, monthly, maximum daily data for some cities are available - Then skipped as default HAZUS-MH-MR3 - Input return period only - Start from discharge - Manual input > option (difficult) - Hourly precipitation data - Together with - Other data are necessary - Total duration, - Upstream area, - Stream length, distance etc. - For identifying hydrographs Figure 5: Input 3: Precipitation data Flood input data is discharge or flow volume per time. Looking at precipitation data from around the world, one can see that polar/ frigid and desert areas do not need to be targeted. Therefore, topographic data will be prepared for the range between N50 and S50 degrees latitude. GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 15

Figure 6: Monthly average precipitation for different climate classes Tropic Desert Temperate Frigid Polar 4.4 Input 4: Hydrograph data - Worldwide data are very limited - Data is available only for some rivers Each river basin has a characteristic hydrograph that specifies the flood pattern for that basin/catchment. Hydrographs are classified by the factors: Q (maximum discharge) and T (duration) Figure 7: Input 4: Hydrograph data - The following parameters will be inputted: - Q (maximum discharge) - T (duration) > default - Season (dry, medium or rainy for the Asian region) - Inlet point of hydrograph - Manual hydrograph input by users will be difficult Typical Hydrographs that are selected by users are classified into the following categories according to past examples from around the world. - Steep and short (like Japan or Flush Flood); - Gentle and long (like Europe or Asia); - Rainy and dry season (like Asia). 16 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

Figure 8: Examples of typical hydrographs from around the world Table 1: The 12 selected cases Type Steep/Short Gentle/Long Season - - - Dry Medium Rainy Q max (m3/day) 15,000 10,000 5,000 15,000 10,000 5,000 15,000 10,000 5,000 15,000 10,000 5,000 T (days) 4 4 4 15 15 15 15 15 15 15 15 15 Figure 9: Hydrographs of different categories GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 17

4.5 Analysis 1: Flood simulation Figure 10: Analysis 1: Flood simulation - Various simulation methods exist - Such as HEC, HAZUS-MH, Unit Hydrograph method, Storage Function method, Discharge Function method, Kinematic Wave method, Tank modeling etc. - These methods require precise and various data - Advanced technologies and calculation tools with expert judgments are also required - FRAT is tutorial - Even less accuracy, easy to access and easy to understand - FRAT has an analysis in flood simulation - Flood flow simulation from hydrograph data by a simple differential modeling with DEM - Without input precipitation data - Manning s roughness coefficients are set as average values - More than 20m than inlet altitude should be non-calculation area Figure 11: Analysis 1: Flood simulation graphs Altitude data are assigned as one value to each grid with 90m width. In this case, rivers sometimes cannot be identified, as in reality the width of many rivers will be represented in one grid size. Ordinary water level at a river or low land will be estimated by the discharge amount during the first day before precipitation. 18 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

4.6 Result 1: Inundation maps Figure 12: Inundation map Figure 13: Result 1: Inundation maps - Inundation map - Distribution of maximum inundation depth - Distribution of inundation depth at four time sections - Time sections will be 4 at start of raining, and every quarter duration of raining - Time for 0.5m inundation depth - Time for maximum inundation depth - Data - Maximum inundation depth at each grid - Inundation depth at a couple of time sections at each grid - Time sections are depending on the duration of hydrograph - Time for 0.5m and maximum inundation depth at each grid - Time starts with the start of precipitation - Percentages of grids for inundation depth (for risk analysis) - Data format should be considered for availability of other maps such as Google to superimpose on GIS - Index of the calculation (name etc.) - Type of discharge, season - Steep/short or gentle/long and dry or medium or rainy season - Inlet point, grid unit length GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 19

Table 2: Example of the simulated results for inundation depths These data can be shown as grids maps and can be used as GIS data. 20 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

4.7 Input 5: Demographic data Figure 14: Input 5: Demographic data To estimate damage and loss due to the estimated flood/inundation, the following demographic data are necessary: - Population - Size of population - Buildings - Critical buildings - Community, school, hospital, police etc. - Residential houses - Field & Crops - Agricultural fields e.g. upland and wet land (paddy fields) - Amount of Products - rice, wheat, corn, beans, etc. However, these data should include the exact locations and characteristics or vulnerability for inundation depth. Taking into account data availability worldwide, these are mostly unknown and difficult to input. Therefore, it is better for this input to be skipped and later, FRAT will provide the inundation results and damage rates. And after calculation, using the results of FRAT, users will estimate these damage or losses. 4.8 Input 6: Economic indices (vulnerability) Figure 15: Input 6: Economic indices (vulnerability) For estimating losses, which are defined as: Loss = number of elements * damage rate * unit cost FRAT provides the damage rate, and users will calculate the loss using the number of elements in the area and the unit cost for them. Thus this input will be default. - Target - Damage to buildings - Effects on human lives - Loss of agricultural crops - Japanese manual for economic effects estimation will be used - rice, wheat, corn, beans, etc. HAZUS-MH-MR - damage & loss estimation - for buildings, lifelines, agriculture, humans etc. - Most of the critical methodologies are in black boxes - only for the US using GIS GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 21

4.9 Analysis 2: Flood risk estimation - Damage rates - estimated according to Japanese manual - for economic effects estimation House (mainly wooden) - damage rate (%) Inundation depth (m) Mud deposit land inclination 0.25 0.5 0.75 1.5 2.5 3 <0.5 >0.5 <1/1000 3.2 9.2 11.9 26.6 58.0 83.4 1/500-1/1000 4.4 12.6 17.6 34.3 64.7 87.0 >1/500 5.0 14.4 20.5 38.2 68.1 88.8 Materials in house - damage rate (%) Inundation depth (m) 43.0 78.5 Mud deposit 0.25 0.5 0.75 1.5 2.5 3 <0.5 >0.5 damage rate (%) 2.1 14.5 32.6 50.8 92.8 99.1 50.0 84.5 Enterprise property/asset - damage rate (%) Inundation depth (m) Mud deposit 0.25 0.5 0.75 1.5 2.5 3 <0.5 >0.5 Depreciation 9.9 23.2 45.3 78.9 96.6 99.5 54.0 81.5 Stock 5.6 12.8 26.7 58.6 89.7 98.2 48.0 78.0 Agriculture/fishery property/asset - damage rate (%) Inundation depth (m) Mud deposit 0.25 0.5 0.75 1.5 2.5 3 <0.5 >0.5 Depreciation 0.0 15.6 23.7 29.7 65.1 69.8 37.0 72.5 Stock 0.0 19.9 37.0 49.1 76.7 83.1 58.0 84.5 Agricultural products - damage rate (%) Inundation depth (m) 0-0.25 0.25-1 >1 Mud deposit (m) Inundated days 1-2 3-4 5-6 >7 1-2 3-4 5-6 >7 1-2 3-4 5-6 >7 <0.5 0.5-1 >1 Damage (%) Figure 16: Analysis 2: Flood risk estimation Wet field rice (peddy) 21 30 36 50 24 44 50 71 37 54 64 74 70 100 100 Dry field rice 20 34 47 60 31 40 50 60 44 60 72 82 Potatoes 11 30 50 50 27 40 75 88 38 63 95 100 Cabbages 42 50 70 83 58 70 83 97 47 75 100 100 Kitchen vegetables 19 33 46 59 20 44 48 75 44 58 71 84 Root vegetables 32 46 59 62 43 57 100 100 73 87 100 100 Gourds 22 30 42 56 31 38 51 100 40 50 63 100 Beans 23 41 54 67 30 44 60 73 40 50 68 81 Average (field) 27 42 54 67 35 48 67 74 51 67 81 91 68 81 100 - Losses - estimated using economic index for local area contributed by users 22 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

4.10 Result 2: Risk (damage/loss) table Based on the damage vulnerability information from above and FRAT inundation results, users can easily calculate damages (see below). And together with the unit cost for each location, losses can be estimated. - Risk (damage/loss) table - Damage rates of Buildings etc. - Loss rates of agricultural crops/fields Figure 17: Result 2: Risk (damage/loss) table GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT) 23

5. Conclusions FRAT has several limitations such as: - FRAT is a Tutorial Tool: - for enhancing awareness of floods at local levels; - for assessing flood risk. - Limited applicability for practical/actual shelter planning at the local level, as: - more advanced technology with more precise data is needed; - need to consider all the related factors, such as river, catchment at upstream, Manning s roughness coefficient etc. Further developments of FRAT will address the following shortcomings yet to be solved: 1. Integration of the following variables, which are currently not considered in FRAT, in order to improve FRAT outputs: - Upstream catchment; - River shape; - Actual Manning s roughness coefficient; - Precipitation; - Etc. 2. Improvement of the time for calculation: - currently 5 to 10 minutes are required for 1 week flood flow simulation. 3. Risk estimation currently uses Japanese criteria: - Damage and loss estimation will need to integrate local knowledge to tailor the tool to local context; - More research is needed to assess FRAT s global applicability. 4. Additional data is needed: - More data for landmarks and area boundaries is needed but there is limited availability, memory size etc. 24 GRIP, UN HABITAT, IFRC SIMPLIFIED FLOOD RISK ASSESSMENT TOOL (FRAT)

Global Risk Identification Programme (GRIP) Bureau for Crisis Prevention and Recovery (BCPR) United Nations Development Programme (UNDP) Empowered lives. Resilient nations. 11-13, chemin des Anémones, Châtelaine, CH-1219 Geneva, Switzerland www.undp.org/cpr