Global Flood Hazard E-Book

Guy J-P. Schumann

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: Geophysical Monograph Series
• Sprache: Englisch

Seitenzahl: 848

Veröffentlichungsjahr: 2018

Beschreibung

Global Flood Hazard Subject Category Winner, PROSE Awards 2019, Earth Science Selected from more than 500 entries, demonstrating exceptional scholarship and making a significant contribution to the field of study. Flooding is a costly natural disaster in terms of damage to land, property and infrastructure. This volume describes the latest tools and technologies for modeling, mapping, and predicting large-scale flood risk. It also presents readers with a range of remote sensing data sets successfully used for predicting and mapping floods at different scales. These resources can enable policymakers, public planners, and developers to plan for, and respond to, flooding with greater accuracy and effectiveness. * Describes the latest large-scale modeling approaches, including hydrological models, 2-D flood inundation models, and global flood forecasting models * Showcases new tools and technologies such as Aqueduct, a new web-based tool used for global assessment and projection of future flood risk under climate change scenarios * Features case studies describing best-practice uses of modeling techniques, tools, and technologies Global Flood Hazard is an indispensable resource for researchers, consultants, practitioners, and policy makers dealing with flood risk, flood disaster response, flood management, and flood mitigation.

Leseprobe

Table of Contents

Cover

Preface

1 The Need for Mapping, Modeling, and Predicting Flood Hazard and Risk at the Global Scale

1.1. INTRODUCTION

1.2. BRIEF OVERVIEW OF RECENT ADVANCES IN GLOBAL FLOOD HAZARD AND RISK MODELING

1.3. GLOBAL FLOOD RISK INFORMATION IN HIGH‐LEVEL DISASTER RISK MANAGEMENT ADVOCACY

1.4. APPLICATIONS FOR INTERNATIONAL DEVELOPMENT ORGANIZATIONS

1.5. APPLICATIONS FOR THE REINSURANCE INDUSTRY

1.6. APPLICATIONS FOR GLOBAL FLOOD FORECASTING AND EARLY WARNING

1.7. COMMUNICATING GLOBAL FLOOD RISK: THE AQUEDUCT GLOBAL FLOOD ANALYZER

1.8. THE WAY FORWARD

REFERENCES

Part I: Flood Hazard Mapping and Modeling from Remote Sensing

2 Rainfall Information for Global Flood Modeling

2.1. INTRODUCTION

2.2. ROLE OF RAINFALL IN LARGE‐SCALE FLOOD MODELING

2.3. GENERAL CONSIDERATIONS AND REQUIREMENTS

2.4. PRECIPITATION INFORMATION SOURCES

2.5. FUTURE DIRECTIONS

2.6. CONCLUSIONS

ACKNOWLEDGMENTS

REFERENCES

3 Flood Risk Mapping From Orbital Remote Sensing

3.1. INTRODUCTION

3.2. MICROWAVE RADIOMETRY FOR MEASURING RIVER DISCHARGE

3.3. PRODUCTION OF SIGNAL/DISCHARGE RATING CURVES

3.4. ASSESSING RIVER WATCH ACCURACY

3.5. SATELLITE GAUGING SITE SELECTION

3.6. FLOOD MAPPING FROM OPTICAL SATELLITES

3.7. REMOTE SENSING‐BASED FLOOD HAZARD QUANTIFICATION

3.8. CONCLUSION

REFERENCES

4 Flood Mapping Using Synthetic Aperture Radar Sensors From Local to Global Scales

4.1. INTRODUCTION

4.2. PRINCIPLES OF SAR: IMPLICATIONS FOR FLOOD MAPPING

4.3. COMMON SAR‐BASED FLOOD MAPPING METHODS

4.4. IMAGE INTERPRETATION: CHALLENGES AND SOLUTIONS

4.5. REPRESENTATION OF UNCERTAINTIES

4.6. CASE STUDIES

4.7. SUMMARY AND PERSPECTIVES

ACKNOWLEDGMENTS

REFERENCES

5 Flood Hazard Mapping in Data‐Scarce Areas

5.1. INTRODUCTION

5.2. STUDY SITE

5.3. METHODOLOGY

5.4. RESULTS

5.5. DISCUSSION

5.6. CONCLUSIONS

ACKNOWLEDGMENTS

REFERENCES

6 Global Flood Monitoring Using Satellite Precipitation and Hydrological Modeling

6.1. INTRODUCTION

6.2. GLOBAL FLOOD MONITORING SYSTEM (GFMS)

6.3. EVALUATION OF THE DRIVE MODEL AND THE GFMS

6.4. A TYPICAL EXAMPLE OF REAL‐TIME FLOOD DETECTION AND INUNDATION MAPPING BY THE GFM

6.5. ONGOING AND FUTURE WORK

6.6. SUMMARY AND CONCLUSIONS

REFERENCES

7 Flood Hazard Mapping for the Humanitarian Sector

7.1. INTRODUCTION

7.2. BACKGROUND INTERFERENCE ISSUES

7.3. THE PROCESS

7.4. CONCLUSION

REFERENCES

Part II: Flood Hazard Modeling and Forecasting

8 Modeling and Mapping of Global Flood Hazard Layers

8.1. INTRODUCTION

8.2. FLOOD MODELING DEVELOPMENTS AT THE GLOBAL SCALE

8.3. CURRENT GLOBAL MODELS

8.4. APPLICATIONS

8.5. OUTLOOK

8.6. CONCLUSIONS

REFERENCES

9 Estimating Change in Flooding for the 21st Century Under a Conservative RCP Forcing

9.1. INTRODUCTION

9.2. METHOD

9.3. RESULTS

9.4. DISCUSSION

9.5. CONCLUSIONS

ACKNOWLEDGMENTS

REFERENCES

10 From Precipitation to Damage

10.1. INTRODUCTION

10.2. MODEL CHAIN

10.3. APPLICATION OF RFM TO THE ELBE RIVER BASIN AND MULDE CATCHMENT

10.4. POSSIBILITIES FOR EXTENDING RFM TO THE CONTINENTAL SCALE

10.5. CONCLUSIONS

ACKNOWLEDGMENTS

REFERENCES

11 Global Flood Risk Modeling and Projections of Climate Change Impacts

11.1. INTRODUCTION

11.2. GLOBAL RIVER HYDRODYNAMIC MODEL FOR FLOOD RISK ASSESSMENT

11.3. GLOBAL FLOOD RISK ASSESSMENT UNDER CLIMATE CHANGE

11.4. SUMMARY AND DISCUSSIONS

ACKNOWLEDGMENTS

REFERENCES

12 Global Flood Forecasting for Averting Disasters Worldwide

12.1. INTRODUCTION

12.2. ORIGINS OF GLOBAL FLOOD FORECASTING

12.3. CURRENT STATUS OF LARGE‐SCALE FLOOD FORECASTING

12.4. RESEARCH ON IMPROVING QUALITY AND USABILITY OF HYDROLOGICAL FORECASTS

12.5. USING GLOBAL FLOOD FORECASTS FOR HUMANITARIAN PREPAREDNESS AND EARLY ACTION

12.6. FUTURE CHALLENGES AND OPPORTUNITIES OF GLOBAL FLOOD FORECASTING

12.7. CONCLUSIONS

ACKNOWLEDGMENTS

REFERENCES

13 Data Assimilation and River Hydrodynamic Modeling Over Large Scales

13.1. INTRODUCTION

13.2. DATA ASSIMILATION

13.3. EXAMPLES OF ESTIMATING FLOOD VARIABLES

13.4. CHALLENGES AND OPPORTUNITIES

ACKNOWLEDGMENTS

REFERENCES

14 Global Flood Hazard Mapping, Modeling, and Forecasting

14.1. CURRENT STATUS

14.2. CHALLENGES AND PERSPECTIVES

14.3. SUMMARY AND OUTLOOK

REFERENCES

Index

End User License Agreement

List of Tables

Chapter 02

Table 2.1 Comparison of the Main Classes of Rainfall Information

Table 2.2 Satellite Multisensor Precipitation Data Sets

Table 2.3 Operational Weather Forecast Systems

Table 2.4 Atmospheric Reanalysis Data Sets

Chapter 03

Table 3.1 Summary of Microwave Discharge Measurement (River Watch) Site Characteristics and Accuracy for Sites Along the Chindwin (108 and 23) and Ayeyarwady (26, 29, 30)

Chapter 04

Table 4.1 Summary of Spaceborne SAR Missions and Sensor Characteristics

Chapter 06

Table 6.1 Metrics for Model Performance in Streamflow Simulation, at Daily and Monthly Time Intervals for Continuous Years, Against 1121 GRDC River Gauges Across the Globe (‐50°S to 50°N)

Table 6.2 Metrics for Model Performance in Streamflow Simulation, at Daily and Monthly Time Intervals for Continuous Years, Against 1121 GRDC River Gauges Across the Globe (‐50°S to 50°N) for Summer Seasons

Table 6.3 Metrics for Model Performance in Streamflow Simulation, at Daily and Monthly Time Intervals for Continuous Years, Against 1121 GRDC River Gauges Across the Globe (‐50°S to 50°N) for Winter Seasons

Chapter 09

Table 9.1 Percentage Change in Estimated Flood Frequency Per Recurrence Interval Per Continent, Comparing 2000s–2100s

Table 9.2 Percentage Change in River Reaches Affected by Flooding Per Flood Magnitude Per Continent, Comparing 2000s–2100s

Chapter 12

Table 12.1 Summary of the Information Transferred Through the Global Flood Partnership Triggered by the January 2015 Malawi Floods

List of Illustrations

Chapter 01

Figure 1.1 National‐level flood risk profile for Turkey, produced for the World Bank using the GLOFRIS model [

World Bank

, 2015].

Figure 1.2 GLOFRIS simulations of the number of people affected per state (expressed as a percentage of the total population per state) for floods of different return periods.

Figure 1.3 Screenshot of an early beta version of ThinkHazard!, a tool developed by the World Bank’s Global Facility for Disaster Reduction and Recovery (GFDRR). The tool utilizes global flood hazard models and gives the development professional easy access to hazard information as well as practical recommendations for risk reduction.

Figure 1.4 Pan‐African flood hazard map for the reference return period of 100 years, as displayed on the GloFAS website (http://globalfloods.jrc.ec.europa.eu/).

Figure 1.5 Screenshot of the Aqueduct Global Flood Analyzer, showing a flood risk assessment at the country level for India. The analyzers displays a flood hazard map, as well as estimates of flood risk under current conditions and under future conditions due to climate change and socioeconomic development.

Chapter 02

Figure 2.1 Comparison of errors in peak discharge estimation resulting from hydrologic simulations using the Iowa Flood Center (IFC) and the research version of the TRMM TMPA satellite‐based data set (3‐hourly, 0.25°resolution) relative to the NCEP Stage IV reference rainfall data set (hourly, 4‐km resolution) for the 24 May–2 June 2013 flood period for the Iowa River (large watershed) and Turkey River (small watershed). Upper panels show boxplots of peak discharge percent errors, categorized by upstream area. Bottom panels show maps of the peak discharge percent error across the river networks. Simulations are performed using the IFC hydrologic model [

Small et al.,

2013].

Figure 2.2 Precipitation information from three data sources over southeastern Africa for the major flood period of January 2015: (a) GPCC “first guess” rain gauge data; (b) MERRA‐2 atmospheric reanalysis; (c) final version of the IMERG satellite multisensor precipitation product.

Figure 2.3 Grid cells containing at least one rain gauge in (a) the 1° daily‐scale GPCC “first guess” and (b) monthly monitoring gridded rain gauge data sets for January 2015. Red indicates that the grid cell contains one rain gauge; blue indicates that the grid cell contains two or more gauges.

Figure 2.4 (a) Daily‐scale TMPA rainfall vs. the ensemble median rainfall from the gridded gauge‐based

Newman et al.

[2015] data set for 100 moderate to heavy rain days in the vicinity of Charlotte, North Carolina, for the 1998–2012 period. (b) Same as (a) but showing the middle 80% of the ensemble spread in addition to the median value. Blue error bars show where the middle 80% of the ensemble spread encompasses the TMPA value; red shows where the TMPA value lies outside the middle 80% of the ensemble spread. Sixty‐five out of 100 TMPA daily values fall within the middle 80% of the ensemble spread.

Figure 2.5 GPM LEO constellation satellites that carry PMW radiometers. Note that the TRMM satellite is no longer operational.

Figure 2.6 (a)–(c): Comparison of monthly rainfall totals from GPCC 1° First Guess, IMERG, and MERRA‐2 reanalysis for January 2015. (d) and (e): Differences from GPCC.

Chapter 03

Figure 3.1 Temporal coverage, 1998 to present, of passive microwave sensors built and operated by NASA and by JAXA (Japanese Space Agency). Each satellite provides daily or near‐daily imaging of the globe.

Figure 3.2 Location of Satellite Gauging Site DFO # 30 over Ayeyarwady River and its floodplain in Myanmar. The site is a single pixel selected from the JRC grid; pixel is 10 km in size and produces the daily M value. The 95th percentile of the highest (driest and warmest) values from a 9 x 9 pixel array in the surrounding area produces the background calibration C value.

Figure 3.3 Scatter plot comparing WBM‐modeled daily discharge over a 5‐year period (January–December monthly daily maximum, minimum, and mean discharges) to the C/M ratio for River Watch site 30. The relationship is empirical and a better curve could be fit to these data, but a straight line is a useful first‐approximation rating curve.

Figure 3.4 Same data as in Figure 3.3, but arranged as time series of (a) maximum, (b) mean, and (c) minimum discharge values. The red line shows the model results and the blue line is the remote sensing as transformed by the rating equation in Figure 3.3.

Figure 3.5 Daily (4‐day forward running mean) discharge values for satellite gauging site 30 on the Ayeyarwady River. The low flow threshold (green line) is the 20% percentile discharge for each day; the flood thresholds use recurrence intervals computed using the Log Pearson III distribution and the annual maximum daily values. Major flooding in 2015 approached the calculated 5‐year recurrence interval; the flood of record, in 2004, was produced by a very damaging tropical storm [

Brakenridge et al

., 2016b]; flooding here exceeded the 25‐year threshold.

Figure 3.6 Model‐ and ground station‐based rating curves for Trinity River, Texas, River Watch site 446 in the Dartmouth Flood Observatory/University of Colorado array. The WBM model was used to produce the black line rating curve, which is fit to widely scattered data. A colocated USGS gauging station was used to produce the red line rating curve, with much better correlation to the remote sensing. Comparison of the two curves indicates a WBM model positive discharge bias increasing with higher discharges. Also, at this location, the WBM model may perform relatively poorly because it does not incorporate upstream river control structures.

Figure 3.7 MODIS remote‐sensing‐produced map of major 2015 flooding in Myanmar (light red), and also the typical annual high water (2014, light blue). Recurrence intervals at the River Watch measurement sites for the 2014 “flood” are approximately 1.5 years. Dark blue is a permanent water mapped by the Shuttle Water Boundary Data (https://lta.cr.usgs.gov/srtm_water_body_dataset) from February 2000 and represents typical winter surface water extent.

Figure 3.8 Recurrence interval and peak discharge estimates for River Watch site 30 along the lower Ayeyarwady River. The microwave measurement site is defined by the white 10 km square. Inundation for (a) a normal winter flow of 6253 m

3

/sec on 11–22 February 2000; (b) observed via MODIS at 250 m spatial resolution, flooded area for a typical summer monsoonal “flood,” r = 1.5 years (27,138 m

3

/sec, observed 2013); and (c) observed via MODIS again, flooded area for major flooding in 2004, r = 24 years (50,579 m

3

/sec. Note that during this flood, a major distributary to the east (the area is at the head of the Ayeyarwady delta) is also flooded.

Figure 3.9 Recurrence interval and peak discharge estimates for River Watch site 26 along the upper Ayeyarwady River. Inundation for (a) a normal winter flow of ~200 m

3

/sec in February 2000; (b) a typical monsoon season flood, recurrence interval of 1.1 years, ~9000 m

3

/sec, in 2002; and (c) inundation during a rare flood, recurrence interval of 21 years, 18,200 m

3

/sec, in 2013.

Chapter 04

Figure 4.1 Distribution of hazard types for Charter activations between 2000 and 2010. From EM‐DAT: The OFDA/CRED International Disaster Database (www.emdat.be – Université catholique de Louvain, Brussels, Belgium).

Figure 4.2 Summary of satellite‐based SAR missions that are applicable for flood studies, with corresponding wavelength bands and frequencies illustrated.

Figure 4.3 Different scattering mechanisms displayed by radar interactions with water and land surfaces. Based on

Martinis et al

. [2015].

Figure 4.4 The image shows example subsets of problem areas in SAR‐based flood mapping taken from a TerraSAR‐X (HH, 3‐m Stripmap) scene acquired on 25 July 2007 covering the Severn River flood event. The urban area shown here lies to the west of Tewkesbury, UK.

Figure 4.5 Swaths of the CSK (different colors) and S1 (turquoise) images of the area hit by the November 2014 Po River basin event. Two predictions of the FEWS‐Po model available through the DEWETRA platform are also shown. The horizontal red line corresponds to the critical reference flood, while yellow and orange lines indicate intermediate levels of warning. Western part of the Po River basin: run of 11 November 2014, h: 12:30 UTC. Eastern part of the Po River basin: run of 17 November 2014, h: 00:30 UTC. From

Boni et al

. [2016].

Figure 4.6 Performance assessment of the NRT classification algorithm for (a) CSK‐derived flood map and (b) the location of zones A (dark blue) and B (light blue) in the reference map. (c) The CSK intensity image (dB) is displayed. In (d), a false color composite of Landsat 8 for the same area is included. From

Boni et al

. [2016].

Figure 4.7 HSBA‐Flood block diagram. RG: Region growing. CD: Change detection. After

Chini et al.

[2017].

Figure 4.8 Contingency maps between (a) Envisat‐WS or (b) TerrasSAR‐X flood map and validation data set. After

Chini et al.

[2017].

Chapter 05

Figure 5.1 The Blue Nile basin and study area.

Figure 5.2 Flood experience of 1‐in‐20‐year event in Africa.

Figure 5.3 Comparison of simulated flood extent using the 1‐in‐20‐year flood derived from in situ data and using (a) REC or (b) PMC.

Figure 5.4 The diagram shows the estimation of the 1‐in‐20‐year design flood estimation depending on the number of observations (ranging from 3 to 26). The black solid line is the value estimated using 26 annual maximum discharges (uncertainty due to flood data and other sources are not shown here); the grey area is the uncertain estimation bounded by tenth and ninetieth percentiles; grey dash line is the median; black dash line is the design flood estimation using PMC, and black dotted line is the design flood estimated using REC.

Chapter 06

Figure 6.1 Outline of the GFMS providing flood intensity estimates and forecasts, time histories, and high‐resolution (1 km) inundation maps. The data flow is indicated by the arrows from top to the bottom. An example of 3‐hourly time step precipitation is given in (a), while the core of the GFMS system is the DRIVE model shown in (b). Examples of output of the GFMS are shown in (c)–(e) including the flood detection and intensity (the routed runoff in depth above threshold derived based on 15‐year retrospective simulation), experimental global flood inundation mapping every 3 hours at 1 km resolution, and time series of model outputs (streamflow and flood detection).

Figure 6.2 DRIVE model is coupled VIC model (left) with DRTR (right) routing model. The DRTR routing model concept on river basin drainage system at (a) grid and (b)–(d) subgrid scales using a real river basin (Mbemkuru River basin, southeast of Tanzania) as example. The light blue lines in (a) are the baseline high resolution (1 km) river network from HydroSHEDS and the red lines are the DRT‐derived coarse‐resolution rivers (1/8th degree in this case). More details for VIC model can be found in

Liang et al.

[1994].

Figure 6.3 Example of the DRIVE model major outputs from the real‐time GFMS with screenshots from http://flood.umd.edu. The examples show the model global outputs of (a) routed runoff, (b) streamflow, and (c) flood detection and intensity (water depth [mm] above flood threshold) at a 3‐hour time interval (15Z01Jul2013). (d) An example of global TMPA 3B42V7 real‐time rainfall input data at the same time interval is shown. (c1‐6) The example shows the spatial‐temporal evolution (at daily interval) of the flood event that happened in North India during 15 June 2013 to 20 June 2013.

Figure 6.4 Snapshots from the real‐time GFMS (online: http://flood.umd.edu) for two major flood waves, covering April to early June 2013, in subbasin rivers upstream of the Mississippi River, including (a and b) the flood detection and intensity (water depth above flood threshold), (c and d) previous 7‐day accumulated precipitation according to TMPA V7RT, and (e and f) streamflow. All data are at 1/8th (~12 km) resolution.

Figure 6.5 (a) The DRIVE‐RT simulated streamflow against observed data from 29 USGS gauges on the rivers of the upper Mississippi River basin for a 2‐year retrospective period (12 June 2011 to 12 June 2013). All USGS gauges are shown in filled circles, while their colors are turned into green when the model estimated positive daily NSCs at the corresponding locations. (b)–(e) show the observed and simulated daily hydrographs for four of the gauges, with locations indicated in (a), during the spring and early summer flooding period (1 April to 9 June 2013).

Figure 6.6 Definition of spatial window for matching between simulated and reported flood events. The blue lines are rivers defined by DRT according to DEM information, while the area in yellow indicates the spatial coverage for a flood event derived by the flood definition algorthm.

Figure 6.7 Example of well‐reported areas (shaded yellow) and their corresponding FAR metrics (according to DRIVE‐RT for all floods with duration greater than 1 day) in the parts of Asia that tend to have more floods.

Figure 6.8 The flood detection metrics (a) POD, (b) FAR, and (c) CSI, across 38 well‐reported areas for DRIVE‐RP and DRIVE‐RT results for all floods with duration greater than 3 days, against DFO flood inventory data during 2001 to 2011. The numbers of dams upstream of each well‐reported area are listed along the x‐axis.

Figure 6.9 DRIVE‐RP model performance (monthly NSC) in reproducing monthly streamflow during 2001–2011, when driven by TMPA RP research precipitation data, at 1121 GRDC streamflow gauges across the globe. All GRDC gauges are shown as filled circles, while at each gauge, if the model performance is of a positive value for monthly NSC, the gauge color turns into green or purple in accordance to the value of NSC.

Figure 6.10 The percentage of gauges in each latitude band (defined in section 6.4.4.2) for which the DRIVE model showed positive daily NSCs using TMPA‐RP and TMPA‐RT precipitation input. The x‐axis values are the central latitude for each band.

Figure 6.11 The daily NSC (a) and (b), and MARE (c) and (d) metrics for the region of South America from (a, c) DRIVE‐RP and (b, d) DRIVE‐RT model results.

Figure 6.12 Malaysia flooding in December 2014.

Figure 6.13 High resolution (1 km) inundation estimates over affected areas in Malaysia.

Figure 6.14 The typical distribution map of the GFMS users for a 3‐day time window (22–24 March 2016 for this snapshot).

Chapter 07

Figure 7.1 Three and‐a‐half decades of global flood effects, CRED‐EMDAT; Data sources include UN agencies, USAID Office of Foreign Disaster Assistance (OFDA), governments, International Federation of Red Cross and Red Crescent Societies, nongovernmental organizations, insurance companies, research institutes, and press agencies.

Figure 7.2 Sample map from the preparedness phase for Super Typhoon Hagupit (Philippines, December 2014) using the GFS precipitation forecast and the ECMWF strike probability ensemble.

Figure 7.3 Risk zones identified by the ECMWF ensemble forecast for Tropical Storm Mahasen (Myanmar, May 2013).

Figure 7.4 Forecasts from ECMWF, NCEP GFS, and COLA were used on a preparedness map for tropical cyclone Giovanna (Madagascar, February 2012). Dark purple highlights on the map indicate flood inundation limits from the Dartmouth Flood Observatory (DFO) historical archive.

Figure 7.5 Pakistan floods from September 2011 with imagery overlay.

Figure 7.6 January 2013 floods in Mozambique; combination of passive (MODIS) and actively tasked analysis.

Figure 7.7 NASA MODIS analysis (purple) for flood‐affected areas in Nigeria July–November 2012.

Figure 7.8 Historical flood inundation limits (in pink) from DFO archives 1985–present. While not exactly a flood‐hazard map, the highlighted areas provided a rough preliminary risk picture, which helped direct some postdisaster damage assessment resources during Typhoon Parma (Philippines, October 2009).

Figure 7.9 Historical flood inundation limits overlaid on affected areas map of December 2014 floods in Sri Lanka.

Figure 7.10 Circles indicate heavily affected areas during the May 2014 Balkans floods. Many of these sites fell outside of the EU JRC modeled areas for flood hazard (red highlight).The JRC model output is a 1:100‐year return map based on uncertain boundary conditions and low accuracy topography [

Alfieri et al.,

2013]. Although the technical limitations are very clear to the scientists, the humanitarian community tends to think of a flood hazard model as a defacto indicator of all areas at risk. Ground‐level expectations from the high level science should be commensurate with its operational usefulness.

Chapter 08

Figure 8.1 Number of disasters, estimated people impacted and estimated damage of disasters that occurred in 2014 (statistics from

Guha‐Sapir et al.

[2015]). Figures are distributed by disaster type.

Figure 8.2 Comparison of (a) raw SRTM DEM with LiDAR and (b) corrected SRTM with LiDAR for (c) western Belize. Cross sections A, B, and C transect the Belize River valley and compare the LiDAR elevation profile (black) with the uncorrected SRTM profile (green) and corrected SRTM profile (purple). From

Sampson et al.

[2016].

Figure 8.3 A description of observed data used in a global flood frequency analysis, shows the Koppen‐Geiger main climate classifications used to partition stations, along with the locations of each of the GRDC stations. (a) The proportion of discharge stations in each climate zone; (b) a histogram of upstream catchment areas. From

Smith et al.

[2015].

Figure 8.4 (a) Aggregated flood results for six models for a 1‐in‐100‐year return period fluvial flood hazard for the African Continent. Color scale indicates how many models predict flooding. (b) Detail for the lower Nile. (c) Detail for the lower Niger, showing areas of strong agreement (narrow confined floodplains at the confluence of Benue and Niger rivers) and areas of disagreement in the Niger coastal delta. From

Trigg et al.

[2016].

Figure 8.5 (a) Flooded area as percentage of continental area for all models and return periods. (b) Percentage of catchment area flooded mapped for all models for a 1‐in‐100‐year return period hazard showing significant spatial differences. From

Trigg et al.

[2016].

Figure 8.6 Examples of defended and undefended hazard layers produced for the entire upper Mississippi basin, using a global model approach. The data have been used to explore the impacts of levees on floodplain connectivity.

Chapter 09

Figure 9.1 Yukon River near its delta in Arctic Alaska. (a) Recurrence intervals with associated simulated peak water discharge for the time period 1975 to 2004, normalized by observations of the same time period using a Log‐Pearson Type III distribution for the Yukon River at Pilot Station (61°56'04″, 162°52'50″); (b) the frequency of each recurrence interval for the 2000s (black) and 2100s (grey) at that same location; (c) near true color LandSat imagery with approximately bankfull conditions represented as black for the Yukon River during June 2003; and (d) near true color LandSat imagery flooding (black) with a recurrence interval between 25 and 50 years at that same location during May 2009. White toward blue colors in (c) and (d) indicate cloud cover.

Figure 9.2 The 2‐year recurrence interval as established from a Log‐Pearson III distribution based on 30‐year daily‐simulated global water discharge defines the “bankfull discharge.” Bankfull discharge less than 30 m3/sec (grey areas) are not taken into consideration.

Figure 9.3 Flooded river reaches for the 30‐year period anchored in the 2000s and 2100s, respectively, wherein flooding defined as a minimum frequency to classify as flooding is one time exceedance of the 10‐year recurrence discharge. Light orange indicates areas that are as such flooded both during the 2000s and 2100s. Blue represents as such flooded areas only during the 2000s, and red indicates areas only flooded as such during the 2100s.

Figure 9.4 Spatial representation of percentage change flood frequency with the 2000s as reference, between the 2000s and 2100s for all floods that have (a) a larger than the bankfull (2‐year) recurrence interval, and (b) larger than 100‐year recurrence interval.

Chapter 10

Figure 10.1 Structure of RFM, Regional Flood Model.

Figure 10.2 (a) The Elbe River Basin (only German territory) and (b) the Mulde catchment.

Figure 10.3 Comparison of inundated areas for the 2002 flood. Red: derived from remote sensing; blue: derived from RFM simulation; green: agreement of remote sensing derived and RFM derived areas. From

Falter et al.

[2016].

Figure 10.4 (a) Derived flood risk approach based on continuous simulation consisting of a cascade of models from precipitation to flood damage. Flood risk is estimated by analyzing the simulated flood damage. The approach is an extension of (b) the derived flood frequency approach based on continuous simulation.

Figure 10.5 Distribution of damage for one of the Mulde subcatchments. During the 10,000‐year simulation, 774 damage events occur in this subcatchment.

Figure 10.6 Two events with the same flood peak for subbasin 995 but different (a) hydrographs and (b) inundation patterns. The colors of the inundation areas show the difference in water depth between the two events.

Figure 10.7 Distribution of (a) return periods and (b) losses for three floods in the Mulde catchment. Each flood results in the same overall damage of approximately 68 million Euro.

Chapter 11

Figure 11.1 Procedures to extract subgrid topography parameters for CaMa‐Flood. (a) Baseline elevation data: SRTM3 DEM elevation (m); (b) baseline hydrography data: HydroSHEDS drainage area (log10 km

2

); (c) unit‐catchment discretization. For the unit‐catchment marked with the orange boundary, the blue vector represents the channel length, the ground elevation is given as the elevation of the point with a red arrow. (d) Subgrid topography parameters:

L

, channel length;

W

, channel width;

B

, channel depth;

Z

, ground elevation;

Ac

, catchment area; and

Df(Af)

, floodplain elevation profile.

Sr

and

Sf

represent river and floodplain water storage, respectively;

Af

represents flooded area,

Dr

and

Df

represent river and floodplain water depth, respectively, and

Qr

and

Qf

represent river and floodplain discharge to down stream, respectively. (e) Floodplain elevation profile function. The black line represents the original cumulative distribution function (CDF) curve of the elevations within a unit‐catchment; the green line represents the approximated floodplain elevation profile using linear interpolation. (f) River network map indicating the downstream unit‐catchment of each unit‐catchment.

Figure 11.2 Diagnostic downscaling of water depth. (a) Simulated water depth at 0.25‐degree resolution. (b) Downscaled water depth at 500‐m resolution.

Figure 11.3 Simulated river hydrodynamics in the Amazon main stem. (a) Daily river discharge at the Obidos gauging station (55.5W, 1.9S). (b) Water surface elevation along the main stem of the Amazon River, averaged for May–June 1996. The blue, red, and green lines represent the Full, Kinematic, and No Floodplain experiments, respectively. The gray line in (a) represents Global Runoff Data Center (GRDC) in situ gauged discharge. The black line in (b) represents the river bed elevation.

Figure 11.4 Simulated and observed flooded area in the Amazon River basin. Top: high‐water season (May–June 1996). Bottom: low‐water season (Sepember–October 1995). (a) Water depth from the Full experiment; (b) water depth from the Kinematic experiment; (c) L‐band radar observations [

Hess et al.,

2003]. Blue: open water; green: inundated floodplains; background grayscale colors represent elevations.

Figure 11.5 Simulated global flood projection. (a) Return period (year) of the 1/100 flood in 20C between 2071 and 2100, that is, the relative frequency of the historical 1/100 flood in the future period. (b) Number of global climate models (GCMs) showing the same direction (decrease or increase) in flood frequency change. RCP8.5 is shown. Sea surface and dry regions (average annual river discharge for 1979–2010 calculated by a retrospective off‐line simulation [

Kim et al.,

2011] forced by observation‐based climate data smaller than 0.01 mm/day) are excluded.

Figure 11.6 Procedures to calculate annual maximum inundation area for 1971–2100.

Figure 11.7 Global flood exposure (in millions) for a 1/100 flood in 20C or greater. (a) Ensemble means of the historical simulations (thick black line) and the future simulations for each climate scenario (colored thick lines). The shading denotes the standard deviation among Atmosphere‐Ocean General Circulation Models (AOGCMs). (b) The maximum and minimum range (whiskers), mean (horizontal thick lines within each bar), standard deviation (height of box), and individual values among AOGCMs (colored markers within each bar) averaged over 2071–2100. (c) Relationship between global flood exposure and global surface temperature over land. The thick black line and colored lines indicate the ensemble mean of all GCMs and the results of each GCM, respectively.

Chapter 12

Figure 12.1 GloFAS forecast of 5 January 2015 showing the probability of exceeding the 20‐year return period (shading) and details of a hydrograph forecast for a location in southern Malawi (the yellow, red, and purple colors refer to the 2‐, 5‐, and 20‐year return periods, respectively).

Figure 12.2 GloFAS forecast over Myanmar on 1 September 2015 showing (a) the accumulated precipitation forecast over the next 10 days (ensemble median) and (b) probability of the streamflow forecast exceeding the 20‐year (purple) and 5‐year (red) return periods. The numbers indicate the corresponding probabilities of threshold exceedance.

Figure 12.3 Comparison of (a) observed precipitation for an intense rainfall event in northern Italy on 13 September 2015 against (b) the median ECMWF NWP ensemble prediction for the same period forecast 24 hours beforehand.

Figure 12.4 Performance of the Global Flood Awareness System in the main world rivers for the period 2009–2010. Shades of blue indicate the maximum lead time (days) of skillful ensemble streamflow predictions in early flood detection (i.e., ROC area > 0.7).

Figure 12.5 Mean daily (1979–2010) precipitation in (a) GPCP and differences with (b) CRU, (c) GPCC, and (d) ERA.

Figure 12.6 Return periods computed from the simulations using the (a) CRU, (b) GPCC, and (c) ERA precipitation forcing based on the streamflow magnitude corresponding to a 20‐year return level of the GPCP simulations.

Figure 12.7 Ensemble Streamflow Prediction (ESP) and seasonal forecasts setup.

Figure 12.8 Example of the reverse‐ESP and ESP behavior with time. The black vertical line represents lead time 0, the starting time of the forecasts. The colored circles on this black line are the respective initial discharges for both systems.

Chapter 13

Figure 13.1 Spatially averaged error decrease in simulated water surface elevation after data assimilation, for Ohio River case study.

Figure 13.2 Spatially averaged error decrease in simulated river discharge after data assimilation, for Ohio River case study.

Figure 13.3 (a) Time series of inundated area for a specific flood event for the truth and open‐loop simulations. (b) Forecast error reduction time series for different observation times (forecast initialization): 11, 5, 1, and 0 days prior to the flood peak.

Figure 13.4 Averaged error decrease in water surface elevation for each river reach of assimilated observations.

Chapter 14

Figure 14.1 World map showing selected supersites and their level of completion (in order to be labeled a “supersite”) in terms of Earth Observation (EO) flood data collection at the Dartmouth Flood Observatory (DFO), University of Colorado Boulder. Only major world rivers are shown. Note that other supersites can be added as needed or requested by the community.

Guide

Cover

Table of Contents

Begin Reading

Pages

ii

iii

iv

vii

viii

ix

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

17

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

157

158

159

160

161

162

163

164

165

166

167

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

Geophysical Monograph Series

180

Arctic Sea Ice Decline: Observations, Projections, Mechanisms, and Implications

Eric T. DeWeaver, Cecilia M. Bitz, and L.‐Bruno Tremblay (Eds.)

181

Midlatitude Ionospheric Dynamics and Disturbances

Paul M. Kintner, Jr., Anthea J. Coster, Tim Fuller‐Rowell, Anthony J. Mannucci, Michael Mendillo, and Roderick Heelis (Eds.)

182

The Stromboli Volcano: An Integrated Study of the 2002– 2003 Eruption

Sonia Calvari, Salvatore Inguaggiato, Giuseppe Puglisi, Maurizio Ripepe, and Mauro Rosi (Eds.)

183

Carbon Sequestration and Its Role in the Global Carbon Cycle

Brian J. McPherson and Eric T. Sundquist (Eds.)

184

Carbon Cycling in Northern Peatlands Andrew J. Baird,

Lisa R. Belyea, Xavier Comas, A. S. Reeve, and Lee D. Slater (Eds.)

185

Indian Ocean Biogeochemical Processes and Ecological Variability

Jerry D. Wiggert, Raleigh R. Hood, S. Wajih A. Naqvi, Kenneth H. Brink, and Sharon L. Smith (Eds.)

186

Amazonia and Global Change Michael Keller, Mercedes

Bustamante, John Gash, and Pedro Silva Dias (Eds.)

187

Surface Ocean–Lower Atmosphere Processes

Corinne Le Quèrè and Eric S. Saltzman (Eds.)

188

Diversity of Hydrothermal Systems on Slow Spreading Ocean Ridges

Peter A. Rona, Colin W. Devey, Jérôme Dyment, and Bramley J. Murton (Eds.)

189

Climate Dynamics: Why Does Climate Vary?

De‐Zheng Sun and Frank Bryan (Eds.)

190

The Stratosphere: Dynamics, Transport, and Chemistry

L. M. Polvani, A. H. Sobel, and D. W. Waugh (Eds.)

191

Rainfall: State of the Science

Firat Y. Testik and Mekonnen Gebremichael (Eds.)

192

Antarctic Subglacial Aquatic Environments

Martin J. Siegert, Mahlon C. Kennicut II, and Robert A. Bindschadler (Eds.)

193

Abrupt Climate Change: Mechanisms, Patterns, and Impacts

Harunur Rashid, Leonid Polyak, and Ellen Mosley‐ Thompson (Eds.)

194

Stream Restoration in Dynamic Fluvial Systems: Scientific Approaches, Analyses, and Tools

Andrew Simon, Sean J. Bennett, and Janine M. Castro (Eds.)

195

Monitoring and Modeling the Deepwater Horizon Oil Spill: A Record‐Breaking Enterprise

Yonggang Liu, Amy MacFadyen, Zhen‐Gang Ji, and Robert H. Weisberg (Eds.)

196

Extreme Events and Natural Hazards: The Complexity Perspective

A. Surjalal Sharma, Armin Bunde, Vijay P. Dimri, and Daniel N. Baker (Eds.)

197

Auroral Phenomenology and Magnetospheric Processes: Earth and Other Planets

Andreas Keiling, Eric Donovan, Fran Bagenal, and Tomas Karlsson (Eds.)

198

Climates, Landscapes, and Civilizations

Liviu Giosan, Dorian Q. Fuller, Kathleen Nicoll, Rowan K. Flad, and Peter D. Clift (Eds.)

199

Dynamics of the Earth’s Radiation Belts and Inner Magnetosphere

Danny Summers, Ian R. Mann, Daniel N. Baker, and Michael Schulz (Eds.)

200

Lagrangian Modeling of the Atmosphere

John Lin (Ed.)

201

Modeling the Ionosphere‐Thermosphere

Jospeh D. Huba, Robert W. Schunk, and George V. Khazanov (Eds.)

202

The Mediterranean Sea: Temporal Variability and Spatial Patterns

Gian Luca Eusebi Borzelli, Miroslav Gacic, Piero Lionello, and Paola Malanotte‐Rizzoli (Eds.)

203

Future Earth ‐ Advancing Civic Understanding of the Anthropocene

Diana Dalbotten, Gillian Roehrig, and Patrick Hamilton (Eds.)

204

The Galápagos: A Natural Laboratory for the Earth Sciences

Karen S. Harpp, Eric Mittelstaedt, Noémi d’Ozouville, and David W. Graham (Eds.)

205

Modeling Atmospheric and Oceanic Flows: Insightsfrom Laboratory Experiments and Numerical Simulations

Thomas von Larcher and Paul D. Williams (Eds.)

206

Remote Sensing of the Terrestrial Water Cycle

Venkat Lakshmi (Ed.)

207

Magnetotails in the Solar System

Andreas Keiling, Caitríona Jackman, and Peter Delamere (Eds.)

208

Hawaiian Volcanoes: From Source to Surface Rebecca

Carey, Valerie Cayol, Michael Poland, and Dominique Weis (Eds.)

209

Sea Ice: Physics, Mechanics, and Remote Sensing

Mohammed Shokr and Nirmal Sinha (Eds.)

210

Fluid Dynamics in Complex Fractured‐Porous Systems

Boris Faybishenko, Sally M. Benson, and John E. Gale (Eds.)

211

Subduction Dynamics: From Mantle Flow to Mega Disasters

Gabriele Morra, David A. Yuen, Scott King, Sang Mook Lee, and Seth Stein (Eds.)

212

The Early Earth: Accretion and Differentiation

James Badro and Michael Walter (Eds.)

213

Global Vegetation Dynamics: Concepts and Applications in the MC1 Model

Dominique Bachelet and David Turner (Eds.)

214

Extreme Events: Observations, Modeling and Economics

Mario Chavez, Michael Ghil, and Jaime Urrutia‐ Fucugauchi (Eds.)

215

Auroral Dynamics and Space Weather

Yongliang Zhang and Larry Paxton (Eds.)

216

Low‐Frequency Waves in Space Plasmas

Andreas Keiling, Dong‐Hun Lee, and Valery Nakariakov (Eds.)

217

Deep Earth: Physics and Chemistry of the Lower Mantle and Core

Hidenori Terasaki and Rebecca A. Fischer (Eds.)

218

Integrated Imaging of the Earth: Theory and Applications

Max Moorkamp, Peter G. Lelievre, Niklas Linde, and Amir Khan (Eds.)

219

Plate Boundaries and Natural Hazards

Joao Duarte and Wouter Schellart (Eds.)

220

Ionospheric Space Weather: Longitude and Hemispheric Dependences and Lower Atmosphere Forcing

Timothy Fuller‐Rowell, Endawoke Yizengaw, Patricia H. Doherty, and Sunanda Basu (Eds.)

221

Terrestrial Water Cycle and Climate Change Natural and Human‐Induced Impacts

Qiuhong Tang and Taikan Oki (Eds.)

222

Magnetosphere‐Ionosphere Coupling in the Solar System

Charles R. Chappell, Robert W. Schunk, Peter M. Banks, James L. Burch, and Richard M. Thorne (Eds.)

223

Natural Hazard Uncertainty Assessment: Modeling and Decision Support

Karin Riley, Peter Webley, and Matthew Thompson (Eds.)

224

Hydrodynamics of Time‐Periodic Groundwater Flow: Diffusion Waves in Porous Media

Joe S. Depner and Todd C. Rasmussen (Auth.)

225

Active Global Seismology

Ibrahim Cemen and Yucel Yilmaz (Eds.)

226

Climate Extremes

Simon Wang (Ed.)

227

Fault Zone Dynamic Processes

Marion Thomas (Ed.)

228

Flood Damage Survey and Assessment: New Insights from Research and Practice

Daniela Molinari, Scira Menoni, and Francesco Ballio (Eds.)

229

Water‐Energy‐Food Nexus ‐ Principles and Practices

P. Abdul Salam, Sangam Shrestha, Vishnu Prasad Pandey, and Anil K. Anal (Eds.)

230

Dawn–Dusk Asymmetries in Planetary Plasma Environments

Stein Haaland, Andrei Rounov, and Colin Forsyth (Eds.)

231

Bioenergy and Land Use Change

Zhangcai Qin, Umakant Mishra, and Astley Hastings (Eds.)

232

Microstructural Geochronology: Planetary Records Down to Atom Scale

Desmond Moser, Fernando Corfu, James Darling, Steven Reddy, and Kimberly Tait (Eds.)

Geophysical Monograph 233

Global Flood Hazard

Applications in Modeling, Mapping, and Forecasting

Guy J‐P. SchumannPaul D. BatesHeiko ApelGiuseppe T. AronicaEditors

This Work is a co‐publication of the American Geophysical Union and John Wiley and Sons, Inc.

This Work is a co‐publication between the American Geophysical Union and John Wiley & Sons, Inc.

This edition first published 2018 by John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA and the American Geophysical Union, 2000 Florida Avenue, N.W., Washington, D.C. 20009

© 2018 the American Geophysical Union

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions

Published under the aegis of the AGU Publications CommitteeBrooks Hanson, Executive Vice President, ScienceLisa Tauxe, Chair, Publications CommitteeFor details about the American Geophysical Union visit us at www.agu.org.

Wiley Global Headquarters111 River Street, Hoboken, NJ 07030, USA

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Limit of Liability/Disclaimer of WarrantyWhile the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials, or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read.

Library of Congress Cataloguing‐in‐Publication data is available.

ISBN: 978‐1‐119‐21786‐2

Cover image: NASA/NOAACover design by Wiley

CONTRIBUTORS

Robert F. Adler Earth System Science Interdisciplinary Center, University of Maryland, College Park, Maryland, USA

Konstantinos M. Andreadis Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA

Heiko Apel GFZ German Research Centre for Geoscience, Section Hydrology, Potsdam, Germany

L. Arnal European Centre for Medium‐Range Weather Forecasts, Reading, UK; Department of Geography and Environmental Science, University of Reading, Reading, UK

Giuseppe T. Aronica Department of Engineering, University of Messina, Messina, Italy

Paul D. Bates Fathom Global Ltd., Engine Shed, Bristol, UK;School of Geographical Sciences, University of Bristol, Bristol, UK

C. A. Baugh European Centre for Medium‐Range Weather Forecasts, Reading, UK

G. Robert Brakenridge Dartmouth Flood Observatory CSDMS/INSTAAR, University of Colorado, Boulder, Colorado, USA

Marco Chini Department of Environmental Research and Innovation, Luxembourg Institute of Science and Technology, Belvaux, Esch'sur‐Alzette, Luxembourg

H. L. Cloke Department of Geography and Environmental Science, University of Reading, Reading, UK

Sagy Cohen University of Alabama, Department of Geography, Tuscaloosa, Alabama, USA

Erin Coughlan de Perez Institute for Environmental Studies (IVM), Vrije Universiteit Amsterdam, Amsterdam, Netherlands; Red Cross Red Crescent Climate Centre, The Hague, Netherlands;International Research Institute for Climate and Society, Columbia University, New York, USA

Antara Dasgupta IITB‐Monash Research Academy, Mumbai, Maharashtra, India; Hydro‐Remote Sensing Applications (H‐RSA) Group, Department of Civil Engineering, Indian Institute of Technology Bombay, Mumbai, Maharashtra, India; Water Group, Department of Civil Engineering, Monash University, Melbourne, Victoria, Australia

Giuliano Di Baldassarre Department of Earth Sciences, Uppsala University, Uppsala, Sweden; Centre for Natural Disaster Science, CNDS, Uppsala, Sweden

Francesco Dottori European Commission, Joint Research Centre, Ispra, Italy

E. Dutra European Centre for Medium‐Range Weather Forecasts, Reading, UK; Instituto Dom Luiz, Faculdade de Ciências, Universidade de Lisboa, Lisbon, Portugal

R. E. Emerton European Centre for Medium‐Range Weather Forecasts, Reading, UK; Department of Geography and Environmental Science, University of Reading, Reading, UK

Daniela Falter GFZ German Research Centre for Geosciences, Section Hydrology, Potsdam, Germany

Balazs M. Fekete The City University of New York, Department of Civil Engineering, New York, USA

Jim Freer School of Geographical Sciences, University of Bristol, Bristol, UK

Zhen Gao Earth System Science Interdisciplinary Center, University of Maryland, College Park, Maryland, USA

Stefania Grimaldi Water Group, Department of Civil Engineering, Monash University, Melbourne, Victoria, Australia

Guojun Gu Earth System Science Interdisciplinary Center, University of Maryland, College Park, Maryland, USA

Björn Guse Department of Hydrology and Water Resources Management, Kiel University, Kiel, Germany

Yukiko Hirabayashi Department of Civil Engineering, Shibaura Institute of Technology, Tokyo, Japan

F. A. Hirpa European Commission Joint Research Centre (JRC), Directorate E, Space, Security and Migration, Ispra, Italy; School of Geography and the Environment, University of Oxford, Oxford, UK

Renaud Hostache Department of Environmental Research and Innovation, Luxembourg Institute of Science and Technology, Belvaux, Esch'sur‐Alzette, Luxembourg

Yeshewatesfa Hundecha Swedish Meteorological and Hydrological Institute, Sweden

Kris Johnson Nature Conservancy, Arlington, Virginia, USA

Brenden Jongman Institute for Environmental Studies (IVM), Vrije Universiteit Amsterdam, Amsterdam, Netherlands; Global Facility for Disaster Reduction and Recovery, World Bank Group, Washington, DC, USA

Melanie Kappes World Bank Group, Washington, DC, USA

Albert J. Kettner University of Colorado Boulder, INSTAAR, Boulder, Colorado, USA

Heidi Kreibich GFZ German Research Centre for Geosciences, Section Hydrology, Potsdam, Germany

Tianyi Luo World Resources Institute, Washington, DC, USA

Patrick Matgen Department of Environmental Research and Innovation, Luxembourg Institute of Science and Technology, Belvaux, Esch'sur‐Alzette, Luxembourg

Bruno Merz GFZ German Research Centre for Geoscience, Section Hydrology, Potsdam, Germany; Institute of Earth and Environmental Science, University of Potsdam, Potsdam, Germany

Jefferey Neal Fathom Global Ltd., Engine Shed, Bristol, UK; School of Geographical Sciences, University of Bristol, Bristol, UK

Dung Nguyen GFZ German Research Centre for Geosciences, Section Hydrology, Potsdam, Germany

Irina Overeem University of Colorado Boulder, INSTAAR, Boulder, Colorado, USA

FlorianPappenberger European Centre for Medium‐Range Weather Forecasts, Reading, UK; College of Hydrology and Water Resources, Hohai University, Nanjing, China; School of Geographical Sciences, University of Bristol, Bristol, UK

Valentijn R. N. Pauwels Water Group, Department of Civil Engineering, Monash University, Melbourne, Victoria, Australia

Rob Porter Canopius Group, London, UK; Songai Hoken Japan Nippon Koa Holdings Kabushiki Kaisha

RAAJ Ramsankaran Hydro‐Remote Sensing Applications (H‐RSA) Group, Department of Civil Engineering, Indian Institute of Technology Bombay, Mumbai, Maharashtra, India

Kashif Rashid Former Head of GIS Operations at the UN World Food Programme and currently Senior Technical Advisor to the Humanitarian and Scientific Community based out of Boston, Massachusetts, USA

B. Revilla‐Romero JBA Consulting, Skipton, UK

Sahar Safaie United Nations International Strategy for Disaster Reduction Secretariat, Geneva, Switzerland

Peter Salamon European Commission Joint Research Centre (JRC), Directorate E, Space, Security and Migration, Ispra, Italy

Christopher Sampson Fathom Global Ltd., Engine Shed, Bristol, UK

Kai Schröter GFZ German Research Centre for Geosciences, Section Hydrology, Potsdam, Germany

GuySchumann Remote Sensing Solutions, Inc., Monrovia, California, USA; School of Geographical Sciences, University of Bristol, Bristol, UK

Alanna Simpson World Bank Group, Washington, DC, USA

Andrew Smith Fathom Global Ltd., Engine Shed, Bristol, UK

P. J. Smith European Centre for Medium‐Range Weather Forecasts, Reading, UK

E. Stephens Department of Geography and Environmental Science, University of Reading, Reading, UK

James P. M. Syvitski University of Colorado Boulder, INSTAAR, Boulder, Colorado, USA

J. Thielen‐del Pozo European Commission Joint Research Centre (JRC), Scientific Development Unit, Ispra, Italy

Mark Trigg School of Civil Engineering, University of Leeds, Leeds, UK

Steffi Uhlemann‐Elmer Aspen Re, Research and Development, Zurich, Switzerland

Sergiy Vorogushyn GFZ German Research Centre for Geosciences, Section Hydrology, Potsdam, Germany

Jeffrey P. Walker Water Group, Department of Civil Engineering, Monash University, Melbourne, Victoria, Australia

Philip J. Ward Institute for Environmental Studies (IVM), Vrije Universiteit Amsterdam, Amsterdam, Netherlands

Satoshi Watanabe School of Engineering, The University of Tokyo, Tokyo, Japan

F. Wetterhall European Centre for Medium‐Range Weather Forecasts, Reading, UK

Daniel B. Wright Civil and Environmental Engineering, The University of Wisconsin‐Madison, Madison, Wisconsin, USA

Huan Wu Guangdong Province Key Laboratory for Climate Change and Natural Disaster Studies, Sun Yat‐sen University, Guangzhou, China; Earth System Science Interdisciplinary Center, University of Maryland, College Park, Maryland, USA

Dai Yamazaki Institute of Industrial Science, The University of Tokyo, Tokyo, Japan; JAMSTEC – Japan Agency for Marine‐Earth Science and Technology, Yokosuka, Japan

Kun Yan Deltares, Delft, The Netherlands

Yan Yan Guangdong Province Key Laboratory for Climate Change and Natural Disaster Studies, Sun Yat‐sen University, Guangzhou, China

E. Zsoter European Centre for Medium‐Range Weather Forecasts, Reading, UK

PREFACE

Floods are among the top‐ranking natural disasters in terms of annual cost in insured and uninsured losses. Since high‐impact events often cover spatial scales that are beyond traditional regional monitoring operations, large‐scale flood hazard modeling as well as remote sensing, in particular from satellites, present attractive alternatives. The complementarity of both can help produce better forecasts and better accuracy models, and render remotely sensed flood products and services more credible.

In the last two decades, there have been significant advances in computing architecture, processor speed, and numerical codes for flood inundation models. At the same time, great research efforts have been spent on making boundary data, especially topography at the global scale, fit‐for‐purpose. This combination has allowed substantial progress in the field of flood hazard modeling at spatial scales that go beyond the traditional reach‐scale applications of these types of numerical models. We have recently entered an era of global‐scale application of a number of flood hazard models and this now necessitates a lot of work on validation, improving model physics, and overcoming deficiencies in process representation.

In terms of remote sensing of flood hazard, there have been many studies in the scientific literature about mapping and monitoring of floods using satellite imagery since the 1970s. The sensors and data processing techniques that exist to derive information about floods from remotely sensed images are numerous. Instruments that record flood events may operate in the visible‐to‐infrared and microwave (radar) range of the electromagnetic spectrum. There is now a general consensus among space agencies, numerous organizations, scientists, and end users to strengthen the support that satellite missions can offer, particularly in assisting flood disaster response activities. This has stimulated more research in this area, and significant progress has been achieved in recent years in fostering our understanding of the ways in which remote sensing can support flood monitoring and prediction, and assist emergency response activities.

This book is a collection of chapters that provide state‐of‐the‐art insight on progress, caveats, and limitations in current efforts to map, model, and predict flood hazard at the global scale. Targeted at decision‐makers, flood response teams as well as scientists and academics active in the field of flood hazard, the general aim of this book is to report on advances in modeling, mapping, and predicting flood hazard and risk at the global scale. It describes different modeling approaches as well as remotely sensed data sets to predict and map flood risk at different scales, ranging from local to global. Recently, many scientist teams have rolled out models and data sets for flood hazard and risk globally with ever‐increasing accuracy and resolution, allowing the research, humanitarian, and development sectors to engage decision‐making processes based on better actionable information. This is precisely what this book outlines, and it concludes with a chapter that critically discusses requirements, challenges, and perspectives for improving global flood hazard mapping, modeling, and forecasting.

Guy SchumannPaul D. BatesHeiko ApelGiuseppe T. Aronica

1The Need for Mapping, Modeling, and Predicting Flood Hazard and Risk at the Global Scale

Philip J. Ward1, Erin Coughlan de Perez1,2,3, Francesco Dottori4, Brenden Jongman1,5, Tianyi Luo6, Sahar Safaie7, and Steffi Uhlemann‐Elmer8

1 Institute for Environmental Studies (IVM), Vrije Universiteit Amsterdam, Amsterdam, Netherlands

2 Red Cross Red Crescent Climate Centre, The Hague, Netherlands

3 International Research Institute for Climate and Society, Columbia University, New York, USA

4 European Commission, Joint Research Centre, Ispra, Italy

5 Global Facility for Disaster Reduction and Recovery, World Bank Group, Washington, DC, USA

6 World Resources Institute, Washington, DC, USA

7 United Nations International Strategy for Disaster Reduction Secretariat, Geneva, Switzerland

8 Aspen Re, Research and Development, Zurich, Switzerland

ABSTRACT

The socioeconomic impacts of flooding are huge. Between 1980 and 2013, flood losses exceeded $1 trillion globally, and resulted in approximately 220,000 fatalities. To reduce these negative impacts of floods, effective flood risk management is required. Reducing risk globally is at the heart of two recent international agreements: the Sendai Framework for Disaster Risk Reduction and the Warsaw International Mechanism for Loss and Damage Associated with Climate Change Impacts. Prerequisites for effective risk reduction are accurate methods to assess hazard and risk, based on a thorough understanding of underlying processes. Due to the paucity of local scale hazard and risk data in many regions, several global flood hazard and flood risk models have been developed in recent years. More and more, these global models are being used in practice by an ever‐increasing range of users and practitioners. In this chapter, we provide an overview of recent advances in global flood hazard and risk modeling. We then discuss applications of the models in high‐level advocacy in disaster risk management activities, international development organizations, the reinsurance industry, and flood forecasting and early warning. The chapter concludes with several remarks on limitations in global flood risk models and the way forward for the future.

1.1. INTRODUCTION

River floods are one of the most damaging forms of natural hazards [Guha‐Sapir et al.,2015], causing economic damage, fatalities, and social hardship all around the world. For example, over the period 1980–2013, flood losses exceeded $1 trillion globally, and resulted in approximately 220,000 fatalities [Munich Re, 2014]. Moreover, flood losses are increasing. While the reported losses due to floods were about USD 7 billion per year during the 1980s (adjusted for inflation), these increased to USD 24 billion per year during the period 2001–2011 (Kundzewicz et al. [2013], based on Munich Re NatCatSERVICE data). These negative impacts of flooding are projected to increase in the future [UNIDSR, 2015a] due to climate change [Arnell and Gosling, 2016; Hirabayshi et al., 2013; Winsemius et al., 2015], urbanization [Güneralp et al., 2015; Jongman et al., 2012], and land subsidence [Brown and Nicholls, 2015; Syvitski et al., 2009].

Flood risk management aims to reduce the negative impacts of floods. The concept of flood risk combines the probability of a flood with its potential consequences. While there are many definitions of risk, it is usually operationalized as being a function of three risk elements, namely: hazard, exposure, and vulnerability [e.g., Kron, 2002; UNISDR, 2011, 2013, 2015a]. As stated by UNISDR [2011], the hazard refers to the hazardous phenomenon itself, such as a flood event, including its characteristics and probability of occurrence; exposure refers to the location of economic assets or people in a hazard‐prone area; and vulnerability refers to the susceptibility of those assets or people to suffer damage and loss (e.g., due to unsafe housing and living conditions, or lack of early warning procedures).

Reducing risk, not only to flooding but also to all natural hazards, is high on the global political agenda. For example, it is at the heart of two recent international agreements: the Sendai Framework for Disaster Risk Reduction (Sendai Framework) [UNISDR, 2015b]; and the Warsaw International Mechanism for Loss and Damage Associated with Climate Change Impacts (Loss and Damage Mechanism) [UNFCCC, 2013]. The Sendai Framework is a 15‐year, voluntary, nonbinding agreement that was adopted at the Third UN World Conference in Sendai, Japan, in 2015. The Sendai Framework aims at the following outcome: “The substantial reduction of disaster risk and losses in lives, livelihoods and health and in the economic, physical, social, cultural and environmental assets of persons, businesses, communities and countries” (UNISDR, 2015b; page 12). To do this, the Sendai Framework sets out four so‐called priorities for action, one of which (Priority 1) is Understanding Disaster Risk. To achieve this at the global level, the framework points to the need to “enhance the development and dissemination of science‐based methodologies and tools to record and share disaster losses and relevant disaggregated data and statistics, as well as to strengthen disaster risk modeling, assessment, mapping, monitoring and multi‐hazard early warning systems” [UNISDR, 2015b; page 16]. The Loss and Damage Mechanism was adopted at the United Nations Framework Convention on Climate Change (UNFCC) Conference of the Parties (COP19) in Warsaw, Poland, in 2013. It promotes the implementation of approaches to address loss and damage associated with impacts of climate change, including extreme events like floods, in developing countries. One of the ways to do this is by “enhancing knowledge and understanding of comprehensive risk management approaches to address loss and damage associated with the adverse effects of climate change” [UNFCCC, 2013; page 6].

To contribute to the aims of the aforementioned agreements, effective risk reduction strategies are required. To achieve this at the global scale requires methods to quantitatively assess global flood risk in a holistic manner [Mechler et al., 2014]. Ideally, this could be achieved by developing detailed, high‐resolution local flood models [Jonkman, 2013] for all parts of the globe. However, in reality, the data required to develop such fine‐scale models do not exist in most locations, and the time required to collect such data and run the models would be prohibitive. Therefore, in the past decade, several global flood risk models have been developed.