Search

Estimating river bank heights is crucial for the accuracy of global flood models. Bank heights determine river-floodplain connectivity, and are used to parametrise channel capacity. Poor bank height estimates can lead to incorrect timings and locations of flood overtopping and erroneous channel capacity, resulting in unsatisfactory flood predictions.In the current implementation of global flood models, bank heights are estimated by extracting elevations from global Digital Elevation Models (DEMs) at river edges. These elevations, even with the latest DEMs, are often noisy and thus need to be heavily filtered and smoothed. Additionally, the surface water masks used to define river edges often do not match the time of acquisition on the DEMs, leading to inconsistencies. These simple methods for estimating bank heights were introduced during the early stages of global flood model development and have not been revisited in depth. With the emergence of new global DEMs (ALOS AW3D, Coperncius, FABDEM), improved surface water masks from multi-temporal, multi-sensor satellite data and novel image processing techniques, we revisit this problem. We present a new method to estimate bank height across scales, comparing estimates derived from global DEMs with high-quality LiDAR. We map the bank height estimates onto a new FABDEM based river network. Using examples from the UK and USA, we demonstrate the impact of bank height estimates on flood inundation.

River bifurcations, multi-thread rivers and artificial channels are not commonly included in global river networks, as they defy the gravity flow assumption elevation-derived networks are based on (e.g. HydroSheds, MERIT Hydro). Yet, these natural and artificial river divergences are important features of the global river drainage system and matter greatly at local to regional scales for various riverine risk assessments. For example, large river deltas are often highly populated regions, in part because of the rivers’ many distributaries. Representing these diverging flows in global river networks will greatly improve the accuracy of many river-based geoscience applications, such as flood forecasting, water availability and quality simulations, or riverine habitat mapping. We developed a vector-based, global river network that not only represents the tributary components of the global drainage network but also the distributary ones, including multi-thread rivers, canals and delta distributaries. We achieve this by merging a 30m, Landsat-based river mask with elevation-generated streams to ensure a homogeneous drainage density outside of the river mask (rivers narrower than approx. 30m). Crucially, this is the first global hydrography derived from a global 30m digital terrain model (FABDEM, based on Copernicus DEM) that shows greater accuracy over the traditionally used SRTM derivatives. OpenStreetMap river centrelines are used to increase the accuracy of the network outside of the river mask. After vectorisation and pruning, directionality is assigned by a combination of elevation, flow angle and continuity approaches. The new global network and its attributes are validated using gauging stations, reference river networks and randomised manual checks. The new network represents ~18 million km of streams and rivers with drainage areas greater than 50km2 and includes ~58 thousand. bifurcations in rivers wider than 30m. The hydrography includes vector river segments, sub-1km reaches and catchments as well as 30m flow direction and accumulation rasters. With the advent of hyper-resolution modelling in the geosciences at the regional and global scale, we expect this river network to be relevant to a broad range of applications in flood protection, hydrology, ecology, fluvial geomorphology and others. The network has been developed as part of the NERC-funded EvoFlood project and will be used to improve global flood models.

Global flood risk maps combine data from hydrodynamic models with gridded population or GDP data to estimate the amount of people or wealth likely to be exposed to future flood events. These estimates rarely incorporate measures of social vulnerability, which is a key source of variation in outcomes for exposed populations. They also use arbitrary return period thresholds, which can disguise potentially catastrophic hazards. To address these limitations, we integrate annual average exceedance probability estimates from a high-resolution (~90m) flood model with gridded population and economic data to create a global vulnerability-adjusted risk index for flooding (VARI Flood). This human welfare-centred approach radically alters how we perceive the geography of risk and could be used as a complement to traditional population or asset-centred approaches. We present global results for both unadjusted and vulnerability-adjusted risk at the subnational (Admin 2) level; country case studies illustrate how accounting for vulnerability changes the perceived subnational geography of risk. Globally, adjusting for vulnerability significantly reduces the total number of people estimated to be at ‘high risk’ from over 575 million to about 117 million. However, this latter estimate is likely an underestimate given the relatively coarse resolution of the economic data available, which disguises variation in vulnerability between communities within large cities and urban regions. Given the increasing concentration of the global human population in cities, there is an urgent need to improve the resolution of vulnerability mapping within large human settlements to inform mitigation measures.

Global flood risk maps combine data from hydrodynamic models with gridded population or GDP data to estimate the amount of people or wealth likely to be exposed to future flood events. These estimates rarely incorporate measures of social vulnerability, which is a key source of variation in outcomes for exposed populations. To address this lacuna, we integrate annual average exceedance probability estimates from a high-resolution (~90m) flood model with gridded population and economic data to create a global vulnerability-adjusted risk index for flooding (VARI Flood). Designed for decision makers, this approach radically alters how we perceive the geography of risk by identifying ‘hotspots’ characterised by high population density and high levels of social vulnerability. VARI Flood could be used as a complement to traditional population or asset-centred approaches. Our results also highlight the urgent need to improve social vulnerability mapping within cities and rapidly urbanizing regions.

In the sparsely investigated region of the Congo Basin (CB), flood seasonality and flood regime shift are established through relative frequency, cluster analysis, directional statistics, and non-overlapping block methods based on block maxima and peak over threshold (POT) series. Two months of significantly rich floods are observed at all gauging stations. The spatial distribution of floods presents three patterns: the north and northwest pattern, south and southeast pattern, and west/east pattern. It is observed that unimodal flood distribution is coherent in the northern and southern parts, as opposed to the bimodal flood distribution observed along the large band of the Equator from west to east of the basin. The time lag of flood indices suggests that the flood regime is not stationary. In addition, the time series show periods of high flooding, with POT frequencies and amplitudes higher during the 1960s and early 1970s than any other time period.

Digital Elevation Models (DEMs) depict the elevation of the Earth’s surface and are fundamental to many applications, particularly in the geosciences. To date, global DEMs contain building and forest artifacts that limit its functionality for applications that require precise measurement of terrain elevation, such as flood inundation modeling. Using machine learning techniques, we remove both building and tree height bias from the recently published Copernicus GLO-30 DEM to create a new dataset called FABDEM (Forest And Buildings removed Copernicus DEM). This new dataset is available at 1 arc second grid spacing (~30m) between 60°S-80°N, and is the first global DEM to remove both buildings and trees.Our correction algorithm is trained on a comprehensive and unique set of reference elevation data from 12 countries that covers a wide range of climate zones and urban types. This results in a wider applicability compared to previous DEM correction studies trained on data from a single country. As a result, we reduce mean absolute vertical error from 5.15m to 2.88m in forested areas, and from 1.61m to 1.12m in built-up areas, compared to Copernicus GLO-30 DEM. Further statistical and visual comparisons to other global DEMs suggests FABDEM is the most accurate global DEM with median errors ranging from -0.11m to 0.45m for the different landcover types assessed. The biggest improvements were found in areas of dense canopy coverage (>50%), with FABDEM having a median error of 0.45m compared to 2.95m in MERIT DEM and 12.95m for Copernicus GLO-30 DEM.FABDEM has notable improvements over existing global DEMs, resulting from the use of Copernicus GLO-30 and a powerful machine learning correction of building and tree bias. As such, there will be beneifts in using FABDEM for purposes where depiction of the bare-earth terrain is required, such as in applications in geomorphology, glaciology and hydrology.

Population projections for alternative socio-economic scenarios are crucial to understand climate change impacts. Current global gridded population projections are only available at coarse resolutions (~1km) that are inconsistent with the latest hazard models. Thus, climate change impact studies often utilise sub-optimum datasets by using coarse resolution gridded population predictions or present day population, and therefore may not adequately represent future population. To fill this gap, we use the latest datasets that align with the policy relevant Shared Socioeconomic Pathway (SSP) Scenarios and CMIP6 projections to create the first gridded population at ~90m resolution globally. We call this new dataset FuturePop. Projections are made at decadal intervals and extend to 2100 for each of the 5 SSP scenarios. Our method uses country level population and % urban projections from the SSP Database, redistributing population based on delineation of rural and urban areas. We add sophistication to our method by considering associated information such as travel time, and also include predictions of urban expansion. Comparison to existing global and regional datasets show FuturePop has considerable skill in predicting plausible population changes and redistribution. Lastly, we demonstrate the importance of using FuturePop for future flood risk compared to existing gridded population projections. Hazard footprints typically have horizontal length scales of tens to thousands of meters, thus it is crucial to depict populations at these scales to accurately estimate future flood exposure.

Dissemination of early warning information and effective preparedness are critical components in flood risk management and shape the dynamics of any successful early response. To enable efficient preparedness and an early response to hazards, early warning information should be simple, usable, and deployed through trusted sources.Successful management of floods are therefore dependent on clear and systematic communication structures which are, in turn, necessary to enable dissemination of such information. However, in many parts of Africa, flood event response is hampered by a lack of information on the inundation and potential exposure, with most large-scale systems limited to river flow forecasting, while local systems may lack coverage.In collaboration with the UK Foreign and Commonwealth Office, we have previously developed a method for fluvial flood inundation and exposure forecasting that combines hydrological forecasts from the Global Flood awareness systems (GLOFAS) with the LISFLOOD-FP global flood model. This was deployed in Mozambique to provide probabilistic inundation maps and exposure estimates to assist humanitarian response for cyclones Idai, Kenneth and Eloise.In this work, the flood models built with globally available datasets, and models augmented with local information, were evaluated for a series of uses and cases. Specifically, the 2020 fluvial flood event in the Nzoia basin in Kenya, the 2019 tropical cyclone flooding in Mozambique from cyclone Idai and recent pluvial flooding in the Zambian city of Lusaka.We discuss the potential and limitations of such information to inform efficient action and build resilient futures via co-production of forecasts in Kenya and the community learning lab component of the FRACTAL+ initiative in Lusaka.

Global flood models have undergone rapid development over the past decade. However, with each new generation of model it is essential to systematically evaluate simulation performance for a range of tests and against multiple sources of data. It is also important to take stock, document lessons learnt and contribute to the formation of better practice and modelling standards in the field. Here we illustrate some of the progress being made in global flood modelling by evaluating the latest 30 m resolution implementation of the LISFLOOD-FP/Fathom global flood model over the Central Highlands of Vietnam, and benchmark it against several previous incarnations of the model.Two independent data sources are used to evaluate the model. The first of these maps recent flood extents using remotely sensed data from the Sentinal-1 missions and compares them to global flood model outputs of commensurate return periods. The second data set identifies land parcels (properties and agricultural fields) that flooded during the same events from a household survey, where uniquely all household land parcels in four villages were sampled. The independence of the date sets also allowed for cross-validation of the observations.Substantial simulation enhancements are associated with the transition from SRTM and MERIT DEM’s at 90 m resolution to FABDEM, a version of Copernicus DEM at 30 m with forests and buildings removed. In addition to improvements derived from the DEM, more accurate river location, river width and discharge estimates combined with the inversion of river bathymetry via gradually varied rather than uniform flow theory also have an impact on performance.

Flood disasters have regularly been reported in the Congo Basin with significant damages to human lives, food production systems and infrastructure. Losses incurred by these damages are huge and re...

The Shuttle Radar Topography Mission (SRTM) has long been used as a source topographic information for flood hazard models, especially in data‐sparse areas. Error corrected versions have been produced, culminating in the latest global error reduced Digital Elevation Model (DEM) ‐ the Multi‐Error‐Removed‐Improved‐Terrain DEM (MERIT). This study investigates the spatial error structure of MERIT and SRTM, before simulating plausible versions of the DEMs using fitted semi‐variograms. By simulating multiple DEMs, we allow modellers to explore the impact of topographic uncertainty on hazard assessment even in data‐sparse locations where typically only one DEM is currently used. We demonstrate this for a flood model in the Mekong Delta and a catchment in Fiji using deterministic DEMs and DEM ensembles simulated using our approach. By running an ensemble of simulated DEMs we avoid the spurious precision of using a single DEM in a deterministic simulation. We conclude that using an ensemble of the MERIT DEM simulated using semi‐variograms by landcover class give inundation estimates closer to a LIDAR based benchmark. This study is the first to analyze the spatial error structure of the MERIT DEM and the first to simulate DEMs and apply these to flood models at this scale. The research work‐flow is available via an R package called demgenerator.

Background: Climate change impacts are felt disproportionately in developing countries, and in particular Southern Asia experiences the most damaging hydrometeorological events in the world, with loss of life from past cyclones in the hundreds of thousands. Despite this, the Bay of Bengal cyclone basin receives far less research attention than many of the others around the world. Here, we study the historical and future impacts of Super Cyclone Amphan, which made landfall in May 2020, bringing storm surges of 2-4 meters to coastlines of India and Bangladesh. Methods: In this modelling study, we combine projections of sea level rise from the Coupled Model Intercomparison Projection, phase 6 (CMIP6), with estimates of storm surge using a dynamic storm surge model. Sampling the spectrum of possible sea level rises, we consider a low, medium and high scenario, based on projections in 2100. We then feed these into a flood inundation model to simulate storm surge-induced flooding, had Cyclone Amphan occurred in these future worlds. Finally, we consider the change in future population growth and urbanisation, thereby calculating the change in population exposure to these future flooding events. Our approach is that of the extreme event attribution community, but projecting into the future rather than interrogating the past. Findings: We find that in 2100, the local sea level rise in the Bay of Bengal during the pre-monsoon cyclone season is between 0.32-0.84 m, depending on which emissions scenario is followed. If a Cyclone Amphan-scale storm surge occurred on top of that sea level rise, the future population of both India and Bangladesh will be more exposed, with India showing >200% increased exposure to extreme (>3 m) and moderate (>1 m) flooding under a high emissions scenario, and Bangladesh showing ~60-80% increased exposure to the same scenarios. The majority of this change in both countries comes from sea level rise rather than population changes, and in Bangladesh the future population change contributes negatively to the change in exposure, as more citizens migrate further inshore. However, if we follow an emissions scenario consistent with meeting the upper Paris Agreement climate goal, we project very little change in exposure. Interpretation: There is an urgent need to reduce carbon emissions to net zero, to prevent the negative impacts of climate change. By far the majority of cyclone research has been undertaken for countries such as America and Japan, with less resilient countries such as those in South Asia, which are more sensitive to changes in climate, seeing far less attention. With Cyclone Amphan occurring at the height of the COVID-19 crisis, we highlight how the risk was compounded and recommend that future climate risk assessments explicitly account for these potential non-linearities. Funding Statement: The main funding is from the Natural Environment Research Council. Declaration of Interests: The authors declare no competing interests.

Freely available Global Digital Elevation Models (GDEMs) are essential for many scientific and humanitarian applications. Recently, TanDEM-X 90 has been released with a global coverage at 3 arc sec resolution. Its release is sure to generate keen interest as it provides an alternative to the widely used Shuttle Radar Topography Mission (SRTM) DEM, especially for flood risk management as for low slope floodplains height errors can become particularly significant. Here, we provide a first accuracy assessment of TanDEM-X 90 for selected floodplain sites and compare it to other popular global DEMs – the Shuttle Radar Topography Mission (SRTM) and the error-reduced version of SRTM called Multi-Error-Removed-Improved-Terrain (MERIT) DEM. We characterize vertical height errors by comparing against high resolution LiDAR DEMs for 32 floodplain locations in 6 continents. Results indicate that the average vertical accuracy of TanDEM-X 90 and MERIT are similar and are both a significant improvement on SRTM. We further our analysis by assessing vertical accuracy by landcover, with our results suggesting that TanDEM-X 90 is the most accurate global DEM in all landcover categories tested except short vegetation and tree-covered areas where MERIT is demonstrably more accurate. Lastly, we present the first characterization of the spatial error structure of any TanDEM-X DEM product, and find the spatial error structure is similar to MERIT, with MERIT generally having lower sill values and larger ranges than TanDEM-X 90 and SRTM. Our findings suggest that TanDEM-X 90 has the potential to become the benchmark global DEM in floodplains with careful removal of errors from vegetation, and at this stage should be used alongside MERIT in any flood risk application.

Many dryland rivers are terminal systems, with small channels undergoing prominent downstream size reductions before ending on channelless floodplains, in wetlands, or at playa margins. Spaceborne Digital Elevation Models (DEMs) provide potential for assessing subtle topographic and hydrodynamic changes in these low-gradient, low-relief settings, but challenges are posed by limitations in vertical and horizontal accuracy. This study evaluates the use of different spaceborne DEMs for topographic characterisation and flood flow modelling of the low-gradient (

Open-access global Digital Elevation Models (DEM) have been crucial in enabling flood studies in data-sparse areas. Poor resolution (>30 m), significant vertical errors and the fact that these DEMs are over a decade old continue to hamper our ability to accurately estimate flood hazard. The limited availability of high-accuracy DEMs dictate that dated open-access global DEMs are still used extensively in flood models, particularly in data-sparse areas. Nevertheless, high-accuracy DEMs have been found to give better flood estimations, and thus can be considered a ‘must-have’ for any flood model. A high-accuracy open-access global DEM is not imminent, meaning that editing or stochastic simulation of existing DEM data will remain the primary means of improving flood simulation. This article provides an overview of errors in some of the most widely used DEM data sets, along with the current advances in reducing them via the creation of new DEMs, editing DEMs and stochastic simulation of DEMs. We focus on a geostatistical approach to stochastically simulate floodplain DEMs from several open-access global DEMs based on the spatial error structure. This DEM simulation approach enables an ensemble of plausible DEMs to be created, thus avoiding the spurious precision of using a single DEM and enabling the generation of probabilistic flood maps. Despite this encouraging step, an imprecise and outdated global DEM is still being used to simulate elevation. To fundamentally improve flood estimations, particularly in rapidly changing developing regions, a high-accuracy open-access global DEM is urgently needed, which in turn can be used in DEM simulation.