|Paper title||Informing regional 1D/2D flood inundation models with Icesat-2: A case study in the Amur river|
|Form of presentation||Poster|
Flooding is one of the most damaging natural hazards, causing economic losses and threats to livelihoods and human health. Predicting flooding patterns in river systems is a challenge in large and remote areas. Climate change has intensified the occurrence of severe flood events, and accurate predictive models are required for flood risk management. However, the available in-situ data in remote areas and ungauged river basins is scarce and often not available in the public domain. The accuracy of hydrodynamic models is limited by the quality of the available observations, which are essential to calibrate unobserved or unobservable model parameters.
Accurate topographic elevation measurements are essential to replicate 1D/2D flood processes. Satellite based DEMs have the advantage of providing large coverage in remote areas. However, high resolution DEMs are not always available and therefore it is necessary to use lower resolution products. Missions such as Shuttle Radar Topography Mission (SRTM) or ALOS-PALSAR offer DEMs down to 1-Arcsec resolution freely available. When used as input to hydraulic models, such DEMs can lead to large errors in simulated water surface elevation and surface water extent, due to the complex topography of floodplains that is normally hard to map precisely. To better integrate this data in the model, a finer resolution product will be needed to map the floodplain topography and river bathymetry. The novel altimetry mission ICEsat-2, operating since 2018, offers a large spatial coverage, with an along track resolution down to 70 cm in its photon cloud product ATL03. This data has shown great potential when mapping river topography, identifying narrow river structures and also when multi channel rivers and braided structures are present. ATL03 can be used as a control point dataset, to correct the biases and refine DEMs.
In this study we use a 1D hydraulic model derived from ICEsat-2 data to characterize the river bed geometry of the main channel. We use ATL03 product to map the topography of the river channel, providing accurate data on river bed geometry. We calibrate depth and manning roughness against ATL13 water surface elevation observations, which is the inland water product from ICEsat-2. The water surface elevation is simulated with an accuracy to decimeter level providing a precise river bathymetry characterization. To study the water surface elevation in the floodplain, we combine the SRTM DEM 1-Arcsec resolution with ATL03 cross-sections, to reduce elevation errors. With the refined DEM, we run Mike Flood 1D/2D inundation module, and validate the simulated inundation areas with flood maps from Global Surface Water Explorer.
The developed workflow is demonstrated for sections of Amur river, which flows in China and Russia. This river is characterized by large floodplains and for having a braided structure, making it a suitable study case to demonstrate our methodology.