Estimating spatial and temporal variations of surface waters is important for water resources management. The upcoming Surface Water and Ocean Topography (SWOT) mission will enhance our understanding on global water cycle by measuring water surface elevations at a high resolution. It will be beneficial to combine SWOT observations to hydrodynamic modelling to overcome its limited observation frequency. But one of the major concern of the hydrodynamic modelling is the uncertainty in the bathymetry of hydrodynamic models. Empirical methods were used to estimate the river bottom in most of the river hydrodynamic models. We performed an observing system simulation experiment for estimating river channel bathymetry from water surface elevation (WSE) measured by SWOT. A Local Ensemble Transform Kalman Filter (LETKF) assimilation algorithm was applied to the CaMa-Flood hydrodynamic model to estimate WSE and bathymetry simultaneously via state-parameter estimation. The assimilation was done once in 21 days in order to utilize at least one SWOT observation. Synthetic SWOT observations were generated from a “true” CaMa-Flood simulation based on the original bathymetry while the corrupted model used channel bathymetry parameters different from the true model. The corrupted channel bathymetry was modelled using the average river water depth of the previous year of assimilation and an exponential correlation function. We found that the assimilation significantly improved river discharge estimations in continental-scale rivers. The LETKF via state-parameter estimation succeeded in improving bathymetry close to true using SWOT observations. Furthermore, the river discharge was also reasonably estimated. These results indicate the potential of the future SWOT mission to spatially and temporally estimate river bathymetry and river discharge on a global scale.
各種検索
サポート
ヘルプ