Deep Temporal Interpolation of Radar-based Precipitation
This work addresses flood risk estimation for hydrological models, but it is incremental as it builds on existing interpolation methods with specific data integration.
The paper tackles the problem of interpolating precipitation at high temporal resolutions for flood risk simulation by using deep neural networks to interpolate radar images, achieving up to 20% error reduction compared to a linear baseline.
When providing the boundary conditions for hydrological flood models and estimating the associated risk, interpolating precipitation at very high temporal resolutions (e.g. 5 minutes) is essential not to miss the cause of flooding in local regions. In this paper, we study optical flow-based interpolation of globally available weather radar images from satellites. The proposed approach uses deep neural networks for the interpolation of multiple video frames, while terrain information is combined with temporarily coarse-grained precipitation radar observation as inputs for self-supervised training. An experiment with the Meteonet radar precipitation dataset for the flood risk simulation in Aude, a department in Southern France (2018), demonstrated the advantage of the proposed method over a linear interpolation baseline, with up to 20% error reduction.