Abstract's details
Joint training of hydrologic and hydraulic models Using Deep Learning and SWOT Pixel Cloud Data for the Torne River
Event: 2025 SWOT Science Team Meeting
Session: Hydrology: Global Hydrology Modeling Working Group
Presentation type: Poster
Floods are among the most devastating natural disasters, affecting both developed and developing regions. However, developing countries often lack sufficient monitoring and early warning systems, making them more vulnerable. The ESA EO4FLOOD project seeks to enhance flood forecasting by integrating satellite data with hydrologic and hydraulic models (Tarpanelli et al., 2025). Within this effort, we introduce a novel, joint modelling framework that couples hydrologic and hydraulic models using a deep learning (DL) approach.
The Surface Water and Ocean Topography (SWOT) mission is the first satellite to provide 2D spatially distributed water surface elevation (WSE) data globally, complimenting the global water gauge network through its ability to produce spatially continuous and consistent WSE measurements (Pavelsky et al. 2014). It has a 21-day orbit, with more frequent revisit times depending on latitude. SWOT’s ability to deliver consistent measurements of WSE, water surface slope (WSS), and river width enables a new era in discharge estimation (Durand et al., 2023).
Our framework leverages physics-informed deep learning to integrate large-scale Earth observation (EO) data while maintaining physical consistency. The hydrologic and hydraulic models are trained against SWOT pixel cloud data, using the output of the hydrologic model as the input to the hydraulic model. Joint training allows both models to benefit from the information contained in the SWOT data (WSE, WSS), and, potentially, satellite earth observations of additional state variables (e.g., soil moisture, evapotranspiration, terrestrial water storage).
We demonstrate this approach on the Torne River, located between northern Sweden and Finland. With extensive in-situ data, Torne provides an ideal case for validation. Our joint model supports accurate water level and discharge forecasting, aiding flood preparedness, informing local adaptation strategies, and enhancing climate resilience. This proof of concept highlights the method’s global potential under the EO4FLOOD initiative.
Durand, M., Gleason, C. J., Pavelsky, T. M., Prata de Moraes Frasson, R., Turmon, M., David, C. H., et al. (2023). A framework for estimating global river discharge from the Surface Water and Ocean Topography satellite mission. Water Resources Research, 59, e2021WR031614. https://doi.org/10.1029/2021WR031614
Pavelsky, T. M., Durand, M. T., Andreadis, K. M., Beighley, R. E., Paiva, R. C. D., Allen, G. H., & Miller, Z. F. (2014). Assessing the potential global extent of SWOT river discharge observations. Journal of Hydrology, 519, 1516–1525. https://doi.org/10.1016/j.jhydrol.2014.08.044
Tarpanelli, A., Schumann, G., and Kittel, C. and the EO4FLOOD team: Earth Observation data for Advancing Flood Forecasting: EO4FLOOD project, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6671, https://doi.org/10.5194/egusphere-egu25-6671, 2025.
Back to the list of abstractThe Surface Water and Ocean Topography (SWOT) mission is the first satellite to provide 2D spatially distributed water surface elevation (WSE) data globally, complimenting the global water gauge network through its ability to produce spatially continuous and consistent WSE measurements (Pavelsky et al. 2014). It has a 21-day orbit, with more frequent revisit times depending on latitude. SWOT’s ability to deliver consistent measurements of WSE, water surface slope (WSS), and river width enables a new era in discharge estimation (Durand et al., 2023).
Our framework leverages physics-informed deep learning to integrate large-scale Earth observation (EO) data while maintaining physical consistency. The hydrologic and hydraulic models are trained against SWOT pixel cloud data, using the output of the hydrologic model as the input to the hydraulic model. Joint training allows both models to benefit from the information contained in the SWOT data (WSE, WSS), and, potentially, satellite earth observations of additional state variables (e.g., soil moisture, evapotranspiration, terrestrial water storage).
We demonstrate this approach on the Torne River, located between northern Sweden and Finland. With extensive in-situ data, Torne provides an ideal case for validation. Our joint model supports accurate water level and discharge forecasting, aiding flood preparedness, informing local adaptation strategies, and enhancing climate resilience. This proof of concept highlights the method’s global potential under the EO4FLOOD initiative.
Durand, M., Gleason, C. J., Pavelsky, T. M., Prata de Moraes Frasson, R., Turmon, M., David, C. H., et al. (2023). A framework for estimating global river discharge from the Surface Water and Ocean Topography satellite mission. Water Resources Research, 59, e2021WR031614. https://doi.org/10.1029/2021WR031614
Pavelsky, T. M., Durand, M. T., Andreadis, K. M., Beighley, R. E., Paiva, R. C. D., Allen, G. H., & Miller, Z. F. (2014). Assessing the potential global extent of SWOT river discharge observations. Journal of Hydrology, 519, 1516–1525. https://doi.org/10.1016/j.jhydrol.2014.08.044
Tarpanelli, A., Schumann, G., and Kittel, C. and the EO4FLOOD team: Earth Observation data for Advancing Flood Forecasting: EO4FLOOD project, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6671, https://doi.org/10.5194/egusphere-egu25-6671, 2025.