5:15 PM - 7:15 PM
[ACG37-P06] Improving CaMa flood model simulations using ICESat-2 satellite observations
Keywords:Satellite, Floods, Remote Sensing, Water Level, River
Hydrological and hydrodynamic models often depend on in situ observations for parameter calibration, which becomes problematic when such data are unavailable. Accurate monitoring of river water surface elevation (WSE) is critical for calibrating these models, particularly in regions with limited gauge data. Additionally, high accuracy in simulating one variable, such as water depth can help in reducing uncertainties in hydrodynamic variables like flooded area extent. To address these challenges, we employed ICESat-2 remote sensing data to develop parameter calibration schemes globally.
Using the ICESat-2 ATL13 data product from 2018-2024, we retrieved river WSE to enhance model calibration, enabling parameter adjustments. The satellite derived WSE offers high spatiotemporal resolution, even in narrower rivers compared to traditional altimetry measurements. This approach improves simulation performance and flood extent validation. The study demonstrates ICESat-2's capability for broader spatial coverage in retrieving river levels and precisely gauging rivers.
The CaMa flood model is run at 6-arc-minute resolution and provides water level elevation from 2018 to 2024. These simulations are compared with ICESat-2 observations for specific matching dates, using mean or median values of multiple satellite observations for individual river network points. Model performance is assessed using standard metrics such as correlations and absolute bias, with outlier points filtered out. Results indicate good correlations and low bias in most Asian river systems, despite the limited availability of matching dates. This research establishes a performance benchmark for models and provides insights into parameter sensitivity. It underscores the importance of leveraging remote sensing data for parameter calibration, thereby refining hydrodynamic simulations and enhancing flood simulation accuracy. Integrating with water levels from the recent SWOT mission will further enhance this study.
Using the ICESat-2 ATL13 data product from 2018-2024, we retrieved river WSE to enhance model calibration, enabling parameter adjustments. The satellite derived WSE offers high spatiotemporal resolution, even in narrower rivers compared to traditional altimetry measurements. This approach improves simulation performance and flood extent validation. The study demonstrates ICESat-2's capability for broader spatial coverage in retrieving river levels and precisely gauging rivers.
The CaMa flood model is run at 6-arc-minute resolution and provides water level elevation from 2018 to 2024. These simulations are compared with ICESat-2 observations for specific matching dates, using mean or median values of multiple satellite observations for individual river network points. Model performance is assessed using standard metrics such as correlations and absolute bias, with outlier points filtered out. Results indicate good correlations and low bias in most Asian river systems, despite the limited availability of matching dates. This research establishes a performance benchmark for models and provides insights into parameter sensitivity. It underscores the importance of leveraging remote sensing data for parameter calibration, thereby refining hydrodynamic simulations and enhancing flood simulation accuracy. Integrating with water levels from the recent SWOT mission will further enhance this study.