Lake geoid correction grids based on the SWOT Raster product

It is well known that the geoid signals over large lakes in mountainous regions are not well modeled. This introduces errors in the satellite altimetry-based water level estimates. The Surface Water and Ocean Topography mission (SWOT) launched in December 2022 provides almost complete spatial coverage, which makes it possible to detect spatial signals over lakes, for example.

Here we apply a state-space model to separate the assumed static signal from a potentially unmodeled geoid signal and the temporal variation in the water level. As data input, we use the SWOT 250m raster product. To evaluate how well the spatial and temporal signals are separated we compare the SWOT-based water level time series with time series based on other altimetry data sets.

The procedure is tested for a set of different lakes, including the African rift lakes and Lake Titicaca. Here, we based the solution on available data at the current time; however, the solutions are expected to improve as more data becomes available. The R code will be available to the user.