Spread/error relationship and spatial error representation in precipitation nowcasting: comparison of STEPS and generative AI

Spread/error relationship and spatial error representation in precipitation nowcasting: comparison of STEPS and generative AI
Martin Bonte, Lesley De Cruz, Fabian Debal, and Stéphane Vannitsem
Weather Clim. Dynam., 7, 1837–1852, https://doi.org/10.5194/wcd-7-1837-2026, 2026
The predictability of the generative AI-based nowcasting model LDCast is evaluated over Belgium, together with the pysteps implementation of the nowcasting algorithm STEPS. It appears that the ensembles of both models correctly estimate the error size through their spread, but fail at spatially representing the error. The analysis is done for two dynamically different types of events, showing how the models adapt their ensembles depending on the situation.

Source: https://doi.org/10.5194/wcd-7-1837-2026