Spread/Error relationship and spatial error structure of precipitation ensemble nowcasting: Comparison of STEPS and generative AI

TL;DR

This study compares the spatial error structures and spread/error relationships of STEPS and LDCast ensemble nowcasting models for precipitation, revealing their limitations in spatial error localization despite informative ensemble statistics.

Contribution

It provides a comparative analysis of traditional and AI-based ensemble nowcasting models, highlighting their similar underdispersion and limited spatial error localization capabilities.

Findings

Both models are slightly underdispersive.

Ensemble spread estimates errors well across scales.

Limited ability to spatially localize ensemble mean errors.

Abstract

The predictability of the generative AI-based nowcasting model LDCast (trained on another region) is evaluated over Belgium, together with the pysteps implementation of the nowcasting algorithm STEPS. STEPS and LDCast are slightly underdispersive, but the ensemble spread provides an estimation of the error at almost all scales. Both models adapt the properties of their ensembles to the type of event, either convective or stratiform. The spatial scores of the STEPS and LDCast ensembles are compared with those of surrogate ensembles having some key properties, revealing that both STEPS and LDCast have very little ability to spatially localise the ensemble mean error vector through their ensemble members. This suggests that the content of STEPS and LDCast ensembles is informative in terms of statistics, but not in terms of dynamics.