Extreme Precipitation Nowcasting using Transformer-based Generative Models

Mircea Tudor Lică

Abstract: This paper presents an innovative approach to extreme precipitation now- casting by employing Transformer-based generative models, namely Now- castingGPT with Extreme Value Loss (EVL) regularization. Leveraging a comprehensive dataset from the Royal Netherlands Meteorological In- stitute (KNMI), our study focuses on predicting short-term precipitation with high accuracy. We introduce a novel method for computing EVL with- out assuming fixed extreme representations, addressing the limitations of current models in capturing extreme weather events. We present both qual- itative and quantitative analyses, demonstrating the superior performance of the proposed NowcastingGPT-EVL in generating accurate precipitation forecasts, especially when dealing with extreme precipitation events. The code is available at https://github.com/Cmeo97/NowcastingGPT.