TempEE: Temporal-Spatial Parallel Transformer for Radar Echo Extrapolation Beyond Autoregression

Meteorological radar reflectivity data (i.e., radar echo) significantly influences precipitation prediction. It can facilitate accurate and expeditious forecasting of short-term heavy rainfall bypassing the need for complex numerical weather prediction (NWP) models. In comparison to conventional models, deep-learning (DL)-based radar echo extrapolation algorithms exhibit higher effectiveness and efficiency. Nevertheless, the development of a reliable and generalized echo extrapolation algorithm is impeded by three primary challenges: cumulative error spreading, imprecise representation of sparsely distributed echoes, and inaccurate description of nonstationary motion processes. To tackle these challenges, this article proposes a novel radar echo extrapolation algorithm called temporal-spatial parallel transformer, referred to as TempEE. TempEE avoids using autoregression and instead employs a one-step forward strategy to prevent the cumulative error from spreading during the extrapolation process. Additionally, we propose the incorporation of a multilevel temporal-spatial attention mechanism to improve the algorithm's capability of capturing both global and local information while emphasizing task-related regions, including sparse echo representations, in an efficient manner. Furthermore, the algorithm extracts spatio-temporal representations from continuous echo images using a parallel encoder to model the nonstationary motion process for echo extrapolation. The superiority of our TempEE has been demonstrated in the context of the classic radar echo extrapolation task, utilizing a real-world dataset. Extensive experiments have further validated the efficacy and indispensability of various components within TempEE.

Automated summary

This article proposes a novel radar echo extrapolation algorithm called TempEE, which tackles the challenges in radar echo extrapolation by avoiding cumulative error, incorporating multilevel temporal-spatial attention mechanism, and using a parallel encoder. Extensive experiments on a real-world dataset have demonstrated the effectiveness and indispensability of TempEE.