What question did this study set out to answer?

This research aims to enhance radar echo extrapolation for better short-term weather forecasting by integrating advanced deep learning techniques.

May 25, 2026

Global‐Frequency Synergy: A Novel Paradigm for Radar Echo Extrapolation via Attention and Fourier Convolution

Puntos clave

This research aims to enhance radar echo extrapolation for better short-term weather forecasting by integrating advanced deep learning techniques.
Proposed a new model called Global-Frequency Spatiotemporal Long Short-Term Memory (GFST-LSTM).
Incorporated global attention mechanisms and Fourier convolutional modules into Spatiotemporal LSTM architecture.
Evaluated performance on Moving MNIST benchmark and Jiangsu Province radar data from 2019 to 2021.
GFST-LSTM improved Critical Success Index by 22.9% and Heidke Skill Score by 13.1% over Predictive Recurrent Neural Network at the 40 dBZ threshold.
Demonstrated a 6.6% reduction in positional bias compared to the Motion Gated Recurrent Unit during 60–120 min predictions.
Ablation studies illustrated the significant impact of both the attention mechanism and Fourier convolution modules.

Resumen

Abstract Accurate radar echo extrapolation is critical for short‐term weather forecasting, yet existing deep learning methods often suffer from echo ambiguity, intensity decay, and insufficient global context utilization. To address these limitations, this paper proposes Global‐Frequency Spatiotemporal Long Short‐Term Memory ( GFST‐LSTM), a novel model that integrates a global attention mechanism and Fourier convolutional modules into the Spatiotemporal LSTM ( ST‐LSTM) architecture. The attention module dynamically weights multi‐scale spatiotemporal features by enhancing channel and spatial correlations, while the Fourier convolution module captures global periodic patterns via frequency‐domain transformations. Evaluated on the Moving Modified National Institute of Standards and Technology database (Moving MNIST) benchmark and Jiangsu Province radar data sets (2019–2021), GFST‐LSTM achieves a 22.9% improvement in Critical Success Index and 13.1% in Heidke Skill Score over Predictive Recurrent Neural Network at the 40 dBZ threshold. Notably, it excels in preserving strong echo regions during 60–120 min predictions, reducing positional bias by 6.6% compared to the Motion Gated Recurrent Unit (MotionGRU). Ablation studies confirm the synergistic effect of both modules, with the full model outperforming variants that lack either component.

Preguntar a la IA

Me gusta

Guardar