LightNet plus : A dual-source lightning forecasting network with bi-direction spatiotemporal transformation

Lightning disaster causes a huge threat to human lives and industrial facilities. Data-driven lightning forecasting plays an effective role in alleviating such disaster losses. The forecasting process usually faces multi-source meteorological data characterized by spatiotemporal structure. However, established data-driven forecasting methods are mostly built on classic convolutional and recurrent neural blocks which processes one local neighborhood at a time, failing to capture long-range spatiotemporal dependencies within data. To address this issue, we propose a dual-source lightning forecasting network with bi-direction spatiotemporal transformation, referred to as LightNet+. The core of LightNet+ is a novel module, namely bidirectional spatiotemporal propagator, which aims to model long-range connections among different spatiotemporal locations, going beyond the constraints of the receptive field of a local neighborhood. Moreover, a spatiotemporal encoder is introduced to extract historical trend information from recent observation data. Finally, all the obtained features are organically fused via a non-local spatiotemporal decoder, which then produces final forecasting results. We evaluate LightNet+ on a real-world lightning dataset from North China and compare it with several state-of-the-art data-driven lightning forecasting methods. Experimental results show that the proposed LightNet+ yields overall best performance.

Automated summary

Lightning disaster poses a significant threat to human lives and industrial facilities. Data-driven lightning forecasting has proven effective in reducing such disaster losses. However, current methods fail to capture the long-range spatiotemporal dependencies within data. To address this issue, we propose a dual-source lightning forecasting network called LightNet+, which utilizes bi-directional spatiotemporal transformation to model long-range connections and extract historical trend information for accurate predictions.