Lightning is a significant natural hazard that poses considerable risks to both human safety and industrial operations.Accurate,fine-scale lightning forecasting is crucial for effective disaster prevention.Traditional...Lightning is a significant natural hazard that poses considerable risks to both human safety and industrial operations.Accurate,fine-scale lightning forecasting is crucial for effective disaster prevention.Traditional forecasting methods primarily rely on numerical weather prediction(NWP),which demands substantial computational resources to solve complex atmospheric evolution equations.Recently,deep learning-based weather prediction models—particularly weather foundation models(WFMs)—have demonstrated promising results,achieving performance comparable to NWP while requiring substantially fewer computational resources.However,existing WFMs are unable to directly generate lightning forecasts and struggle to satisfy the high spatial resolution required for fine-scale prediction.To address these limitations,this paper investigates a fine-scale lightning forecasting approach based on WFMs and proposes a dual-source data-driven forecasting framework that integrates the strengths of both WFMs and recent lightning observations to enhance predictive performance.Furthermore,a gated spatiotemporal fusion network(gSTFNet)is designed to address the challenges of cross-temporal and cross-modal fusion inherent in dual-source data integration.gSTFNet employs a dual-encoding structure to separately encode features from WFMs and lightning observations,effectively narrowing the modal gap in the latent feature space.A gated spatiotemporal fusion module is then introduced to model the spatiotemporal correlations between the two types of features,facilitating seamless cross-temporal fusion.The fused features are subsequently processed by a deconvolutional network to generate accurate lightning forecasts.We evaluate the proposed gSTFNet using real-world lightning observation data collected in Guangdong from 2018 to 2022.Experimental results demonstrate that:(1)In terms of the ETS score,the dual-source framework achieves a 50% improvement over models trained solely on WFMs,and a 300% improvement over the HRES lightning forecasting product released by the European Centre for Medium-Range Weather Forecasts(ECMWF);(2)gSTFNet outperforms several state-of-the-art deep learning baselines that utilize dual-source inputs,clearly demonstrating superior forecasting accuracy.展开更多
Thunderstorms are severe convective weather systems generating lightning,which can lead to various catastrophic weather when a large amount of lightning is produced.In the past decade,high spatiotemporal resolution li...Thunderstorms are severe convective weather systems generating lightning,which can lead to various catastrophic weather when a large amount of lightning is produced.In the past decade,high spatiotemporal resolution lightning detection technology has been developed,which has laid a solid foundation for investigating the propagation and development mechanism of lightning as well as associated physical effects.Based on the Doppler dual polarization weather radar and high-resolution numerical models,thunderstorm dynamics,microphysics,electrical processes,and their interactions have been well investigated,and some new insights into the thunderstorm charge distribution and its relation to the thunderstorm structure have been obtained.All these have promoted the lightning forecasting and lightning data assimilation.This paper reviews the recent research progress in detection,mechanism,and forecasting of thunderstorms and lightning in China in the last decade from four aspects:1)high-resolution three-dimensional(3D)lightning mapping technology and application,2)lightning in different thunderstorms and its relationship with cloud dynamics and microphysics,3)observation and simulation of lightning charge structure in thunderstorms,and 4)lightning prediction and lightning data assimilation for thunderstorm forecasting.Major challenges and the cuttingedge research directions in lightning and thunderstorms studies are also highlighted.展开更多
基金supported by the Open Grants of Key Laboratory of Lightning,China Meteorological Administration(Grant Nos.2023KELL-B002 and 2024KELL-A001)the National Natural Science Foundation of China(Grant Nos.62306028,42075088 and U2342215)the Science and Technology Program of Shenzhen,China(Grant No.KJZD20240903102742055)。
文摘Lightning is a significant natural hazard that poses considerable risks to both human safety and industrial operations.Accurate,fine-scale lightning forecasting is crucial for effective disaster prevention.Traditional forecasting methods primarily rely on numerical weather prediction(NWP),which demands substantial computational resources to solve complex atmospheric evolution equations.Recently,deep learning-based weather prediction models—particularly weather foundation models(WFMs)—have demonstrated promising results,achieving performance comparable to NWP while requiring substantially fewer computational resources.However,existing WFMs are unable to directly generate lightning forecasts and struggle to satisfy the high spatial resolution required for fine-scale prediction.To address these limitations,this paper investigates a fine-scale lightning forecasting approach based on WFMs and proposes a dual-source data-driven forecasting framework that integrates the strengths of both WFMs and recent lightning observations to enhance predictive performance.Furthermore,a gated spatiotemporal fusion network(gSTFNet)is designed to address the challenges of cross-temporal and cross-modal fusion inherent in dual-source data integration.gSTFNet employs a dual-encoding structure to separately encode features from WFMs and lightning observations,effectively narrowing the modal gap in the latent feature space.A gated spatiotemporal fusion module is then introduced to model the spatiotemporal correlations between the two types of features,facilitating seamless cross-temporal fusion.The fused features are subsequently processed by a deconvolutional network to generate accurate lightning forecasts.We evaluate the proposed gSTFNet using real-world lightning observation data collected in Guangdong from 2018 to 2022.Experimental results demonstrate that:(1)In terms of the ETS score,the dual-source framework achieves a 50% improvement over models trained solely on WFMs,and a 300% improvement over the HRES lightning forecasting product released by the European Centre for Medium-Range Weather Forecasts(ECMWF);(2)gSTFNet outperforms several state-of-the-art deep learning baselines that utilize dual-source inputs,clearly demonstrating superior forecasting accuracy.
基金Supported by the National Natural Science Foundation of China(42230609)Chinese Academy of Sciences Strategic Leading Science and Technology Project(XDB0760300)。
文摘Thunderstorms are severe convective weather systems generating lightning,which can lead to various catastrophic weather when a large amount of lightning is produced.In the past decade,high spatiotemporal resolution lightning detection technology has been developed,which has laid a solid foundation for investigating the propagation and development mechanism of lightning as well as associated physical effects.Based on the Doppler dual polarization weather radar and high-resolution numerical models,thunderstorm dynamics,microphysics,electrical processes,and their interactions have been well investigated,and some new insights into the thunderstorm charge distribution and its relation to the thunderstorm structure have been obtained.All these have promoted the lightning forecasting and lightning data assimilation.This paper reviews the recent research progress in detection,mechanism,and forecasting of thunderstorms and lightning in China in the last decade from four aspects:1)high-resolution three-dimensional(3D)lightning mapping technology and application,2)lightning in different thunderstorms and its relationship with cloud dynamics and microphysics,3)observation and simulation of lightning charge structure in thunderstorms,and 4)lightning prediction and lightning data assimilation for thunderstorm forecasting.Major challenges and the cuttingedge research directions in lightning and thunderstorms studies are also highlighted.
