2025-08-26 ワシントン州立大学（WSU）
＜関連情報＞
- https://news.wsu.edu/press-release/2025/08/26/wsu-study-projects-increases-in-lightning-wildfire-risk/
- https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025EF006108
- https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2024JD042147
米国西部における落雷起因山火事リスクの予測 Projections of Lightning-Ignited Wildfire Risk in the Western United States
Dmitri A. Kalashnikov, John T. Abatzoglou, Frances V. Davenport, Zachary M. Labe, Paul C. Loikith, Danielle Touma, Deepti Singh
Earth’s Future Published: 26 August 2025
DOI:https://doi.org/10.1029/2025EF006108
Abstract
Cloud-to-ground (CG) lightning is a major source of summer wildfire ignition in the western United States (WUS). However, future projections of lightning are uncertain since lightning is not directly simulated by most global climate models. To address this issue, we use convolutional neural network (CNN)-based parameterizations of daily June-September CG lightning. CNN parameterizations of daily CG lightning occurrence at each grid cell use fields of three thermodynamic variables—ratio of surface Moist Static Energy (MSE) to 500 hPa saturation MSE, 700–500 hPa lapse rate, and 500 hPa relative humidity. Applying these parameterizations to the Community Earth System Model version 2 Large Ensemble, we find widespread increases in CG lightning days across much of the region by the mid-21st century (2031–2060) under a moderate warming scenario. Projected increases are pronounced in the northern WUS where many grid cells experience 4–12 additional CG lightning days compared to 1995–2022 and are driven by increases in all three thermodynamic variables. To assess the risk of lightning-ignited wildfire (LIW) ignition, we also quantify the concurrence of CG lightning with high Fire Weather Index (FWI) days. By 2031–2060, CG lightning will coincide more frequently with high FWI, but the magnitude of increases relative to CG lightning days varies across the region. Future projections of CG lightning and LIW risk can be useful for understanding the changing risks of associated hazards, and guide wildland fire management and suppression planning.
Key Points
- Increases in cloud-to-ground lightning days are projected across the western U.S. in CESM2-LENS2, particularly in the interior Northwest
- Increases in lightning are driven by rises in moisture and instability indices used for convolutional neural network-based lightning models
- Lightning-ignited wildfire risk grows as high Fire Weather Index days increasingly align with lightning days
Plain Language Summary
Cloud-to-ground lightning is a major source of wildfire ignition during the summer in the western United States, but future projections of lightning and lightning-ignited wildfire (LIW) have been limited. We use a machine learning technique—convolutional neural networks–to predict lightning based on three meteorological variables. These variables describe aspects of atmospheric moisture and vertical instability and therefore capture conditions favorable for lightning occurrence. We then apply these machine learning-based models to output from GCM simulations to project cloud-to-ground lightning days in the future. Our projections show an increase in cloud-to-ground lightning days in the mid-21st century (2031–2060), especially in the interior northwestern United States. These increases are driven by widespread projected increases in all three meteorological variables used for lightning prediction. We also project an increased likelihood of cloud-to-ground lightning occurring on days with meteorological conditions favorable for wildfires, thus increasing the risk of LIWs. These findings are important for understanding changes to LIW risk, and for planning wildland fire management and suppression needs in a warming climate.
説明可能なニューラルネットワークを用いた米国西部における大規模環境からの雲地間落雷予測 Predicting Cloud-To-Ground Lightning in the Western United States From the Large-Scale Environment Using Explainable Neural Networks
Dmitri A. Kalashnikov, Frances V. Davenport, Zachary M. Labe, Paul C. Loikith, John T. Abatzoglou, Deepti Singh
Journal of Geophysical Research: Atmospheres Published: 22 November 2024
DOI:https://doi.org/10.1029/2024JD042147
Abstract
Lightning is a major source of wildfire ignition in the western United States (WUS). We build and train convolutional neural networks (CNNs) to predict the occurrence of cloud-to-ground (CG) lightning across the WUS during June–September from the spatial patterns of seven large-scale meteorological variables from reanalysis (1995–2022). Individually trained CNN models at each 1° × 1° grid cell (n = 285 CNNs) show high skill at predicting CG lightning days across the WUS (median AUC = 0.8) and perform best in parts of the interior Southwest where summertime CG lightning is most common. Further, interannual correlation between observed and predicted CG lightning days is high (median r = 0.87), demonstrating that locally trained CNNs realistically capture year-to-year variation in CG lightning activity across the WUS. We then use layer-wise relevance propagation (LRP) to investigate the relevance of predictor variables to successful CG lightning prediction in each grid cell. Using maximum LRP values, our results show that two thermodynamic variables—ratio of surface moist static energy to free-tropospheric saturation moist static energy, and the 700–500 hPa lapse rate—are the most relevant CG lightning predictors for 93%–96% of CNNs depending on the LRP variant used. As lightning is not directly simulated by global climate models, these CNNs could be used to parameterize CG lightning in climate models to assess changes in future CG lightning occurrence with projected climate change. Understanding changes in CG lightning risk and consequently lightning-caused wildfire risk across the WUS could inform fire management, planning, and disaster preparedness.
Key Points
- Convolutional Neural Networks (CNNs) trained on seven meteorological variables show skill at predicting CG lightning days in the WUS, performing best in the Southwest
- Year-to-year variation in lightning activity is robustly captured by the CNNs (domain median r = 0.87)
- MSE ratio and 700–500 hPa lapse rate are the top CG lightning predictors in most grid cells
Plain Language Summary
Lightning is a major source of wildfire ignition in the western U.S. We use a machine learning technique called “Convolutional Neural Networks,” or CNNs, to predict the occurrence of cloud-to-ground lightning across the western U.S. Our CNN models use seven meteorological variables that are known to be important for lightning activity and are trained over the summer season (June–September) during 1995–2022. CNN models are trained at each individual latitude by longitude grid box and show high skill at predicting cloud-to-ground lightning days across the western U.S., especially in the interior Southwest where summertime lightning is most common. Further, we show that CNNs realistically predict year-to-year variation of cloud-to-ground lightning across the region. We also quantify the relevance of each of the seven meteorological variables to successful cloud-to-ground lightning predictions at each grid cell. Our results show that two variables, describing aspects of atmospheric moisture and vertical instability, are most relevant for successful cloud-to-ground lightning predictions by the CNNs. Since lightning is not directly simulated by global climate models, the CNNs can be applied to climate model output to quantify future changes in cloud-to-ground lightning occurrence, and consequently lightning-caused wildfire risk across the western U.S.
