An official website of the United States government
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.
more »
« less
Predicting Cloud‐To‐Ground Lightning in the Western United States From the Large‐Scale Environment Using Explainable Neural Networks
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 (medianr = 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.
more »
« less
- Award ID(s):
- 2019762
- PAR ID:
- 10597104
- Publisher / Repository:
- AGU
- Date Published:
- Journal Name:
- Journal of Geophysical Research: Atmospheres
- Volume:
- 129
- Issue:
- 22
- ISSN:
- 2169-897X
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Cloud‐to‐ground lightning with minimal rainfall (“dry” lightning) is a major wildfire ignition source in the western United States (WUS). Although dry lightning is commonly defined as occurring with <2.5 mm of daily‐accumulated precipitation, a rigorous quantification of precipitation amounts concurrent with lightning‐ignited wildfires (LIWs) is lacking. We combine wildfire, lightning and precipitation data sets to quantify these ignition precipitation amounts across ecoprovinces of the WUS. The median precipitation for all LIWs is 2.8 mm but varies with vegetation and fire characteristics. “Holdover” fires not detected until 2–5 days following ignition occur with significantly higher precipitation (5.1 mm) compared to fires detected promptly after ignition (2.5 mm), and with cooler and wetter environmental conditions. Further, there is substantial variation in precipitation associated with promptly‐detected (1.7–4.6 mm) and holdover (3.0–7.7 mm) fires across ecoprovinces. Consequently, the widely‐used 2.5 mm threshold does not fully capture lightning ignition risk and incorporating ecoprovince‐specific precipitation amounts would better inform WUS wildfire prediction and management.more » « less
-
While overall impacts of heatwaves have been extensively studied, the connection between heatwaves and wildfire activity remains relatively underexplored. We analyze links between heatwaves and both wildfire occurrence and growth across the western United States (WUS) and find that 42% of burned area during 2001–2024 occurred during and immediately following heatwaves. Heatwaves facilitate significant increases in daily burned area through meteorological and fuel flammability conditions that promote new ignitions and exacerbate ongoing fire activity, with effects persisting post-heatwave in most regions. In addition, heatwaves co-occur with increased cloud-to-ground lightning that can potentially increase ignitions. Last, we observe a 2.5-fold increase in burned area in WUS forests since 2001, with ~64% of this increase coinciding with heatwaves, but without corresponding increases in nonforests. The growing influence of heatwaves in shaping burned area in WUS forests has important implications for fire management and public health and can improve predictions of wildfire risk.more » « less
-
Abstract. The annual area burned due to wildfires in the western United States (WUS) increased bymore than 300 % between 1984 and 2020. However, accounting for the nonlinear, spatially heterogeneous interactions between climate, vegetation, and human predictors driving the trends in fire frequency and sizes at different spatial scales remains a challenging problem for statistical fire models. Here we introduce a novel stochastic machine learning (SML) framework, SMLFire1.0, to model observed fire frequencies and sizes in 12 km × 12 km grid cells across the WUS. This framework is implemented using mixture density networks trained on a wide suite of input predictors. The modeled WUS fire frequency matches observations at both monthly (r=0.94) and annual (r=0.85) timescales, as do the monthly (r=0.90) and annual (r=0.88) area burned. Moreover, the modeled annual time series of both fire variables exhibit strong correlations (r≥0.6) with observations in 16 out of 18 ecoregions. Our ML model captures the interannual variability and the distinct multidecade increases in annual area burned for both forested and non-forested ecoregions. Evaluating predictor importance with Shapley additive explanations, we find that fire-month vapor pressure deficit (VPD) is the dominant driver of fire frequencies and sizes across the WUS, followed by 1000 h dead fuel moisture (FM1000), total monthly precipitation (Prec), mean daily maximum temperature (Tmax), and fraction of grassland cover in a grid cell. Our findings serve as a promising use case of ML techniques for wildfire prediction in particular and extreme event modeling more broadly. They also highlight the power of ML-driven parameterizations for potential implementation in fire modules of dynamic global vegetation models (DGVMs) and earth system models (ESMs).more » « less
-
Abstract Lightning has a transformative role in our environment and society. To understand the patterns of lightning incidence in the dry U.S. Southwest, where lightning is a prominent source of wildfires, we analyze 22 years of lightning incidence data in New Mexico (NM) collected by the U.S. National Lightning Detection Network (NLDN). Seasonality patterns indicate that most lightning happens in the afternoons of the months of July and August in NM and that lightning activity has been consistent over the span of two decades, displaying a minor decrease over time, correlated with a decrease in average precipitation. We present maps cross-referencing lightning occurrence with the state’s geography. Our analysis indicates lightning peaks in two types of regions. First, highest rates of cloud-to-ground (CG) lightning (including of positive polarity) seem to happen in a cluster of counties near the border with Texas in a geographic region known as the Great Plains. Roosevelt County (bordering Texas) has the highest CG lightning density at 4.2 flashes per square kilometer per year, while San Juan County in the northwest corner of the state has the lowest CG lightning density at 1.3 flashes per square kilometer per year. Second, secondary peaks in CG lightning density can also be found in forested mountains throughout the state, which include several national forests. The Gila National Forest and Lincoln National Forest have the highest rate of CG lightning incidence. Thus, they may be at greatest risk of lightning-ignited wildfires. Across most of the state, precipitation rates reasonably explain trends in lightning activity, but this correlation breaks down in NM’s lightning hotspot, the southeast corner of the state.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript
- Journal Article:
- https://doi.org/10.1029/2024JD042147
