Lightning is the major cause of natural wildfires worldwide. These fires, ignited by lightning strikes (Figure 1), can quickly spread depending on weather conditions and the availability of fuel, releasing substantial amounts of carbon, nitrogen oxides, and other trace gases into the atmosphere.
Several studies suggest that the frequency and distribution of lightning may shift in the future. However, there is still substantial uncertainty regarding the impact this may have on the likelihood of lightning-ignited wildfires. The susceptibility of lightning-caused wildfires to climate change can vary based on changes in lightning frequency and meteorological factors that impact fuel availability and fire spread conditions. Furthermore, we suggest that future patterns of lightning-ignited fires may also be influenced by variations in the duration of continuing currents in lightning strikes.
This work aims to explore the variation of lightning with continuing currents and meteorological conditions in order to predict future patterns of lightning-caused wildfires. Firstly, we combine lightning measurements and wildland fire database over the Continental United States to investigate the role of continuing currents in lightning ignitions. Our research indicates that lightning strikes with continuing currents have a greater likelihood of igniting wildfires compared to those without continuing currents. Secondly, we introduce a previously developed parameterization of lightning with long-continuing currents in a chemistry-climate model to estimate the variation of lightning-ignited wildfire patterns under climate change.
Continuing currents and lightning-induced wildfires
We combined lightning measurements provided by the Geostationary Lightning Mapper (GLM) aboard the Geostationary Operational Environmental Satellite-16 (GOES-16) and wildland fire database provided by the U.S. Department of Agriculture to investigate the role of continuing currents in lightning ignitions. We identified the most probable candidates for each lightning-induced wildfires in the GLM database based on the distance and holdover time between the reported wildfire and observed lightning flashes. Our analysis found 5,254 out of 5,574 wildfires (90%) were preceded by at least one long-continuing current lightning flash within 10 km of the ignition point within 14 days prior. To further verify the cause of ignition, we applied a threshold of 0.7 on the proximity index and found that 1,703 (29%) wildfires could have been ignited by long-continuing current lightning. This is higher than the ratio of long-continuing current lightning to total flashes observed in North America during the summer, which is slightly below 10%. Hence, we conclude that the likelihood of ignition by long-continuing current lightning is greater than ignition by lightning without continuing currents.
We performed two simulations by using the atmosphere-chemistry model EMAC (ECHAM/MESSy Atmospheric Chemistry) together with a recently developed parameterization of long-continuing current lightning. We simulated the 2090s decade under the Representative Concentration Pathway 6.0 (RCP6.0) scenario to estimate the future occurrence of total lightning, long-continuing current lightning and their relationships with the preferential meteorological conditions of lightning-ignited wildfires. We compared the results with a present-day simulation between 2009 and 2011 to set the sensitivity of lightning-induced wildfires under climate change.
Our findings showed a 43% and 41% increase in total and long-continuing current lightning globally, respectively. Notably, we observed a 47% rise in long-continuing current lightning over land, which could raise the risk of lightning-induced wildfires in the future.
Our analysis considered the predicted changes in total and long-continuing current lightning, in conjunction with variations in meteorological conditions favorable for wildfire spread and fuel availability, such as accumulated rainfall during thunderstorms, vapor pressure deficit, relative humidity, and air temperature at 2 meters altitude (Figure 2). Our findings predict a decreased risk of lightning-caused wildfires in polar regions in the 2090s, except in some small areas in Scandinavia, Alaska, and Siberia where the risk could be elevated due to an increase in long-continuing current lightning. On the other hand, we predict a heightened risk of lightning-ignited wildfires in Southeast Asia, South America, Africa, and Australia, and a notable change in regional patterns in North America and Europe. Specifically, we estimate a large increase in lightning-caused wildfires along the Mediterranean basin and on the western and central coasts of North America in the 2090s.
Despite limitations in our study from the use of lightning parameterizations and other uncertainties related to the chemistry-climate model, these results highlight the importance of considering long-continuing current lightning in climate modeling for response studies.
Read the full article, now out in Nature Communications.