Human-driven greenhouse gas and aerosol emissions cause distinct regional impacts on extreme fire weather

Wildfire severity and area have significantly increased in recent decades across many regions. To understand these changes, global climate model experiments have been used to identify the human impact on observed fire-weather conditions (warm, dry, and windy conditions leading to dry fuels and fire spread). Studies show that compared to a world without anthropogenic climate change, higher temperatures and lower precipitation have made extreme fire-weather conditions significantly more likely over the past four decades, particularly in regions like the western US, western Canada, and Australia. These conditions contributed to unprecedented fire severity and area in these regions. Future anthropogenic climate change is expected to continue increasing the frequency and severity of wildfires. By the end of the 21st century, the frequency of extreme fire-weather conditions is projected to increase significantly in many regions, potentially becoming the new norm by 2100 for two-thirds of the global burnable area, regardless of emission scenarios. Understanding regional-scale fire risk is crucial for mitigation and planning. However, the interplay of greenhouse gases, aerosols (from industrial and biomass burning), and land-use/land-cover change (LULC) on fire-relevant climate conditions historically and in the future remains unclear. The mitigating effects of aerosol cooling or albedo changes due to LULC on greenhouse gas warming have not yet been fully quantified. Greenhouse gases, aerosols, and LULC have distinct impacts on precipitation, relative humidity, and surface wind, but their combined effects on regional wildfire risk are ambiguous. This study aims to use large ensemble simulations to quantify the distinct influences of these anthropogenic factors on regional wildfire risk, both historically and under projected future climate change.

Numerous studies have highlighted the increasing trend in wildfire severity and extent globally, citing various contributing factors including climate change. Abatzoglou & Williams (2016) demonstrated a strong link between anthropogenic climate change and increased wildfire risk in western US forests. Subsequent research has confirmed these findings across other regions, such as western Canada (Kirchmeier-Young et al., 2017), Australia (van Oldenborgh et al., 2020), and Southern Europe (Barbero et al., 2020). These studies utilized climate model experiments to isolate the anthropogenic signal from natural variability. Projections consistently point towards a future with more frequent and severe wildfires, underlining the importance of understanding the complex interactions between various anthropogenic drivers. However, a comprehensive quantification of the individual and combined effects of greenhouse gases, aerosols, biomass burning, and land-use/land-cover change on regional wildfire risk has been lacking. This gap in knowledge motivated the current study.

This study utilized large ensemble simulations from the Community Earth System Model version 1 (CESM1) to analyze the impacts of various anthropogenic forcings on extreme fire weather. The Canadian Forest Fire Weather Index (FWI) was used to quantify fire weather conditions, considering maximum temperature, precipitation, relative humidity, and surface wind speed. Extreme fire weather was defined as FWI exceeding the 95th percentile of the baseline distribution. The researchers employed both fully forced (ALL) simulations and all-but-one forced (X) simulations, where one forcing (greenhouse gases, aerosols, biomass burning, or land-use/land-cover change) was held constant at 1920 levels. This allowed for the isolation of the individual effects of each forcing. The risk ratio (RR) was calculated as the probability of exceeding the extreme fire weather threshold in the ALL ensemble divided by the probability in the corresponding X ensemble. Time of emergence (TOE) analysis was performed to determine when the forced changes in extreme fire weather emerged from the background variability. Further analysis was conducted to isolate the contributions of meteorological variables (temperature, precipitation, relative humidity, and wind speed) to changes in extreme fire weather risk by removing the forced signal for each variable individually and recalculating the RR. Eight fire-prone regions were selected for regional analysis to investigate spatial variations in the effects of anthropogenic forcing. The CESM-LE and CESM-LE-SF simulations provided the datasets, and MetPy Python package aided in the calculation of relative humidity.

The study found that anthropogenic greenhouse gas emissions have already substantially increased extreme fire-weather risk in several regions, particularly the Amazon, where the risk had doubled by 2005 and is projected to increase seven-fold by 2080. In other regions, such as the Mediterranean, the risk increased by 50% by 2005. Increases in maximum temperature are primarily responsible for the increase in extreme fire weather risk, but decreases in relative humidity and increases in wind speed also contributed, especially in the Amazon. Industrial aerosols have historically reduced extreme fire-weather risk in regions like the Amazon and Mediterranean by partially offsetting greenhouse gas warming. However, projected decreases in industrial aerosol emissions eliminate this mitigating effect by 2080. Biomass burning aerosols, unlike industrial aerosols, amplified extreme fire-weather risk, particularly in the Amazon and western North America. Land-use/land-cover change also contributed, albeit to a lesser extent. The time of emergence (TOE) analysis indicated that for much of the global burnable land area, the forced changes in extreme fire weather will emerge from natural variability before 2080, particularly in regions like the Amazon, Eastern North America, Mediterranean, and Southern Africa. However, in some regions, such as the Western US, this emergence is delayed due to a counteracting effect of increased atmospheric moisture.

The findings highlight the complex interplay between various anthropogenic factors in shaping regional extreme fire weather risk. While greenhouse gas warming is the dominant driver, the historically offsetting effect of industrial aerosols underscores the significance of considering multiple factors in assessing wildfire risk. The projected disappearance of the aerosol mitigating effect in the future highlights the urgency of mitigation efforts to reduce greenhouse gas emissions. The substantial drying trend over the Amazon, amplified by greenhouse gas-induced changes in relative humidity and wind speed, underscores the region's vulnerability to future wildfires. The varying responses of precipitation, relative humidity, and wind speed to different forcings demonstrate the complex regional variations in the impacts of anthropogenic factors on extreme fire weather. The study provides crucial insights into the spatial and temporal variations in extreme fire-weather risk driven by human activities, informing more targeted mitigation and adaptation strategies.

This study presents a novel quantification of the competing anthropogenic influences on extreme fire weather risk, both historically and under future climate change scenarios. The large ensemble simulations allowed for a robust assessment of the forced response, highlighting the significant and regionally variable impacts of greenhouse gases, aerosols, and biomass burning. The projected decline in aerosol mitigation emphasizes the substantial increase in extreme fire weather risk in the 21st century under continued greenhouse gas emissions. This work underscores the importance of considering the combined and nuanced effects of different anthropogenic factors in shaping regional fire risk and calls for collaborative international efforts for wildfire management in a rapidly changing climate.

The study's findings rely on a single climate model (CESM1), which may have inherent structural uncertainties. While the model reasonably captures the spatial patterns of historical FWI, analyzing results from additional models would help to better quantify the uncertainties associated with the results. The study focuses primarily on the impacts of climate forcing on fire weather and does not explicitly consider changes in fuel sources and amounts, ignition rates, or fire suppression efforts, which can also significantly influence wildfire occurrence and spread. Furthermore, the limited ensemble size and simulation length for the all-but-LULC experiment might affect the accuracy of LULC's quantified impacts on extreme fire weather.