With warming temperatures and severe droughts, the state of California has seen an increase of fire activity. With the decadal summer average temperature increasing by 0.8°C during 1984-2014, historical observations show that the total size of large fires has expanded by a factor of 2.5. These large fires have caused a substantial loss in both the ecological and economic sectors. Due to rising global temperatures, this upward trend is expected to continue in areas with available fuel, stressing the need for reliable estimates of future fire statistics. Accurate projections are vital for improving fire management, decision-making, and long-term adaptation and mitigation planning.
In a new Stochastic Environmental Research and Risk Assessment article, authors Shahrbanou Madadgar, Mojtaba Sadegh, Felicia Chiang, Elisa Ragno, and Amir AghaKouchak proposes a multivariate probabilistic approach for quantifying changes to fire risk given different climatic conditions. This approach focuses on changes of annual fire size distribution given a one-degree Celsius increase in summer temperature or deficit in annual precipitation, as examples of valuable information for risk assessment and planning. The statistical model builds a conditional probability distribution function of fire size for a given expected climatic condition and provides a wide range of distribution functions and calculates the probability of exceeding critical thresholds for different precipitation and temperature conditions.
Using this model, researchers found that the risk of large ﬁres in California increases substantially in response to unit degree changes in summer temperature. The probability of annual mean ﬁre size exceeding its long-term average increases by 30% when summer temperature anomaly increases by 1 C (from -0.5 C to 0.5 C). Furthermore, the probability of annual average ﬁre size exceeding its long-term average doubles when the annual precipitation decreases from the 75th (wet) to the 25th (dry) percentile.
Comparison of anomalies in a rolling decadal average of annual and summer temperatures, and the corresponding number of fires in 1984–2014.
Variation of annual precipitation versus annual average fire size during 1984–2014. The vertical dashed lines associate the largest fires with annual precipitation.
Researcher Amir Aghakouchak said about the study "While most fires are ignited by humans, weather and climate patterns have significant influence on severity and spread of wildfires. Our recent study shows that the risk of large fires in California increases substantially (~30%) in response to even 1°C warming. Our recent study is one of many out there that shows even a degree or two (let alone more) of long-term temperature change will have a significant impact on extreme events and their impacts."
"Many people still think a degree or two of warming won't have a significant impact on their lives. But our study shows that the risk of large fires in California has increased substantially (~30%) in response to just 1°C of warming. We expect more worming in the coming years which will likely exacerbate the severity and frequency for future fires."
This multivariate model provides probabilistic information to support risk assessment and decision-making. Such risk analyses provide additional information on ﬁre distribution in different climate conditions and may allow for seasonal forecasts to better inform wildﬁre resource and response decision-making, especially in ﬁre-prone regions where future climate projections will lead to more favorable ﬁre conditions. This study was partially funded by the MAPP program.
Read the full paper here