Understanding the cause-effect relationship between heat waves and wildfires within a Eurasia longitudinal gradient

Understanding the cause-effect relationship between heat waves occurrence and wildfires spread represent a key priority in global change studies due to the significant threats posed on natural ecosystems and society. Predominately, previous studies have not explored the spatial and temporal mechanism underlying the relationship between heat waves and wildfires occurrence, especially over large geographical regions. In this study, we investigate such a relationship with the goal to assess the potential effects of heat waves occurrence on wildfire spread and regime with a focus on 37 ecoregions within a Eurasia longitudinal gradient. In terms of methodology, we based our analysis on the wildfire dataset provided by the Global Wildfire Information System (GWIS) and the meteorological dataset ERA5-land from Copernicus Climate service. In both cases we focused on 2000-2021 timeframe. Using a 12 km square grid, we computed several wildfire metrics as proxy of fire regime, such as density, seasonality, and severity of wildfires. For the same period, we also characterized heat waves in terms of frequency, duration, seasonality, and intensity. As general definition, we considered “heat wave” as a period of at least 6 consecutive days with temperature greater than the 90th percentile, while heat wave intensity was computed as the total amount of degrees exceeding the 90th percentile during the heat wave occurrence period. We performed several statistical tests to evaluate the different patterns of heat waves occurrence and wildfires spread in the 37 ecoregions within the area of study. Using Geographically Weighted Regression (GWR) we explored the spatial and temporal coupling effect of heat waves characteristics and wildfires metrics. As expected, our results suggest that the 37 ecoregions identified within the Eurasia longitudinal gradient differ in terms of fire regime. However, the occurrence of heat waves did not show significant differences among ecoregions, suggesting that in the selected timeframe (2000-2021) heat waves have rather been equally distributed across the study area. The outcome of the GWR analysis allowed us to identify the spatial locations (i.e., hotspot areas) where the coupling effect between heat waves and wildfires is more significant. In other words, in these hotspot areas, the temporal occurrence of heat waves can be seen a driver of wildfire spread in forest and steppe ecosystems. Our findings could support a more comprehensive assessment of the wildfire patterns in this region, thus supporting cross-regional prevention strategies for disaster risk mitigation.