Wildfires exhibit a strong seasonality that is driven by climatic factors and human activities. Although the fire seasonality is commonly determined using burned area and fire frequency, it could also be quantified using biomass consumption estimates that directly represent biomass loss (a combination of the area burned and the fuel loading). Therefore, in this study a data set of long-term biomass consumed was derived from geostationary satellite data to explore the interannual variation in the fire seasonality and the possible impacts of climate change and land management practices across the Contiguous United States (CONUS). Specifically, daily biomass consumed data were derived using the fire radiative power retrieved from Geostationary Operational Environmental Satellites series with a pixel size of 4–10 km from 1995 to 2011. Annual fire seasonality metrics including the fire season duration, the timing of the start, peak, and end of the fire season, and interannual variation and trends were derived from the 17 year biomass consumed record. These metrics were associated with climatic factors to examine drivers and mediators of fire seasonality. The results indicate that biomass consumed significantly increased at a rate of 2.87 Tg/yr; however, the derived fire season duration exhibited a shortening trend in various states over the western CONUS and no significant trend in most other regions. This suggests that the frequency of extreme fire events has increased, which is perhaps associated with an observed increase of extreme weather conditions. Further, both the start and the end of the fire season exhibited an early shift (1.5–5 d/yr) in various eastern states although a late shift occurred in Arizona and Oregon. The interannual variation and trend in the fire seasonality was more strongly related to temperature in the western CONUS and to precipitation in the southeast. The Palmer Drought Severity Index was found to effectively reflect interannual variations in total biomass consumed although it was poorly correlated to the fire seasonality metrics. The results indicate that across the CONUS, the spatial patterns of the start, peak, and end of the fire season shift regularly in various regions in response to latitudinal gradients of temperature variation.
Available at: http://works.bepress.com/xiaoyang-zhang/14/