Joint Analysis of Bulk Wildfire Characteristics from Multiple Satellite Retrievals

Abstract:

Biomass burning significantly impacts atmospheric composition, as well as regional and global climate. Here, we investigate the spatiotemporal trends in fire characteristics in several major fire regions using combustion signatures observed from space. Our main goals is to identify key relationships between the trends in co-emitted constituents across these regions, as well as linkages to main drivers of change such as meteorology, fire practice, development patterns, and ecosystem feedbacks. Our approach begins with a multi-species analysis of trends in the observed abundance of CO, NO2, and aerosols over these regions and across the time period 2005 to 2014. We use MOPITT multi-spectral CO, OMI tropospheric NO2 column, MODIS AOD, and MODIS FRP retrievals. The long records from these retrievals provide a unique opportunity to study atmospheric composition across the most recent decade. While several studies in the past have reported trends over these regions, most of these studies have focused on a particular constituent. A unique aspect of this work involves understanding co-variations in co-emitted constituents to provide a more comprehensive look at fire characteristics, which are yet to be fully understood. Here, we introduce a derived quantity (called smoke index) to represent bulk fire characteristics (e.g., flaming versus smoldering). The smoke index is calculated as the ratio of the geometric mean of CO and AOD fire enhancements to that of NO2 fire enhancements. Our initial results, which focused on the Amazon region, show that: 1) deforestation fires are dominantly flaming fires while non-deforestation fires are more likely to be dominantly smoldering fires; and 2) droughts have larger influence on non-deforestation (possibly understorey) fires than deforestation fires. Here, we will present an extension of this analysis to other fire regions around the globe (tropical, temperate and boreal fires) and explore other measurements available during this period for comparisons. We will also compare with current fire emission models, such as GFED and FINN, to test the robustness of our findings. We note that this exploratory work provides a unique perspective of fire characteristics that will be useful to improve predictive capability of fire emission and atmospheric models.