ABSTRACT
In this study, we estimate the global CO2 emissions from biomass burning using satellite data. A polynomial regression function calibrated the traditional vegetation index (NDVI) for assessing the aboveground biomass (AGB) using a process-based model (VISIT). Also, soil organic matter (SOM) modeled values from VISIT were also integrated into emissions calculation. MODIS’s finer spatial resolution at 1 km enabled to detect of small fires (in spatial scales of several m to km) than the Global Fire Emissions Database (GFED) 4.1s product at 25 km. Fire radiative power (FRP) of MODIS was used as a dynamic combustion completeness ratio since it provides variable values for every fire event. We delineated the monthly dynamics of fire emissions (FEs) in detail in terms of scale and strength. We addressed the following objectices: (1) estimate global FE considering dynamic combustion completeness by using FRP of MODIS, (2) characterize seasonal and regional patterns of wildfires, and (3) report uncertainties in GFED4.1s compared to our estimates. Our results on monthly time-series FEs agreed well with the GFED data. The annual averages of carbon (C) emissions from FEs were 1.25 ± 0.12 PgC/year compared to 1.68 ± 0.15 PgC/year from the GFED.
