Constraining CO2 emissions from open biomass burning by satellite observations of co-emitted species: a method and its application to wildfires in Siberia

A method to constrain carbon dioxide (CO2) emissions from open biomass burning by using satellite observations of co-emitted species and a chemistry-transport model (CTM) is proposed and applied to the case of wildfires in Siberia. CO2 emissions are assessed by means of an emission model assuming a direct relationship between the biomass burning rate (BBR) and the Fire Radiative Power (FRP) derived from the MODIS measurements. The key features of the method are (1) estimating the FRP-to-BBR conversion factors (α) for different vegetative land cover types by assimilating the satellite observations of co-emitted species into the CTM, (2) optimal combination of the estimates of α derived independently from satellite observations of different species (CO and aerosol in this study), and (3) estimation of the diurnal cycle of the fire emissions directly from the FRP measurements. Values of α for forest and grassland fires in Siberia and their uncertainties are estimated by using the IASI carbon monoxide (CO) retrievals and the MODIS aerosol optical depth (AOD) measurements combined with outputs from the CHIMERE mesoscale chemistry transport model. The constrained CO emissions are validated through comparison of the respective simulations with the independent data of ground based CO measurements at the ZOTTO site. Using our optimal regional-scale estimates of the conversion factors (which are found to be in agreement with the earlier published estimates obtained from local measurements of experimental fires), the total CO2 emissions from wildfires in Siberia in 2012 are estimated to be in the range from 262 to 477 Tg C, with the optimal (maximum likelihood) value of 354 Tg C. Sensitivity test cases featuring different assumptions regarding the injection height and diurnal variations of emissions indicate that the derived estimates of the total CO2 emissions in Siberia are robust with respect to the modelling options (the different estimates vary within less than 10% of their magnitude). The obtained CO2 emission estimates for several years are compared with the independent estimates provided by the GFED3.1 and GFASv1.0 global emission inventories. It is found that our "top-down" estimates for the total annual biomass burning CO2 emissions in the period from 2007 to 2011 in Siberia are by factors of 2.3 and 1.7 larger than the respective bottom-up estimates; these discrepancies cannot be fully explained by uncertainties in our estimates. There are also considerable differences in the spatial distribution of the different emission estimates; some of those differences have a systematic character and require further analysis