New eastern China agricultural burning fire emission inventory and trends analysis from combined geostationary (Himawari-8) and polar-orbiting (VIIRS-IM) fire radiative power products

Open burning of agricultural crop residues is widespread across eastern China, and during certain post-harvest periods this activity is believed to significantly influence air quality. However, the exact contribution of crop residue burning to major air quality exceedances and air quality episodes has proven difficult to quantify. Whilst highly successful in many regions, in areas dominated by agricultural burning MODIS-based fire emissions inventories such as GFAS and GFED are suspected of significantly underestimating the magnitude of biomass burning emissions due to the typically very small, but highly numerous, fires involved that are quite easily missed by coarser spatial resolution remote sensing observations. To address this issue, we here use twice daily fire radiative power (FRP) observations from the ‘small fire optimised’ VIIRS-IM FRP product, and combine it with fire diurnal cycle information taken from the geostationary Himawari-8 satellite. Using this we generate a unique high spatio-temporal resolution agricultural burning inventory for eastern China for the years 2012-2015, designed to fully take into account small fires well below the MODIS burned area or active fire detection limit, focusing on dry matter burned (DMB) and emissions of CO2, CO, PM2.5 and black carbon. We calculate DMB totals 100 to 400% higher than reported by GFAS and GFED4.1s, and quantify interesting spatial and temporal patterns previously un-noted. Wheat residue burning, primarily occurring in May-June, is responsible for more than half of the annual crop residue burning emissions of all species, whilst a secondary peak in autumn (Sept-Oct) is associated with rice and corn residue burning. We further identify a new winter (Nov-Dec) burning season, hypothesised to be caused by delays in burning driven by the stronger implementation of residue burning bans during the autumn post-harvest season. Whilst our emissions estimates are far higher than those of other satellite-based emissions inventories for the region, they are lower than estimates made using traditional ‘crop yield-based approaches’ (CYBA) by a factor of between 2 and 5x. We believe that this is at least in part caused by outdated and overly high burning ratios being used in the CYBA approach, leading to the overestimation of DMB. Therefore we conclude that that satellite remote sensing approaches which adequately detect the presence of agricultural fires are a far better approach to agricultural fire emission estimation.