Biomass burning emission inventory of multi-year PM10 and PM2.5 with high temporal and spatial resolution for Northern Thailand

Abstract

The biomass burning emission inventory reported by the Global Fire Emissions Database is at a coarse temporal (monthly) and spatial resolution (0.25°-0.25°) and may not be appropriate for a model simulation or disease burden investigation. This study estimated an emission inventory of PM10 and PM2.5 caused by biomass burning in the nine provinces of Northern Thailand during 2012-2016 based on daily, monthly, and annual. The Visible Infrared Imaging Radiometer Suite (VIIRS) fire counts (375 m-resolutions), land uses, emission factors, and activity data were applied for the bottom-up estimation. According to the findings, Mae Hong Son (29%), Chiang Mai (20%), and Chiang Rai (38%), respectively, had the highest proportion of forest fires, savanna and grassland fires, and agricultural fires. There was a consistent trend between estimated emission and measured PM. The difference in emission ratio of PM2.5 between the current study and GFED4 was 1.2-2.6, 1.4-2.2, and 1.4-2.5 for forest, savannas and grasslands, and agricultural lands, respectively. The uncertainty range of PM2.5 and PM10 emissions from the three types of biomasses were the same with relative errors of -15% to 27%, -8% to 7%, and -17% to 10%, respectively. The VIIRS fire count can be used for estimation of biomass burning with finer resolution in both temporal and spatial terms.