Improved estimation of fire particulate emissions using a combination of VIIRS and AHI data for Indonesia during 2015–2020

Smoke aerosol emissions from Indonesian fires frequently cause adverse environmental consequences across southeast Asia. Satellite observations provide us with a great opportunity to monitor such at large scales. However, existing satellite-based estimates of fire vary considerably in magnitude, differing by factor four. Here, we aim improve through improved calculations radiative energy (FRE: time-integrated power (FRP)) and smoke emission coefficients (Ce) using multiple new-generation satellite observations. Specifically, peatland non-peatland Ce values were derived FRP rates aerosols based on Visible Infrared Imaging Radiometer Suite (VIIRS) active products. FRE was calculated the diurnal cycle that reconstructed fusing cloud-corrected retrievals high temporal-resolution Himawari-8 Advanced Himawari Imager (AHI) those spatial-resolution VIIRS. Then, fuel type-specific fused AHI-VIIRS used produce hourly daily data 2015 2020 Indonesia. To evaluate estimates, generated reference dataset total particulate matter (TPM) computing Moderate Resolution Spectroradiometer (MODIS) observation differences successive Terra Aqua satellites overpasses over set manually-selected plumes. AHI-VIIRS-based correlated significantly MODIS (r = 0.84; p < 0.001). The interannual time series analysis showed extreme variability (~26 fold), greatest amount (6.09 Tg) least (0.23 Tg), during study period. Mean annual distributed Indonesia's islands primarily Kalimantan (43.8%) Sumatra (41.7%), followed Sulawesi (5.6%), Java (4.5%), Papua (4.4%). Additionally, contributions peatland, forest, cropland, savanna/grassland types 51.4%, 37.8%, 7.0%, 3.8%, respectively, season strong El Niño event. suggests majority are very likely associated large-scale land use conversion agriculture, as well prolonged droughts induced events.