| Open biomass burning is an important source of aerosols and trace gases in the atmosphere,and satellite remote sensing is an important tool to detect biomass burning.Based on the Himawari-8 AHI geosynchronous satellite spectrum,we developed an active fire detection algorithm suitable for China,and use synchronous satellite active fire monitoring data to estimate the total emissions of various pollutants in China from 2016 to 2020.We presented a preliminary assessment of the environmental effects of open biomass burning in China.The premise of active fire detection is accurate cloud mask information.In this study,the reflectance was corrected based on image enhancement method,and a fast daytime cloud detection algorithm was constructed based on the optical properties of clouds.The performance of the daytime cloud detection algorithm was verified by TROPOMI effective cloud retrieval results and MODIS cloud mask products.The PCT of the daytime cloud detection algorithm constructed in this paper is about 85%,which meets the requirements of active fire detection.The night cloud mask information is obtained by using the processing steps in the existing active fire detection algorithm.After obtaining accurate cloud mask information,this study conducted a series of spatio-temporal tests on land clear pixels to extract active fire pixels.In the step of extracting potential fire points,we tested a reasonable fixed threshold of brightness temperature of mid-infrared channel that applies to China.In the extraction step of background fire spots,the background test conditions suitable for fire events of different scales are used to further screen suspected fire pixels.In the time series analysis step,the Prophet model was applied to the fire point retrieval algorithm for the first time.For the pixels that have been affected by clouds for a long time,we conducted additional tests on the relevant suspected pixels based on the statistical characteristics of the time series of brightness temperature values.Through the above three steps,the fire point pixels are extracted.The active fire retrieval algorithm constructed in this study is more suitable for China.Based on MODIS active fire products,the PCT of the active fire retrieval results of this study in China is about 10%higher than that of JAXA active fire products.We combine Himawari-8 and MODIS fire radiation power products to construct the daily fire radiation energy distribution in China and further calculates the particulate matter and trace gases emitted by China in 2016-2020 by biomass burning based on the fire radiation energy.From 2016 to 2020,the average annual TPM emissions from biomass burning to the atmosphere in China were about 225 Tg,CO is about 13403.5 Gg,NMVOC is about 2015.7 Gg,and PM2.5 is about 1619.4 Gg,PM10 is about 2313.8 Gg,OC is about 709.2 Gg,BC is about 95.1 Gg,and CO2 is about 310016.1 Gg.Among them,the burning of crop stalks in the Northeast is the biggest contributor to the above-mentioned pollutants.The trend of biomass burning emissions calculated in this study is similar to that of GFAS inventory,but the difference is significant.The application of geostationary satellite observation fire dataset in biomass burning emission estimation and environmental effect exploration in China still needs further exploration. |
