Straw Burning Monitoring Based On Multi-Source Satellite Data

Environmental protection departments take various measures to supervise straw burning behavior during the harvest period each year in some big agricultural provinces.With regard to straw burning monitoring,remote sensing technique can make up for the inadequacy of traditional artificial method,improving the monitoring efficiency and reducing costs.Based on relevant fire detecting theories and application home and abroad,by comprehensively using multiple satellites data and other auxiliary data,this paper takes Henan province as the study area to conduct straw burning monitoring research during the autumn harvest period in 2015.The concrete contents are as follows.(1)Research on straw burning point extraction methodBased on FY-3 Medium Resolution Spectral Imager(MERSI)and Landsat 8 Operational Land Imager(OLI)images,etc.,this paper firstly uses supervised classification to extract arable lands in Henan,then gives the modified straw burning point extraction method suitable for FY-3 Visible and Infrared Radiometer(VIRR)images,and finally takes Oct.1st,2015 for instance,respectively comparing the extracted straw burning points with the visual interpretation fire characteristics on VIRR images,known burning regions,visual interpretation burning marks characteristics on HJ-1B Charge Coupled Device(CCD)images,as well as Straw Burning Monitoring Daily data to validate the feasibility of the algorithm.The result indicates that in terms of spatial distribution,the experimental extracted straw burning points,and the validation data are in good consistency.The straw burning mainly concentrates in Zhoukou,Zhumadian and Kaifeng cities on that day,and the total number of the fire points in these cities accounts for 64.63% of the whole province.And the most serious straw burning is located in Zhoukou,where straw burning point number accounts for 30.49% of the province.(2)Research on straw burning dynamic monitoringBased on the modified straw burning point extraction method and research in Part(1),this part firstly utilizes available FY-3 VIRR images to conduct straw burning dynamic monitoring during the 2015 harvest period in Henan province;at the same time,Moderate Resolution Imaging Spectroradiometer(MODIS)thermal anomaly product is used for the contrastive analysis.And the visual interpretation burning marks characteristics on higher resolution images like Landsat 8 OLI and HJ-1B CCD images,etc.are used to verify the extracted results.This paper then uses buffer analysis method to conduct correlation studies between the straw burning space-time distribution data and the Air Quality Index(AQI)data in Zhoukou region,which is the most serious straw burning region among the whole province,and finally analyzes the possible influence factors.The result shows that the straw burning dynamic monitoring results and the validation data are in good consistency,which indicates that the modified method in this paper shows a good rationality and validity.In terms of temporal distribution,the straw burning phenomenon is mainly concentrated on Sep.28 th solstice on Oct.2nd,and Oct.9th solstice on Oct.14th;as for the spatial distribution,the straw burning mainly concentrates in the weak regulation administrative border zone;and the total number of fire points in Zhoukou,Zhumadian,Kaifeng and Nanyang cities accounts for 72.18% of the whole province.In addition,straw burning point space-time distribution and AQI in Zhoukou city have moderate correlation;and the city’s air pollution has 1 to 2 days lag due to the air diffusion and the influence of some meteorological factors such as rainfall,wind,etc.