Study On Thermal Infrared Remote Sensing Inversion Of Surface Temperature In North China Plain Based On FY-3D

Under the background of global warming,agricultural droughts are increasingly harmful to the development of the national economy,threatening China’s food security and ecological security.The North China Plain is the region with the largest,most severe and long-lasting drought in China.Land Surface Temperature(LST)is an important parameter for remote sensing monitoring of agricultural drought disasters.Using thermal infrared remote sensing to accurately obtain land surface temperature information is an important foundation for improving agricultural drought monitoring and early warning.It is of great significance to build capacity and maintain the safe production of food in our country.At present,the research on the surface temperature has been more in-depth,and the split window algorithm has been derived for MODIS data analysis and reasoning,and the improved split window method calculation for AVHRR-16,-17,etc,for different data sources.The corresponding inversion method.However,there are few researches on inversion algorithms for the data of China’s new-generation polar-orbiting FY-3meteorological satellite(FY-3D/MERSI-Ⅱ),and many problems have not been resolved.This research selects the most widely used universal single The channel algorithm obtains the surface temperature data of the two thermal infrared channels of MERSI-Ⅱ,and selects the surface temperature of towns,vegetation,and water bodies in the inversion results for data fitting,and compares the simulated curve model with the real temperature Verify and analyze the accuracy of FY-3D/MERSI-Ⅱ data using the single-channel method to invert the surface temperature of the North China Plain,and finally analyze the temporal and spatial changes of the surface temperature of the North China Plain.The main research conclusions of this article are as follows:(1)Based on the surface temperature inversion results of MERSI-Ⅱ 24 and 25 data,the surface temperature data is fitted according to the types of towns,vegetation,and water bodies.When the type of town is fitted,the R~2of channel 24 is 0.731,which is fitted by the data in this paper.The best model,and the R~2of channel 25 is 0.683;the R~2of channels 24and 25 are closer to 0.61 and 0.60 when the water body and land types are fitted;the R~2of the vegetation type curve fitting results is the worst of 0.464 and 0.382,both In contrast,the analog correlation of channel 24 is better.(2)Use real temperature and various curve models to compare and analyze,the RMSE of real temperature and model fitting temperature are both within 4.5℃,and the correlation coefficient R~2between the fitting value of water surface temperature and the real temperature obtained by using 25 channel inversion reaches 0.910,it is concluded that the single-channel method has the best inversion result for the water body type of the North China Plain,the inversion result of the channel 24 is better for the vegetation land type,and the two-channel inversion result of the urban land type is the worst.(3)Through the analysis of the spatial and temporal changes of the surface temperature in the North China Plain in 2020,the highest value of the surface temperature appeared in June earlier than in other regions,recognizing that the surface type has a significant impact on the surface temperature at the same time;throughout the year In the North China Plain,the temperature around Yanshan Mountain and Jundu Mountain in northern Hebei Province is often low.It is recognized that because of many mountains,hills and unused land,the vegetation coverage is high,which will affect the lower surface temperature in this area;Affected by the ocean,the average temperature is relatively mild compared to the central part of the North China Plain,and the change trend is not obvious for more than a year.Based on the classification study of the surface temperature retrieved from the North China Plain,this paper will compare and analyze the retrieval results of the three types of land,and develop a land surface temperature retrieval algorithm suitable for the FY-3D meteorological satellite.Provide reliable services in related fields such as agriculture,environment and disaster monitoring.