Remote Sensing Thermal Infrared Surface Temperature Inversion System And Its Application

Surface temperature is not only an important physical and chemical parameter,but also an important variable in many research fields,such as ecological environment,phenology,global energy cycle and so on.Remote sensing thermal infrared data can be used for retrieving land surface temperature products at different regional or global scales.The purpose of this paper is to integrate the analysis of a variety of remote sensing thermal infrared data source surface temperature inversion algorithms.Based on the MATLAB platform,a multi data source surface temperature inversion system is designed and developed.At the same time,the application of surface temperature retrieval system in the direction of information extraction of multi temporal land surface temperature anomalies is also studied.Through the study,the following results are obtained:1.the basic theory of thermal infrared remote sensing and the retrieval algorithm of thermal infrared land surface temperature are analyzed and studied.The theoretical basis,hypothetical conditions,algorithm parameters and restrictive conditions of single band algorithm,split window algorithm and multi band algorithm are discussed respectively.On this basis,the thermal infrared data of TM,ETM+,TIRS and ASTER are selected as the data source of the surface temperature inversion system.According to the above data,the design of system architecture,the design of the algorithm module,the design of the control interface and the realization of the inversion algorithm are carried out.According to the characteristics of the above data sources,the single band algorithm,the split window algorithm and the multi band surface temperature emission separation algorithm are respectively integrated,and the surface temperature product production under the visual operation of the system is realized.The MODIS standard temperature products using the same phase can control the root mean square error of ETM+ and ASTER data within 1K by regression analysis,and the root mean square error of Landsat8 TIRS temperature products is within 2K.The accuracy of the land surface temperature retrieval system is MODIS.2.the distribution of solar radiation on the surface of the complex terrain is analyzed,and the study of the surface average temperature based on the distribution of the surface solar radiation(surface background temperature distribution)is carried out.First,the distribution of solar radiation energy on the surface is simulated under the conditions of topography,latitude and time.Then by Boltzmann law,the relationship between the solar radiation energy and the background temperature of the ground surface is established in a day,and the temperature distribution of the ground surface environment is simulated.Finally,through the ArcGIS platform design,modeling and implementation of multi scripts integrated ground surface environment temperature estimation model.Through visual interpretation of the temperature estimation results of the surface background environment under different phase,it is found that the temperature distribution of the regional surface environment is mainly controlled by the terrain,and the altitude elevation plays a decisive role in the regional temperature distribution contour.The details of slope surface and slope influence surface temperature distribution.3.choose the Gonghe basin area of Hainan Autonomous Prefecture in Qinghai as the research area,and use the land surface temperature inversion system to produce the products of multi temporal land surface temperature.The surface background temperature model is used to estimate the background temperature of the corresponding phase.The difference method is used to extract the surface temperature anomalies under single time.According to the known surface temperature anomaly points,the training samples are set up,and the BP neural network algorithm is used to integrate the surface temperature anomaly information under the multi time phase to achieve the effect of surface temperature abnormal information enhancement.The regression analysis is used to test the simulation results with the known point data,and the regression coefficient R2 reaches 0.81.It is proved that it is feasible to use BP neural network algorithm to extract the abnormal surface temperature of multi phase.