| Land surface temperature is a key parameter of the physical process of land surface temperature on both regional and global scales,which integrates all results of the interactions and energy exchanges between surface and atmosphere.It has an important influence on the growth of vegetation,climate change and human daily production and life.The Tibetan Plateau is the plateau with the highest average altitude,largest area,and most complex terrain in the world.The study of the surface temperature variation over the Tibetan Plateau is not only important for itself,but also affects the atmospheric circulation in the East Asia and even global climate change.Landsat 8 equipped with the thermal infrared sensor TIRS which has set up two thermal infrared channels for the first time,and it can be better used to the accurate inversion of surface temperature.In this paper,the surface radiance of the Tibetan Plateau measured by the multi-spectral infrared radiometer CE312 was used to study the emissivity of different surface types in the Tibetan Plateau.The radiance transmission model MODTRAN was used to simulate the radiance datasets of the TIRS two-heat infrared channels.The inversion accuracy of he mono-window algorithm of Qin Zhihao,the generalized single-channel algorithm,the split-window algorithm of Jimenez-Munoz&Sobrino,and the split-window algorithm of Rozenstein&Qin was analyzed.With the sensitivity of emissivity,the surface temperature inversion algorithm suitable for the Tibetan Plateau was determined.Finally,Landsat 8/TIRS data was used to invert the surface temperature of the Nierong County in the Tibetan Plateau and validated with the measured data.(1)Using surface radiance data measured from the August-September 2016 multi-spectral infrared radiometer CE312 on the Tibetan Plateau,the surface thermal infrared emissivity of different coverage types was calculated and compared with the ASTER spectral library.Comparison and Analysis of Measured Thermal Infrared Emissivity of the Same Species at Different Measuring Points.The results show that the measured surface emissivity of the five channels of the CE312 is similar to the emissivity of the equivalent channel of the spectral library,and the measured infrared emissivity of the same kind of ground object is consistent,indicating that the CE312 can accurately measure the emissivity of the ground object.(2)Combining with the observed thermal infrared emissivity characteristics of the Tibetan Plateau,different atmospheric conditions under different geographical conditions,and the conditions of surface emissivity,the atmospheric radiative transfer model MODTRAN was used to simulate the establishment of the top of the atmosphere of the Landsat8/TIRS dual thermal infrared channel.Apparent radiance and brightness temperature datasets.In addition,the accuracy evaluation and parameter sensitivity analysis were carried out for the mono-window algorithm of Qin Zhihao,the generalized single-channel algorithm,the split-window algorithm of Jimenez-Munoz&Sobrino,and the split-window algorithm of Rozenstein&Qin.The average of the improved universal single channel algorithm was improved.The absolute error MAE is 0.644 and the root mean square error is 0.749,which are the lowest of the four algorithms,and the emissivity sensitivity is second only to he mono-window algorithm of Qin Zhihao.It is determined that the improved generalized single-channel algorithm is the surface temperature of the Tibetan Plateau.The best inversion algorithm.(3)Based on the improved generalized single-channel algorithm,Landsat8/TIRS 10th channel data was used to invert the surface temperature of Nierong County in the Tibetan Plateau,and the inversion results were tested using the ground temperature data observed in the region.The MAE of the inversion result and the measured value is 2.062 K,and the RMSE is 2.302 K.This indicates that the improved generalized single-channel algorithm has higher inversion accuracy in the Tibetan Plateau. |
PDF Full Text Request