| The Qinghai Province is located in the northeastern part of the Qinghai-Tibet Plateau and serves as a significant node along the "Belt and Road" economic corridor.Its unique geographic and climatic conditions result in the widespread distribution of permafrost and seasonally frozen ground within the province,making it an integral part of the permafrost region on the QinghaiTibet Plateau.The study of permafrost distribution in Qinghai Province holds great significance for ecological conservation,resource development,climate change adaptation,and engineering construction.Traditional mapping methods based on observational station data are limited by the spatial distribution and quantity of these stations,making it difficult to accurately reflect the true distribution of permafrost.However,remote sensing techniques provide a means to obtain parameters related to permafrost distribution over large areas and multiple time periods,enabling the creation of permafrost distribution maps based on remote sensing data that better reflect the actual situation of permafrost distribution in Qinghai Province.In this paper,multi-source remote sensing data combined with the permafrost roof temperature model TTOP are used to identify the permafrost distribution in Qinghai Province.Firstly,the harmonic analysis algorithm(HANTS)is used to fill the missing value of MODIS LST data,and the MODIS LST data at 0cm of the meteorological station in Qinghai province and the corresponding location are used,and the MODIS LST data at four moments in a day are given different weights by multiple linear regression to obtain the daily average LST.The Kersten empirical model of soil thermal conductivity was used to take soil moisture related remote sensing data as input parameters,and the soil thermal conductivity based on remote sensing data was obtained,which was input into the TTOP model to draw the latest 1km resolution permafrost distribution map QH-2020 of Qinghai Province.Compared with the frozen soil distribution map based on the ground freezing number model and the previous frozen soil distribution map,its reliability is verified.The above content mainly obtains the following conclusions:1.In the ground freezing number model,with the increase of E value,the area of seasonal frozen soil identified by the model in Qinghai province is increasing,which mainly reflects the area along the Tuotuo River around Tanggula-shan Town.Compared with TTOP model,it is found that when E is 0.9,the simulation results of permafrost distribution are similar to those of TTOP model.2.Using the soil thermal conductivity based on remote sensing data,the area of permafrost in Qinghai Province simulated by TTOP model was 346513.063 square kilometers,accounting for 49.59% of the total area of permafrost in Qinghai Province.Compared with the empirically selected E-value freezing number model,the TTOP model using the new soil thermal conductivity is more certain.3.QH-2020’s classification accuracy was verified by comparing it with 60 measured borehole points.The results show that out of a total of 60 measured borehole data,57 points were correctly identified,and 3 points were misclassified,resulting in an accuracy rate of 95%for QH-2020.4.QH-2020 was compared with the 1:300 000 permafrost map of the Qinghai-Tibet Plateau(TP-1996)and the newly drawn permafrost distribution map of the Qinghai-Tibet Plateau(TP-2017).The distribution results of permafrost of QH-2020 and TP-2017 are the closest,with a kappa coefficient of 0.84,because the same distribution model of permafrost is used,but there are differences in the calculation method of permafrost thermal conductivity and the time of remote sensing data.The similarity with TP-1996 is the lowest,kappa coefficient is only 0.54,because TP-1996 uses the traditional mapping method and the mapping time is relatively long,and the boundary of permafrost will change with the change of time. |
