Abstract
The degradation of the frozen soil in the Qinghai–Tibetan Plateau (QTP) caused by climate warming has attracted extensive worldwide attention due to its significant effects on the ecosystem and hydrological processes. In this study, we propose an effective approach to estimate the spatial distribution and changes in the frozen soil using the moderate-resolution imaging spectroradiometer (MODIS) land surface temperature products as inputs. A comparison with in-situ observations suggests that this method can accurately estimate the mean daily land surface temperature, the spatial distribution of the permafrost, and the maximum thickness of the seasonally-frozen ground in the source region of the Yellow River, located in the northeastern area of the QTP. The results of The Temperature at the Top of the Permafrost model indicates that the area of permafrost in the source region of the Yellow River decreased by 4.82% in the period from 2003 to 2019, with an increase in the areal mean air temperature of 0.35 °C/10 years. A high spatial heterogeneity in the frozen soil changes was revealed. The basin-averaged active layer thickness of the permafrost increased at a rate of 5.46 cm/10 years, and the basin-averaged maximum thickness of the seasonally-frozen ground decreased at a rate of 3.66 cm/10 years. The uncertainties in calculating the mean daily land surface temperature and the soil’s thermal conductivity were likely to influence the accuracy of the estimation of the spatial distribution of the permafrost and the maximum thickness of the seasonally-frozen ground, which highlight the importance of the better integration of field observations and multi-source remote sensing data in order to improve the modelling of frozen soil in the future. Overall, the approach proposed in this study may contribute to the improvement of the application of the MODIS land surface temperature data in the study of frozen soil changes in large catchments with limited in-situ observations in the QTP.
Highlights
LST products to simulate the permafrost distribution on the Qinghai–Tibetan Plateau (QTP), and the results showed simulate the permafrost distribution on the QTP, and the results showed that the application that the application of moderate-resolution imaging spectroradiometer (MODIS) LST products could improve the accuracy of the mapping
The areas of permafrost estimated in exp2 and exp3 (Figure 8c,d) were larger than the results of the normal run (Figures 2 and 8a), which led to an incorrect identification of the frozen soil type near the YNG-2 borehole (Table 4 and Figure S2c,d). These results indicate that the spatial distributions of the permafrost were sensitive to the regression coefficients for the estimation of the mean daily LST
Comparisons with the in-situ observations indicate that the method proposed in this study can accurately estimate the spatial distribution of permafrost and the changes in the maximum thickness of the seasonally-frozen ground (MTSFG) using the remote sensing LST data
Summary
‘Frozen soil’ refers to soil and rock containing ice, with a temperature at 0 ◦ C or below [1]. According to the time length of the frozen state, frozen soil can be generally divided into two types: seasonally-frozen ground and permafrost. Due to its unique thermal and hydraulic characters, frozen soil plays an important role in the energy and water exchange between the land surface and the atmosphere system [2]. In response to global warming, the degradation of frozen soil has exhibited significant effects on regional hydrological processes, biochemical cycles and land ecosystems [2,3,4].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
