The impact of the terrain effect on land surface temperature variation based on Landsat-8 observations in mountainous areas

In mountainous areas, land surface temperature (LST) exhibits a high spatial heterogeneity due to the influences from the changes in surface topographic factors (elevation, aspect, and slope), vegetation coverage, surface solar radiation, and other factors. Two small river basins located in the northeast and southwest of China were selected as the study areas to perform quantitative analysis and specify the terrain effect on the LST. Both summer and winter images were acquired from Landsat-8 for each study area. The LST variability was assessed by applying a linear regression against the key factors to discriminate the significance of them. The results illustrated the strong inverse relationship between the LST and the elevation for both study areas, with the LST lapse rate ranging from 7.6 to 13.8 degrees C km(-1), and the summer images had a higher value than the winter images. Furthermore, regarding the effects from the changes in surface aspect and slope, the analysis at different aspects indicated a similar directional pattern for the lapse rate and showed that the southern aspect usually had the highest rate in both study areas. In addition, the lapse rate had an obviously negative relationship with the increase in the local solar zenith angle. Moreover, the condition of the vegetation cover was also a key factor that influenced the lapse rate by changing the thermal property of the land surface. Finally, the solar radiation effect on LST was investigated by analysing the relationship between the LST and the cumulative incoming solar radiation (CSR) at different elevation ranges. The average statistics from the linear regression between the LST and the CSR clearly indicated a different function of solar radiation in different seasons. The CSR effect on the LST was stronger in winter than in summer, and its impact decreased with the rise in the elevation according to the changing rate derived by the linear regression. Hence, this study provides valuable insights into the complexity of the spatio-temporal variability in LST and its driving factors, which can be used in future studies to analyse the response of the ecosystem to the changing climate.