Spatiotemporal heterogeneity analysis of land surface temperature on the Tibetan Plateau from 2003 to 2020

ABSTRACTLand Surface Temperature (LST) is a crucial indicator of the Earth's energy balance and a significant remote sensing index for assessing regional ecological and environmental quality. LST exhibits substantial temporal fluctuations and pronounced spatial variations. This study, based on the mean LST data of Terra and Aqua satellite products in the Tibetan Plateau from 2003 to 2020, explores the LST changes in the study area, their spatiotemporal differentiation characteristics, and driving mechanisms.Using landscape pattern analysis, spatial centroid models, and geographical detectors, this study yields the following findings:(1) The average annual rate of LST change on the Tibetan Plateau is 0.016°C per annum, with nighttime LST trends showing more pronounced warming compared to daytime, indicating an overall warming trend.(2) Spatially, the centroids of high and low temperatures oscillate in a southeast-northwest direction, with the migration trajectory of cooler regions being more distinct than that of warmer regions, suggesting greater spatial heterogeneity in cooler areas. Moderate and sub-high temperature zones, compared to other temperature zones, cover larger areas and demonstrate higher degrees of spatial clustering, with simpler and more regular patch shapes. Other temperature zones also display varied spatial proportions, patch shapes, and clustering degrees, indicating significant spatial heterogeneity.(3) Altitude is identified as the primary driving factor, with other factors also contributing significantly in conjunction with altitude.This paper provides a scientific reference for analyzing the plateau's surface thermal environment and addressing climate change