The water and energy in the land surface and lower atmosphere have a strong coupling relationship. Apart from the land surface temperature (Ts) and air temperature (Ta), the land surface-air temperature difference (Ts-Ta) is also an essential parameter reflecting the coupling process. However, the global spatiotemporal variations and influencing factors of Ts-Ta remain not well explored. Here, ERA5-land reanalysis data, GIMMS NDVI data, and elevation data were used to analyze the global spatiotemporal heterogeneity and influencing factors of Ts-Ta. It was found that annual mean Ts-Ta exhibited a decreasing trend from the equator to polar areas. And the annual Ts-Ta increased at 0.009 degrees C/ 10a from 1981 to 2020. The variations of global net radiation mainly determined the spatiotemporal heterogeneity of global Ts-Ta. The different properties of the land surface and near-surface atmosphere were the main factors affecting the Ts-Ta, including soil moisture, vegetation, snow cover, and the water vapor content in the atmosphere. In addition, Ts and Ta also affected each other. These findings are conducive to a better understanding of the land-atmosphere coupling, and it is of great significance to take better measures to adapt the global climate change.

The Tibetan Plateau (TP), known as earth's Third Pole, influences regional and even global weather and climate systems through its mechanical and thermal-dynamical forcing. Near-surface (2 m) air temperature (T-a) and surface (skin) temperature (T-s) are two crucial parameters of land-atmosphere interactions and climate change. Their difference (Delta T = T-s - T-a) determines the heating source over the TP that drives the Asian summer monsoon. This study focuses on climatology, inter-annual variability, and long-term trend of Delta T over the TP in the last four decades (1979-2018), based on four latest reanalysis datasets including ERA-Interim, ERA5, MERRA2, and JRA55, along with observational data. We show that Delta T-based different datasets display fairly different climatology in terms of seasonality, spatial distribution, and long-term trend. Delta T exhibits a clear seasonality with negative value in winter and positive ones in summer despite different strengths and timings presented by the reanalyses. Along with global warming, all reanalyses except JRA55 exhibit obvious downwards trends of Delta T in a spatially non-uniform way. The median Delta T among the four reanalyses features uniform decreases in all seasons, with the most distinct area on the northern TP, as well as the largest and least decreases in autumn and spring, respectively. Further analysis shows that the differences in Delta T are most likely associated with discrepancies in radiation forcing, snow cover, wind speed, and boundary layer height within the reanalyses. The present findings highlight the difficulty for the state-of-the-art reanalyses to represent the climate change over the TP and point to possible factors behind the deficiencies identified.