Recent studies have demonstrated a persistent decreasing trend in the spring sensible heat (SH) source over the Tibetan Plateau (TP) during the past three decades. By comparing simulations from nine state-of-the-art atmospheric general circulation models (AGCMs) driven by historical forcing fields with both observational data and five reanalysis datasets, the authors found that the AGCMs are unable to reproduce the change in the SH flux over the TP. This deficiency arises because the observed decreasing trend in SH flux depends primarily on the change in surface wind speed according to the bulk formula, whereas in the models it is also influenced largely by changes in the land-air temperature difference related to the systematic cold bias. In addition, an obvious discrepancy exists in other aspects of the diabatic heating simulated by the models, suggesting that a significant improvement is required in the physical schemes associated with land surface processes and diabatic heating over the complicated topography.