Sensitivity of NWP Over the Tibetan Plateau to Soil Moisture Uncertainties in Summer

Abstract The Tibetan Plateau (TP), consisting of complex high mountains, is a key regional heat source where soil moisture (SM) modulates surface energy exchange and boundary layer processes. However, significant uncertainties (biases/errors) exist in SM data sets over the TP due to complex terrain. It is not known if and how the numerical weather prediction (NWP) of the surface meteorological fields over the TP is sensitive to these SM uncertainties. This has been investigated by using the WRF model. Two simulations for JJA 2023 were conducted, initialized with ERA5 (EXP_CTL) and GFS (EXP_GFS) SM data, representing typical errors/biases. Results show that SM uncertainties significantly alter the surface energy budget, generating differences in sensible and latent heat fluxes exceeding 300 W m −2 . This induced a pronounced “dry soil‐warm, wet soil‐cool” effect on surface air temperature (differences up to 8°C) and modified local winds (up to 10 m/s). For precipitation, EXP_GFS reduces it in most central/northeastern areas and increases it in the southwest, altering JJA rainfall by up to 40% in some areas. SM errors/biases also modified cloud formation and shifted heavy rainbands over the eastern slope of the TP. A positive feedback loop associated with drier (wetter) lands is found over the TP that amplifies (minimizes) surface warming, significantly affecting local circulations, in contrast to the adjacent Sichuan Basin with abundant moisture in the air. This study demonstrates the high sensitivity of NWP to SM uncertainties over the TP and underscores the need to improve the accuracy of SM products in the region.