Accurate atmospheric water vapor information is crucial for improving the performance of numerical weather prediction (NWP) models. However, current models are limited by insufficient vertical layering structures that fail to capture the spatial variability of water vapor, as well as outdated and low-resolution underlying surface data. To address these issues, this study proposes a hybrid vertical layering method that integrates Global Navigation Satellite System (GNSS)-derived precipitable water vapor (PWV) to enhance short-term precipitation forecasting. The method introduces a vertically weighted water vapor equalization principle and uses spatially continuous water vapor data to adaptively optimize the vertical structure within the key precipitation height range (1-7 km). In addition, high-resolution (30 m) land cover data from the 2020 global dataset are used to replace the default USGS data in the WRF model. Xi'an and surrounding areas are selected as the study region to evaluate the impacts of surface data updating and vertical layering optimization. Results show that the combined approach improves 1-hour precipitation forecasts, reducing RMSE, MAE, and bias by 15.2%, 12.5%, and 55.2%, respectively, with the correlation coefficient increasing to 0.53. Significant improvements are also observed in temperature, pressure, and relative humidity, demonstrating enhanced forecasting capability.

GNSS, precipitable water vapor, WRF model, hybrid vertical layering, precipitation forecasting