Simulation of summer precipitation diurnal cycles over the Tibetan Plateau at the gray-zone grid spacing for cumulus parameterization

The Tibetan Plateau (TP) is often referred to as the "water tower of Asia" or the "Third Pole". It remains a challenge for most global and regional models to realistically simulate precipitation, especially its diurnal cycles, over the TP. This study focuses on evaluating the summer (June–August) precipitation diurnal cycles over the TP simulated by the Weather Research and Forecasting (WRF) model. The horizontal resolution used in this study is 9 km, which is within the gray-zone grid spacing that a cumulus parameterization scheme (CU) may or may not be used. We conducted WRF simulations with different cumulus schemes (CU experiments) and a simulation without CU (No_CU experiment). The selected CUs include the Grell-3D Ensemble (Grell), New Simplified Arakawa-Schubert (NSAS), and Multiscale Kain-Fritsch (MSKF). These simulations are compared with both the in-situ observations and satellite products. Results show that the scale-aware MSKF outperforms the other CUs in simulating precipitation in terms of both the mean intensity and diurnal cycles. In addition, the peak time of precipitation intensity is better captured by all the CU experiments than by the No_CU experiment. However, all the CU experiments tend to overestimate the mean precipitation and simulate an earlier peak of precipitation frequency when compared to observations. The frequencies and initiation timings for short-duration (1–3 h) and long-duration (> 6 h) precipitation events are well captured by the No_CU experiment, while these features are poorly reproduced by the CU experiments. The results demonstrate simulation without a CU outperforms those with a CU at the gray-zone spatial resolution in regard to the precipitation diurnal cycles.