The Benefits of Kilometre-scale Simulations for Extreme Summertime Precipitation in the Eastern Valleys of Qinghai

Abstract This study demonstrates the potential benefits of km-scale convection-permitting modelling for short-duration extreme precipitation forecasting in the valley areas of eastern Qinghai. Forecasting such weather events in this region of complex terrain can be challenging, even with the 3-km grid spacing used operationally by the China Meteorological Administration. We conduct experiments focused on a case study day (13 August 2022) with the Weather Research and Forecasting (WRF) model, using nested domains with three different resolutions (9-, 3- and 1-km grid spacing) and configurations with and without a convection parameterization scheme. The character of the precipitation forecast is highly sensitive to the resolution used and application or otherwise of the convection parameterization scheme. The precipitation magnitude and timing are substantially better forecast, compared to observations, in the 1- and 3-km simulations with explicitly represented convection than in other simulations. The use of a convection parameterization scheme leads to a premature onset of weak parameterized precipitation, followed by a delayed onset of the main explicitly represented precipitation peak in simulations at all three resolutions. The thermodynamic pre-convective environment is similar in all simulations, but there is evidence that the more realistic precipitation forecast produced by the high-resolution explicit simulations arises from a better representation of the low-level valley flows. Although this study has focused on a single case study and used a single numerical weather prediction model, the findings are translatable to other heavy precipitation events occurring over complex terrain and imply that there may be benefits from using resolution finer than that used operationally in this region.