Diurnal Cycle of Summer Precipitation Over Mainland Southeast Asia Revealed by Observations, Reanalysis, and Dynamic Downscaling

The diurnal cycle of precipitation plays a crucial role in regulating Earth's water cycle, energy balance, and regional climate patterns. However, the diurnal precipitation across mainland Southeast Asia (MSEA) and the factors influencing its spatial variations are not fully understood. In this study, we investigated diurnal precipitation patterns in summertime (June–August) from 2002 to 2005 over MSEA using ground-based observations, satellite products, the global ERA5 reanalysis, and high-resolution simulations from the Weather Research and Forecasting (WRF) Model at 9- and 3-km grid spacing forced by ERA5 hourly data on ∼0.25° grids. Various observation-based data sets including GHCN-Daily, Multi-Source Weighted-Ensemble Precipitation (MSWEP), Asian Precipitation - Highly-Resolved Observational Data Integration Towards Evaluation of Water Resources (APHRODITE), and Integrated Multi-satellite Retrievals for Global Precipitation Measurement (IMERG) were used. In evaluating daily precipitation over MSEA, MSWEP, and APHRODITE data sets show similar patterns in precipitation amount, frequency, and intensity, while IMERG tends to produce higher amounts but with less frequency. ERA5 overestimates light precipitation compared to the other data sets. The WRF simulations generally produce heavier but less frequent light precipitation, with the 3-km simulation producing less intense precipitation than the 9-km simulation. A k-means classification of IMERG data revealed five distinct spatial regimes with varying diurnal precipitation cycles. The WRF simulations closely match these regimes, capturing key diurnal cycles missed by ERA5 over mountainous regions and coastlines. Additionally, convective activities and near-surface winds influence these cycles, with WRF simulations better representing coastal and mountain precipitation patterns than ERA5. High-resolution WRF simulations, especially the 3-km simulation, capture diurnal precipitation more accurately than ERA5, highlighting the importance of employing convection-permitting models to simulate precipitation diurnal cycles over complex terrain.