Abstract/Summary

This study investigates the impact of model resolution on simulating South Asian monsoon rainfall, focusing on the Ganges-Brahmaputra-Meghna (GBM) basin. By comparing high- and low-resolution versions of four CMIP6 HighResMIP model families against reference datasets (MSWEP and ERA5), we emphasize the advantages of high-resolution models in accurately simulating key monsoon characteristics, including annual rainfall, timing, intensity, and duration. Our results show that the high-resolution models align more closely with observed data, outperforming their low-resolution counterparts. Between 1979 and 2014, the high-resolution model ensemble (HR-models) captures key shifts in monsoon timing, such as delayed onset and withdrawal, leading to a slight increase in monsoon duration. In contrast, the low-resolution ensemble (LR-models) showed more pronounced delays in onset. The observational datasets, MSWEP and ERA5, indicate earlier (7±3 days) and later (3±1.2 days) onsets, respectively, with both showing delays in withdrawal, indicating extended monsoon duration. Notably, the increase in monsoon duration is more pronounced in MSWEP observations than in the model simulations, particularly for LR-models. Regarding rainfall trends, the HR-models more accurately reflect observed changes in both total rainfall and extreme rainfall from 1979-2014 compared to LR-models. Future projections (2015-2050) indicate further delays in monsoon onset, with HR-models projecting larger increases in total rainfall and extreme events (up to 4.5%/decade for the 95th percentile of rainfall) compared to LR-models, which show smaller increases and higher variability in total and extreme rainfall. These findings highlight the critical role of model resolution in improving the accuracy of monsoon simulations, with HR models offering more reliable simulations of historical monsoon behaviour and therefore likely more robust projections of future monsoon behavior. These are essential for informed water management and agricultural decision-making over the complex topography of the GBM basin.