Self-recording rain gauges hourly rainfall data from 1969 to 2010 have been utilized to identify rain events at a sub-daily scale. At the sub-daily scale, a significant decrease in the frequency of heavy rainfall events (HREs) is observed over central India and northeast India, while an increase is observed over the northern west coast of India. Frequency of short-duration HREs over central India and long duration HREs over northern west coast of India is increased in the recent decades than in earlier decades. Incongruity with the observations, CMIP6 historical and AMIP high temporal resolution models are not able to simulate the short-duration HREs and, in turn, the observed trends at a sub-daily scale over the India landmass. The inability of CMIP6 models to predict short-duration HREs suggests caution in predicting future projections of extreme precipitation at a sub-daily scale and highlights the need for further improvements in climate models.

自 记录 1969 年至 2010 年的雨量计每小时降雨数据已被用于识别次日尺度的降雨事件。在次日尺度上，印度中部和印度东北部的强降雨事件 （HRE） 频率显着降低，而印度西北部海岸的降雨频率有所增加。近几十年来，印度中部的短时 ORE 和印度西北海岸的长时 HRE 的频率比前几十年有所增加。与观测结果不一致，CMIP6 历史和 AMIP 高时间分辨率模型无法模拟短期 HRE，进而无法模拟印度陆地上亚日尺度的观测趋势。CMIP6 模型无法预测短期 HREs，这表明在预测未来低于日尺度的极端降水预测时要谨慎，并强调需要进一步改进气候模型。