资源环境科技发展态势分析平台

Hourly precipitation is important in climate research because it is closely related to convective synoptic systems and short-term weather forecasts. In this study, characteristics like climatology, inter-annual variability and diurnal cycle are quantified from 1979 to 2015 over eastern China using hourly gauge observations and ERA5 precipitation dataset. The results show that the reanalysis can reproduce the spatial pattern of precipitation climatology, but it generally overestimates the frequency and underestimates the intensity, resulting in a better estimate of the precipitation amount. The mean relative biases of ERA5 compared to gauge observations are 137.04%, −49.09% and 16.30% for precipitation frequency, intensity and amount, respectively. ERA5 has positive correlation coefficients with gauge observations in inter-annual variations of hourly precipitation quantities, but with lack of long-term trend estimation. At national scale, the trend of the annual mean hourly precipitation amount calculated by ERA5 is −0.009 mm/h/decade, which is much smaller than the 0.002 mm/h/decade based on gauge observations. The trends of the annual mean hourly precipitation frequency and intensity calculated by ERA5 (−0.901%/decade and − 0.008 mm/h/decade) are also smaller than those based on the gauge observations (−0.044%/decade and 0.023 mm/h/decade). ERA5 can also reproduce the diurnal cycle of the gauge observations, but largely overestimates the precipitation amount during daytime. The reanalyses agree with gauge observations best over the south China region, whereas the worst performance occurs over northern southwest China. The results can provide a useful reference for the hydrological and meteorological research about the hourly precipitation.