内蒙古 ERA5 再分析降水数据性能评估与极端降水时空特征分析

ERA5 is a new generation reanalysis product launched by the European Center for Medium-Range Weather Forecasts that can provide a new source of precipitation data for areas with few ground observation stations. Based on the daily precipitation data of 45 ground stations in Inner Mongolia, China from 2008 to 2017, we evaluated the accuracy of ERA5 reanalysis precipitation data on multiple temporal and spatial scales using multiple evaluation indicators and applied a comprehensive weighting model to construct an extreme precipitation danger index by integrating 13 extreme precipitation indices. We then used principal component analysis, Sen’s slope method, and a Mann-Kendall trend test to analyze the temporal and spatial changes of extreme precipitation and extreme precipitation danger in the region from 1981 to 2021. The results show that: (1) ERA5 can reproduce the precipitation process better; the ERA5 precipitation reanalysis dataset performs better at a monthly time scale than at a daily time scale, and its accuracy is highest in summer and lowest in winter. ERA5 precipitation data are highly correlated with observed data at monthly and seasonal scales (correlation coefficient>0.85) and are strongly correlated with observed data at a daily scale (correlation coefficient=0.68). The detection accuracy of ERA5 data is better for the eastern stations than for the western stations. (2) Extreme precipitation indices showed a downward trend except for daily wet precipitation intensity, total heavy precipitation, and continuous dry days. The annual total wet day precipitation declined fastest, and the Sen’s slope value was −15.74 mm·(10a)^{− 1}. (3) The extreme precipitation indices show clear regional spatial differentiation. Extreme precipitation shows an increase in intensity in western Inner Mongolia, a decrease in frequency, intensity, and duration in mid-Inner Mongolia, and an increase in intensity, frequency, and duration in eastern Inner Mongolia. (4) The extreme precipitation danger index has high central values and shows significant upward trends in Ordos City, Hulun Buir City, Bayannur City, Hinggan League, and other cities/leagues with relatively dense populations and rapid economic development and should therefore receive special attention. The results of this study can facilitate the discovery of better datasets for the analysis of climate factors in Inner Mongolia, and the study provides a theoretical basis for the formulation of climate change adaptation measures and future climate prediction.