ERA5 Reproduces Key Features of Global Precipitation Trends in A Warming Climate

The largest impact of future climate changes on societies and ecosystems will likely come from precipitation variability and change. Using the ERA5 dataset, this global study examines precipitation trends using many precipitation parameters across five main components: precipitation amount, precipitation frequency, precipitation type, wet and dry spells, and precipitation extremes. Global trends are summarised by land and ocean areas, by climate region, and then zonally averaged to identify broader precipitation patterns and interactions that may not be apparent in local and regional scale studies, especially with a reanalysis dataset. We find that the ERA5 dataset was able to reproduce key features of precipitation change: the near‐ubiquitous increase in extreme precipitation, the increase in Arctic precipitation, the transition from snowfall to a rainfall regime in mid‐to high latitudes, and the contrasting sign of change in precipitation amount and frequency between land and ocean. Two noteworthy findings from the ERA5 dataset are that (1) spatial intensification of extreme precipitation around the warmest locations (equatorial region) is not matched by temporal intensification around the warmest time of year (summer months) in the northern hemisphere, and (2) the Himalayas show altitudinal stratification of precipitation phase changes. Finally, consistent with other studies, we find that synoptic weather types may influence the scaling of extreme precipitation with temperature and should be explored in future research. Additionally, ERA5 results are compared to those from GPCC, GPCP, and MERRA‐2 precipitation datasets to evaluate the robustness of the results. Daily, annual, and seasonal means, including inter‐annual variability of ERA5 precipitation estimates align strongly with those of the three validation datasets; however, trends show mixed results, with minimal to moderate agreement. In general, trends in GPCC precipitation are most similar to those of ERA5.