Assessment of precipitation reanalysis product ERA5 on summer extreme precipitation in the Yellow River Basin

Abstract

Based on the rain gauge-based gridded precipitation data from the CN05.1 dataset, the applicability of ERA5 in representing summer extreme precipitation in the Yellow River Basin (YRB) was evaluated from three aspects: climatology, deterministic and probabilistic assessment, and the spatiotemporal characteristics of the leading modes. The results showed that the ERA5 reanalysis demonstrates capability in capturing climatological characteristics of precipitation and extreme precipitation over the YRB, albeit with intensified climatological means and standard deviations, coupled with opposing linear trends. In terms of interannual variability, its precipitation estimates demonstrate notably higher accuracy in the recent decade compared to earlier periods. Deterministic evaluation reveals superior representation of general precipitation compared to intense precipitation. Probabilistic assessments indicate that uncertainties predominantly originate from the upper and northern middle reaches of the YRB, accompanied by systematic biases in probabilistic estimations of increased extreme precipitation amounts and decreased extreme precipitation days. While ERA5 successfully reproduces spatial patterns of the first two dominant modes of precipitation and extreme precipitation across the YRB, it amplifies anomalous signals. ERA5 accurately simulates interannual cycles associated with the first mode, but introduces spurious decadal signals in the second mode's temporal evolution.