Evaluation of the ERA5 Reanalysis Snowfall Product in China

Abstract

Snowfall data is crucial for climate, hydrological and disaster research. However, long-term comprehensive ground-based snowfall observations are limited. In this context, reanalysis datasets such as ERA5 provide promising alternatives, but require careful evaluation prior to application. In this study, mainland China was divided into eight subregions based on terrain and climate characteristics. Using daily precipitation and weather phenomenon records from 2145 meteorological stations during 1980–2023, we assessed the accuracy of ERA5 snowfall data at multiple time scales and analysed its performance relative to snowfall intensity. Key findings include: (1) ERA5 performs best at the monthly scale, followed by the annual scale, but shows relatively low accuracy for daily snowfall detection, necessitating corrections for broader application. (2) ERA5 struggles to accurately capture snowy days (SDs) during light-to-moderate snowfall events and average snowfall on snowy days (ASSD) during blizzard-to-extreme blizzard events, which reduces daily accuracy. (3) ERA5's performance varies regionally, with better results in Northern China (NC) and Changjiang (CJ) regions, and relatively poor performance in the Tibetan Plateau (TP) region. These results provide important insights into the strengths and limitations of the ERA5 snowfall product and offer valuable guidance for its application in China.