Assessing the Performance of the Copernicus European Regional Reanalysis Dataset in Reproducing Extreme Weather Events in Poland

ABSTRACT Reliable representation of extreme weather events is essential for the climate impact studies and risk assessment. However, global reanalyses, for example ERA5 and ERA5‐Land, often struggle to capture local‐scale extremes due to the limited spatial resolution. This study evaluates the performance of the high‐resolution Copernicus European Regional ReAnalysis (CERRA) and CERRA‐Land in representing extreme temperature, precipitation, and wind conditions over Poland. CERRA, as well as ERA5 and ERA5‐Land, are compared against the long‐term in situ observations from the period 1991–2020. The assessment combines a climatological analysis of standardized extreme indices with detailed case studies of high‐impact weather events, including an extreme windstorm associated with Cyclone Kyrill in January 2007, a regional heavy rainfall episode leading to a flooding in May 2010, and a severe heatwave in July 2010. Statistical metrics, spatial distributions of the mean biases, probability density functions, and Taylor diagrams were used to quantify biases and overall agreement with observations. Results show that all reanalyses tend to underestimate temperature and precipitation extremes, with the strongest temperature biases occurring at coastal and mountainous stations. However, CERRA consistently exhibits smaller mean biases and improved correlations for extreme temperature and precipitation indices compared to ERA5 and ERA5‐Land. These improvements are particularly pronounced for localized and short‐lived precipitation extremes, where CERRA substantially outperforms the global reanalyses. In contrast, wind extremes remain challenging for all datasets, with only marginal improvements provided by CERRA and persistent large errors at high‐elevation sites. Overall, the findings demonstrate that CERRA provides a clear and consistent improvement over ERA5 and ERA5‐Land for temperature and precipitation extremes in Poland. These advances are relevant especially for applications such as flood risk assessment or heatwave impact analysis, where an accurate representation of extreme conditions is essential.