Assessment of the Copernicus Arctic Regional Reanalysis (CARRA) dataset for Iceland

Abstract

Abstract Reanalyses are essential to provide scientists with high-resolution, continuous gridded datasets that help improve the understanding of weather and climatology. This is particularly important in regions with sparse weather station coverage. At the Icelandic Meteorological Office (IMO), a reanalysis dataset extending further than the timespan covered by the Icelandic Reanalysis (ICRA; 1979–2017) has long been needed. This study provides a first assessment of the Copernicus Arctic Regional Reanalysis (CARRA) dataset that extends from 1990 until 2026. While more than 20 variables were downloaded and processed through the Copernicus Climate Data Store (CDS), this study focuses on three of them: the 2-m temperature, 10-m wind speed and precipitation. In this report, details on how this dataset was retrieved from CDS are provided, as well as the methods used to create time series at the locations of 49 automatic weather stations distributed all around Iceland. Overall, the results show a good agreement between observations and simulated data for all three variables. The largest discrepancies arise in areas of complex orography, such as narrow fjords and valleys, where orography is not fully resolved by the CARRA grid, which has a 2.5-km resolution. The RMSE for the mean, maximum, and minimum 2-m temperature is below 2°C for most stations, with warming trends well reproduced by the reanalysis. Daily 10-m wind speed is well simulated in flat areas and inland stations, while simulated values are overestimated for stations located in complex terrain. For daily precipitation, RMSE is below 3 mm for nine stations, above 5 mm for 11 stations, with all other stations having an RMSE in between, with results improving significantly when only summer values are considered. Annual Maximum Series (AMS) based on daily precipitation are also evaluated, with most stations having mean AMS within a 5% difference when based on observations and simulations. Key words: CARRA, reanalysis, 2-m temperature, 10-m wind speed, precipitation.