Trends of ERA5 atmospheric climate variables in the Arctic

Arctic Amplification implies a stronger warming of the Arctic compared to the rest of the world (Rantanen et al., 2022). There are many different interacting factors that can be considered as the cause of this phenomenon (Thoman et al., 2020; Goosse et al., 2018). The rapid changes in the Arctic could also have implications on other areas of the world, making it an interesting object to observe (Cohen et al., 2014). Therefore, quantifying the changes in meteorological variables over the last decades can provide important information about the interactions between the different variables, but also about possible changes in the future. Due to the sparseness of in situ measurements and observations in the Arctic, reanalysis models that combine model and measurements are often used for the analysis of environmental variables. In this study, a trend analysis was performed for a total of 13 meteorological variables using the European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis v5 (ERA5) dataset. In this process, trends of monthly and yearly values were calculated for each variable for the period from 1979 to 2021. Depending on the characteristics of the data and the fulfilled assumptions of trend methods, simple linear regression, Mann-Kendall and Theil-Sen methodology or a Mann-Kendall Theil-Sen algorithm with prewhitening (3PW, Collaud Coen et al., 2020) was used for the trend analysis. Temperature as the indicator of Arctic Amplification shows higher positive trends (increases) in the Arctic circle than in lower latitudes, whereas higher trends can be detected especially around Svalbard, Barents and Kara Sea. Those trends are mainly driven by the autumn and winter months. Longwave radiation and boundary layer height showed positive trends in the same regions, whereas shortwave radiation is decreasing in those parts. In the region Europe, decreases in cloud cover and relative humidity but an increase in shortwave radiation can be detected. Another important variable is the sea ice concentration, which is showing negative trends (decreasing) around the sea ice border in all of the months, reaching the maximum negative trend in autumn. A lot of variables like cloud formation, sea surface temperature or large-scale rain rate show a correlation to the observed trends in sea ice concentration. The results of the analysis show that the variables all interact or are dependent of each other and further analysis should be carried out to identify correlations between variables but also to identify potential roots of the changes. According to the analysis of the 13 studied variables, the region around Svalbard, as well as certain areas of the Arctic Ocean and Europe were identified as the most changing environments. Furthermore, an analysis of differences between the results when using different trend analysis methods was done. The analysis showed that depending on the method used, the results may vary, especially when using prewhitening. This