ИСТИНА

The Arctic is one of the most vulnerable to climate change regions in the world. On average, warming in the Arctic occurs two to four times faster than in the entire globe [Screen et al., 2012; Serreze et al., 2009; Solomon et al., 2007]. Also, the Arctic is one of the most important economically significant regions - both for Russia and for the whole world, since the Northern Sea Route passes through the Russian Arctic. The rapid climate change in the Arctic has a significant impact on the plans for the socio-economic development of the Arctic region, maritime shipping, the development of oil and gas shelf areas. Freeing the surface of the Arctic Ocean from ice contributes to an increase in the recurrence of extreme winds [Laffineur et al., 2014; Moore, Renfrew, 2005]. Information about extremely high wind speeds is very important - data on wind loads should be considered when designing and operating antennas, masts, bridges, and other structures on land, as well as ships and drilling rigs in the offshore area, which is why this study focused on wind speed. Considering the growing number of dangerous hydrometeorological phenomena, and the prospects for the development of the Arctic coast and the Northern Sea Route, it becomes necessary to study in detail the observed changes in the Russian Arctic, as well as to provide the region with relevant detailed hydrometeorological and climatic information with high spatial resolution. To make detailed assessments of climate change in the Russian Arctic becomes possible using the new model archive COSMO-CLM (COSMO-CLM Russian Arctic hindcast), covering the period from 1980 to 2016. with a spatial resolution of ~ 12 km [Platonov, Varentsov, 2021]. This hindcast, which includes about a hundred different hydrometeorological characteristics at both surface and model levels (50 levels), was created by hydrodynamic modeling using the COSMO-CLM regional atmospheric model. The calculations were carried out for the region that includes the Barents Sea, Kara Sea and Laptev Sea (Fig. 1). Currently, data are available from 1980 to 1992, 1994, 2000–2008, 2010–2016 (30 years), so the calculations were carried out using the data of these years. To assess the data on the wind speed of the model archive and conduct a comparative analysis, urgent data from the meteorological stations of Roshydromet from the site meteo.ru were used. For the study were selected 95 stations located within the model domain (Fig. 2). For comparison, station data were selected for the period from 1960 to 2016. Further, the model grids closest to the stations were found according to the coordinates. Then the statistics were calculated at these points: the differences between the values of the parameters in the model grids and at the stations, mean errors, standard deviations, and correlation coefficients. To study the extreme values of meteorological parameters, the quantile method was used. We calculated the quantile values for the stations and the corresponding model grids, as well as the differences between them. To estimate the maximum wind speeds, 95, 99, 99.9% quantiles were taken. All calculations were carried out using the MatLab software package. The research showed that the average wind speed is well reproduced by the COSMO-CLM Russian Arctic hindcast. However, there are stations where the model and the station data are not equal: in most of these stations hindcast is overestimating wind speed (Fig. 3a, 3b). The spatial distribution of the extreme values is well displayed on both model and station data. The study of the quantile difference showed that, in contrast to the average wind speed, the extreme speeds according to the hindcast data are underestimated - the error ranges from 2 to 10 m / s (Fig 4). For a more detailed and accurate study of the reproduction of wind speeds by the COSMO-CLM model archive, it is necessary to investigate the behavior of statistics of average and extreme values of wind speed on smaller time scales: seasonal, monthly, and daily. It is planned to estimate the studied statistics using satellite data in the future, and to compare the trends of the obtained characteristics using different data sources.