ИСТИНА

Simulation and prediction of severe weather events including extreme winds is very important and actual problem. These events have many dangerous consequences for navigation, coastal ports infrastructure, oil-and-gas marine platforms and environment in the Arctic basin. Extreme winds are closely connected with many mesoscale processes and its interactions with synoptic-scale circulations; therefore, its fine resolving and investigation is determined and possible by high-resolution atmospheric modeling only. The COSMO-CLM model was used for many case-studies simulation of the most extreme winds observed over the Russian Arctic basin during the last 15 years. These extreme cases were sorted out previously from observations samples, based on the 0.99 quantiles of Weibull and Pareto distributions, i.e. the wind speed exceeded 20 – 25 m/s. Model runs were performed as for the unified ‘large’ domain with spatial resolution of ~15 km, as for the different ‘small’ domains (resolution of ~2 – 3 km), using the downscaling technology. These domains cover Barents Sea, part of Kara Sea and the surrounding water areas. Driving conditions came from ERA-Interim reanalysis (~0.750). Standard configuration of COSMO-CLM 5th version was applied: Runge-Kutta integration scheme with 5th advection order; 50 vertical levels; prognostic TKE-based scheme for turbulence; standard Tiedtke convection scheme. Runs continued for a week for the most cases, with the extreme situations observed near the middle period. Experiments has shown the 15 km resolution model runs could underestimate the maximal 10 m wind velocity up to 4 – 5 m/s, but it reproduces different synoptic situations and spatial distribution of wind speed and gusts well. However, after the downscaling to ~2.8 km, the reproduction of mesoscale fields was significantly better according to extreme wind speeds and gusts. The mean error and the root mean square error decreases to 2 – 3 m/s. The spatial distribution of wind fields it these cases turns more detailed and complicate, it is affected by the land-sea distribution, complex terrain and, perhaps, non-hydrostatic effects. Extreme coastal winds were observed mainly when the close isobars are parallel to the coast, or it is intensifying dynamically over the narrow sea belts (e.g., Kara Strait). Future tasks are to search an optimal model configuration for this region, applicate it for the extreme wind fields reproduction, and use many additional opportunities for adaptation (e.g., spectral nudging or other parametrizations). Besides, an attempt of the physical interpretation of these severe events will be done.