ИСТИНА

This study evaluated surface heat fluxes in reanalyses (ERAInterim, NCEP/NCAR, ASR) in the Arctic ocean during last summer and fall period. Several surface conditions are compared: very new ice cover during a period of low temperature, ice-free conditions, ice with leads and melt ponds, pack ice and marginal ice zone. Our meteorological and micrometeorological observations were used to validate the temperature profiles and surface heat fluxes in the major reanalyses. We use data from Nansen and Amundsen basins observation system (NABOS) experiment to evaluate the performance of reanalyses for the Arctic Ocean. The NABOS field experiment was deployed in the central part of the Arctic and in the eastern Arctic seas during last summer and fall period 2004-2009, 2013, 2015. These data are independent and detailed enough to evaluate process descriptions. With its explicit treatment of the ice concentration, ERA-Interim generally reproduces the surface heat transfer, while NCEP/NCAR, based on a 55% concentration threshold, shows obvious disagreement with the observations in highly ice-covered and ice-free situations. The spatial and temporal patterns of the resulting flux fields will be investigated and compared with those derived from satellite observations such as HOAPS, from blended data such as AOFLUX (in the open water cases). Also a computation of the sensible heat flux at the surface is formulated based on the spatial variations of the surface temperature estimated with satellite. Based on the comparison of the data of field experiments, satellite data and reanalysis investigated the causes underestimation of the values of turbulent heat fluxes in the Arctic modern reanalysis. Obtained differences are related to the temperature and structural inhomogeneity of the surface and the development of space-organized convection fields. Reanalyses data are sometimes used to calculate the surface heat budgets over polynyas to estimate ice production in polar/sub-polar oceans. In particular, the near-surface air temperature and wind fields, which are difficult to observe using satellites or with in-situ measurements, are key parameters for estimating turbulent heat fluxes. If the sea-ice concentration and SST in reanalyses are not treated appropriately, careful attention is needed when using the resultant air temperature for such calculations. The work was supported by RSF (grant № 14-37-00053) and RFBR (grant № 17-05-01221).