Вестник МГТУ. 2017, том 20, № 1/2.

Вестник МГТУ. 2017. Т. 20, № 1/2. С. 231–241. DOI: 10.21443/1560-9278-2017-20-1/2-231-241 241 I. F. Zaporozhtsev, D. V. Moiseev Simulation of spatio-temporal variability of temperature in the Taganrog Bay with MITgcm model The goal of the paper is to analyze efficiency of MITgcm in simulation of temperature fields' variability in the Taganrog Bay. Authors are the first to consider hydrodynamic modeling approach based on MITgcm for this bay. In situ temperature values to be compared with the model data have been obtained during two Murmansk Marine Biological Institute Kola Scientific Center RAS and Southern Scientific Center RAS coupled expeditions in summer and autumn of 2005. The step of calculation grid is agreed with the step of direct measurements stations grid. The obtained during cruises temperature and salinity data cover the Taganrog Bay with 2' latitude and 3' longitude steps (and with 4' latitude and 6' longitude steps correspondingly for thermohaline homogeneous areas). Depth step is 0.5 m. Data for initialization and atmospheric forcing have been taken from public reanalysis databases and atlases, datasets limitations are discussed. To simplify boundary conditions simulation has been carried out for the whole Azov Sea. Numerical experiments series has been fulfilled to determine the optimal start date of simulation and initial constant temperature field. In view of significant experiment time cost optimization task has been solved for restricted parameters values set and with doubled grid steps (4' latitude and 6' longitude steps). The determined values have been used to solve original task of model data verification with the measured ones. As far as the results obtained by the authors with the particular workstation PC are concerned, the conclusion about possibility of MITgcm simulation in real areas without specialized high- performance computers has been given. Key words: model MITgcm, Taganrog Bay, Azov Sea, simulation of spatio-temporal variability of temperature, model verification with in situ data.