We used a satellite-derived global daily sea surface temperature (SST) dataset with resolution 0.25 × 0.25${}^{\circ }$ to analyze interannual changes in the Arctic Shelf seas from 2000 to 2020 and to reveal extreme events in SST distribution. Results show that the second decade of the 21st century for the Siberian Arctic seas turned significantly warmer than the first decade, and the increase in SST in the Arctic seas could be considered in terms of marine heatwaves. Analyzing the spatial distribution of heatwaves and their characteristics, we showed that from 2018 to 2020, the surface warming extended to the northern deep-water region of the Laptev Sea 75${}^{\circ }$ to 81${}^{\circ }$N. To reveal the most important forcing for the northward extension of the marine heatwaves, we used three-dimensional numerical modeling of the Arctic Ocean based on a sea-ice and ocean model forced by the NCEP/NCAR Reanalysis. The simulation of the Arctic Ocean variability from 2000 to 2020 showed marine heatwaves and their increasing intensity in the northern region of the Kara and Laptev seas, closely connected to the disappearance of ice cover. A series of numerical experiments on the sensitivity of the model showed that the main factors affecting the Arctic sea-ice loss and the formation of anomalous temperature north of the Siberian Arctic seas are equally the thermal and dynamic effects of the atmosphere. Numerical modeling allows us to examine the impact of other physical mechanisms as well. Among them were the state of the ocean and winter sea ice, the formation of fast ice polynias and riverine heat influx.

There is considerable uncertainty about the methane emission from Arctic shelf seas. Methane fluxes in this region can be underestimated and play a significant role due to the large volume of gas contained in bottom sediments in the permafrost and gas hydrates. We analyzed the model sensitivity to the parameterization of gas exchange processes on the sea surface based on the numerical modeling of the transport of dissolved methane in Arctic seas. A dissolved methane transport model is included into the basic model of the ocean and sea ice SibCIOM developed at Institute of Computational Mathematics and Mathematical Geophysics, Siberian Branch, Russian Academy of Sciences. Methane emissions into the atmosphere were estimated with various parameterization of the gas exchange process in the “water–atmosphere” and “water–ice–atmosphere” systems with NCEP/NCAR reanalysis data. The uncertainty of the estimate of annual methane emission amounted to 6–12% when considering different dependencies of gas exchange on wind. The scheme which considers the ice cover has a more pronounced influence on methane flux calculations: the uncertainty increased to 50–130%. Parameterization of the relation between ice cover and gas exchange can have a great effect on the calculated methane fluxes and lead to underestimation of its emission from Arctic shelf seas.