Estimating the timing of the complete clearance from ice of the Russian Arctic seas in summer period

The article discusses a statistical method for forecasting the complete ice clearance in September of the seas of the Russian Arctic, the information basis of which was the data on the area of sea ice extent (SIE) for the period 1979-2021, borrowed from the archives of the AARI and NSIDC. The physical basis of the method is the assumption of stationarity (invariability) of climatic changes in the “ocean-sea ice-atmosphere” system, that is, their preservation for the entire forecasting period. The statistical basis is the calculation of linear trends and significant cyclical fluctuations. It is shown that the contribution of linear trends to the variance of the annual SIE values is significant for all seas; moreover, as it moves from west to east, it increases and reaches a maximum in the Chukchi Sea. According to both archives, the earliest clearing of ice in September occurs in the Chukchi Sea and is associated with its small area and the direct influence of the influx of warm waters from the Pacific Ocean through the Bering Strait. According to NSIDC data, reaching the ice-free regime occurs much later than according to the AARI data. Errors (uncertainties) in the SIE forecast depend on the degree of stability of linear trends over time. Therefore, 10 trends were calculated for previous years, and the maximum and minimum trends were selected for each sea. The difference in the years when the ice-free regime was reached according to these trends shows the area of uncertainty in the prognostic estimates. According to NSIDC data, even the minimum spread (Chukchi Sea, 11 years) is greater than the maximum according to the AARI data (East Siberian Sea). A small range of uncertainties for the SIE of the seas AARI data is associated with the peculiarities of the frequency structure of their time series after 2005. However, even the uncertainties according to the NSIDC data are small and significantly less than the uncertainties arising in the forecast of SIE based on climate models.