*Alexander Sokolov1、Tatiana Moiseenko2、Natalia Gashkina2
(1.Institute for information transmission problem RAS、2.Vernadsky Institute of Geochemistry and Analytical Chemistry RAS)
キーワード:pollution transformation, nickel, cuprum, subarctic water and terrestrial ecosystems, mathematical modeling, balanced identification
The problem of choosing an optimal accuracy model that describes heterogeneous data (atmospheric emissions of heavy metals, pollution of lake water, soil and bottom sediments) associated with the operation of the Pechenganickel plant (Kola Peninsula) is considered. Balanced identification technology is used: by minimizing the mean square cross-validation error, a model with an optimal balance of complexity and proximity to the data is selected. The model was used to describe the spatial dynamics of the state of small lakes and their catchments. The retrospective state of the simulated objects (from 1946) is assessed and a forecast until 2030 is proposed. The issues of assessing the reliability of forecasts are discussed.
Funding: This work was supported by the Russian Science Foundation under grant no. 22-17-00061.