ИСТИНА

While substantial research has been devoted to methane emissions from water-saturated soils over the past four decades, the absorption of CH4 by non-saturated soils, despite their expansive coverage, has received less focus. There's also limited data concerning methane consumption in temperate non-saturated soils, which are prevalent in forests, grasslands, steppes, and croplands. This data scarcity has result-ed in estimated uncertainty: methane consumption ranges between 1% to 15% of the global methane sink attributed to photochemical degradation in atmosphere. In the proposed study, based on an analytical review, four simulation mathematical models were selected (Dörr et al. [1993] modified by Glagolev and Filippov [2011]; Curry [2007]; DLEM model [Tian et al., 2010]; and the MeMo model without autochtho-nous CH4 sources [Murguia-Flores et al, 2018]) to implement an ensemble approach to modeling methane uptake on the example of chernozem soils of various agrarian and natural ecosystems in south of the European Russia (croplands, broad-leaved forests, fallows, and forest-steppe). Own data on soil characteristics, land use, vege-tation and weather conditions were used as input variables. To verify the quality of modeling, we conducted instrumental measurements of methane fluxes in the stud-ied biotopes at the height of the growing season. The highest soil methane uptake rates were recorded in broadleaf forests (1.3-3.4 mg CH4 m-2 d-1) and the lowest in croplands (0.6-1.0 mg CH4 m-2 d-1). The average across all sites relative simulation error was 36% and the average variation in experimental data was 26%. Checking various ways of combining the results of individual models in an ensemble showed that the best results, according to Theil's criterion of inequality, are demonstrated by the simplest operators: the half-sum of extreme terms and the arithmetic mean. The ensemble approach has not been previously applied to solve the problem of methane uptake by soils.