Parametrization of the model DNDC for evaluating components of carbon biogeochemical cycle in the European part of Russia

Abstract

The aim of this research was to parameterize the simulation model DNDC (DeNitrificationDeComposition) to estimate components of the carbon biogeochemical cycle in arable soils of the European part of Russia. Methodological recommendations for DNDC use were developed on the basis of official statistical data, literature sources, as well as our results obtained earlier during initial approbation of the model. Principles of preparation of input information such as natural (climatic and soil conditions) and anthropogenic (tillage, fertilization, crops yield) factors were established. Internal parameters of the model including soil characteristics (depth of top soil, C : N ratio, and soil organic carbon fractions) and crops biology (heat and water demand, C : N ratio, and biomass fractions) were corrected; also a standardized description of the main cultivation technologies was developed in the model’s format. A method for evaluation of modelling efficiency was offered; it is a complex consisting of five criteria: the Nash-Sutcliffe coefficient, the Theil coefficient, the Pearson correlation coefficient, ANOVA, and the F-test of equality of variance. For verification of the adapted model DNDC to Russian conditions, data of two field experiments aimed at measuring СО2 emissions in Kursk and Moscow regions were used, as well as published information about СО2 emission and carbon balance in arable soils in Kursk, Moscow, Orel, Samara, and Vladimir regions. At the step of verification, the accuracy of obtained results was confirmed by a set of relevant statistical criteria: there was medium correlation between experimental values of CO2 emission and modelled ones; as results of ANOVA showed, the average measured value of emission and the calculated one are similar; Theil coefficients less than 0.3 and positive Nash-Sutcliffe coefficients confirmed the correctness of changes inputted into the model. It was proved that the developed parametrization method is effective for estimating components of carbon biogeochemical cycle in the European part of Russia.