Аннотация:A multimodal approach is offered – the joint use of simulation models, carbon calculators and statistical methods to analyze sources, causes and intensity of GHG emissions in crop production and animal husbandry, as well as to create a strategy and practical recommendations for reducing them in agriculture. Such a complex approach is important because each model or calculator is characterized by specific mathematical structure, different equations, mistakes, distinct input parameters, time scales and based on various field experiments. We chose three models that were successfully approbated and applied at the national and local levels around the world: DNDC, RothC, T&P. DNDC was parameterized to estimate components of the carbon cycle in arable soils of the European Russia. It was applied for retrospective analysis (1990-2017) of soil organic carbon, soil respiration, and net ecosystem exchange in the Central Forest zone. This model was for the first time applied in the Central Asia for the estimation of CO2 balance of pastures. It successfully reproduced net ecosystem exchange in Uzbekistan for the period of 1998-2001. The method of preparing input data for RothC for arable soils of the European Russia was developed. Later DNDC and RothC were used to reconstruct dynamics of soil organic carbon and СО2 emission from arable chernozems under five crops in Kursk region over 1990–2018. The use of GHG calculators to estimate carbon footprint at farm level is especially relevant in animal husbandry. The widely used Cool Farm Tool, AgRE-Calc, Farm Carbon Calculator, and EX-Act were analyzed for their effectiveness to assess emissions from livestock in Russia by the examples of three farms in order to develop recommendations to reduce GHG footprint. The study was supported by the Russian Science Foundation project no. 20-76-00023 and the state assignment for the Institute of Geography RAS no. 0148-2019-0009.