In Russia, the need for forecasting development in all areas of the national economy has recently increased. It is especially true in the field of agricultural land use management. Forecasting of agricultural land use is the process of establishing possible promising directions for the development of agriculture and determining the means to achieve the forecasted result in organizing the rational use of agricultural land in the future [1].
To achieve the goals set for forecasting sustainable agricultural land use most effectively, the selected forecasting method must correspond to these goals to the maximum extent.
There are two main classifications of these methods: by the degree of formalization and by the central idea underlying the method. Forecasting methods can be divided into two large groups: formalized (objective) and expert (subjective). Formalized (objective) forecasting methods are based on the use of mathematical statistics, including mathematical modelling. Formalized forecasting methods best highlight the relationships between the predicted indicators and the factors that influence them. Formalized methods include, for example, regression models, extrapolation, and others.
Let us consider in more detail the application of extrapolation methods to use agricultural land. Extrapolation methods are usually understood as the spread of past trends to the future, i.e. all extrapolation methods are based on the assumption that the future will be completely similar to the past and all trends observed in the past and existing now will remain unchanged in the future [4].
The basis for applying extrapolation methods is the concept of a time series. A time series (or dynamics series) is a set of values of an indicator arranged in chronological order. The values of this indicator (levels of time series) vary, and this variation is due to the influence of a whole set of factors. First, most time series have an increasing or decreasing trend. It reflects the cumulative longterm impact of many factors on the dynamics of the indicator under study. Factors taken separately may have a multidirectional effect on the result, but their cumulative effect, and their resultant, form a positive or negative trend. Secondly, the studied indicator can be subject to cyclical fluctuations. If these fluctuations have a stable variability over the seasons, then this cyclicity is called seasonal fluctuation.