Modeling of the expected yield in the steppe district of the Cis-Ural region taking into account the lag variables and changes of the distance from the mass center of the Solar system to the Earth

Abstract

A more precise method of modeling of the long-term yield forecast by the example of spring wheat was used. The close connection of the long-term dynamics of the spring wheat yield is specified with the changes in time of the distance from the barycenter of the Solar system to the Earth, which permits using the value of the predictors beyond the available yield range. The necessity to consider the influence of the lag variables while developing the models of forecast of this index is shown. The methods of the standard multiple regression and regression in the neural networks are applied.