The goal of the research is to design and to evaluate parameter identification algorithms for Agrotool productivity model adapted for the Altai region. It is required to determine permissible limits of the model parameters changes and to perform the model validation for predicting crop yield. In the paper, problems of Agrotool productivity model adaptation for conditions of the Altai region are discussed. Namely, there are identification of model parameters in accordance with agro-meteorological data of the region and study of the model sensitivity to variations and accuracy of model parameters. Optimal values of parameters are evaluated by methods of global optimization with diagonal approach to a solution of multidimensional problems. The scheme of the diagonal algorithm is provided. Results of identification for blocks of soil moisture regime, phenological development, and crops productivity are analyzed. Also, the accuracy criteria for setting the permissible limits of the model parameters are elaborated.