Theoretical foundations of probabilistic and statistical forecasting of agrometeorological risks

Еach model for forecasting agrometeorological risks based on the analysis of one-dimensional time series is effective a certain range initial information. In addition, values observations can differ significantly each specific case, respectively, widespread use one method arbitrary information lead to significant inaccuracies. Thus, problem choosing set data arises. this regard, universal adaptive probabilistic-statistical approach predicting proposed, which makes it possible solve model. The article presents results first stage research carried out with financial support Ministry Education and Science Russian Federation brief overview current state in direction presented, theoretical foundations onset drought frost have been developed, including task generating information, description basic models, also direct proposed presentation general structure an intelligent system, basis corresponding algorithm be developed automated as directions further work.