On the Forecast of Long-Term Changes in the Hydrological Regime of Rivers Using the Results of Climate Modeling

Abstract

A method is proposed for probabilistic forecasting of river flow under non-stationary conditions based on the Bayesian approach and using the results of climatic system modeling. A forecast is given for the minimal discharge of the Volga basin rivers based on the correlation between variations of winter runoff and increments in winter air temperatures. The effect of the inter-model dispersion of the predicted air temperature on the results of river flow forecasting is examined.