ИСТИНА

Flood forecasting methods have been developed in order to prevent dangerous floods on the rivers Mzymta, Sochi, Zapadnyi Dagomys, Kuapse, Tuapse and Vulan, which flow from the southern slope of Caucasus the Black sea. Flood forecasting technique is based on a regional physical-statistic model of runoff formation. Such model allows providing river flow forecasts with use of available data from the only hydrological gauge and the only meteorological station within the basin also considering absence of data on snow cover. Runoff formation processes are simulated with the time-step of 1 day. Model parameters have been evaluated for each month separately. Verification methods for independent data showed sufficiently high accuracy and efficiency of the developed methods. The system of methods of probabilistic forecasting of floods with 1 day lead-time was proposed. Depending on the forecasted discharge value it lets evaluate probability of exceeding the critical discharge and water levels for the next day. A scheme for the computation of daily critical precipitation amount at the meteorological stations, which cause the exceeding with a given probability of critical discharges and water levels during the expected day, has been developed. Since 2013 the proposed forecasting system has been applied in an automated way in Hydrometcenter of Russia and is used for operational warning of dangerous floods on the chosen rivers of the Black Sea coast.