Process automation of the long-term forecasting of ice break-up and ice disappearance dates at the Dnipro reservoirs

The article is devoted to the creation of an automated system "Ice-Spring" for long-term forecasting of dates of ice break-up and ice disappearance at the Dnipro Cascade reservoirs by writing a computer program for its further use in the operative work of the Ukrainian Hydrometeorological Center of the State Emergency Service of Ukraine (UkrHMC). To automate the forecasting process, the "Ice-Spring" system software was created, which is written in the C# and C++ programming languages in the Visual Studio 2022 Community Edition environment in the view of Windows forms and console applications. The "Ice-Spring" system allows for long-term forecasting of the dates of ice break-up (February 20) and ice disappearance (March 5) at the Dnipro Cascade reservoirs at 36 water gauges each year. Forecasting is carried out using 12 forecast dependencies based on regression relationships between dates on indicator water gauges and teleconnection indicators. In this case, the mean monthly values of 11 teleconnection indicators are used, which affect the processes of ice break-up and ice disappearance at the Dnipro Cascade reservoirs, which are determined by the National Weather Service (NWS) of the National Oceanic & Atmospheric Administration USA (NOAA). Forecasting on other water gauges is carried out using 60 regression dependencies between the dates at these gauges and the dates at the indicator water gauges. The created software for the automated long-term forecasting system "Ice-Spring" has successfully passed testing and has shown its performance for independent data for 2021. At the same time, carried out evaluation of forecasting efficiency has shown acceptable results, since of 12 forecasts, only 2 forecasts did not come true. The "Ice-Spring" system was transferred to the UkrHMC for testing and operative use.