A Long-Term Forecast Model of Spring Runoff: The Case of the Belaya River

Abstract

The results of long-term forecasting of the spring runoff in the Belaya River basin based on the water-balance model are presented. The structure and parameters of the water-balance model equations are optimized by the Levenberg-Marquardt algorithm with restrictions imposed on the input data values. The obtained values of the equation coefficients are checked according to the Δ/σ criterion adopted in the hydrometeorological service. The reliability of the forecasting procedure is assessed by statistical calculations of the stability of their parameters and test calculations on the independent sample. All equations obtained during the numerical experiment can be suitable for making forecasts.