Methods for the streamflow forecast correction / Borsch S.V., Simonov Y.A., Khristoforov A.V. // Hydrometeorological Research and Forecasting, 2020, no. 1 (375), pp. 162-175.

Three methods for the streamflow forecast correction are described.

The first method is based on the conversion of forecasts using the linear regression equation. The method is recommended for all types of hydrological forecasts. The efficiency of the method depends on a degree of correlation between the actual values of hydrological characteristics and their forecasts.

The second method is based on replacing forecasts that are within the certain intervals of values by the average values of the predicted characteristic. This method prevents from unreasonably large or small forecast values and is recommended to reduce the error of forecasting techniques with a doubtful efficiency.

The third method is widely used in the practice of hydrological forecasting. It takes into account the errors of previous forecasts based on the autoregressive model that describes their sequence. The effectiveness of the method depends on a degree of correlation between the forecast errors for close dates. This method is recommended for correcting daily short-term forecasts of hydrological characteristics.

The potential of the considered correction methods is demonstrated by examples of refining various short-term and long-term river flow forecasting techniques.

Keywords: error of streamflow forecasting techniques, correction of hydrological forecasts, linear regression, forecast variability range, partial averaging, autoregressive model

DOI: https://doi.org/10.37162/2618-9631-2020-1-162-175

Tab. 1. Fig. 1. Ref. 15 .