Recent developments in simple adaptive flow forecasting models in Poland

Abstract

ABSTRACT Real-time flow hydrograph forecasting models have been operational in Poland as a routine service since 1975. A trend in recent years is a shift away from complex conceptual catchment models to simple ones of the storage function type. Another approach has been to forecast the mass flow curve (S-curve) using an ARMA-type model the parameters of which are estimated via a linear Kalman filter algorithm. Examples are shown of the application of the latter approach to models of both the rainfall-runoff category and the channel routing category including estimation of ungauged lateral inflow in the latter models. Previous experience with the application of a linear Kalman filter to flow hydrograph forecasts is discussed. Derivation of the forecast flow hydrograph from the S-curve is done by converting flow volume increments into flow discharges assuming a linear shape for the flow hydrograph within the time step and a trend corresponding to the last observed trend. In the vicinity of the hydrograph pe...