About the HHFS hydrological forecast graph

The hydrological forescast graph consists of two parts. The first part presents observed water level values of the last 14 days (with blue color), the second part shows water level forecasts for the next 6 days (with red color). The time of the observed and forecasted values is applied for 1, 7, 13 and 19 if available. The graph contains the height of some critical levels such as HHW, LLW, 1st, 2nd and 3rd alert levels. In the lower left corner the time of issue can be seen, while in the lower right corner the applied meterological model and its running time are shown. The uncertainty of the hydrological forecast is indicated by the error interval (domain between the two purple lines). The error interval is defined based on the last years' root-mean-square error. Water level values are expected to remain within this interval with the probability of 70%.

At certain gauges close to the spring or gauges on high altitudes water level rising usually occurs within a short time. Consequently, hydrological regime at these gauges is defined by the performance of the meteorological models, mainly precipitation forecast. Since uncertainties of both precipitation and hydrological forecasts can increase significantly with lead time, hydrological forecasts with several days lead time can contain higher uncertainties than the average uncertainties indicated by the error interval and thus they are rather intented to draw attention to the possibility of the predicted hydrological situation.

In order to highlight the uncertainties of precipitation forecasts, lead time of hydrological foreacasts is divided into two parts. The first part (marked with light grey color) presents shorter lead time hydrological forecasts caused by meteorological events occured in the past period. In the second part (marked with dark grey color) longer lead time hydrological forecasts are presented which mainly rely on meteorological forecasts, thus their accuracy strongly depends on the performance of the meterological forecast model. The proportion of the two parts basically depends on the concentration and propagation times at the given gauge.

For gauges situated in backwater reaches another significant error source can be the artificial water regime resulted from the unknown working program of power plants.