Four methods of forecasting: „no-change", LOESS, local linear regression and Holt-Winters were applied to annual minimum water levels observed at ten cross-sections of two tributaries of the Vistula river. The 1-, 2-, ..., 5-year forecasts were made for each year after some initial year, and four quality measures: bias, root mean square error, mean absolute error and maximum absolute error were calculated for each time series and lead time. The naïve model turned out to be always the worst in it bias and almost always very good, sometimes the best regarding the other measures.
...
The analysis of time variability of riverbed elevation recorded at four cross-sections of two rivers: Raba and Dunajec was made. The causes of very intense erosion were pointed out, namely: intensive sediment extraction, inadequately carried out river training works including shortcuts and narrowings of riverbeds, and constructions of hydro-technical facilities, disrupting the continuity of sediment transport along the river length. Forecasting was carried out using a local weighted regression method, LOESS. In order to examine how far the length of the historical series influences the quality of forecasts, a comparison was made between the 5-year forecasts based on the 5-year history and the 5-year forecasts based on the 10-year history. The results suggest the need for preliminary data analysis before using a particular version of the model in or-der to take into account the type of variation exhibited by the historical series. ...
