Non-homogeneity of hydrometric data and estimating the rating curve

One of the main tasks of contemporary hydrology is to plot the best possible rating curve, which has numerous applications in practical hydrological problems. The quality of the curve depends to a significant degree on the precision of hydrometric measurements and the precision in determining the relationship between discharge and the stage (water level). In this study, we engage the multidimensional problem of non-homogeneity of hydrometric data. The distances between a given point of measurement and the given cross-section proved to be highly diverse. Hence, we propose data transformation so that every measurement can be treated as taken in a given cross-section. The transformation allows partial homogenization of measurement data. An automatic non-linear method for estimating the rating curve is also proposed. The method has been tested on selected representatives of mountain, submontane, and lowland rivers of the Oder River catchment basin (Lower Silesia, Poland). Cross-sections selected for analysis were: Łąki on the Barycz River, Lądek Zdrój on the Biała Łądecka River, Jelenia Góra on the Bóbr River, Jakuszyce on the Kamienna River, and Łozy on the Kwisa River.