Land cover and its characteristics influence actual and long term water balance of landscape segment, particularly evapotranspiration, infiltration, surface and hypodermic runoff. Thus description and modelling of the particular events allow determining basic requirements for land cover classification effective to rainfall-runoff (further in the text as RR) condition assessment. The cooperation “remote sensing – GIS” together with hydrological models proposes solution for comprehension of the investigated phenomenon.

LANDSAT ETM+ data represents a valuable and frequently used data source for description of LULC data as a significant input to rainfall-runoff modelling (further in the text as RRM). LC classified according RRM requirements were compared to the CORINE LC data. The paper shows results comprising very important differences and proves irreplaceableness of remote sensing data for the purpose especially in conditions such as rapid landscape changes (e.g. geological hazards), seasonal changes in vegetation cover, and significant changes in agricultural areas during the year; due to its capability of providing the current LC information.

Land cover data sources (CORINE, LANDSAT ETM+) were use for CN-curve value association with particular land cover and soil conditions and the traditional and new approach of CN-value determination were tested in RRM for 2 small catchment areas (Bela and Olse). Such process requires an evaluation of individual data sources, their processing and evaluation of outcomes from RRM.