The Use of Remote Sensing Data for the Application of a Hydrological Catchment Model

M. Wegehenkel (Germany)


Hydrological modelling, validation, remote sensing, evapotranspiration, NDVI


In our study, we tested simple methods for the evaluation of spatial input data from optical remote sensing data such as Landsat-TM5 for the application of a hydrological catchment model. Calculations of discharge and water balance dynamics were carried out for two catchments using a semidistributed conceptual hydrological model. For both catchments, land cover data were obtained from the classification of Landsat-TM5-scenes. The model was calibrated and validated comparing measured and simu lated discharge rates. In one catchment, six Landsat-TM5 subsets were used to calculate the Normalized Difference Vegetation Index (NDVI). These NDVI-distributions were compared with the corresponding simulated spatial patterns of actual evapotranspiration (ETr) rates. The co efficients of correlation between NDVI and ETr ranged within 0.10-0.88. A strong relationship between simulated ETr and NDVI, reported from other studies, could not be observed due to the higher spatial variability of NDVI in comparison with ETr. This indicated, that the model and the data in the present state have to be improved.

Important Links:

Go Back