Detecting the influence of land use changes on discharges and floods in the Meuse River Basin - the predictive power of a ninety-year rainfall-runoff relation?

Quantifying how changes in land use affect the hydrological response at the river basin scale is a current challenge in hydrological science. A daily discharge record (1911-2000) of the river Meuse (21000 km 2; Western Europe) has been simulated with a semi-distributed conceptual model (HBV). The model has been calibrated and validated with a data set for the period 1968-1998. In this study the performance of the model for the period prior to 1968 has been analysed. The observed and simulated discharge records are compared in terms of annual average discharge. summer and winter average discharge, annual maximum daily discharge, and annual maximum 10-day average discharge. The results are discussed with reference to land use change (i.e. forest type change) and shortcomings of the available precipitation and discharge records. The general agreement between the observed and simulated discharge records is good (Nash-Sutcliffe efficiency: 0.89-0.93). in particular flood volumes and the highest flood peaks are simulated well but the model has problems with the medium floods (shape and peak value). However, there are some systematic deviations between the observed and simulated discharges during specific periods. The simulation result could somewhat be improved by taking the historical land use into consideration. But the systematic overestimation of the discharge for the period 1933-1968 could not be attributed to observed changes in land use. It is concluded that the overall impact of land use changes in the Meuse basin is too small to be detected given the uncertainties in the available records.