Assessment of the influence of land use data on the water balance components of a peri-urban catchment using a distributed modelling approach

• We compared long term modelling outputs for 5 land use maps on a peri-urban catchment.
• Total discharge is not very sensitive to the map used for large or rural zones.
• Flow components (surface runoff, interflow, baseflow) are the most sensitive.
• Precision of land use information is important for small urbanised sub-catchments.

SummaryThis paper addresses the impact of the source and processing method of land use information for hydrological simulations on the long-term water balance of the Yzeron peri-urban catchment (150 km2), located near Lyon, France. A customised version of the distributed hydrological model J2000 was used to perform simulations at a daily time step. Five land use data sets obtained from aerial photographs BDOrtho@IGN and satellites Quickbird and Spot for the year 2008 are compared. The paper presents the methodology for model setup and the simulation results for the main water balance components of the catchment: total runoff at several gauging stations, runoff components, evapotranspiration and soil moisture. The model evaluation against discharge measurements at six locations shows a reasonable agreement between simulated and observed values, in particular for general seasonal variations, low flow periods and simulation of runoff components (surface runoff, interflow and base flow), with Nash–Sutcliffe efficiencies ranging from 0.25 to 0.51 at the daily time step and 0.46–0.82 at the monthly time step. The comparison of the model outputs for the various land use maps shows that the total discharge is not very sensitive to the data set used (−4.88% to 4.65% at the catchment outlet), except in a small and more densely urbanised sub-catchment for which a significant impact of image resolution on simulated flow is detected (+25.81%). For all gauges, the results also highlight the sensitivity of the modelled flow components, in particular regarding the amount and seasonal dynamics of surface runoff generation (8–44% of total flow at the catchment outlet depending on the data set used). As a conclusion, land use information should be selected and processed with care, with respect to the objectives of a given study, and the sizes and urbanisation rates of the target sub-catchments.