Résumé
This study analyses the sensitivity of a hydrological model to different ways of interpolatingclimate forcing on the Elqui basin (5660 km2) in the Chilean Andes. A 36-year period (1976–2011) waschosen in order to account for the hydro-climatic variability. Precipitation and using the inverse distanceweighted methods were interpolated on a 5 × 5 km grid based on 12 and eight stations, respectively. Elevation effects on precipitation and temperature distribution were considered using a digital elevationmodel. Two precipitation datasets (with and without a mean altitudinal gradient) and three temperaturedatasets (using constant or monthly lapse rates based on altitudinal bands) were computed. All datasetcombinations were assessed through the calibration of the GR4j model including a snow reservoir.Calibration was performed by the succession of Rosenbrock and simplex algorithms using a multi-objectivefunction. Results show that the dataset based on a constant lapse rate of 6.5°C/km for temperature and noelevation effects for precipitation is sufficient to accurately simulate discharge and the snowmelt regime ofthe catchment over the last 30 years.