Effect of the accuracy of spatial rainfall information on the modeling of water, sediment, and NO3–N loads at the watershed level

In a given watershed, the accuracy of models in predicting the hydrologic and erosion behavior depends, to a large extent, on the quality of the knowledge in respect of the spatial rainfall. The hydrologic and erosion aspects of rainfall are often discussed without due regard to any resulting improvement in watershed modeling. Thus, there is a real need for streamlining raingauge networks in order to reflect rainfall variability and its effect on the prediction of water, sediment and nutrient fluxes at the watershed scale. In this study, such an impact was analyzed using 9-year data collected at the outlets of two watersheds encompassing a range of climates, surface areas and environmental conditions. The Soil and Water Assessment Tool (SWAT) was applied using as input data that collected from 1 to 15 precipitation gauges per watershed. At both sites the highest densities of raingauges were used for SWAT calibration. The differences between the highest gauge concentration and lower concentrations used for the estimation of sediment loads led to the conclusion that a high gauge concentration is necessary. At both watersheds, predictions using rainfall records from the national service stations produced inaccurate estimations. This was probably because the gauge concentration was too sparse. Finally, the general applicability of these results is proposed by displaying the possibilities of extrapolation to other watersheds or models.