Impact of spatial rainfall variability on hydrology and nonpoint source pollution modeling

Rainfall is regarded as the most important input for the hydrology and nonpoint source (H/NPS) models and uncertainty related to rainfall is generally recognized as a major challenge in watershed modeling. In this paper, we focus on the impact of spatial rainfall variability on H/NPS modeling of a large watershed. The uncertainty introduced by spatial rainfall variability was determined using a number of commonly-used interpolation methods: (1) the Centroid method; (2) the Thiessen Polygon method; (3) the Inverse Distance Weighted (IDW) method; (4) the Dis-Kriging method; and (5) the Co-Kriging method. The Soil and Water Assessment tool (SWAT) was used to quantify the effect of rainfall spatial variability on watershed H/NPS modeling of the Daning watershed in China. Results indicated that these interpolation methods could contribute significant uncertainty in spatial rainfall variability and the carry-magnify effect caused even larger uncertainty in the H/NPS modeling. This uncertainty was magnified from hydrology modeling (stream flow) into NPS modeling (sediment, TP, organic nitrogen (N) and dissolved N). This study further suggested that H/NPS prediction uncertainty relating to spatial rainfall variability was scale-dependent due to the averaging effect of spatial heterogeneity. From a practical point of view, a global interpolation method, such as IDW and Kriging, as well as elevation data derived from a digital elevation model (DEM), should be included into the H/NPS models for reliable predictions in larger watersheds.

► The impact of spatial rainfall variability on H/NPS modeling was evaluated.
► The uncertainty of rain input would ne magnified into H/NPS modeling.
► The spatial interpolation method is highlighted in large-scale watersheds.
► Global interpolation method and DEM data could benefit H/NPS modeling.
► The selection of interpolation methods for H/NPS models is scale-dependent.