4.5 Article

The NHDPlus dataset, watershed subdivision and SWAT model performance

Journal

Publisher

TAYLOR & FRANCIS LTD
DOI: 10.1080/02626667.2014.916408

Keywords

SWAT simulation; NHDPlus; watershed subdivision; HRU definition; runoff; sediment; total nitrogen

Funding

  1. United States Environmental Protection Agency (USEPA) through its Office of Research and Development

Ask authors/readers for more resources

The Soil and Water Assessment Tool (SWAT) has been developed to evaluate the effectiveness of agricultural management practices on watershed water quality. Many studies have indicated that watershed subdivision can affect the accuracy of model predictions. Most of them used the minimum drainage area (MDA) to delineate sub-watersheds, and varied the value of MDA depending on the size of the watershed being modelled. Instead of MDA, we use the National Hydrography Dataset Plus (NHDPlus)-an integration of the best features of the National Hydrography Dataset (NHD), Watershed Boundary Dataset (WBD), National Elevation Dataset (NED), and the National Land Cover Dataset (NLCD)-to delineate the watershed. The Kaskaskia River watershed in Illinois, USA, was selected to investigate the individual effects of sub-watershed and hydrologic response unit (HRU) delineations on predicted streamflow, total suspended sediment (TSS) and total nitrogen (TN) losses at two USGS gauges. In addition, an MDA of 3000 ha, and four levels of stream (the 2nd, 3rd, 4th and 5th order) were evaluated. Three levels of HRU threshold (5%, 10% and 15%) were used for each stream order model. The results show that stream order had little effect on predicted streamflow, but a great impact on TSS and TN losses, and the impact of HRU delineation became greater when a higher stream order was used to delineate the watershed. For higher stream order, fewer streams were recognized in SWAT simulations, which resulted in less sediment routing and channel processes, which, in turn, led to less deposition in the channels; thus high sediment losses were obtained at the watershed outlet. However, fewer channel processes led to less in-stream N processes; thus lower TN losses. Overall, the SWAT simulations performed the best when the 2nd stream order was used for delineations comparing with USGS observed data, followed by the 3rd stream order. Therefore, to fully depict the watershed characteristics to perform SWAT simulations, a stream order higher than 3rd order is not recommended for watershed delineation.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.5
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available