Key points are not available for this paper at this time.
Abstract The latest version of the Soil and Water Assessment Tool (SWAT+) features several improvements compared with previous versions of the model, for example, the definition of landscape units that allow for a better representation of spatio‐temporal dynamics. To evaluate the new model capabilities in lowland catchments characterized by near‐surface groundwater tables and extensive tile drainage, we assess the performance of two SWAT+ model setups in comparison to a setup based on a previous SWAT model version (SWAT 3S with a modified three groundwater storage model) in the Kielstau catchment in Northern Germany. The Kielstau catchment has an area of about 50 km 2 , is dominated by agricultural land use, and has been thoroughly monitored since 2005. In both SWAT+ setups, the catchment is divided into upland areas and floodplains, but in the first SWAT+ model setup, runoff from the hydrologic response units is summed up at landscape unit level and added directly to the stream. In the second SWAT+ model setup, runoff is routed across the landscape before it reaches the streams. Model results are compared with regard to (i) model performance for stream flow at the outlet of the catchment and (ii) aggregated as well as temporally and spatially distributed water balance components. All three model setups show a very good performance at the catchment outlet. In comparison to a previous version of the SWAT model that produced more groundwater flow, the SWAT+ model produced more tile drainage flow and surface runoff. Results from the new SWAT+ model confirm that the representation of routing processes from uplands to floodplains in the model further improved the representation of hydrological processes. Particularly, the stronger spatial heterogeneity that can be related to characteristics of the landscape, is very promising for a better understanding and model representation of hydrological fluxes in lowland areas. The outcomes of this study are expected to further prove the applicability of SWAT+ and provide useful information for future model development.
Wagner et al. (Sun,) studied this question.