SWAT+ vs. SWAT2012

The current FORTRAN codes of SWAT (SWAT2012) are being rebuilt to ease new model and module development by incorporating object-oriented programming techniques into defining spatial objects and calculating variables. SWAT+, the new SWAT code, is expected to offer flexible watershed representation with linked spatial objects. Ideally, SWAT+ produces outputs that are comparable to SWAT2012, if not identical, when the two versions of the SWAT model are set with the same input for a watershed. However, the drastic restructuring of the source code may cause unexpected results in the output especially compared with the current SWAT2012 model that has a long history of model validation published in numerous research papers. In this paper, we compare sub-daily, daily, and monthly hydrographs provided by the SWAT2012 and SWAT+ models that were prepared for a small urban watershed located in Austin, Texas. The model parameters were calibrated to 15-min and daily runoff observed in the watershed. Then, the performance of the models was compared to each other to evaluate the differences in the simulated hydrographs. This paper provides ideas of how SWAT+ is organized and if a SWAT modeler can keep consistency in modeling results when shifting from the original to the new codes. In addition, this study makes recommendations for improved urban SWAT modeling.