Application of SWAT and APEX Using the SWAPP (SWAT-APEX) Program for the Upper North Bosque River Watershed in Texas

Environmental models are used to assess and evaluate various best management practices (BMPs) at field (e.g., Agricultural Policy/Environmental eXtender, APEX) and watershed (e.g., Soil and Water Assessment Tool, SWAT) levels. However, models such as SWAT and APEX are only capable of simulating, mechanistically, a limited number of BMP scenarios individually. Therefore, this study was conducted to: (1) develop an automated program referred to as SWAPP to convert SWAT files to and from APEX format and simulate SWAT and APEX simultaneously, and (2) to evaluate this program using measured data from the upper North Bosque River watershed in central Texas. In this study, SWAT (version 2000) and APEX (version 2110) were utilized within SWAPP. Simulations from SWAT alone (SWAT-A) and the combined SWAT and APEX models in SWAPP were compared with the historical measured data collected in the upper North Bosque River (UNBR) watershed. In the first phase of this study, SWAT-A was applied to simulate all land uses in the UNBR watershed. In the second phase, stream flow and losses of sediment and nutrients from all crop and pasture lands within the UNBR watershed were simulated by APEX. The results obtained from APEX, and remaining land uses in the UNBR watershed (e.g., urban) simulated by SWAT, were then routed to the outlet of the watershed using the SWAT routing function. The measured flow, sediment, and nutrients at three sites within the UNBR watershed were simulated by SWAT-A and the combined SWAT and APEX within SWAPP for the period of January 1994 through July 1999. The results obtained from the three sites within the UNBR watershed indicate that the pattern (Nash-Sutcliffe efficiencies, E) and average monthly values of flow and loadings predicted by the combined SWAT and APEX within SWAPP were similar, and in some cases (mainly for sediment and nutrients loadings) closer to the measured values, as compared to SWAT-A. For instance, the E values for flow obtained from SWAPP ranged from 0.65 to 0.74, while those from SWAT-A ranged from 0.55 to 0.75. The E value for sediment loading at three sampling sites ranged from 0.55 to 0.74 for SWAPP, as compared to 0.41 to 0.72 for SWAT-A, and the E values for nutrient loadings ranged from -0.04 to 0.88 for SWAPP, compared to -0.74 to 0.82 for SWAT-A. The results obtained from this study indicated that the SWAPP program not only improved the modeling results, but also provides an opportunity to utilize APEX within the SWAPP program for simulating management practices, such as multicropping or filter strips, at the field level, whereas SWAT alone currently has limited capability to simulate these practices. In addition, SWAPP can be used to rapidly convert SWAT data files, generated from Geographical Information System (GIS) layers, to APEX data file format.