Evaluating best management practices to lower selenium and nitrate in groundwater and streams in an irrigated river valley using a calibrated fate and reactive transport model

A calibrated fate and reactive transport model is applied to evaluate alternative water and land best management practices (BMPs) in Colorado's intensively irrigated Arkansas River Valley to attenuate nonpoint source pollution and more closely meet regulatory standards for selenium (Se) and nitrogen (N) in groundwater and streams. Reduced irrigation (RI), lease fallowing (LF), canal sealing to reduce seepage (CS), reduced fertilizer application (RF), and enhanced riparian buffers (ERB) are explored as stand-alone BMPs, and in combination, at basic to more aggressive levels of implementation. The distributed-parameter model, which couples MODFLOW-SFR with RT3D-OTIS, predicts impacts that vary significantly over a region encompassing about 500 km(2) and across time. Results suggest that, over the course of several decades, average Se and nitrate-nitrogen (NO3-N) groundwater concentrations within the region could be lowered by as much as 23% and 40%, respectively, using combined BMPs. Average Se concentration in the river could be decreased by up to 56% with combined BMPs, and NO3-N concentrations by up to 32% by using ERB. The CS-RF-ERB combination type may be the most promising for simultaneously lowering both Se and NO3-N concentrations. To insure compliance with Colorado water rights and the Arkansas River Compact with Kansas, measures must be taken to compensate for altered return flow patterns that will be a consequence of BMP implementation. Results also point to the need to consider the targeting of BMPs to specific locations within the region to maximize their effectiveness and efficiency.