Journal of Environmental Management 161 (2015) 252e260
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Research article
Spatial optimization of watershed management practices for nitrogen load reduction using a modeling-optimization framework Guoxiang Yang a, *, Elly P.H. Best b a
Oak Ridge Institute for Science and Education (ORISE) Postdoctoral Research Associate, U.S. Environmental Protection Agency, National Risk Management Research Laboratory, 26 West Martin Luther King Drive, Cincinnati, OH 45268, USA b U.S. Environmental Protection Agency, National Risk Management Research Laboratory, 26 West Martin Luther King Drive, Cincinnati, OH 45268, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 5 November 2014 Received in revised form 1 June 2015 Accepted 29 June 2015 Available online xxx
Best management practices (BMPs) can be used effectively to reduce nutrient loads transported from non-point sources to receiving water bodies. However, methodologies of BMP selection and placement in a cost-effective way are needed to assist watershed management planners and stakeholders. We developed a novel modeling-optimization framework that can be used to find cost-effective solutions of BMP placement to attain nutrient load reduction targets. This was accomplished by integrating a GISbased BMP siting method, a WQM-TMDL-N modeling approach to estimate total nitrogen (TN) loading, and a multi-objective optimization algorithm. Wetland restoration and buffer strip implementation were the two BMP categories used to explore the performance of this framework, both differing greatly in complexity of spatial analysis for site identification. Minimizing TN load and BMP cost were the two objective functions for the optimization process. The performance of this framework was demonstrated in the Tippecanoe River watershed, Indiana, USA. Optimized scenario-based load reduction indicated that the wetland subset selected by the minimum scenario had the greatest N removal efficiency. Buffer strips were more effective for load removal than wetlands. The optimized solutions provided a range of trade-offs between the two objective functions for both BMPs. This framework can be expanded conveniently to a regional scale because the NHDPlus catchment serves as its spatial computational unit. The present study demonstrated the potential of this framework to find costeffective solutions to meet a water quality target, such as a 20% TN load reduction, under different conditions. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Best management practices Nitrogen-loading Wetland restoration Buffer strips Multi-objective optimization Genetic algorithm Modeling-optimization
1. Introduction Non-point source (NPS) pollution from agriculture-related activities has become one of the largest causes of water quality degradation within the United States. Watershed-scale best management practices (BMPs), including conservation, restoration and altered management practices, may effectively reduce NPS pollutants in receiving water bodies and enhance local and regional ecosystem services. However, methodologies of selection and placement of BMPs in a cost-effective way are needed to assist watershed management planners and other stakeholders involved in solving water issues at various spatial and temporal scales to
* Corresponding author. Now at U.S. Geological Survey contracted through CNTS, Richmond, VA 23228, USA. E-mail addresses: [email protected] (G. Yang), [email protected] (E.P.H. Best). http://dx.doi.org/10.1016/j.jenvman.2015.06.052 0301-4797/© 2015 Elsevier Ltd. All rights reserved.
facilitate actual implementation. Wetland restoration is a typical BMP being employed in numerous watershed management plans in USA, because of the vital water-related physical, ecological, and economic functions and benefits it may provide, e.g., in the form of water quality improvement and stormwater mitigation. Dahl (1990) conducted a study showing wetlands losses in the United States from 1780's to 1980's. Indiana is one of the states with the greatest wetland loss with over 85% loss of its wetland area relative to the total wetland surface area present in the landscape 200 years ago. Much of these changes in the drainage system was designed to increase agricultural area. As a result most of the watersheds in Indiana have reduced capacity to remove excess pollution and reduce peak flows. Therefore, wetland restoration has gained a vast interest among stakeholders in Indiana as well as those studying the Mississippi River system. Riparian buffer implementation has been widely recommended
G. Yang, E.P.H. Best / Journal of Environmental Management 161 (2015) 252e260
as a BMP by multi-level agencies to remove excess nutrients entering water bodies (Lowrance et al., 1997; Passeport et al., 2013). Nitrogen removal through buffer strips tends to increase with buffer width. Mayer et al. (2007) showed that wider buffers (>50 m) consistently remove a larger portion of nitrogen than narrow ones (
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Research article
Spatial optimization of watershed management practices for nitrogen load reduction using a modeling-optimization framework Guoxiang Yang a, *, Elly P.H. Best b a
Oak Ridge Institute for Science and Education (ORISE) Postdoctoral Research Associate, U.S. Environmental Protection Agency, National Risk Management Research Laboratory, 26 West Martin Luther King Drive, Cincinnati, OH 45268, USA b U.S. Environmental Protection Agency, National Risk Management Research Laboratory, 26 West Martin Luther King Drive, Cincinnati, OH 45268, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 5 November 2014 Received in revised form 1 June 2015 Accepted 29 June 2015 Available online xxx
Best management practices (BMPs) can be used effectively to reduce nutrient loads transported from non-point sources to receiving water bodies. However, methodologies of BMP selection and placement in a cost-effective way are needed to assist watershed management planners and stakeholders. We developed a novel modeling-optimization framework that can be used to find cost-effective solutions of BMP placement to attain nutrient load reduction targets. This was accomplished by integrating a GISbased BMP siting method, a WQM-TMDL-N modeling approach to estimate total nitrogen (TN) loading, and a multi-objective optimization algorithm. Wetland restoration and buffer strip implementation were the two BMP categories used to explore the performance of this framework, both differing greatly in complexity of spatial analysis for site identification. Minimizing TN load and BMP cost were the two objective functions for the optimization process. The performance of this framework was demonstrated in the Tippecanoe River watershed, Indiana, USA. Optimized scenario-based load reduction indicated that the wetland subset selected by the minimum scenario had the greatest N removal efficiency. Buffer strips were more effective for load removal than wetlands. The optimized solutions provided a range of trade-offs between the two objective functions for both BMPs. This framework can be expanded conveniently to a regional scale because the NHDPlus catchment serves as its spatial computational unit. The present study demonstrated the potential of this framework to find costeffective solutions to meet a water quality target, such as a 20% TN load reduction, under different conditions. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Best management practices Nitrogen-loading Wetland restoration Buffer strips Multi-objective optimization Genetic algorithm Modeling-optimization
1. Introduction Non-point source (NPS) pollution from agriculture-related activities has become one of the largest causes of water quality degradation within the United States. Watershed-scale best management practices (BMPs), including conservation, restoration and altered management practices, may effectively reduce NPS pollutants in receiving water bodies and enhance local and regional ecosystem services. However, methodologies of selection and placement of BMPs in a cost-effective way are needed to assist watershed management planners and other stakeholders involved in solving water issues at various spatial and temporal scales to
* Corresponding author. Now at U.S. Geological Survey contracted through CNTS, Richmond, VA 23228, USA. E-mail addresses: [email protected] (G. Yang), [email protected] (E.P.H. Best). http://dx.doi.org/10.1016/j.jenvman.2015.06.052 0301-4797/© 2015 Elsevier Ltd. All rights reserved.
facilitate actual implementation. Wetland restoration is a typical BMP being employed in numerous watershed management plans in USA, because of the vital water-related physical, ecological, and economic functions and benefits it may provide, e.g., in the form of water quality improvement and stormwater mitigation. Dahl (1990) conducted a study showing wetlands losses in the United States from 1780's to 1980's. Indiana is one of the states with the greatest wetland loss with over 85% loss of its wetland area relative to the total wetland surface area present in the landscape 200 years ago. Much of these changes in the drainage system was designed to increase agricultural area. As a result most of the watersheds in Indiana have reduced capacity to remove excess pollution and reduce peak flows. Therefore, wetland restoration has gained a vast interest among stakeholders in Indiana as well as those studying the Mississippi River system. Riparian buffer implementation has been widely recommended
G. Yang, E.P.H. Best / Journal of Environmental Management 161 (2015) 252e260
as a BMP by multi-level agencies to remove excess nutrients entering water bodies (Lowrance et al., 1997; Passeport et al., 2013). Nitrogen removal through buffer strips tends to increase with buffer width. Mayer et al. (2007) showed that wider buffers (>50 m) consistently remove a larger portion of nitrogen than narrow ones (
