Decision Support Model for Generating Optimal Alternative Scenarios of Watershed Best Management Practices

Elias G. Bekele, Laura Keefer, Sangeetha Chandrasekaran

Research output: Book/Report/Conference proceedingTechnical report


Surface and subsurface agricultural runoff has been the main cause of water quality problems in Lake Decatur, which is the major s ource of public water supply for the City of Decatur and the Village of Mt. Zion, serving a total population of more than 80,000. The lake has a watershed area of 925 square miles and was created by building a dam on the Sangamon River in 1922 with a modification in 1956 to incr ease its capacity. Extensive siltation is another critical issue, causing loss of significant storage volume. Nearly 90 percent of the Lake Decatur watershed is cropland, of which corn and soyb ean account for 44 and 39 percent, respectively. The watershed is extensively tile -drained to lower the water tabl e, creating favorable conditions for agricultural production. Hydr ologic and water quality monitoring has been conducted from 1993 to 2008 by the Illinois State Wate r Survey (ISWS) with support fr om the City of Decatur in an effort to alleviate the wa ter quality problem in Lake Decatur through watershed management alternatives. Additional waters hed monitoring was carried out from 2005 to 2008 by ISWS for a United States Environmental Protection Agency (USEPA) targeted watershed study with the goal of addressing economic and environmental as pects of nutrient management in the Upper Sangamon River watershed. The Illinois Environmental Protection Agency (IEPA) added Lake Decatur to the Illinois 2004 Section 303(d) list as impaired for nitr ogen-nitrate and total phosphorus (IEPA, 2004). Consequently, a Total Maximum Daily Load (TMDL) assessment was completed for the Sangamon River/Lake Decatur watershed in 20 07 and was approved by the USEPA. The TMDL study provided an overview of implementation alte rnatives that reduce nitrate and phosphorous loads, including nutrient manage ment, conservation tillage, conser vation buffers, and restriction of livestock. In addition, practices that limit losses from private sewage discharges and sedimentation were also proposed to reduce phosphorus loading (IEPA, 2007). Most cropland in the Lake Decatur watershed has been extensively tile-drained and therefore, the effectiveness of surface water-based best management practices (BMPs) for reducing nitrate may be limited. Specific placement areas for implementation of thes e alternatives have not been identified, which is the focus of this study. Two tributary watersheds of Lake Decatur we re identified for developing alternative implementation scenarios of selected BMPs that are designed to reduce nonpoint source pollutants (NPS) from agricultural sources. The watersheds are Big/Long Creek and Big Ditch watersheds, as illustrated in Figure 1. The Big/Long Creek watershed is located in the downstream portion of the Lake Decatur watershed, draining directly into the lake. In contrast, the Big Ditch watershed is located about 50 miles fr om the lake in the nort heastern edge of the Lake Decatur watershed. Both are agricultu rally dominated watersheds and their areas considered in this study correspond to the drainage areas of ISWS monito ring stations, which are close to the respective watershed outlets. The objective of this research was to evaluate the water quality benefits of selected BMPs at a watershed scale, generati ng alternative scenarios for implementation in Big Ditch and Big/Long Creek watersheds. This was accomp lished through the development of decision support models (DSMs) for each watershed. The DSMs were developed based on an integrated modeling approach, coupling a watershed si mulation model known as the Soil and Water Assessment Tool (SWAT) with an Archived-B ased Micro-Genetic Algorithm 2 (AMGA2) - a multi-objective optimization algorithm. Such integrated modeling approach, which involves interfacing a simulation model with an optimization algorithm, ha s been extensively applied to solve complex problems in watershed manageme nt (Bekele et al., 2013; Bekele et al., 2011), reservoir operations (Nicklow and Mays, 2000), groundwater monitoring design (Reed and Minsker, 2004), and others. The DSM was design ed to generate cost-effective implementation scenarios of selected conventional and newly em erging BMPs that include nutrient management, cover crops, perennial crops, constructed wetlan ds, drainage water management, bioreactors, saturated buffers, and filter strips. It is capable of providing optimal BMP placement scenarios that result in maximum re duction of NPS pollutants for a prescribed level of BMP implementation. BMP scenarios that strike a ba lance between NPS reducti on and total cost of implementation are identified as best tradeoff so lutions and are recommended for preparation of watershed implementation plans.
Original languageEnglish (US)
StatePublished - 2014

Publication series

NameISWS Contract Report


  • ISWS


Dive into the research topics of 'Decision Support Model for Generating Optimal Alternative Scenarios of Watershed Best Management Practices'. Together they form a unique fingerprint.

Cite this