Simulation of Streamflow and Water Quality to Determine Fecal Coliform and Nitrate Concentrations and Loads in the Mad River Basin, Ohio

The Hydrological Simulation Program–Fortran (HSPF) was used to simulate the concentrations and loads of fecal coliform and nitrate for streams in the Mad River Basin in west-central Ohio during the period 1999 through 2003. The Mad River Basin was divided into subbasins that were defined either by the 14-digit Hydrologic Unit (HU) boundaries or by streamflow-gaging-station locations used in the model. Model calibration and simulation processes required the formation of nine meteorologic zones to input meteorologic time-series data and water-quality data. Sources of fecal coliform and nitrate from wastewatertreatment discharges and combined sewer overflow discharges (CSOs) within the City of Springfield were point sources simulated in the model. Failing septic systems and cattle with direct access to streams were nonpoint sources included in the study but treated in the model as point sources. Other nonpoint sources were addressed by adjusting interflow and groundwater concentrations in the subsurface and maximum storage capacities and accumulation rates of the simulated constituents on the land surface for each meteorologic zone. Simulation results from the calibrated model show that several HUs exceeded the water-quality standard of 1,000 colony-forming units per 100 mL for fecal coliform based on the maximum 30-day geometric mean. Most HUs with high fecal coliform counts were within or downstream from the City of Springfield. No water-quality standard has been set for instream nitrate concentrations; however, the Ohio Environmental Protection Agency (Ohio EPA) considered a concentration of 5 mg/L or greater to be of concern. Simulation results indicate that several HUs in the agricultural areas of the basin exceeded this level. The calibrated model was modified to create scenarios that simulated loads of fecal coliform and nitrate that were either reduced or eliminated from selected sources. The revised models included the elimination of failing septic systems, elimination of direct access of cattle to streams, decrease in fecal coliform loads from the CSOs and selected wastewater-treatment facilities, and decrease in nitrate loads from land surfaces. The fecal coliform source-reduction model decreased the fecal coliform concentrations below a target concentration of 1,000 colonies per 100 milliliters for all HU outlets and decreased the load at the mouth of the Mad River by 73 percent. The nitrate source-reduction model decreased some HU mean concentrations to 5 milligrams per liter or less and decreased the load at the mouth of the Mad River by 52 percent. Other reduction scenarios may be run by Ohio EPA with the intent of identifying a management strategy that will attain a target concentration for the Mad River Basin. Introduction Stream waters of the Mad River Basin (fig. 1) are used for recreation, agricultural and industrial water supply, and support of aquatic life. Long-term availability of water for some of these uses is threatened, however, because several segments of the Mad River Basin are listed under Section 303(d) of the Federal Clean Water Act as not being in compliance with Ohio Water Quality Standards (WQS). An extensive evaluation of the Mad River Basin by Ohio Environmental Protection Agency (Ohio EPA) (2005a) found that, throughout the basin, ambient-water-quality standards for fecal coliform bacteria are exceeded (geometric mean of five or more samples within a 30-day period exceeds 1,000 col/100 mL and (or) more than 10 percent of samples within the 30-day period exceed 2,000 col/100 mL). Other causes of impairment to specific stream segments include nutrient and organic enrichment resulting from agricultural activities, urban runoff, or wastewater-treatment plants (Ohio Environmental Protection Agency, 1998, 2005a). Habitat alteration due to channelization also has degraded several stream segments in the watershed (Ohio Environmental Protection Agency, 1998, 2004, and 2005a). To bring all streams in the basin into compliance with Ohio WQS will require quantification of contamination loads contributed by various sources, information regarding the effects of different land covers and other land-surface characteristics on water quality, and documentation of the response Simulation of Fecal Coliform and Nitrate in the Mad River Basin, Ohio of contaminant loads to precipitation events and various flow conditions. Such information will serve as a basis for waterresource-management decisions in the basin. To quantify loads and concentrations of nitrate and fecal coliform in the Mad River Basin and to estimate these concentrations over a range of hydrologic conditions, the U.S. Geological Survey (USGS), in cooperation with the Ohio EPA, used a watershed model called Hydrological Simulation Program–Fortran (HSPF). HSPF simulates transport and storage of water and associated water-quality constituents, as well as instream chemical reactions (Bicknell and others, 1997). HSPF is included as part of the U.S. Environmental Protection Agency (USEPA) Better Assessment Science Integrating point and Nonpoint Sources (BASINS) program. HSPF is also one of several principal models currently recommended by the USEPA Office of Water for determining the Total Maximum Daily Load (TMDL) of a pollutant that a stream can receive from point, nonpoint, and background sources and still meet state water-quality standards with an adequate margin of safety. The process of developing a TMDL for a pollutant helps the Ohio EPA identify the amount by which both point and nonpoint sources in impaired stream segments must be reduced. Subsequently, scientifically based restoration solutions can be implemented with the ultimate goal of reaching full attainment of biological and chemical WQS within each stream segment and, thereafter, removal of the waterbody or waterbodies from the 303(d) “impaired” list. Purpose and Scope This report describes the development and calibration of an HSPF model to simulate streamflow and the transport of fecal coliform bacteria and nitrates. The model was developed under the USEPA 319 grant program in support of a TMDL that will be prepared by the Ohio EPA. The model simulation period is January 1999 through December 2003, based on availability of both streamflow and water-quality data. Current water-quality conditions in the Mad River Basin are described. Locations of point sources and their fecal coliform and nitrate loads are presented, as well as estimates of fecal coliform and nitrate loads from nonpoint sources, including groundwater discharge to streams. The calibrated model is used to calculate the loads of fecal coliform bacteria and nitrate in the Mad River Basin and to evaluate where these loads exceed the targets established by Ohio EPA. Load-reduction scenarios prescribed by Ohio EPA are simulated, and the resulting loads are presented. Loads are converted to mean annual nitrate and 30-day geometric mean fecal coliform concentrations to assess whether the source-reduction scenarios will achieve Ohio EPA target concentrations. Several topics are beyond the scope of this report, and will be addressed instead in the Ohio EPA TMDL report for the Mad River Basin. For example, stream impairment caused by habitat alteration is not addressed in this report. Designation of specific load-reduction scenarios as the TMDLs for fecal coliform and nitrate in the Mad River Basin will be addressed by Ohio EPA. Similarly, an implementation plan for achieving TMDL targets is excluded from this report but will be addressed by Ohio EPA in their TMDL report to USEPA. Previous Studies The information base for the Mad River Basin is relatively rich. Several studies on a variety of water-resources topics have focused on all or part of the Mad River Basin (table 1). Because it has long been documented that the Mad River has an unusually large base-flow component (Leverett, 1902), the interaction of ground water and surface water has been of special interest in the area. One reason for the abundance of information in the area is the presence of the Miami Conservancy District (MCD). The MCD was established in 1915 to provide flood protection for citizens in the Great Miami River Basin (which includes the Mad River Basin) (Miami Conservancy District, 2005c). With this goal in mind, the USGS and MCD currently operate streamflow-gaging stations (“streamflow gages” hereafter) at four locations on the Mad River. In addition, a crest-stage gage (which is used to determine peak streamflow) is operated on the Mad River near Urbana. Since its founding, MCD has assumed the additional responsibilities of preserving surfaceand ground-water resources and enhancing river corridors. In support of these expanded responsibilities, a surface-waterquality monitoring station was operated for 3 years at one of the four active streamflow gages (the Mad River near Eagle City, in Clark County), ground-water quality and quantity have been regularly monitored since 1997, and precipitation amounts have been recorded at precipitation gages across the basin (Miami Conservancy District, 2000, 2002, 2003, 2004, 2005b). Another source of information in the Mad River Basin is the National Water-Quality Assessment (NAWQA) Program. This USGS program was established to describe current waterquality conditions in major river basins and aquifer systems across the Nation, assess how water quality is changing over time, and investigate factors that affect water-quality conditions. Intensive water-quality analysis of the Mad River at St. Paris Pike was done from October 1998 through 2004 as part of the NAWQA program (Shindel and others, 2000, 2001, 2002, 2003, 2004; and 2005, p. 134–138). Water quality in the Mad River Basin has also been studied by Ohio EPA, Ohio Department of Natural Resources, and other agencies and researchers. Results of these studies are detailed in the “Water-Quality Characterization” section of this report.