Flow records for this watershed date back to 1903.  Many United States Geological Society (USGS) gage stations started collecting flow information before 1950 with a few newer stations initiated in the 1980s and 1990s.  There are currently 28 active discharge gages with approximately 53 additional sites that have been active for some period of time in the past.  The data from these stations will be compiled and evaluated using various statistical methods to determine seasonality and to develop the hydrologic budget for the watershed.  Seasonality will be determined by evaluating the monthly flow data and by examining statistical differences using nonparametric techniques, such as Kruskall-Wallis.  Regression analysis techniques will be used to look for relationships between USGS flow measurement stations.  Data will be normalized to parameters, such as drainage area, to aid in determining the type and source of flow at each station.  The hydrologic budget will present the major contributors and, if appropriate, compare their contribution based on seasonality of the flow regime.  The focus of this task will be on the contributing drainage area below Angostura Reservoir, with the exception of the Belle Fourche River and Rapid Creek Watersheds.  The final report will summarize the contributions of the area above Angostura Reservoir and the Belle Fourche River and Rapid Creek Watersheds based on the previous work completed in these areas.

Flow records at the long-term stations and temporary stations will be analyzed using regression techniques to confirm the flow relationships between stations.  This relationship will be critical to understand the water quality data when flows were not measured (majority of the historic South Dakota Department of Environment and Natural Resources (SD DENR) data does not have associated flows).  Using this information, flow will be estimated.

Water samples were collected and analyzed by many organizations in the past.  USGS and SD DENR have collected the most samples.  USGS and SD DENR have been collecting water quality samples from streams and lakes during different time periods for over 30 years at several locations throughout the watershed.  This incredible data source provides a great foundation to compare other “grab sample type” water quality results within the watershed and to look for relationships that further our understanding of the nonpoint source pollution and potential remediation measures. The focus of this task will be on the contributing drainage area below Angostura Reservoir, with the exception of drainage areas where the total maximum daily load (TMDL) process has already occurred and remediation is already occurring.  This includes Belle Fourche River Watershed and most of Rapid Creek Watershed.  The final report will summarize the contributions of the area above Angostura Reservoir and Belle Fourche River and Rapid Creek Watersheds based on the previous work completed.

Water quality data will be compiled from the Internet and by contacting the appropriate people at the responsible agencies.  There is a great historical record that can be used to develop statistical comparisons of precipitation and flow with water quality monitoring stations. 

Statistical comparisons will be developed for each water quality parameter listed for impairment in the Integrated Report, versus seasonality and flow.  Flow, concentration, and load diagrams will be developed for the period of record at stations with large data sets.  Annual trends will be investigated as well as trends over time.  Statistical analysis will occur for each parameter beginning at its furthest downstream point of listing and covering the upstream contributing drainage area.  For example, total suspended solids (TSS) and fecal coliform will be analyzed from the mouth of the Cheyenne River and cover the entire watershed, while conductivity will be analyzed from the mouth of Cherry Creek and the upstream drainage area for Cherry Creek, since conductivity is only listed for this area.  Nutrient analysis will be performed upstream of lakes listed in the Integrated Report as impaired, where data are available, and where stream assessments show evidence of impairment due to nutrient loading. 

To develop the relationship between flow and concentration, FLUX, a computer program developed by United States Army Corps of Engineers (USACE), will be used.  FLUX is designed for use in estimating the loadings of suspended solids or other water quality components passing a tributary sampling station over a given period of time.  A relationship of flow and concentration is developed using the available samples.  This relationship is used to develop an extrapolated concentration using the long-term daily flow records for the USGS stations.  Finally, a loading estimate is developed by multiplying the daily flow and concentration for the USGS stations resulting in a tons/month estimates.  Seasonality and precipi­tation relationships will be evaluated using a regression analysis.  The data analysis at these sites will be a cornerstone for the comparison analysis using the data from other tributary sites.  It is expected that there will be seasonality in flow data as well as a good relationship between flow, TSS, fecal coliform, and other water quality parameters.  Thus it is expected that there will be a seasonality relationship using the water quality data.  The relationship between water quality and precipitation will be evaluated to determine the contribution of surface runoff.

In addition, statistical regression analysis will be performed between various water quality parameters.  For example, recent literature has documented cases where a strong relationship between TSS and fecal coliform exists.  Stepwise regression analysis will be preformed using the water quality data to investigate a relationship between TSS, fecal coliform, and other water quality parameters. 

Based on the analysis and literature from similar studies, qualitative conceptual water quality budget diagrams will be prepared.  The diagrams will present potential sources of impairment and relative contributions for each water quality parameter of concern in the system.

Flow, concentration, and load diagrams will be developed for the period of record for the Water Quality Monitoring (WQM) and USGS gage stations where sufficient data exists.  A combination of FLUX and regression analysis will be used to investigate these relationships.