This section summarizes data collected during Phase II. Each subsection highlights and describes only those constituents that either (1) require description because they were identified as data needs during Phase I of the MMSD Corridor Study or (2) demonstrate a statistical relationship, data distribution, or other consistent response that enhances the understanding of water quality as part of Phase II. Stream and harbor data are both discussed within each section: harbor sites are compared to other harbor sites, stream sites are compared to other stream sites, and general comparisons are made between stream and harbor sites.

Manually collected sampling events were assigned a category with respect to flow and season. Although personnel targeted high-flow events by selecting events in which the instantaneous discharge was greater than the 10-percent-flow-duration discharge at gages representing the major basins draining to the harbor, temporal variations in precipitation as well as the relative flashiness of smaller urban streams occasionally meant that a fraction of the streams were not at high flow at the time of targeted high-flow-event sampling. In addition, quarterly sampling was scheduled without regard for flow conditions, and therefore occasionally entailed the collection of samples during high-flow events. Therefore, flow condition at the time of sampling was assigned by examining 10-day hydrographs bracketing the sampling date (5 days before and after)(appendix 4). For stream sites without stream-discharge data, flow-condition estimates were based on comparisons to records from nearby streams. Seasonality was roughly based on the original sample purpose: quarterly samples were assigned to their season (spring, summer, autumn, and winter), snowmelt events were assigned to spring season, and summer storm events were assigned to summer. Specific seasonality assignments are listed in appendix 4. Overall, seasonal responses were explored for both stream and harbor sites. Only stream sites were considered with respect to flow.

Automatically collected samples were used for load determinations of chloride, total phosphorus, and suspended sediment at 4 of the 12 gaging stations that were continuously monitored for stream discharge. Samples were collected from June through September 2004 and for water year 2005. Daily loads of constituents at the four automatic-sampling sites were summed to monthly and annual totals. Loads are described with respect to total load, yield, yield per day, and volumetrically weighted (VW) concentration. Yield data (load data normalized to drainage area) were computed to allow for comparisons among the sites, whose drainage basins differ greatly in area: Milwaukee River near Cedarburg, 607 mi²; Milwaukee River at Milwaukee, 690 mi²; Menomonee River at Wauwatosa, 123 mi²; and Kinnickinnic River, 18.8 mi². VW concentration was determined by dividing the total load by the total flow then multiplying by a conversion factor; this computation allowed for comparison of loads among various flow conditions.

General comparisons between Phase I and Phase II median concentrations were made where data were available for both phases and were sufficient to discern an overall pattern. Phase I median concentrations were calculated from available data; depending on the constituent, the amount of data ranged from a small dataset with only a few results to a large dataset with numerous results. Comparisons could not be made for a number of Phase II constituents (for example, the WWCs) as technology to analyze for many of these compounds has only recently been developed. The heterogeneity of the Phase I historical data set prohibited strict statistical comparisons (Schneider and others, 2004). The Phase I report (p. 11–23) describes how the Phase I data set was compiled and how the data can be used. Phase II sampling sites within subwatersheds were not necessarily located at the most downstream point of the subwatershed (fig. 1), and therefore, all data collected from the sites did not integrate water and sediment from the entire subwatershed. Additionally, the three Phase II sites in the Lower Milwaukee River subwatershed (Milwaukee River near Cedarburg, Milwaukee River at Milwaukee, and Milwaukee River at Mouth at Milwaukee) were not considered adequately representative of their subwatershed due to their location and the size and complexity of the Lower Milwaukee River subwatershed (table 8). Although all data are summarized in appendix 5, comparison of these sites to the subwatershed was omitted in the discussion of general comparisons between Phase I and Phase II data; however, contamination of this subwatershed with synthetic organic contaminants (especially PCBs) is well documented and is discussed in this report in general terms using the data from the Milwaukee River at Milwaukee site. Correlation between Phase I subwatersheds and Phase II sites is described in table 8.

Table 8. Comparison of the Phase I subwatersheds to Phase II sites of the Milwaukee Metropolitan Sewerage District (MMSD) Corridor Study.

   ¹ Three of the Phase II sites sampled were in the Lower Milwaukee River Subwatershed. Comparisons are drawn individually between each site and the Phase I subwatershed.

Where possible, percent differences in median concentrations were used to help quantify changes between Phase I and Phase II. Otherwise, comparisons were in terms of general increases or decreases in median concentrations. For this discussion, percent differences of 20 percent or greater were considered noteworthy. Quartile ranges used in maps were based on Phase I data (Schneider and others, 2004).

In addition, water- and bed-sediment-quality results were compared to USEPA criteria and (or) observed results (Robertson and others, 2006; Corsi and others, 1997; and MacDonald and others, 2000) where possible.

return to top

Hydrologic Conditions During the Study Period

Data for precipitation, stream discharge, and runoff at selected sites in the MMSD planning area are described in the following sections.

Precipitation

Precipitation data from the U.S. Department of Commerce, National Oceanic and Atmospheric Administration, and National Climatic Data Center (2003, 2004, 2005) were available at three sites within the MMSD planning area: Germantown Department of Public Works, Mount Mary College, and General Mitchell International Airport. Data collected at these sites during water years 2004 and 2005 were compared to monthly and yearly normals for 1971–2000 (table 9).

Table 9. Monthly and water-year totals for rain gages in the Milwaukee Metropolitan Sewerage District planning area, Wis. (U.S. Department of Commerce, National Oceanic and Atmospheric Administration, and National Climatic Data Center, 2003, 2004, 2005).

[All data in inches; M, total precipitation computed with 1–9 daily values missing]

Total precipitation in water year 2004 was above normal at all sites: 7.64 in. above normal at Germantown, 1.32 in. above normal at Mount Mary College, and 0.23 in. above normal at General Mitchell Airport. At all three sites, the monthly precipitation totals in May 2004 were the highest recorded during the study period, and had the greatest departures from normal (ranging from 5.12 to 9.80 in.).

Total precipitation in water year 2005 was below normal at all sites: -9.29 in. at General Mitchell Airport, -6.64 in. at Mount Mary College, and -1.93 in. at Germantown. Monthly precipitation totals in March and April were among the lowest observed at all sites during water year 2005. The greatest departures from normal were in August (-2.74 in. at General Mitchell Airport and -2.42 in. at Mount Mary College) and April (-2.21 in. at Germantown).

Discharge and Runoff

Discharge is defined as the volume of water passing by a location over a given period of time. Runoff is defined as the depth to which the entire drainage area would be covered in water if the discharge for a given time period were evenly distributed across the area. Continuous stage measurements (that is, water-surface elevation above the datum of the stream gage) were collected and used to determine discharge and runoff during the study period at 12 of the 15 stream sites. At 8 of these sites, discharge records spanned more than 20 years, enabling comparisons of water year 2004–05 data with the long-term means for both discharge (appendixes 6A and 6C) and monthly runoff (appendixes 6B and 6C)(Waschbusch and others, 2005, 2006). At the remaining four sites, discharge records spanned shorter time periods, and are therefore provided without period of record comparisons (appendix 6A).

Annual mean discharges in water year 2004 were greater than long-term annual mean discharges for all eight long-term stream-gaging stations (appendix 6C). Percent differences from long-term annual means for water year 2004 ranged from 1 percent at Underwood Creek to 84 percent at the Milwaukee River at Milwaukee (appendix 6C). Sites with drainage basins in the southern portion of the MMSD planning area had the lowest percent differences: Underwood Creek (1 percent), Kinnickinnic River (6 percent), Oak Creek (7 percent), and Root River near Franklin (8 percent)(appendix 6C). Sites with drainage basins in the northern portion of the MMSD planning area had the highest percent differences: Milwaukee River at Milwaukee (84 percent), Milwaukee River near Cedarburg (56 percent), Menomonee River at Menomonee Falls (47 percent), and Menomonee River at Wauwatosa (44 percent)(appendix 6C). Runoff during May, June, and July, 2004, was greater than respective long-term at all eight long-term stream-gaging stations, with percent differences ranging from 16 in July at Kinnickinnic River to 538 percent in June at Milwaukee River at Milwaukee (appendixes 6B and 6C).

Annual mean discharges in water year 2005 were less than long-term annual mean discharges for all eight long-term stream-gaging stations (appendix 6c). Percent differences from long-term annual means for water year 2005 ranged from -50 percent at Menomonee River at Menomonee Falls to -16 percent at Milwaukee River at Milwaukee (appendix 6C). No spatial patterns in discharge data were observed in water year 2005.

Runoff from March through August, 2005, was less than respective long-term means at all eight long-term stream-gaging stations, with percent differences ranging from -86 in June at Menomonee River at Menomonee Falls to -11 percent in March at Menomonee River at Menomonee Falls (appendixes 6B and 6C).