|
Memo
U.S.
Geological Survey
1820 Midpark Drive
Knoxville, TN 37920
To: Robert
Emmott, National Park Service, Asheville, NC
From: Greg C.
Johnson, Hydrologist, Knoxville, Tennessee
Date: 9/9/03
Subject: Description of the USGS Retrospective
Analysis of Water Quality Data for the Obed Wild and Scenic River (OBRI)
Below
is a summary of available water quality data for the OBRI. These
data were retrieved from multiple databases and were collected by a number
of state and federal agencies. In some cases, multiple stations were located
at the same site, and the data from these stations were combined. Sixty-two
sites were identified with adequate data for analysis. Water quality data
for selected parameters are organized by watershed and summarized in graphs
and maps. Some data are censored at different levels due to analytical
constraints. In some cases, both total and dissolved concentrations are
presented. Trend analyses were preformed at sites with sufficient data.
Databases
All available
historic water quality data were collected from the following sources:
Environmental
Protection Agency (EPA) STORET database that includes
data from multiple agencies, including EPA and Tennessee Valley Authority
(TVA) U.S.
Geological Survey (USGS) National Water Information System (NWIS)
database National
Park Service (NPS) local Access database Microsoft
Excel spreadsheet from Tennessee Department of Environment and Conservation
(TDEC) OSM
Access Database
The period
of record ranged from February 1965 to January 2002. The largest database
(dB) is the STORET dB, which is divided between the
STORET Legacy data center dB of data before 1999, and
the Modernized database. No current data were found in the Modernized
STORET dB. Sometimes the same agency appears twice
on the same plot due to different agency codes. The agency name on the
plots is generally the reported name in STORET.
Site Selection
All sites
sampled by each agency in each dB were combined (about 128 sites) (OBRI-sites-all.xls). Sites were checked
to see that the latitude and longitude and stream name matched, and the
reported latitude and longitude were adjusted for a few sites. The
NPS station ID is listed for sites that were included
in the baseline water quality data inventory and analysis. Sites were
then combined where multiple agencies sampled the same site and numbered
from 1 to 62 (Obed-site-names.xls).
These sites
were then checked to minimize duplication between dBs.
Sometimes TVA and USGS data were reported separately,
but the data were identical. In addition, the NPS,
OSM and TDEC data had to be compared
between the STORET dB and the local data files. The
most current and complete data were kept, and the sites with less complete
data were deleted. Sites that had less than five samples or only field
parameters were also deleted. All sites were grouped by watershed and
were associated with the Emory River, Obed River, Daddy’s Creek or Clear
Creek.
Water Quality
Data
Data are
presented graphically for selected constituents in the River Flow and
Quality section. For each of the water quality parameters there is
a
graph of the temporal distribution of data by collecting agency; a
graph showing the distribution of data by site in each basin by collecting
agency; and a
box plot showing the distribution of data for each site in each basin.
Water quality
parameters are grouped into the following sections: field parameters,
bacteria, nutrients and sediment, major ions and trace elements. Water
Quality data for the 62 sites were combined into Microsoft Excel files
for selected parameters. For each parameter, the data are plotted on a
map to show ranges of average concentrations. Time series data are plotted
by agency to show when data was collected and by whom. Scatter plots and
box plots were generated for each parameter. The data are grouped by watershed
in the box and scatter plots.
Censored
Values
Censored
values are those values with associated comments such as greater or less
than a specified detection limit. For graphing purposes, these were plotted
at the detection limit. For parameters that were heavily censored at various
detection limits, a plot was generated showing the values that are censored.
For trend analysis, the amount of censored data determined what regression
technique was utilized; if the dataset was heavily
censored, no trends could be calculated.
Dissolved
vs. Total Constituents
Where possible,
the dissolved constituents were utilized in this exercise. The total constituent
includes the suspended fraction in the water column and is less representative
of what will affect aquatic life than the dissolved fraction.
Most of the OSM data is total constituent, except where
noted as dissolved. The NPS data in STORET
is generally reported as dissolved. The USGS data generally has both dissolved
and total constituents.
Trend
Analysis
Trend analyses
for sites with adequate data were computed using S-ESTREND
USGS software on the S-plus statistics program. The minimum required data
is 50 observations and a minimum of five years of data. Table 3 in the
trends section shows the period of record utilized in the trend analysis,
the p value to indicate the level of significance, and a note if the trend
was up, down or none. No sites were determined to have sufficient
flow data to flow-weight the trend analysis.
OBRI_Report
Information
|