United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
EPA-600/2-79-156
November 1979
Research and Development
Dissolved Oxygen
Impact from Urban
Storm Runoff
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service. Springfield, Virginia 22161.
-------
EPA-600/2-79-156
November 1979
DISSOLVED OXYGEN IMPACT
FROM URBAN STORM RUNOFF
by
Thomas N. Keefer, Robert K. Simons, and
Raul S. McQuivey
The Sutron Corporation
Arlington, Virginia 22209
Contract No. 68-03-2630
Project Officer
John English
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmen-
tal Research Laboratory, U.S. Environmental Protection Agency,
and approved for publication. Approval does not signify that
the contents necessarily reflect the views and policies of the
U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or
recommendation for use.
-------
FOREWORD
The Environmental Protection Agency was created because of
increasing public and government concern about the dangers of
pollution to the health and welfare of the American people.
Noxious air, foul water, and spoiled land are tragic testimony
to the deterioration of our natural environment. The complexity
of that environment and the interplay between its components re-
quire a concentrated and integrated attack on the problem.
Research and development is that necessary first step in
problem solution and it involves defining the problem, measuring
its impact, and searching for solutions. The Municipal Environ-
mental Research Laboratory develops new and improved technology
and systems for the prevention, treatment, and management of
wastewater and solid and hazardous waste pollutant discharges
from municipal and community sources, for the preservation and
treatment of public drinking water supplies, and to minimize the
adverse economic, social, health, and aesthetic effects of pol-
lution. This publication is one of the products of that re-
search, a most vital communications link between the researcher
and the user community.
This report investigates the correlation between storm run-
off and dissolved oxygen deficits downstream of urban areas.
Several hundred station years of water quality monitor flow and
rainfall data obtained by the U.S. Geological Survey, the U.S.
Environmental Protection Agency, the National Weather Service
and other government agencies are used in the study.
Francis T. Mayo
Director
Municipal Environmental
Research Laboratory
111
-------
ABSTRACT
The primary objective of the research reported here was to
determine if on a national basis a correlation exists between
strength of dissolved oxygen (DO) deficits and the presence of
rainfall and/or storm runoff downstream of urban areas. A
secondary objective was to estimate the magnitude and extent of-
the problem.
One hundred and four water quality monitoring sites in and
downstream of urban areas throughout the country were considered
for inclusion in the study. These were screened from over 1000
monitors maintained by federal and state agencies such as the U.S.
Geological Survey, Environmental Protection Agency (EPA), Ohio
River Valley Sanitation Commission and Wisconsin Department of
Natural,Resources. Daily data were obtained and processed for
83 of the 104 candidate sites. Of the ,83 monitors considered,
42 percent or roughly one monitor in two had data which demon-
strated a 60 percent or greater probability of a higher than
average DO deficit occurring at times of higher-than-average
stream flow or on days with rainfall. This result was obtained
by considering daily data for entire water years. Not all years
at any given station exhibited a 60 percent probability. One to
three years out of five is typical. DO levels fell to less than
75 percent saturation at most of the sites where 60 percent or
greater probability existed. Levels of 5 mg/1 or less were not
uncommon.
Detailed hourly data analysis was made .at 22 of the sites
with high correlation between flow and DO deficit. Typically,
at times of steady low flow the DO fluctuates widely on a daily
cycle. These cyclic changes range from 1 to 7 mg/1. When a
storm event occurs and the flow increases, the diurnal cycle
disappears. The minimum DO drops from 1 to 1.5 mg/1 below the
minimum values observed during steady flows and remains constant
there for periods ranging from one to five days. As the flow
event subsides, the DO level resumes its cyclic behavior. Of
the 22 monitors examined on an hourly basis, 11 would not meet
a 5.0-mg/l DO standard. Six of the 11 would not meet the EPA-
suggested 2.0-mg/l-for-four-hour standard. Streeter-Phelps
analysis indicated that two additional monitor sites at which
hourly data were examined would not have met the EPA standard
had they been properly located. An additional two sites at which
hourly data could not be obtained would also not have met the
EPA standard.
iv
-------
In general, the data examined here indicate that 19 percent
of the 104 cnadidate monitors might not meet a 5.0-mg/l standard
and 15 percent might not meet a 2.0-mg/l standard. Frequency of
violations was not tabulated exactly but appears to be zero to
five times per year at sites with correlations.
The data base from which this study was drawn is not geo-
graphically homogeneous. By far, the largest number of monitors
is in the midwest and northeast. Less than one-half of the
states are represented. No conclusions can thus be drawn on the
national scope of either correlations or standards violations.
Further study of monitor sites with both strong correlations and
standards violations is recommended to specifically identify
causes. The frequency of standards violation appears to be low,
but the causes should be identified to prevent the frequency
from increasing.
This report was submitted in partial fulfillment of
Contract No. 68-03-2630 by the Sutron Corporation under the
sponsorship of the U.S. Environmental Protection Agency. This
report covers a period from November 18, 1977 to May 1, 1979
and work was completed as of May 1, 1979.
v
-------
CONTENTS
Foreword . iii
Abstract iv
Figures viii
Exhibits viii
Tables ix
English to Metric Conversion Units xi
Acknowledgments xii
1. Introduction 1
2. Summary of Findings 6
3. Conclusions 9
4. Recommendations 11
5. Site Location and Screening 13
6. Daily Correlation Analysis 26
7. Detailed Site Analysis 57
8. Assessment of Extent and Causes of Problem .... 76
References 94
Appendices
A. Water quality data collection agencies 96
B. Monitor sites considered for analysis 119
C. Results of daily correlation analysis 128
D. Results of detailed site analyses 152
Vll
-------
FIGURES
No. Page
1. Card format used by National Climatic Center for
weather data 33
2. Typical daily flow and DO records illustrating
high degree of correlation between high flow
and low DO 41
3. Box or compartment method used to analyze daily
data for correlation between high flow and low DO. . 43
4. Streeter-Phelps analysis results for Scioto River
at Chillicothe, OH 69
5. One month of hourly data for the Scioto River
at Chillicothe, OH 74
6. Hourly data for one-month period, Scioto River
at Chillicothe, OH 75
7. DO/flow correlation on the Little Miami River
near Spring Valley, OH 79
8. Probability of low DO at high flow versus per-
centage of urban area 87
EXHIBITS
No. Page
1. Tape File Structure Description, ORSANCO Data ... 39
2. Site Analysis Form - Streeter-Phelps ........ 59
3. Site Analysis Form ..........••••••. 60
Vlll
-------
TABLES
No. Paqe
1. OWDC Regions and Number of Agencies Collecting
Water Quality Data 15
2. List of All Sites Considered for Daily or Hourly
Analysis 23
3. Study Code Classifications 24
4. USGS, 1,656-Byte, Magnetic-Tape Output Format .... 28
5. Record Arrangement on Disk File for STORET
MORE = 4* 32
6. Card Image Content for National Climatic Center
Weather Data 34
7. Data Block Structure, ORSANCO Data 38
8. Eligible Sites not Analyzed and Reasons 47
9. Monitor Sites Exhibiting a 60 Percent or Greater
Probability of Low DO During High Flow or Periods
of Rainfall 49
10. Monitor Sites with 60 Percent or Greater Proba-
bility of Low DO at Times of High Flow . 50
11. USGS Monitor Sites with 60 Percent or Greater
Probability of Low DO on Days with Rainfall 53
12. Streeter-Phelps Analysis Sites 68
13. Sites at Which Hourly Data Were Processed . 70
14. Monitor Sites at Which DO Levels Below 5.0 mg/1
Were Observed ........... 81
15. Monitor Sites at Which EPA DO Standards Were Not
Met During Runoff Events ........ 81
IX
-------
TABLES (CONTINUED)
No. Page
16. Monitor Sites with Potential for EPA Standard
Violations If Properly Located 82
17. Problem Sites Per State Versus Number of Monitors
Per State 85
18. Sites Recommended for Further Study 90
-------
ENGLISH TO METRIC CONVERSION UNITS
cfs x 0.02832 = m 3/g
ft x 0.3048 = m
in. x 2.54 = cm
mile x 1.609 = km
sq. miles x 2.590 =
XI
-------
ACKN OWLEDGMENT S
Sutron gratefully acknowledges the cooperation of the U.S.
Geological Survey in obtaining data for this study. Mr. Robert
Wall of the Automatic Data Processing Unit at the national head-
quarters in Reston, Virginia, was particularly helpful in obtain-
ing and using the daily data. Mr. I. D. Yost of the Texas
District Office in Austin, Mr. R. 0. Hawkinson of the Ohio
District Office in Columbus, and Mr. T. G. Ross of the Pennsyl-
vania District Office in Philadelphia were very cooperative in
obtaining hourly data.
XII
-------
SECTION 1
INTRODUCTION
BACKGROUND
Dissolved oxygen (DO) levels have long been taken as indi-
cators of the health of a water body. The aquatic plant and
animal life depend on the presence of minimum DO levels. A
stream's ability to assimilate waste is largely a-function of
available DO. .
- In the initial efforts to clean up the nation's waters and
meet DO standards, attention was focused on point source waste
discharges. It was generally believed that upgrading sewage
treatment plants and industrial outfalls to a high treatment
level would restore water quality. There is little question
that these efforts were beneficial. However, as more and .more
point sources have come under control, it has become evident
that nonpoint waste sources are also significant.
The exact effect of urban nonpoint runoff on DO levels is
difficult to quantify for several reasons. The phenomenon is
highly transient in nature. How and what enters a receiving
water is a function of the type of urban activity, land use,
time between storms, distribution of rainfall, collection
methods, and numerous other factors. Urban runoff may enter a
receiving water directly through storm sewers or through com-
bined storm/sanitary sewers.
The combined sewers represent a particularly complex prob-
lem. These route both storm runoff and sanitary waste to a
central treatment facility. At times of peak flow, the sewer
is designed to overflow directly into a receiving water. These
overflows may occur at the treatment plant or at a variety of
places in the urban drainage system.
Studies in Durham, North Carolina (1), indicate that direct
urban runoff is no more desirable than sanitary sewage. Urban
surface waters receive substantial amounts of organics, solids,
nutrients, heavy metals, and microorganisms. It is not surpris-
ing then that stream quality may be adversely affected downstream
of urban' areas after a storm event even in the presence of ad-
vanced wastewater treatment.
-------
The research presented here is,designed to determine, as
far as possible, the behavior of DO levels downstream of urban
areas after storms.
SCOPE AND OBJECTIVES
The original scope of work divided the study into six in-
terrelated areas. The first and major area was to examine
available, continuously recorded water quality information in
data banks such as U.S. Geological Survey (USGS) and the U.S.
Environmental Protection Agency's (EPA) STORET system. Data
from other state and local agencies such as the Ohio River
Valley Sanitation Commission (ORSANCO), Wisconsin Department of
Natural Resources (WDNR), Greater Metropolitan Chicago Sanita-
tion District and others were to be considered when possible.
The data bases were to be reviewed to obtain a five-year his-
torical picture of DO, temperature, and flow in the receiving
waters of the United States at the location of all reliable
continuous water quality monitoring stations. Concurrent with
this review, five years of hourly rainfall data for the areas
in which the monitoring stations are located were to be obtained
from the National Climatic Center in Asheville, North Carolina,
or other sources. The DO and temperature records were to be
examined to find all locations downstream of urban areas in
which the DO during the warm weather portion of the year fell
to less than 75 percent of saturation. At these locations, an
analysis was to be made to determine whether a correlation
exists between wet weather discharge and DO sag and how sig-
nificant the wet weather impact is.
The second area of study recognized the fact that existing
monitors are not necessarily located to study urban runoff. The
location of the monitoring stations and rain gages relative to
the urbanized areas and the relative sizes of urban and nonurban
drainage areas contributing to the receiving water were to be
considered in making judgments about urban rainfall/runoff and
DO deficits.
The third study area was to determine:the accuracy and re-
liability of the monitors used in the research. Inoperative
periods and frequency of maintenance were variables to be con-
sidered.
The fourth component of the Scope of Work was a reemphasis
of the need for studying hourly variations in the variables of
interest. Study, of daily averages to determine whether a moni-
tor was to-be included in the study,was permissible ,but hourly
values were to be used in the final analysis.
The fifth study area was to.perform an in-depth analysis of
sites with a strong high flow, low DO correlation. The tradi-
tional Streeter-Phelps (Thomann (2) p. 110) technique was to be
-------
used to determine if severe oxygen sags could be caused by the
particular urban area and how many stream miles (Kilometers —
see conversion factors page 238)are affected.
The-final study area was to analyze the study results on a
national basis and attempt to estimate the overall impact of
urban storm runoff. The analysis was to be stated in terms of
stream miles affected, if possible.
For the most part, the Scope of Work was carried out as
originally intended. Modifications were made primarily to rec-
ognize the very large number oi: monitors available with less
than the five years of continuous data1and those without hourly
rainfall data. The need for this modification became apparent
early in the study. For instance, the USGS maintains over 100
water quality monitors. Only 12 had five years'of record, were
in close-proximity to a stream gage, and were within reasonable
distance of a primary weather station where hourly rainfall data
were available. By modifying Area 1 of,the Scope of Work to in-
clude monitors with five years or less of data,'the number to be
considered increased significantly. This required a greater
emphasis on daily analysis to keep the total amount of data
values manageable. As stated in the fourth area of the Scope,
of Work, each station was examined on a daily average basis for
inclusion in the study. Periods of hourly data ranging in
length from 240 to 720 hours at sites with strong correlation
between flow or rainfall and low DO were then examined.
The major objective of the study was to determine on a
national basis the impact of urban runoff on DO levels in re-
ceiving waters. Data from points-specific studies such as the
Triangle J Joint Council of Governments "208" waste management
plan (3) have indicated a strong correlation between periods of
low DO and high stream flow. Similar, results have been obtained
in other locations (Texas (4) ,> Missouri (5)) . It was hoped that
by examining existing water quality monitor records that a broad
national picture of the problem-could be obtained if, indeed,
one existed.
DEFINITION OF PROBLEM DO LEVELS
Tne exact definition of a low DO level is subject to some
debate. The definition is largely a function of the intended
use. Thomann (2) summarizes several representative standards.
The Ohio River Sanitation Commission requires that the dissolved
oxygen should not be less than 5 milligrams per liter (mg/1)
during at- least 16'hours of any 24 nor less than 3 mg/1 at any
time in the upper basin areas. In the industrialized lower
basin' and estuary region, a level of 3.5 mg/1 must be maintained
on a daily average basis. The National Academy of Sciences Com-
mittee on' Water Quality Criteria (6) recommends a minimum level
3
-------
of 4 mg/1 at all times and-other higher criteria based on the
"natural seasonal oxygen minimum."
The U.S. Environmental Protection Agency has supported con-
siderable research on DO levels "and their effect on aquatic life.
The DO criteria proposed by EPA for freshwater aquatic life are
given as follows:
"Freshwater aquatic life: A minimum concentration
of dissolved oxygen to maintain good fish popula-
tions is 5.0 mg/li The criterion for salmonid
spawning beds is a minimum of 5.0 mg/1 in the in-
terstitial water of the gravel." (7)
Research has established that DO:concentrations as low as .
2 to 3 mg/1 can be tolerated for short duration. Thus, minimum
DO levels and allowable frequency of exceedance have been estab-
lished. The DO level-duration relationship is based on measured
survival times of juvenile brook trout when subjected to lethal
DO levels. Juvenile brook trout are a very sensitive indicator
species.
The EPA DO criteria based on the survivability of brook
trout are as follows:
"The minimum receiving water dissolved oxygen
concentration shall not average less than 2.0
mg/1 for more than 4 consecutive hours; nor
shall the minimum receiving water dissolved
oxygen concentration average less than 3.0
mg/1 for more than 72 consecutive hours (3
days) . In addition, the annual 'average re-
ceiving water dissolved oxygen concentration
shall be greater than 5.0 mg/1 for all waters
which will support warm water species and
shall be greater than 6.0 mg/1 for all waters
which will support cold water (salmonid) spe-
cies." (7)
Simulation studies indicate that if the criterion of 2.0 mg/1
for 4 consecutive hours is met, then, in general, all other
criteria will be met. - An allowable frequency of exceedance of
the 2.0 mg/1 criteria of one four-hour period per year is the
basis for biological oxygen demand (BOD) removal requirements.
For the purposes of the this study, DO levels of less than
5.0 mg/1 will be referred to as low DO or poor water quality.
Sites at which the standard of 2.0 mg/1 for more than four consecu-
tive hours is violated will be referred to as problem sites.
The question of what constitutes a problem is important in
an economic sense. Since the existence of a problem implies the
-------
necessity of a cure, in this case treatment of urban runoff,
considerable dollars are involved. It was an important part of
the study, if not a secondary objective, to carefully examine
"problems" before concluding the cause to be urban runoff.
-------
SECTION 2
SUMMARY OF BINDINGS
CORRELATION OF LOW DO WITH FLOW AND RAINFALL
Daily correlation analyses using either flow or rainfall
or both were attempted at 104 stations. These included USGS,
STORET, and WDNR data. Of these, 83 had sufficient data to
produce results. These included 55 USGS monitors, 17 STORET
monitors, and 11 WDNR monitors.
Of the USGS monitors, 19 exhibited a 60 ^percent or greater
probability of low dissolved oxygen at times of high flow.
Eighteen exhibited a 60 percent or greater probability of low DO
on days with rainfall. Eight of the stations correlated with
both flow and rainfall. The total number of stations exhibiting
either type correlation was 30.
Of the 17 STORET monitors, three exhibited a 60 percent or
greater probability of low DO at times of high flow. Of the 11
WDNR monitors, three exhibited the 60 percent or greater proba-
bility.
Out of 104 candidates for analysis, 36 gave positive re-
sults in the correlation analysis. Twenty-one of the monitors
could not be correlated because of data problems. Thus, 42
percent of the monitors successfully examined or 36 percent of
the likely candidates gave positive correlation results. For
discussion purposes, it is thus concluded that four monitors
in ten placed near an urban area might indicate lower than aver-
age DO at times of storm runoff.
MAGNITUDE OF DO DEFICITS
Because of the coarse time resolution of the daily data, no
direct conclusions could be drawn concerning the severity of the
DO deficits which sometimes accompany high flow or rainfall.
Examination of hourly data at sites with strong daily correla-
tion indicated that water quality violations can occur. Of the
30 USGS sites, 11 clearly indicate a severe DO deficit at times
of high flow. Eleven additional sites clearly indicate an
effect from storm runoff but could not be classified as severe.
-------
Several of these were judged to be too close to the urban area
to detect the maximum deficit in the Streeter-Phelps sense.
Six of the sites did not have water quality problems, at least
concerning DO. Some depression during storm events would
usually be seen.
CHARACTERISTICS OF DO VARIATIONS
Although the eleven sites with severe problems varied
widely in size and physical setting, all demonstrated remarkably
similar hourly data records. The period prior to a storm event
is characterized by fairly large diurnal cycles in the DO level;
fluctuations of 4 to 5 mg/1 are not unusual. Periods of
supersaturation are often indicated. As the flow increases,
the diurnal cycles in the DO record disappear. This may be due
to increased turbidity and depth cutting off the sunlight to
the aquatic plant life at times of high flow. '
At the time of peak flow, deficit levels at least equal to
and. occasionally 10 to 15 percent higher than the peak diurnal
cycle value are reached.
The effect of the storm flow on the DO level lasts quite a
long time. This long-term effect gives added validity to the
daily analysis procedure. Sites at which a single hydrograph
peak was examined usually recovered in three to five days.
COMPARISON OF STORM FLOW DEFICITS AND QUALITY STANDARDS
When measured against existing stream quality standards,
the DO deficits accompanying storm flow may cause water quality
violations. Hourly data at monitor sites with strong correla-
tion consistently show deficits below 5.0 mg/1 extending over
several days. The Streeter-Phelps analysis of 10 sites, in-
cluding several with strong correlations, consistently indi-
cated that monitors are not ideally placed to sense maximum
effects. In many cases, deficits 30 to 50 percent stronger
could theoretically be found.
Storm-flow-related deficits at some sites consistently vio-
late 2.0 mg/1 - 4 hour standards. In fairness, it must be
pointed out that at these sites the water quality is marginal
at all times. Storm events merely push the level down further.
General improvements in water quality at all the critical sites
would help alleviate the problem.
NATIONAL DISTRIBUTION OF SITES
The question of whether dissolved oxygen deficits caused
by urban runoff is a national problem is difficult to answer.
One item to consider is the geographic coverage of this study.
-------
The USGS data base maintained at the Reston, Virginia,
Headquarters contains records for 150 water quality monitors.
These monitors are located in 30 of the 48 conterminous states
or 63 percent coverage. The distribution of states containing
monitors is highly nonuniform. Thirteen states are east of the
Mississippi River and 17 are west. Only 47 of the 150 monitors
are in western states. If the dividing line between east and
west is considered to be along the western boundaries of Louisi-
ana, Arkansas, Iowa, and Minnesota, the distribution becomes
even more unbalanced. Only 19 monitors are then located in the
west.
The distribution of monitors by state is highly nonuniform
also. Ohio alone has 32 monitors, followed by New Jersey with
13 and Louisiana with 11.
It was ultimately concluded that the existing monitor net-
work is probably inadequate to define a national problem. The
probability is approximately one in three of detecting a corre-
lation between flow and DO deficit. Many states with signifi-
cant urban areas on streams have no monitor records at all in
the major data banks.
-------
SECTION 3
CONCLUSIONS
Several hundred station-years of DO, stream-flow, and temp-
erature data from locations around the United States have been
analyzed to determine if a correlation exists between DO deficit
and either flow and/or rainfall downstream of urban areas. In
the first portion of the study, daily average data were used to
select stations where a high probability existed of the simul-
taneous occurrence of higher than average flow and/or a rainfall
event and DO deficit. In the final portions of the study,
hourly flow, rainfall, and DO data were examined at those sites
exhibiting a high probability. The hourly data examination was
used to more clearly identify the nature of the correlation. In
addition, a Streeter-Phelps analysis was made at selected sites
to determine if a particular water quality monitor was in the
best location to detect DO deficits from an urban area. The
following conclusions were drawn from the study:
• The probability is approximately one in three
that analysis of data from a currently existing
water quality monitor in or near an urban area
will show a correlation between high flow and/
or rainfall events and high DO deficit.
• In close examination of sites that exhibited
daily correlation between flow and low DO or
both rainfall and flow and low DO, strong
visual evidence was usually found of the cor-
relation in hourly data records. Sites that
exhibited daily correlation with rainfall
seldom showed any visible evidence of correla-
tion in hourly data records. Flow was judged
the better correlation parameter because it is
a direct rather than a secondary indicator of
stream conditions.
• At stations where correlation exists, maximum
deficits observed during periods of high flow
are equal to or 10 to 15 percent greater than
the maximum deficits observed during diurnal
cycles at times of steady low flow.
-------
© At locations where water quality is already
marginal (e.g., 5-7 mg/1 DO), a storm event
can result in DO levels less than 4-5 mg/1
for periods of several days or longer. Occa-
sional violations of 2 mg/1 - 4 hour standards occur,
a Most water quality monitors are located too
close to the urban areas to detect the maximum
possible DO deficit in a Streeter-Phelps sense.
• Flow and DO deficit are correlated in a wide
variety of urban situations ranging from
towns of 20,000 population to urban mega-
lopolises of greater than 1 million population.
• The absence of a correlation between flow and
DO deficit does not correlate well with
the percentage of contributing urban areas in
the drainage basin. There are exceptions in
extreme cases where 50 percent or more of the
contributing area is urban.
• There is no reliable way to extend the results
of this study to an estimate of total affected
stream miles nationally. No government agency
currently publishes data on the distribution
of population along major rivers.
• The probability of a high DO deficit occurring
at times of high flow is not demonstratably
greater during the summer months.
• In most instances, the exact cause of the in-
crease in DO deficit is not obvious. In some
cases, there are heavy concentrations of in-
dustry along the stream channel. In other
cases, a sewage treatment plant is close
upstream. Reintrainment of benthic material
is a likely cause in such locations. In other
instances, there is no industry or sewage
treatment shown on the site maps. Here, the
problem may be strictly due to BOD of urban
runoff, a chemical oxygen demand, or other
problems.
10
-------
SECTION 4
RECOMMENDATIONS
The resources of this study were directed primarily at de-
termining if a correlation exists between increased flow and/or
rainfall events and increased DO deficit below urban areas.
This fact was reasonably well established at a number of loca-
tions. At several of these locations the dissolved oxygen levels
were low enough to violate water quality standards. The follow-
ing recommendations are made in an effort to determine why such
a relationship exists:
More-detailed investigations should be undertaken
at three to five of the sites with strongest cor-
relation. These investigations should include:
(1) a field study of the river reach through the
urban area to the monitor; (2) studies of the
industrial waste and sewage treatment practices
along the reach; (3) details on the drainage
systems from the urban areas; (4) population
density, climate, and other pertinent data.
Sites specifically identified as worthy of study
include the Cuyahoga River between Akron and
Cleveland, Ohio; the Scioto River between Colum-
bus and Chillicothe, Ohio; the Mahoning River
below Youngstown, Ohio; the Sandusky River below
Upper Sandusky, Ohio; and six other sites with
hydraulically complex conditions.
The data base created as part of this study has
considerable value. Consideration should be
given to publishing a short report on the data
base and making it available along with digital
magnetic tapes of the data.
A project should be undertaken at one site with
a strong correlation to develop an adequate data
set to support an unsteady flow water quality
model. Streeter-Phelps analysis is not an accu-
rate representation of water quality behavior
under storm conditions. "Adequate" data set
does not mean historical data collected by random
11
-------
agencies at random times and places. It means
a concerted effort to obtain synoptic or near-
synoptic values of pertinent water quality
variables in a way designed for use with a
model. Such a data set would allow meaningful
investigation into the exact behavior of the
DO deficit and the oxygen-demanding material
during storm flow.
Even in the absence of a modeling effort,
further cause-effect type data would be valu-
able at sites with strong flow-DO correlation.
It would be particularly valuable to determine
if the cause is nonpoint runoff, treatment
plant bypass, reintrainment of settled indus-
trial waste, combined sewer overflow (CSO), or
other causes.
12
-------
SECTION 5
SITE LOCATION AND SCREENING
Locating agencies that maintain data bases of water quality
information and screening for relevant material required a
significant amount of effort. This section identifies those
sources that were contacted, describes the type of data avail-
able, and outlines the procedure for selecting sites for inclu-
sion in this study.
The following paragraph from the Scope of Work clearly
describes the required tasks:
''Available, continuously recorded water quality
information in data banks such as EPA's STORET
system and USGS's NAWDEX system and from other
sources and agencies such as the Ohio River
Sanitation Commission (ORSANCO) that operate
monitoring stations shall be reviewed to obtain
a five (5) year historical picture of dissolved
oxygen, temperature, and flow in receiving waters
of the United States at the locations of all re-
liable continuous water quality monitoring
stations. Concurrent with this review five (5)
year hourly rainfall data for the areas in which
the monitoring stations are located shall be
obtained from the National Climatic Center in
North Carolina and the National Weather Service
in Maryland, or from other available sources."
The requirement for five years of data was subsequently relaxed,
but initially the project proceeded as directed above.
In order to comply with this segment of the Scope of Work,
a search of the listed agencies was begun. During the course
of the search, other sources were discovered and in turn searched
for pertinent data. The quantity of information available is
quite large and cannot practically be presented here, even as
an appendix. Instead, each agency that provided useful input
to the study will be discussed individually. Relevant names and
addresses are provided so that the interested reader can obtain
more-detailed information if desired . Idiosyncrasies of the
system, if any, are mentioned. After describing the various
information banks, the screening procedure will be described.
13
-------
AVAILABLE DATA BASES
Office of Water Data Coordination (OWDC)
The Office of Water Data Coordination is under the Depart-
ment of the Interior. It is part of the Water Resources Divi-
sion of the USGS. Its nominal mission is to "coordinate
federal activities in the acquisition of certain water data."
The OWDC is located in the USGS National .Headquarters at Reston,
Virginia. Inquiries may be mailed to the USGS Headquarters,
Mail Stop 417. At the time of this writing, Russell H. Langford
was in charge of the operation.
The OWDC compiles a catalog of water data collection sta-
tions throughout the United States. Water quality and quantity
monitors for streams, lakes, reservoirs, estuaries, and ground
water are included. It publishes a 21-volurae set of "Catalog of
Information on Water Data," the most recent dated 1974. Another
set was under review but was not available for this study.
These 21 volumes correspond to 21 regions of the country. These
21 regions, along with the number of federal, nonfederal, and
Canadian agencies contained in each, are listed in TABLE 1.
The OWDC catalogs draw information from many sources in-
cluding federal agencies, state and local agencies, and Canadian
agencies. The following statement appears on page 2 of each
catalog:
"Although not all non-federal agencies acquiring
water data were contacted, the district offices
attempted to include those agencies most active
in water-data acquisition in each state, thereby,
deriving extensive, though not necessarily com-
plete, coverage."
All of the agencies in each region are listed in Appendix A.
A quick glance at Appendix A should convince even the most
casual reader that an, enormous number cf agencies collect water
quality data. It also should explain why the OWDC catalog was a
primary source of information for this study. Over 350 non-
federal (state and local) agencies are listed, plus all the
relevant federal agencies. The writers are confident that no
more complete list of.such information exists.
The reader should be aware that OWDC^does not collect or
disperse data. It merely keeps track of who does. The OWDC
will supply upon request names and addresses of all the agencies
listed,as well-as-a'personal contact ,and telephone number, if
known^ This was invaluable for the present study.
14
-------
TABLE 1. OWDC REGIONS AND NUMBER OF
AGENCIES COLLECTING WATER QUALITY DATA
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Region
New England
Mid-Atlantic
South Atlantic-Gulf
Great Lakes
Ohio
Tennessee
Upper Mississippi
Lower Mississippi
Souris-Red-Rainy
Missouri Basin
Arkansas-White-Red
Texas-Gulf
Rio Grande
Upper Colorado
Lower Colorado
Great Basin
Pacific Northwest
California
Alaska
Hawaii and other Pacific islands
Caribbean
Federal
agencies
8
10
10
10
9
6
7
7
5
9
8
7
7
6
9
10
14
13
9
6
4
Non-federal
agencies
5
14
32
15
19
13
31
21
9
36
17
2
2
6
9
14
28
61
4
10
1
Canadian
agencies
2
2
0
2
0
0
0
0
2
2
0
0
0
0
0
0
2
0
1
0
0
15
-------
One unfortunate idiosyncrasy of the OWDC catalogs is that
the information is not stored in computer data banks. In order
to locate the continuous monitors, it was necessary to go
through all 21 volumes one page at a time and decipher the codes.
National Water Data Exchange (NAWDEX)
The National Water Data Exchange is also a program of the
Department of the Interior. It is administered through the
Water Resources Division of the USGS. While OWDC is a catalog-
ing agency, NAWDEX actually deals in data. When fully opera-
tional, NAWDEX will'be capable of searching the complete USGS
water data records as well as the EPA STORET data base. Contact
with NAWDEX is established through the Office of the Assistant
Chief Hydrologist for Scientific Publications and Data Manage-
ment at the USGS National Headquarters in Reston, Virginia,
Mail Stop 440. At the time.of this writing, Mr. George W. Whet-
stone was in charge of the office. The writers dealt with Mr.
Melvin D. Edwards.
At the time the present study was undertaken, the NAWDEX
program was only capable of scanning the USGS records. A list
of 45 continuous water quality monitors at which flow data were
available was obtained. It was subsequently determined that
these could have been found in the OWDC catalogs. The advantage
of NAWDEX, if there is one, is more current information on the
USGS files. Better results, however, can be obtained from the
USGS Automatic Data Processing (ADP) Unit, which is described next.
The cost of a NAWDEX search is roughly $50 to $100.
USGS Automatic Data Processing Unit
The USGS ADP Unit is located in the National Headquarters
in Reston, Virginia. Its purpose is to provide standardized
processing programs and procedures for the water resources data
collected by USGS. The ADP Unit may be contacted ,by writing
the.National Headquarters, Mail Stop 485. Mr.,Robert B. Wall
is in charge.
This unit is the most up-to-date source of information on
the contents of the USGS water quality and flow data base. It
is also"the place to contact to obtain data in digital format
such as cards or magnetic tape. For very reasonable prices ($35
to $70) any of the daily parameter values stored by the ADP
Unit can be obtained along with excellent instructions on how
to read and translate the cards or tape. The personnel of USGS
ADP Unit are some of the most friendly and cooperative of any
encountered in the course of this study. The work would have
been very difficult without them.
16
-------
EPA's STORET System
STORET is a computerized water quality data base. It was
conceived and initiated in the early 1960s by the U.S. Public
Health Service. Its sole purpose is to allow storage and re-
trieval of water quality information by government users. Con-
tact with STORET is initiated by calling the User Assistance
Center in Washington, D.C., at (202) 426-7792. Mr. Phil Lindens-
truth was very helpful to this study.
The potential STORET user is in for a difficult time if he
does not have connections with a government agency. While any-
one can walk into the USGS and receive data for a fee, this is
not true of STORET. Some kind of government account number is
required. For the purpose of this study it was necessary to
obtain a user identification number through the project officer.
Finding continuous monitor records in STORET was not easy.
A number of the people at EPA Headquarters were convinced that
no more than half a dozen continuous moni'c ji records were in
STORET. Over 1100 were eventually located. Approximately 200
of these monitors were located on lakes, over 100 in pipes, al-
most 100 beside the ocean, and about 50 in intakes and estuaries.
This left about 500 monitors located along rivers. Data from all
the river sites below urban areas were transferred to a digital
tape for use in this study.
The water quality data in STORET is in an unusual format.
Daily and hourly observations are mixed together in a continuous
string. This complicates processing to some degree. The poten-
tial user should be aware that STORET contains no flow informa-
tion.
U.S. Army Corps of Engineers (COE)
The U.S. Corps of Engineers maintains on its own or
through USGS a great many water quality monitors. These are
listed for the most part in the OWDC catalog described earlier.
Virtually all are located at the inlet or outlet of reservoirs
and were of no interest to this study. No direct data input
from COE is included here.
State and Local Agencies
Little actual contact was made with state and local agen-
cies during the course of this study. There are two reasons for
this. First, the OWDC catalogs are a definitive source of which
state and local agencies collect water quality data. The very
great percentage of those that do, did not have continuous data
bases. They concentrate on random observation. Very few data on
flow are available from such sources. Second, those rare
17
-------
state agencies that actually collect continuous data report the
data to the Environmental Protection Agency' s STORET data base..
Thus, it was not necessary to spend a great deal of time coiv-
tacting many places. Two agencies that do maintain their own
data bases are the Wisconsin Department of Natural Resources and
the Metropolitan Sanitation District of Greater Chicago. They.
are described below. .
Wisconsin'Department of Natural Resources
The Wisconsin Department of Natural Resources maintains a
network of 11 water quality monitors. It also records flow
continuously. The WDNR may be contacted at Box 7921, Madison,
Wisconsin. Mr. Mitchell S. Nussbaum of the Water Quality Eval-
uation Section was quite helpful.
For a very reasonable amount, the WDNR will copy all 11
stations of daily data onto a magnetic tape. Hourly data are
also available. The WDNR is as helpful and courteous to deal
with as the USGS ADP Unit.
Metropolitan Sanitary District of Greater Chicago
The Chicago Sanitation District maintains a number of flow-
water quality monitor stations in and around Chicago. Many are
located in canals. Hourly data are available, but in digital
forms. Hand-copying of the numbers in person at its offices
in Chicago is the only way to obtain information. Even this can
only be arranged with difficulty. A letter from a government
project officer explaining the need for the data is required.
Because of the manpower and travel requirements, no data were
used from this source in this study.
At the time of this writing, the point of contact is Dr.
Cecil Lue King, Director of Research and Development, Metropoli-
tan Sanitation District of Greater Chicago, 100 E. Erie Street,
Chicago, Illinois, 60611.
Ohio River Valley Water Sanitation Commission
The ORSANCO is an interstate compact agency. It was cre-
ated in 1948 by the states bordering the Ohio River. Its pri-
mary purpose is to monitor and prevent pollution along the river.
To achieve this, ORSANCO maintains a network of water quality
monitors. This network reports continuously to the headquarters
in Cincinnati, Ohio, where a complete picture of the quality of
the river at any time is available. The address of the commis-
sion is 414 Walnut Street, Cincinnati, Ohio, 45202.
At various times, 56 monitor sites have been placed on the
river. Half of these have been in continuous operation since
1961; the DO and temperature data from these stations are
18
-------
available on an hourly basis. However, no flow data are avail-
able. For less than $100, ORSANCO will copy all the data onto
a digital tape.
While the data are readily available, there are serious
drawbacks. Most important is the data format. Because of
ORSANCO1s interest in an "overall" look at the basin, the num-
bers are not grouped by station. Instead, each tape record
contains data for one day for all stations reporting that day-
The records are random length and structured in such a way that
hundreds of tape reads are required to sort the data by station.
A second drawback is the age of the ORSANCO tape drives. ORSAN-
CO' s system is an IBM 11/30. The tapes it provides produce
numerous read errors on later-model IBM equipment. Processing
cost was so high for the ORSANCO data that none is included in
this report. For example, it cost $45 to transfer three months
of a single station to a disk pack and plot the results.
National Climatic Center
Precipitation data for the present study were obtained
from the National Climatic Center in Asheville, North Carolina.
Its address is Federal Building, Asheville, North Carolina 28801.
Telephone inquiries can be made to (704) 258-2850, ext. 683.
The Climatic Center is a data bank for the parent agency,
the National Oceanic and Atmospheric Administration (NOAA),
which is part of the Department of Commerce. A very large
quantity of atmospheric data is available there. Literally
thousands of rain gages are scattered at small airports and
other locations throughout the United States. Most of these
report daily. In addition, it maintains nearly 300 primary
weather stations that record rainfall, temperature, wind, and
relative humidity on an hourly basis. This voluminous infor-
mation is available in monthly and annual reports and on digital
tapes or cards.
For this study, only two forms of data were used. In the
initial phases a digital tape containing daily precipitation and
other variables was obtained. This type of data is very costly.
For reasons that were never clear to the writers, the Climatic
Center charges $33 to mount a tape. An additional charge of
roughly $50 per single station is encountered as each station
requires a separate tape mount. These charges are very high
compared to other government agencies on a per-number basis.
They are also much higher than private computer centers. How-
ever, the Climatic Center is the only available source. Digital
data have a large lead time of six to eight weeks.
The large money requirement ruled out obtaining the hourly
data for the later phases of the study in digital form. Instead,
free copies of "Local Climatologic Data," a monthly publication,
19
-------
were obtained from the National Weather Service (NWS) in.' Silver
Spring, Maryland. These contain the hourly .precipitation for
the month of issue. The normal single copy price of the local
climatologic data is $0.20. These may be ordered from the
National Climatic Center in Asheville.
SITE SELECTION PROCESS
The process of selecting monitors for inclusion in the
study was quite time consuming. It consisted of four steps.
First, the data bases discussed above were scanned to discover
which of the water quality stations were actually continuous <
monitors. Next, the continuous monitor sites were located.
Following location, each monitor was classified as to its use-
fulness to this study. Finally, those with the best classifi-
cation ratings were selected and data obtained. Each of these
steps will now be described in detail.
Monitor Discovery
The process of discovering monitor data for use in this
study was complicated by the number of agencies involved and the
variety of cataloging systems. Many of the possible sources of
information (exclusive of the large federal groups such as USGS
or EPA) were not known to the writers at the outset. It became
a matter of detective work to go from known specific monitor
listings to more general information.
The first government agency contacted was the Water Re-
sources Division of the USGS. The NAWDEX provided a list that
supposedly contained all the continuous USGS water quality
monitors where flow was also recorded. There were only 45.
Two of the writers had worked for USGS for over 10 years and
were aware of monitors that did not appear on the list. Further
inquiries led to the Office of Water Data Coordination and in-
dividual USGS district offices in the various states.
The Office of Water Data Coordination was the best source
of monitor sites discovered in the entire study. Its 21-volume
catalog contains thousands of locations where sampling activi-
ties occur. There was only one serious drawback. Continuous
monitors are not separated out except by letter codes. Each of
the several thousand pages could contain from zero to 10 moni-
tors useful to the study. These had to be located by going
through the 21-volume one page at a time, reading the codes, and
writing down potentially useful sites and the governing agency.
This process produced a very comprehensive list of likely sites.
The USGS maintains a district office in almost every state
in the union. These offices conduct water quality monitoring
20
-------
through cooperative agreements with many state, local, and other
federal agencies. Most monitors are designed to specifications
set by the USGS. The writers initially set out to call all 50
district offices one by one to determine if monitors were avail-
able that did not appear on the NAWDEX list. Inquiries were
also made concerning activities in the cooperative program.
During this slow process, several districts suggested contacting
the USGS ADP Unit and obtaining a list of the monitors on the
backfile records. This provided a list of all water quality
monitors maintained by USGS as well as flow stations. The list
was far more comprehensive and up to date than the one obtained
from NAWDEX. No further searches were conducted for USGS moni-
tors .
No difficulty was encountered in determining how many moni-
tors were available from ORSANCO or EPA. ORSANCO merely pro-
vided a comprehensive list. Cross-checks verified that the same
information was available through the OWDC catalogs. At first,
it was difficult to gain access to the STORET system. Once a
user number was obtained, very little time was required to
obtain lists of the monitors contained in the data base. A
great many of these were state and local agency monitors that
were also listed in the OWDC catalogs.
In all, over 1500 monitors that recorded continuous DO and
temperature plus other parameters were located. The next steps
were to locate, classify, and sort them to determine which were
near urban areas and had flow and rainfall data available.
The location, classification, and sorting processes are
described in the following sections. Each of the 1500 monitors
was processed in the same way unless it became obvious for some
reason that it did not apply to the study (for instance, a moni-
tor 50 miles from the nearest town in the middle of a swamp).
Monitor Site Location
Most monitors were located by means of latitude and longi-
tude coordinates. All the station lists with the exception of
ORSANCO reported an accurate latitude and longitude for each
site. ORSANCO locates its stations by COE river mile markers.
In the early parts of the study, USGS 7.5-minute topo-
graphic sheets were purchased and used to locate monitors and
flow gages. This proved unwieldy and expensive. To avoid
buying so many maps, most locating was done at the National
Headquarters Library of the USGS. Topographic maps covering 7.5
minutes, 15 minutes, and at a scale of 1:250,000 for the entire
country are available there. Each monitor was found and its
proximity to an urban area noted. As mentioned in the previous
section, a number of monitors were rejected from consideraton at
this point if no urban area was in a position to affect the stream.
21
-------
Locating the monitors on 7.5-minute quadrangles required
one intermediate step. The appropriate quad sheet name had to
be determined. This was done by obtaining index sheets for all
50 states from the USGS Cartographic Information Center at the
National Headquarters in Reston, Virginia. These were free.
Where possible, the town after which the monitor was named was
located and the quad sheet identified. If the town could not be
found, rough latitude-longitude was used and several quad sheets
picked. The quad sheet names and other pertinent information
were entered on a classification form, which will be discussed
shortly. The format is presented in TABLE 2.
Monitor Site Classification
Classification of the monitor sites for suitability was an
important step. Several factors had to be considered. These
were
• proximity to urban drainage area,
• presence of continuous record,
• length of record,
• proximity of flow gage,
• proximity of rain gage, and
• length of time monitor was in operation.
There were several hundred monitors to consider, which made it
impossible to remember even a small portion. The previously
mentioned standard classification sheet was developed and filled
out at the time the monitors were being located. An example of
the completed form is shown in TABLE 2. Although the form indi-
cates that it'was used only for monitors in the OWDC catalog, it
was actually used for all monitors in the study. A complete
list of slightly over 100 monitors considered for analysis is
contained in Appendix B.
The state column was used to identify both the state in
which" the monitor was located and the 7.5-minute quad sheet
required to locate it precisely. The station name, agency,
latitude, and longitude columns are self-explanatory. The site
column identified the type of monitor location. These were
based on the OWDC catalog description or the writer's observation
of the topographic map. ' The date established is self-explana-
tory. This information came from either the OWDC catalog or
the agency that operated the station. The column'labeled water
discharge indicates whether or not a stream gage was located
near the monitor. Information obtained in this regard was occa-
sionally misleading. The USGS, for example, assigns different
22
-------
TABLE 2. LIST OF ALL SITES CONSIDERED FOR DAILY OR HOURLY ANALYSIS
State
Alabama
Colorado
Georgia
Illinois
Louisiana
Station name
Coosa R. at Verbena
Coosa R. at Gadsen
S. Platte R., 60 ave.
S. Platte R., 88 ave.
Burlington Ditch at
York St.
Sand Cr. at
Burlington Ditch
Chattahoochee R. at
Atlanta
Peachtree Cr. near
Atlanta?
Ocmulgee R. near
Warner-Robins?
Calumet R. STWt
Chicago R. Bridge?
Chicago Sanitation
and Ship Canal
at Lockport?
Bayou Tech at
Olivier
Houma Nav. Canal
near Dulac?
Ouachita R.at
Monroe
Agency*
GS
EPA
EPA
EPA
EPA
EPA
EPA
GS
GS
GDI
G01
GDI
GS
GS
GS
Lat.
324756
340057
394826
395115
394802
394837
335132
335133
324017
413946
415333
413408
295718
292306
323019
Long.
862602
855843
1045730
1045615
1045730
1045659
842716
842716
833611
873940
873832
880441
914254
904347
920732
Site
Stream
Stream
Stream
Stream
Canal
Stream
Stream
Stream
Stream
Canal
Stream
Canal
Stream
Stream
Stream
Established
1974
1971
1968
1968
1967
1967
1960
_
1970
1969
1969
1969
1972
1973
1954
Water discharge
available?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
Yes
No
Yes
Primary
rain
gage
Montgomery
Gadsen
Denver
Denver
Denver
Denver
Atlanta
Atlanta
Macon
Chicago
Chicago
Chicago
-
-
Shreveport
Applicability
Distance to
, , to study
rain gage (mi)
code
4
4
<10 1
<10 1
<10 1
<10
10 1
10 1
20 4
20 2
5 1
35 2
- -
-
90 4
The agency codes, taken from OWDC, identify the agency that was in charge of the data ; specifically, GS = Geological Survey, EPA = Environmental Protection
Agency, and GO1 = Metropolitan Sanitation District of Greater Chicago.
was later dropped.
-------
names and station numbers to monitors and gages that are not
exactly in the same place, even if only a few hundred yards
separates them. Thus, the monitor Lonely River at Outback,
Montana, may be listed by OWDC or the USGS as having no flow.
A stream gage named Lonely River near or below Outback, Montana,
may be quite nearby. This problem was minimized by checking
USGS state reports to determine if what OWDC indicated was true.
The rain column gives the location of and distance to the near-
est weather bureau primary weather station (those with, hourly
rainfall data) . The .distances are in miles. The last column,
applicability to study codes, was used to rate each site as to
its usefulness to this research. These ratings tended to be
subjective and based heavily on how the monitor's location on
the map appeared in relation to an urban area.
The classification codes are listed in TABLE 3. The amount
of urban area was judged by eye. In general, towns were judged
in relation to the size of the river. A town of 1/4 square mile
on the Mississippi was ignored. The same town on a small river
was included. Thus, the terms "significant urban area" and
"small town" were relative rather than absolute measures. The
criterion of 15 miles to a weather station was based on judgment
and experience with other studies.
TABLE 3. STUDY CODE CLASSIFICATIONS
Code Definition
1 Significant urban area, NWS primary station within 15 miles
2 Significant urban area, NWS primary station farther than 15 mites
3 Some urban area, NWS primary station within 15 miles
4 Some urban area, NWS primary station farther than 15 miles
5 Small town(s)
6 No urban area
7(1-6) Reservoir or lake
Site Selection
By the time the study had progressed to site selection, it
became apparent that there were very few places that met the
criteria in the Scope of Work. • These criteria were the existence
of five years of research including hourly flow, DO, and temper-
ature plus 15-mile proximity to a rain gage with hourly data.
24
-------
There were 12 in the USGS, none in ORSANCO, none in EPA'sSTORET,
and only a few with other agencies.
A decision was subsequently reached through discussion with
the project officer to broaden the Scope of Work. It was decided
to accept sites with five years or less of data, to examine
sites that had only daily data, and to consider sites that were
not near hourly rain gages. At the same time, it was decided to
concentrate initially on daily correlations and use the hourly
data for a detailed look at sites with strong daily flow-DO
deficit correlations.
Summary
A review of the process of site selection is now in order.
Initially, slightly more than 40 USGS monitors were found in the
NAWDEX system. A search of the OWDC catalog series added 250
more. Several additional USGS monitors were located through the
ADP Unit. ORSANCO added 56. A search of the STORET files pro-
duced 1100. Thus, there have been nearly 1500 monitor sites
established along the nation's streams, lakes, and estuaries.
Each monitor was located on USGS maps and evaluated for proximity
to urban areas. They were further rated on proximity to rainfall
and flow gages and for length of record. Those in proximity to
an urban area of significant size and which were near a flow
gage or primary weather station were selected for daily correla-
tion analysis. The number of such sites was slightly greater
than 100. The process of examining these on a daily basis is
the subject of the following section.
Some explanation is in order on why out of 1500 monitors
less than 10 percent were considered. The most common reason
was that the monitor was not in or near any urban area. The
vast majority of EPA's monitors (at least those in the STORET
files) are at industrial outfalls or in canals or lakes. This
is also true of the USGS and ORSANCO networks. A small minority
of monitors were rejected because no flow records were available.
Some were rejected because an examination of the daily monitor
records indicated no quality problems at any time. Some sites
were incorrectly listed as continuous monitors when in fact they
were just grab-sample locations.
The list of all monitors considered is too large to conven-
iently publish here, even as an appendix. Appendix B expands
TABLE 2, presenting 104 sites that were seriously considered for
analysis. The following section describes how the 104 sites
were analyzed. Note that the ORSANCO sites considered for anal-
ysis are included even though they were not analyzed because of
the awkward data storage format.
25
-------
SECTION 6
DAILY CORRELATION ANALYSIS
GENERAL CONSIDERATIONS
The necessity for basing initial correlation efforts on
daily data stemmed from broadening the Scope of Work to accept
monitors with less than five years of data or located more than
15 miles from an hourly rain gage. Hourly data are difficult to
obtain and process. For example, hourly data from the USGS are
not stored on permanent files in a computer or on magnetic tape.
They are available only on listings called primary computation
sheets. In order to analyze a year of this type of data, one
must read 8760 numbers from the sheets and either punch the
values on cards or have them put on tape. Performing hourly
correlation analysis on 100 sites with an average of 2.5 years
record and four parameters (DO, rainfall, temperature, and stage)
would require 8,760,000 numbers.
Another important factor restrict:", the use of hourly data.
Hourly precipitation data are only available at about 300 sites
throughout the United States. Therefore, many dissolved oxygen
monitors are located at such a distance from these primary
weather stations to make an hourly correlation between DO and
rainfall meaningless.
The necessity of using daily data to supplement the hourly
values was anticipated in the Scope of Work:
"Some of the available data banks have been sum-
marized and contain only daily average values.
These can be utilized to determine whether the
particular stream segment should be included in
the analysis, but it will be necessary to obtain
the hourly parameter variations from the organi-
zation responsible for the water quality monitor-
ing or rainfall station in carrying out the
analysis."
The method of examining daily data first saved a great deal of
computation by allowing use of readily available computer tapes.
The USGS, STORET, and WDNR data bases were all able to provide
daily values in this form.
26
-------
There were three distinct phases to the daily correlation
analysis. First, the daily data had to be acquired and converted
to a standard format. These values were stored on a computer
direct-access file. Second, the stored values had to be re-
trieved and processed with suitable correlation programs. Final-
ly, sites for hourly and detailed analysis were selected. Each
phase will be discussed in turn. The computer processing and
results of the analysis will be presented first. The process
used for selecting sites for hourly analysis completes the sec-
tion .
PRELIMINARY DATA ACQUISITION PROCESSING
Four major computer programming efforts were required to
accomplish the first phase of data processing. Each type of
taped data (USGS, NWS, STORET, and WDNR) required different
recovery and processing methods. A short discussion o± the data
formats and the processing problems encountered with each will
now be presented.
USGS Data
The importance of USGS data to this study cannot be over-
emphasized. When the site selection process was complete, 62
of its monitors were of interest. USGS flow data were often
required to supplement DO data from other sources. For these
reasons, the USGS data were obtained and processed first.
The USGS ADP Unit is highly cooperative and accustomed to
dealing with the public. To request data, all that is required
is a letter or memorandum stating what stations and years of
record are of interest. In from three days to a week the re-
quested data arrives along with a detailed set of instructions
on how to decode it. The price for 100 station-years of record
including a magnetic tape is around $65. The appendix of the
instruction book contains a program for reading and listing
either a data tape or cards. Anyone familiar with programming
can have the program operational in less than a day.
The data are stored on tape such that data covering a year
are easily accessible at a particular site. The USGS uses water
years, which begin October 1 and end on September 30 the follow-
ing year. The data values are stored in a two-dimensional,
12 x 31 array. This is preceded by header information such as
the state code, latitude and longitude, year of record, and other
information. Days that do not exist (such as February 31) or
missing data are marked by a 999999 no-value indicator. TABLE 4
presents the 9-track 1600 bpi tape format used by the USGS.
The program provided by USGS to read and list the tape was
modified slightly to write the 12 x 31 arrays, the station name,
the year of record, and the latitude-longitude onto a direct
27
-------
TABLE 4. USGS. 1.656-BYTE, MAGNETIC-TAPE OUTPUT FORMAT
K)
00
Byte
position
1-2
3-4
5-9
10-24
25-28
29-32
33-36
37-38
39-40
41-44
45-1532
1533-1535
1536-1537
1538-1540
1541-1588
1589-1592
1593-1596
1597-1600
Data
type
Description
CHAR (2)
CHAR (2)
CHAR (5)
CHAR (15)
FLOAT (6)
FLOAT (6)
FIXED BIN (31)
FIXED BIN (31)
FIXED BIN (15)
FLOAT (6)
FLO AT (6)
CHAR (3)
CHAR (2)
CHAR (3)
CHAR (48)
FLOAT (6)
FLOAT (6)
FLOAT (6)
Blank.
Code for the state in which the station is located.
Code for the agency collecting the data.
Station identification number (right justified).
Cross-section locator. Location of sampling point in the
cross-section. A value of '999999' indicates no value
stored.
Sampling depth. Depth at which sample was collected. A
value of '999999' indicates no value stored. A value of
'111111' indicates a top sample and '888888' indicates a
bottom sample.
Parameter code.
Calendar year date.
Statistic code.
No value indicator - value that is stored in place of a
missing daily value. A value of '999999' is stored as the
no value indicator for most data. The no value indicator
is also stored for non-existent days in the array, such as
February 30.
Two-dimensional array (12, 31) containing all of the daily
values for a 12-month period. For additional information,
see page 12.
Blank.
Code for the state in which the Geological Survey office
that operates the station is located.
Code for county in which the station is located.
Station name.
Drainage area of the site in square miles. A zero indicates
no value stored.
Contributing drainage area of the site in square miles. A
zero indicates no value stored.
Depth of well. The greatest depth at which water can
enter the well. Applicable only to sites coded as 'GW. A
value of '-99999' indicates no value stored.
(continued)
-------
TABLE 4 (continued).
fO
Byte
position
1601-1604
1605-1608
1609-1610
1611-1612
1613-1614
1615-1620
1621-1627
1628-1629
1630-1637
1638
1639
Data
type
Description
FIXED DEC (7, 2)
FIXED BIN (31)
FIXED BIN (15)
FIXED BIN (15)
CHAR (2)
CHAR (6)
CHAR (7)
CHAR (2)
CHAR (8)
CHAR(l)
CHAR(l)
Datum of gage to nearest hundredth of a foot.
Hydrologic unit code.
Retrieval sequence number for data control of batched
retrievals.
Month number that contains the beginning date of the
12-month period. For additional information, see page 12.
Site code. The codes are:
SW = Stream
SP = Spring
ES = Estuary
GW = Well
LK = Lake or reservoir
ME = Meteorological
1640-1656
CHAR (9)
Latitude of the site location in degrees, minutes and
seconds.
Longitude of the site location in degrees, minutes and
seconds.
Sequence number—used differentiate between stations
having the same latitude and longitude and to identify
the site location in relation to the Earth's spherical
quadrants.
Geologic unit code.
Reserved.
Aquifer type. The codes are:
U = Unconfined single aquifer.
N = Unconfined multiple aquifer.
C = Confined single aquifer.
M = Confined multiple aquifer.
X = Mixed (confined and unconfined) multiple
aquifers.
Blank.
Source: "Data Formats for U.S. Geological Survey Computer Files Containing Daily Values for Water Parameters," by Charles R. Showen, 1976.
-------
access disk file. This allowed any year at any station to be
called into the computer without reading any others. Because of
the relatively large number of USGS monitors that were to be
analyzed and because of the convenience of the 12 x 31 array
data format, daily data from other agencies was converted to
this form. By building the disk files in a single format, only
one set of plotting and analysis programs will be discussed
following the presentation of the data formats of the other
agencies.
Data are stored in the Daily Values File as a data array
that comprises a water year of data (October 1 through September
30). The water-year date is designated by the calendar year in
which the 12-month period ends. The daily values data array is
always represented as a two-dimensional array (12,31), with the
month as the first dimension and the day as the second dimension.
The data are physically stored in the array such that the first
data value (01,01) is for January 1, the second data value (01,
02.) is for January 2, etc., on through December 31 (12,31).
The data within each daily values array may be chronologi-
cally ordered for output to comprise any type of year. How-
ever, the physical order will always begin with the values for
January and end with those for December even though two calendar
years are spanned. The type of year stored within a daily
values record is defined by a beginning month number stored in
the record. Unless specifically requested, the daily values
output records will be comprised of water years (beginning
month = 10). For example, data for the 1974 water year are
physically stored in the data array such that the first data
value'(01,01) would be for January 1, 1974, the 180th data
value (10,01) would be the data value for October 1, 1973, and
the data value in the array (12,31) would be for December 31,
1973.
In terms of number of monitors the STORET data base was
second to USGS. The monitor records were from a variety of
federal, state, and local agencies. The STORET data were some
of the last to be processed because of the difficulty in gaining
access to the system for private organizations. Once in, how-
ever, the system works well..
The personnel of the STORET user assistance facility at EPA
Headquarters in Washington, D.C., created a number of useful
products for this study. First, a. scan was done on the data
base to produce a list of all the continuous water quality moni-
tors. This was mentioned previously, in the site location
section. After this listing had been evaluated and the monitors
of interest selected, all the daily and hourly data available
at each monitor were transferred to a magnetic tape.
30
-------
STORET intermixes daily, hourly, and random observations.
This makes processing more difficult. STORET, as with USGS,
provides the user with a sample reading and translation program.
The only difficulty with it was sorting out what variable went
with what variable name. This was necessary to know so that
useless data could be discarded. Roughly one week of effort
was required to get a working program.
The data came in a format somewhat similar to the USGS but
not in 12 x 31 yearly arrays (TABLE 5). The sequence of records
(Delimiter-station-data) shown in TABLE 5 is repeated until all
stations retrieved have been exhausted. An IBM "end-of-file"
is encountered following the last data record. FORTRAN program-
mers must code an "END-" control transfer in the read statement
for the data records to avoid a program interrupt at end-of-file.
The data were read from the tape, sorted, restructured, and
then written on the disk file in 12 x 31 water-year arrays. The
initial cost of the data is unknown since it was supplied by EPA.
Processing costs were moderate, that is, $200-300.
National Climatic Center (U.S. Weather Bureau) Data
Daily precipitation data used in the study came from the
National Climatic Center. Daily precipitation data are avail-
able for many locations in the United States and almost all DO
monitors were located within a few miles of these stations.
However, only about 300 hourly precipitation stations are avail-
able and many water quality monitors are too far away from pre-
cipitation stations to make hourly analysis meaningful.
The daily precipitation data are readily obtained from the
National Climatic Center, but as mentioned earlier, they are
very expensive. The numbers are stored on the tape as computer
card images. A standard computer card has 80 columns in which
numbers and symbols can be punched. The same format is used on
tape and is given in a publication from the U.S. Weather Bureau
(1009 Daily Observations 486). The card image and column-by-
column description are presented in Figure 1 and TABLE 6 , respec-
tively. Only the first 33 columns are shown in Figure 1 since
these are the ones used in this study.
WDNR Data
The WDNR is one of the few state agencies that collects and
stores its own DO data instead of reporting them to STORET.
This data source was found in the OWDC catalogs. Daily data
from 11 stations were ordered for the years 1972-1977. The data
were similar in format to the U.S. Weather Bureau data. Station
identification, date, DO level, water temperature, and flow were
31
-------
TABLE 5. RECORD ARRANGEMENT ON DISK FILE FOR STORET JV1ORE=4*
Group Description
1 Parameter heading records. Each record is 145 characters long. The first 25 characters of each
record may be ignored. The remaining 120 characters are a printable alphanumeric line. These
lines, in groups of four, would comprise the columns printed by the standard retrieval. With
MORE=4, there will be five groups of four lines each, a total of 20 parameter heading records.
2 Dummy data record. This is a delimiter record in Group 4 format, containing the print charac-
ters '99' in columns 26 and 27.
3 Station heading records. Each record is 145 characters long. The agency code is in column 1-8,
the station code is in column 9-23. Column 24 is blank. Column 25 contains a line number (1
through 9, in order). The remaining 120 columns of each record are a printable alphanumeric
line. There are nine such records in this group. These nine lines are normally printed at the top
of each page of retrieval output. The content is affected by use of the "SHIFT" and "HEAD"
parameters in the retrieval deck.
4 Data records. Each record covers one sample, either grab or composite, and its remarks codes.
Each record is 305 characters long. Agency code is in columns 1-8, station code is in columns
9-23, columns 24 and 25 contain '99', columns 26-31 contain the date as YYMMDD, columns
32-35 contain the time (0000-2400, with blank time stored as 2500). The next 200 columns
(36-235) contain the values of the water quality parameters retrieved, in IBM four-byte words.
The FORTRAN programmer must read this field as 50 A4, into an array whose type is REAL-
*4. Missing entries in this field, which occur either when fewer than 50 parameters are re-
trieved or when no data were stored for a particular parameter within the sample, are key-coded
with a binary number approximately equal to 1.E-20. The next 50 columns (236-285) contain
the STORET alphabetic remarks codes to go with the parameters retrieved. Column 286 will
be blank if this represents a grab sample, and the next four columns may be ignored. If this is
a composite sample, column 286 will contain "S", "T", or "B" for space, time, or both. Col-
umn 287 will be coded "A" for average, "L" for minimum, "H" for maximum. Columns 288-
293 contain a date in the form YYMMDO. This will be an initial date for comprehensive com-
posite samples, and a final date for a regular composite. Columns 294-297 are time for the same
point (0000-2400). Columns 298-299 are the number of samples comprising the composite.
Columns 300-340 are sample depth, and column 305 is coded "B" to indicate bottom sample,
and is otherwise blank.
5 Delimiter data record with year coded '99' in date area (columns 26-27).
6 Station heading records for next station, formatted as in Group 3.
7 Data records - See Group 4.
*This is the broadest type of retrieval possible, allowing for as many as 50 parameters and more than one
station.
Source: STORET Handbook Supplementary Manual, Volume 2, "Advanced Retrieval."
32
-------
UJ
IT
1-
fe
0 0
1 2
1 1
2 2
3 3
TATION
UMBER
ALPHA
ORDER
NUMBER
0000
3456
1111
2222
YR
0 0
7 E
1 1
2 2
3 3 3 3J3 3
4 4J4 4 4 4 4 4
5 5
6 S
7 7
8 8
9 9
1 2
5 5 5 515 5
6 S S 6
7777
8888
9999
6 6
7 7
8 8
9 9
345678
DATE
MO
0 0
9 10
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
9 1C
DAY
0 0
11 12
1 1
2 2
3 3
4 4
z
DIVISIO
0
13
1
2
3
4
5 515
e 6
7 7
8 8
9 9
n 12
5
7
8
9
13
TEMPERATURE ° F
MAX.
3
0 0 0
14 15 16
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
6 S 6
7 7 7
8 8 8
9 9 9
1* 15 16
MIN.
3
0 0 0
17 18 19
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
6 6 B
7 7 7
8 8 8
9 9 9
17 18 19
AT
TIME
OF OBS
3
0 0 0
20 21 22
1 1 1
2 2 2
3 3 3
1 4 4
5 5 5
6 6 6
7 7 7
8 8 3
9 9 9
PRECIP.
(INCHES)
E
0 00 0
23 24|25 26
1 1M 1
1
2 2\2 2
I
3 313 3
1
i
4 44 4
I
5 515 5
1
8 BJ6 6
1
7 717 7
i
8 8J8 8
I
9 919 9
20 21 22 23 2«|25 X
SNOW
24 HR
FALL
(IN.)
tl
0 00
27 28j29
1 111
1
2 2|2
i
3 313
1
i
4 44
!
5 515
1
i
6 61
l
7 717
1
8 8J8
i
3 919
n J»!M
DEPTH
ON
GRD.
(IN.)
H
0 0 0
30 31 32
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
6 6 6
7 7 7
8 8 8
9 9 9
30 31 32
DAY WITH
EC
33
S
S
0
z
34
13
S
<
SMOKE h
35
C/7
0
O
36
is
DRIZZLE
3!
S
jjj
38
2
0
39
S
CC
THUNDE
40
"
I
41
UJ
IDSTMJ
Z
I-'
D
Q
42
ft
o
BLWG Srv
43
UJ
z
HIGH Wll
44
iS
o
TORNAC
45
LJJ
*"
WB FORMS
1009, 1006 AND 1024
CLIMATOLOGICAL
RAINFALL & EVAPORATION
DAILY OBS
REVISED JULY 1,1949
33
34
36
5J|» M|W
41
4*
44
45
MAX
DEP
3
0 0 0
4« 47 4°
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
6 6 6
7 7 7
8 8 3
9 9 9
4647 '8
MAX
CHG
3
0 0 0
<9 50 51
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
6 6 6
7 7 7
8 8 8
9 9 9
MIN.
DEP.
3
0 0 0
52 53 54
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
6 6 6
7 7 7
8 8 8
9 9 9
49 50 51] \>2 53 54
MIN.
CHG.
3
0 0 0
55 56 57
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
6 6 6
7 7 7
8 8 8
9 9 9
K 55 51
cc
D
z i-
< <
LU CC
d 1
0 0 0
58 59 CO
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
S 6 8
7 7 7
8 8 8
9 9 9
« 59 «
Q
DEGRE1
I
0 0 0
6' 62 C-
1 1 1
2 2 1
3 3 3
4 4 4
5 5 5
6 6 6
7 7 7
8 8 6
9 9 9
51 R K
Q t-
z z —
- li/ w
£ 5 uj
0 0 0
i4 6? Sf
1 1 1
2 2 2
3 3 3
4 J 4
5 5 5
6 6 C
7 7 7
8 8 8
9 9 9
EVAPO 1
1 HUN- I
INCHES} f
u- o
<*§
00 0
S7I6? 5'
1| ! 1
1
l\l 2
1
313 o
1
i
44 4
1
515 5
1
6J6 6
.
717 7
I
88 8
1
919 9
64 SS 6E 67'[S8 69
PAN
WATER
TEMP.
X
<
C 0
70 ;i
1 1
2 2
3 3
4 4
5 5
6 G
7 7
3 8
9 9
70 71
z
0 [
72 :•
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
72 ?•
>£
0 z
WTR.E
SNOW 0
0 0 0
74 75 7
1 1 1
2 2 2
3 3 :
4 4 4
5 5 5
C 6 f
7 7 1
8 8 f
9 9 J
Jt 75 7
0 0 i t)
'? 7S 79 60
1111
2222
J)
m
3 3 3 3 1
O
C
4444 <
S
5 5 5 5 g
B 6 6 6
7777
8888
9999
)7 It -o 5i
Figure 1. Card format used by National Climatic Center for weather data.
-------
TABLE 6. CARD IMAGE CONTENT FOR NATIONAL CLIMATIC CENTER WEATHER DATA
Column Item or element
1-33 Missing data
35-76
1-2 State index
number
3-6 Station index
number
7-8 Year
9-10 Month
w 11-12 Day
** 13 Intrastate-
division and
time of
observation
14-16 Maximum
temperature
17-19 Minimum
temperature
Symbolic
. .. Card code
letter
B
x/
1-99
0001-
9999
00-99
01-12
01-31
0-9
X/
TXTXTX 000-199
X01-X99
X/
TnTnTn 000-099
X01-X99
Card code definition
Blank
X overpunch
State index number
Station index number
within state
Last 2 digits of year
January to December
Climate Division within
state 0= 10
X overpunch for a.m.
observations
Whole degrees F
X punch in column 14
indicates minus ( - )
X overpunch in column
16 indicates estimated
maximum temperature
Whole degrees F
X punch in column 17
indicates minus ( - )
Remarks
Columns blank if data missing or not
reported. See supplemental card content
for 1949 and prior years, page 5.
Assigned alphabetically. See list on page
6, maps pages 2 and 3.
Number assigned in proportion to its
relative position in "Index of Cities and
Towns" in Rand McNally Atlas, 65th
edition.
80-99 for 1880-1 899.
Indicates division number with p.m.
observation.
Indicates division number with a.m.
observations. X overpunch for a.m.
observations discontinued 1 July 1967.
For 24 hours ending at time of
observations.
Note: When temperature was rounded
to whole degrees (-0.4 F rounded =
0° F punched 000) (-0.5 = -1° F
punched X01).
For 24 hours ending at time of
observations.
See note columns 14-16.
(continued)
-------
TABLE 6 (continued).
Column
Item or element
Symbolic
letter
Card code
Card code definition
Remarks
17-19
(continued)
20-22
23-26
Temperature at
observation time
Precipitation
TTT
RRRR
OJ
27-29
Snowfall
includes
sleet
sss
X/
000-199
X01-X99
0001-
9999
OOOX
XBBB
OOXB
X/
001-999
XBB
X overpunch in column
19 indicates estimated
minimum temperature
Whole degrees F 0 in
column 20 = + value
X punch in column 20
indicates minus ( -)
00.01-99.99 inches to
hundredths
Trace
None, X in column 23
Amount included in
subsequent measurement
X overpunch in column 24
indicates accumulated value
00.1-99.9 inches to
tenths
X in column 27 no
snowfall
See note columns 14-16.
For 24 hours ending at time of
observation.
Less than 0.005 inches.
Blank in columns 24-26. Some
stations use 0000 for none.
Blank in column 26.
Note: Beginning Dec. 56 estimates of
daily precipitation were derived from
snowfall by the 10 to 1 ratio method
and identified by "X over 9" punch in
column 33. However, state climatologists
have been authorized to use X/9 punches
in column 33 to indicate estimates of
daily precipitation derived by other
methods.
Amount for 24 hours ending at
observation time.
Note: Hail included with snowfall from
July 1948 - December 1955. Hail
(continued)
-------
TABLE 6 (continued).
U)
CTl
Column Item or element
27-29
(continued)
30-32 Snow depth
on ground
includes
sleet and
ice
33 Estimated
precipitation
Symbolic
. Card code
letter
OXB
OOX
X/
sss 001-999
OOX
XXB
X/9
Card code definition
0 in column 27, X in
column 28 amount
included in subsequent
measurement
Trace
X overpunch in column
28 for 100.0-199.9
1-999 inches (whole
inches)
Trace
X in column 30 = none
When precipitation
amount in columns 23-26
estimated from snowfall by
Remarks
occurring alone was not included with
snowfall and snow on ground before
and after that period.
Less than .05 inch.
Depth of snow at time of observation .
See Note Columns 27-29.
Depth of less than 0.5 punched as
trace.
Blank in columns 31-32.
Columns 33 and 34 not punched for
"day with," effective 1 January 1949.
See note columns 23-26 above.
the 10 to 1 ratio
-------
in specified locations on the tape. The data were written in
ASCII characters, which were converted to integers and then to
floating point numbers as was the Weather Bureau data. These
were again stored on the permanent file in 12 x 31 water-year
arrays.
The tapes of data from WDNR were obtained at a reasonable
cost of about $60. Processing was similar to that of the Weather
Bureau.
ORSANCO Data
Data from ORSANCO were not used in this study primarily
because of difficulties related to the way it is stored for
computer processing and secondarily because the Ohio is a large
river and few of the urban areas along it have drainage areas
significant to that of the river itself. This section is in-
cluded because other investigators may wish to examine ORSANCO
data. The data format and the difficulties it causes are de-
scribed here.
Two magnetic tapes of data were obtained from ORSANCO. The
first tape contained the ORSANCO robot monitor data for 1961
through 1974. The second tape covered 1975 to July 31, 1977.
This includes hourly data from 28 of the 56 sites it had oper-
ated since beginning operation. The data were stored on tape
such that all stations reporting data on a particular day were
stored in one block. The next block would then contain all
hourly data at all stations for the next day. This is a con-
venient and efficient method of storing data if one wished to
see the water quality of the Ohio River basin on a day-by-day
basis. However, it is very cumbersome to obtain data on a
particular site for a year. A typical data block is presented
in TABLE 7. The structure of the data tapes is described in
Exhibit 1.
In TABLE 7, words 1 through 10 constitute the block header
and the last word of the block is its word count. These are
always present and in some cases represent the entire block.
This happens when no stations transmitted data for a day. There-
fore, block sizes of 11 words should be ignored.
This example illustrates a day (block) when only two sta-
tions transmitted data. Each station requires 240 contiguous
words (station segment), which are subdivided into 24 groups of
10 words reflecting 24 hours of data each day. Each word in a
group has a fixed definition only its value may change. This
example shows the first station segment occupying words 11
through 250 with the second station segment at words 251 through
490.
37
-------
TABLE 7. DATA BLOCK STRUCTURE, ORSAISICO DATA
Word Meaning - one word integers 1 & 2
1 Block number (day of year)
2 Type of block (always equals three)
3 Year of record
4 Month
5 Day of month
6 Storage date - day of year
7 Storage date - year
8 Number of words in this block
9 Zero
10 Zero
11 Station number (first station segment)
12 Hour
13 Test signals
14 Oxidation reduction potential
15 Temperature
16 Conductivity
17 pH
18 Solar radiation intensity
19 Dissolved oxygen
20 Chloride
21 Station number (same value as at word 11)
250 Chloride
251 Station number (second station segment)
252 Hour
253 Test signal
254 Oxidation reduction potential
255 Temperature
256 Conductivity
257 pH
258 Solar radiation intensity
259 Dissolved oxygen
260 Chloride
261 Station number (same value as at word 251)
490 Chloride
491 Number of words in this block
38
-------
Exhibit 1. Tape file structure description,
ORSANCO data.
Each block within a tape file constitutes one day of data for all stations having transmitted
data on this day. Each file contains as many blocks as there are days in the year (either 365
or 366).
Blocks are variable length ranging from 11 words (no stations transmitted just block header)
to a maximum of 6731 words (28 stations max.).
Block size is determined as follows:
SIZE = Header + (Numst * Segment) + Word Count,
where
Header
Numst
= first 10 words of the block and are always present,
= the number of stations that transmitted data on this day.
Segment = 240 contiguous words required to house a station-day (24
hours at 10 words per hour), and
Word Count = the last word of a block, containing the value of 'SIZE'. This
word is always present.
ORSANCO IBM 1130 magnetic tape routines write data by cal-
culating the lowest physical core address of the block to be
transmitted. Data are written into tape as a single block. The
contents of the lowest physical location are written first, with
bits 0-7 becoming the first byte on the tape and bits 8-15 be-
coming the second byte. Succeeding 16-bit words are transferred
in a similar fashion.
IBM 1130 arrays are arranged in core storage in descending
storage addresses. Element 1 of a five-element array will be at
storage address 508 and element 5 at byte 500. This is the re-
verse of 360 and 370 allocation and results in element 5 being
the first element transmitted to tape. Reading these data to an
IBM 360 or 370 array requires that the array be reversed in
storage.
Thus, unusual structure of the tape required the following
manipulation to recover data at a single station:
• initiate a read, which picked up the number of
words in a tape block;
backspace the tape
39
-------
• read the entire data block;
• search the input data buffer from the highest
element down (remember they are stored back-
wards) to discover the station of interest;
• transfer 24 values (one day per block) to a
storage array;
• translate the 24 values from ASCII to floating
point; and
• read the next block.
Backspacing is very inefficient, as are the formatted reads that
were required to recover the binary data. The ORSANCO tape
drives are old and did not produce a "clean" copy. Many tape
read errors were encountered. The cost of transferring three
months of hourly data to the disk pack for a single station was
$45. Simply figuring out how to read the tape cost over $600.
No one at ORSANCO could provide much guidance. Apparently, the
people who work the tape software have gone on to other jobs.
The people currently on board just use it and did not fully
understand it.
The quality of the data on the tape appeared very poor. No
flow information was available. For these reasons, the ORSANCO
data were abandoned.
This completes the discussion of the computer processing
necessary to transfer data from the various agencies to the disk
pack for use in the correlation analysis. The following section
describes the actual correlation programs and their development.
Data Retrieval and Correlation
The first step in the daily correlation analysis after the
required data were stored was to develop a suitable correlation
technique. This is not as straight forward as it might seem.
The objective of the correlation program development was to
build a program that was a sensitive indicator of the problem in
question. That is, it should give a strong numerical indication
whenever a site was found where the DO level dropped with in-
creasing flow. At the same time, it should not be costly to run
or difficult to implement. These goals were never entirely
achieved, but workable methods were obtained with experimenta-
tion.
Before undertaking program development, several test data
bases were created for Sutron's desktop calculator system.
40
-------
These test data consisted of a year's daily rainfall, tempera-
ture, and DO data from Wilson Creek near Springfield, Missouri.
Wilson Creek was selected as a standard because a very well
documented problem existed there. Over 80 percent of the Wilson
Creek drainage is from Springfield, Missouri. If a correlation
program could not clearly identify the problem there, it could
not be expected to work elsewhere.
The type of data typical of daily analysis is illustrated
in Figure 2. Illustrated are the flow and DO signals. They
have been plotted to illustrate the rather obvious correlation
of high flow and low DO at Wilson Creek. The problem was to
detect this mathematically.
O)
E
10-
X
o
o
LU
>
-J
O
00
00
r600
WILSON'S CREEK NEAR
SPRINGFIELD, MO.
WATER YEAR 1975
PERIODS OF LOW
DO AND HIGH
FLOW
Uoo
DISSOLVED OXYGEN
210 240
DAY OF YEAR
270
300
Figure 2. Typical daily flow and DO records illustrating
high degree of correlation between high flow and low DO
The first and most obvious correlation method attempted was
standard statistical cross-correlation analysis. This is de-
scribed in several textbooks such as Bendat and Piersol (8).
Initially developed in the communications industry, cross-corre-
lation is a means of determining how two such signals "look
alike" at different time offsets. A cross-correlation of 1.0
(for properly prepared signals) at zero time lag indicates per-
fect correlation or identical values at all times. A correlation
41
-------
of zero would mean no relationship. A correlation of 1.0 at an
offset of five hours would mean identical signals displaced in
time.
Cross-correlation proved unworkable for daily data. The DO
signal is more or less random in time with some long-term (low
frequency) components. The flow is an intermittent highly
skewed signal. Low correlations (0.6 or less) were obtained at
all reasonable time lags for the Wilson Creek data. This indi-
cated that other methods must be developed.
The next correlation method used was based on a comparison
of the DO levels on "wet" and "dry" days. Various definitions
of "wet" and "dry" were tried with both rainfall and flow data.
The different definitions of "wet" and "dry" are described as
follows. When correlating the DO sag with rainfall, a dry day
was when the daily precipitation was zero and a wet day was a
day with rainfall greater than zero. When correlating the DO
sag with flow, a dry day was when the discharge was less than a
given percentage of mean annual discharge and a wet day was when
the discharge was greater than the same percentage of mean annual
discharge. The percentage was varied from 50 percent to 400 per-
cent by 50 percent increments in order to determine the effect
of the wet/dry cutoff. A wet/dry cutoff equal to the mean
annual discharge was found to be a reasonable value.
The method was implemented on one year of data at a time.
First, all the DO deficits were summed for wet days and for dry
days. DO deficit is the difference between the actual DO level
and the saturation DO level for a given water temperature. The
number of wet and dry days was also computed. The average dry
DO sag was then calculated.by dividing the sum of all dry DO
sags by the number of dry days, and vice versa. The average
annual wet DO sag was then divided by the average annual dry DO
sag. If this ratio was greater than one, the DO sag was worse
during wet weather. This method proved sufficiently sensitive
to indicate a problem at Wilsons Creek and was retained for use
in the analysis. It was not considered the best method, however.
The final, and most sensitive correlation measure was de-
veloped by viewing the occurrence of a correlation in a proba-
bilistic sense. The "following questions were asked: (1) "Is it
more likely for the DO deficit to increase or decrease when the
flow changes or a rainfall event occurs?" (2) "How does this
likelihood compare during periods of low and high flow?" The
answers to these questions were found by developing a computer
counting scheme. Figure 3 illustrates what may be thought of as
a four-compartment box. The compartments are labeled A through
D. Numbers are assigned to the four compartments by graphing
DO deficit divided by a moving average deficit versus discharge
divided by average discharge. The averaging period was initially
not specified and various values were tried in order to find one
42
-------
"DRY"
-». "WET"
o
U.
ill
Q
ui
cc
LU
1.0
u.
LU
Q
CC
CC
O
0.0
0
© -
0 @.GL - ©
©
0
TENDENCY FOR DO DEFICIT
TO INCREASE AS FLOW
INCREASES
"WORSE"
'BETTER'
0.0 1.0
CURRENT DAY'S DISCHARGE/
MOVING AVERAGE DISCHARGE
Figure 3. Box or compartment method used to analyze daily
data for correlation between high flow and low DO.
most suited to this study.
mately chosen as the best.
The value of seven days was ulti-
The reasoning behind the graph is as follows. First, some
means was necessary to determine whether the DO deficit on any
given day was "better" or "worse" than on previous days. The
method selected compares the value observed "today" with the
average of the values which occurred for the previous seven days
Thus, a value of (deficit/moving average deficit) greater than
1.0 indicates "worse" and a value less than 1.0 indicates
"better." Points on Figure 3 that lie in the upper half (com-
partments A&B) represent days when the water quality was worse
than it had been for the previous seven days. Points that lie
below the line (compartments C&D) represent days on which the
water quality improved.
43
-------
Next, some means was required to determine if the water
quality consistently decreased on days with rainfall or increased
flow. Identifying days with rainfall was easy. Identifying
days with increased flow was again done by moving average. The
average daily flow "today" was compared with the average for the
previous seven days. A value of (flow/moving average flow)
greater than 1.0 indicates "increased" flow or "wet" and a value
less than 1.0 indicates "decreased" flow or "dry-" Points on
Figure 3 that lie to the left of center (compartments A&C) rep-
resent days on which the flow decreased. Points that lie to the
right of center (compartments B&D) represent days with increases.
The logic of Figure 3 can now be discussed. For any year
of DO and flow or rainfall data up to 358 points (365 - 7 used
for moving average) could be plotted on a graph. By counting
the number of points which fell in each of the four compartments
it was possible to compare the number of days when DO levels
decreased and flow increased (or rain fell) (compartment B) with
the number of days when DO levels increased and flow increased
(compartment D). If the contents of compartment B were signifi-
cantly greater than the contents of compartment D, then a corre-
lation was considered to be established between flow (or rainfall)
and decreased DO. A number of other comparisons can also be
made. These will be discussed later.
The implementation of the probability method was as follows.
First, a year of data was retrieved from the disk file. Next,
the seven-day moving average was computed beginning with the
first days data. The eighth day was then examined to see if the
flow had increased or decreased compared to the average of the
last seven days and whether the DO deficit increased or decreased
compared to the average for the last seven days. The answer
gave a point on the graph in Figure 3. The procedure was then
repeated with a new moving average beginning at the second day.
The number of points that fell in each quadrant of the graph
(that is, the compartments of the box) were counted as the
process was repeated. At the same time, the strength of the DO
deficit in each quadrant (compartment) was computed and summed.
This was used to determine how strong the deficits were under
high- and low-flow conditions. This was important because even
if a correlation between flow/rainfall and low DO existed, a
problem did not exist if the DO level did not fall significantly
below saturation. At the end of a year's data the number of days
in each of the following categories could be obtained by count-
ing the contents of the four compartments:
• wet days with worse than average deficit,
compartment B;
• wet days with better than average deficit,
compartment D;
44
-------
• dry days with worse than average deficit,
compartment C;
• dry days with better than average deficit,
compartment A;
• wet days with worse than average DO,
compartment B, using DO values in moving
average instead of DO deficit values;
• wet days with better than average DO,
compartment D, using DO values in moving
average instead of DO deficit values;
• dry days with worse than average DO,
compartment C, using DO values in moving
average instead of DO deficit values; and
• dry days with better than average DO,
compartment A, using DO values in moving
average instead of DO deficit values.
The probability of a worse-than-average deficit during wet
weather was determined by dividing the number of days of worse-
than-average deficit during wet weather by the total number of
wet days. Similarly, the probability of a worse-than-average
deficit during dry weather was determined by dividing the number
of days of worse-than-average deficit during dry weather by the
total number of dry days. In these calculations, wet weather
is increased flow or day with rain and dry weather is decreased
flow or day with no rain.
These two probabilities were compared to see if a worse-
than-average DO sag is more likely during wet (day with rain or
higher average flow) or dry weather. Similar probabilities were
also computed for the DO level itself. The average deficit and
percentage of saturation DO level was computed for each of the
possible situations. From this, the relative strength of the
sag during wet (day with rain or higher than average flow) and
dry weather and the relative percentage saturation level of DO
could be determined.
The probability method was tested on data from Wilsons
Creek near Springfield, Missouri. The method was sensitive
enough to indicate the known correlation at that site. A
greater than 70 percent probability of low DO at times of
higher than average (seven-day moving) flow and on days with
rainfall was found for several years of data.
45
-------
After determining that the probability method would identify
a site with a known correlation between flow or rainfall events
and DO deficit, it was necessary to set criteria for selecting
sites for more-detailed examination. Many sites exhibited cor-
relation between low flow and DO deficit as well as high flow
and DO deficit. Only those sites at which low DO could clearly
be identified with high flow were desired.
A cutoff level of 60 percent was finally selected. That is,
only sites at which the probability of a greater than seven-day
moving average DO deficit at times of greater than seven-day
moving average flow were chosen for further study. In an inde-
pendent review of this report, Meta Systems of Cambridge, Massa-
chusetts, verified the 60 percent cutoff level as being statis-
tically valid. A Chi-squared test was used to demonstrate that
a probability of greater than 60 percent or less than 40 percent
is required in order for nonrandom distributions to exist be-
tween two categories at the 95 percent confidence level. Thus,
for a given station year if a 60 percent probability of low DO
exists at times of high flow or on days with rainfall, then low
DO is significantly associated with these events.
Consideration was also given to the absolute DO level in
selecting sites. Even though a correlation existed between low
DO and flow or rainfall, a"problem" did not necessarily exist.
In general, sites were selected where DO levels less than 75
percent of saturation were present at times of high flow.
Results of Daily Correlation Analysis
'Presentation of the results of the daily analysis is com-
plicated by the volume of numbers involved. As mentioned in the
previous section on Site Selection, the potential locations to
be analyzed was 104. Three primary factors reduced the number
to 83. First, data could not be obtained for some sites. Second,
information in the OWDC catalog was not always accurate. Final-
ly, the list of 104 sites included a number of ORSANCO monitors
that were not analyzed because of the difficulty with the data.
The final list includes 55 USGS monitors, 17 STORET monitors
(recall that STORET contains data from numerous state and local
agencies), and 11 WDNR monitors. Periods of record range from
one to- five years. TABLE 8 lists those sites contained in the
original group of 104 that were not analyzed. The reason or
reasons for their rejection are also listed. Appendix B con-
tains the list of all monitors considered for analysis.
Daily rainfall correlations were completed at most of the
USGS monitor sites. The prohibitive expense of the National
Climatic Center data discouraged the completion of the rainfall
correlation at the STORET and WDNR sites.
46
-------
TABLE 8. ELLIG1BLE SITES NOT ANALYZED AND REASONS
State
Name of site
Data
base*
Reason(s) not included
Georgia
Illinois
Louisiana
Massachusetts
New Jersey
Ohio
Oregon
Pennsylvania
West Virginia
Peachtree Cr. near Atlanta
Ocmulgee R. near Warner-Robins
Calumet R.STW
Chicago R. Bridge
Chicago Sanitation & Ship Canal at Lockport
Houma iMav. Canal near Oulac
Connecticut R. at Agawam
Merrimac R. above Concord R. at Lowell
Passaic R. at Little Falls
Cuyahogu R. at Superior St. Bridge
Ohio R. at West End (Cincinnati)
Ohio R.at Andersons Ferry
Willamette R. at Portland
Willamette R. at Oregon City
Willamette R. above Oregon City
Allegheny R. at Oakmont
Beaver R. at Beaver Falls
Kiskiminetas R.at Vandergrift
Monongahela R. at S. Pittsburgh
Ohio R. at Huntington
GS
GS
G01
G01
G01
GS
GS
GS
GS
GS
R02
R02
EPA
EPA
EPA
R02
R02
R02
R02
R02
Unable to locate records in archive files
Unable to locate records in archive files
Chicago Sanitation District would not provide data at reasonable cost
Chicago Sanitation District would not provide data at reasonable cost
Chicago Sanitation District would not provide data at reasonable cost
Unable to obtain rain or flow data - marginal site
Unable to obtain flow or rainfall for correlation
Unable to obtain flow or rainfall for correlation
Unable to locate records in archive files
Unable to locate records in archive files
ORSANCO site - data processing cost too high
ORSANCO site - data processing cost too high
Data not found in STORET files
Data not found in STORET files
Data not found in STORET files
ORSANCO site - data processing cost too high
ORSANCO site - data processing cost too high
ORSANCO site - data processing cost too high
ORSANCO site - data processing cost too high
ORSANCO site - data processing cost too high
"GS = U.S. Geological Survey, G01 = Metropolitan Sanitation District of Greater Chicago, R02 = ORSANCO, and EPA = STORET.
-------
The results of the daily correlation study for all the
sites examined are presented in Appendix C. TABLE 9 lists those
USGS, STORET, and WDNR monitors that were considered for hourly
examination. As mentioned previously, these are sites for which
the probability of lower than average dissolved oxygen reached
60 percent on days with higher than average flow or rainfall.
For convenience in identifying the sites with strong cor-
relations, TABLE 9 lists the names with no accompanying data.
Listed first in TABLE 9 are those USGS sites that exhibited a
strong correlation (60 percent or greater) between higher than
average flow and low DO. Listed second are those STORET moni-
tors that exhibited a strong correlation with flow. The WDNR
monitors with strong flow correlation are listed next. The
final group of stations in TABLE 9 are those USGS sites that
exhibited a strong correlation (60 percent or greater) between
days with rainfall and periods of lower than average DO.
In TABLE 9 some stations, such as the Mad River at Dayton,
exhibited strong correlations between both flow and rainfall and
low DO. Such stations appear tv/ice in the listing.
TABLES 10 and 11 provide daily correlation analysis data
for the stations listed in TABLE 9. The information in TABLES
10 and 11 is taken from Appendix C. Data are presented only for
those years that exhibited a 60 percent probability of low DO at
times of higher than average flow (TABLE 10) or days with rain-
fall (TABLE 11). These data are included here for the conveni-
ence of the reader in determining why the sites were chosen for
possible hourly analysis. Three types of statistics are in-
cluded in each table: the probability of a worse-than-average
DO deficit, the strength of the deficit, and the average per-
centage saturation occurring when the DO deficit is worse than
average. The three types of statistics answer the following
six questions about a given monitor site:
• What is the probability at this station that a
worse-than-average (seven-day moving) DO deficit
will occur on a wetter-than-average (seven-day
moving) day?
• What is the probability at this station that a
worse-than-average (seven-day moving) DO deficit
will occur on a dryer-than-average (seven-day
moving) day?
• If the DO deficit on a particular wetter-than-
average day is worse than average, how much
worse is it?
48
-------
TABLE 9. MONITOR SITES EXHIBITING A 60 PERCENT OR
GREATER PROBABILITY OF LOW DO DURING HIGH FLOW
OR PERIODS OF RAINFALL
Agency
Correlation
Monitor sites
USGS
Flow
EPA
(STORED
WDNR
USGS
Flow
Flow
Rainfall
North Nashua R. near Lancaster (Leominster), MA.
Westfield R. at Westfield, MA.
Wilsons Cr. near Springfield, MO.
Delaware R. at Trenton, I\IJ.
Manasquan R. near Squankum, NJ. (later rejected for lack
of urban area)
Raritan R. near South Bound Brook, NJ.
Ashtabula R. at Ashtabula, OH.
Hocking R. below Athens, OH.
Little Miami R. at Miamiville (Milford), OH.
Little Miami R. near Spring Valley, OH.
Mad R. near Dayton, OH.
Maumee R. at Defiance, OH.
Portage R. at Woodville, OH.
Sandusky R. near Upper Sandusky, OH.
Scioto R. at Chillicothe, OH.
S. Umpqua R. near Brockway, OR
Delaware R. at Bristol, PA.
Delaware R. at Chester, PA.
Trinity R. below Dallas, TX.
S. Platte R. at Denver, CO.,
Burlington Ditch at York St., MDSDD #1
S. Platte R. at Denver, CO.,
S. Platte R. at 60th ave., MDSDD #1
Wisconsin R. at Biron, Wl.
Wisconsin R. at Dubay Dam, Wl.
Wisconsin R. at Petenwell, Wl.
Connecticut R. at W. Springfield, MA.
Westfield R. at Westfield, MA.
Center Cr. near Carterville, MO.
James R. near Boaz, MO.
Wilsons Cr. near Battlefield, MO.
Wilsons Cf. near Springfield, MO.
Delaware R. at Trenton, NJ.
Ashtabula R. at Ashtabula, OH.
Blanchard R. near Findlay, OH.
Cuyahoga R. at Independence, OH.
Cuyahoga R. at Old Portage, OH.
Grand R. at Painesville, OH.
Little Miami R. near Spring Valley, OH.
Mad R. near Dayton, OH.
Mahoning R. at OH.-PA. State Line, OH.
Sandusky R. near Upper Sandusky, OH.
Scioto R. at Chillicothe, OH.
Lehigh R. at Easton. PA.
49
-------
TABLE 10. MONITOR SITES WITH 60 PERCENT OR GREATER PROBABILITY
OF LOW DO AT TIMES OF HIGH FLOW
Strength of DO deficit
Monitor site
by state and agency*
OHIO (USGS)
Ashtabula R. at Ashtabula
Hocking R. below Athens
Little Miami R. at Miami-
ville (Milford)
Little Miami R. near Spring
Valley
Mad R. near Dayton
tn
0 Maurnee R. at Defiance
Portage R. at Woodville
Sandusky R. near Upper
Sandusky
Water
Year
(19-)
75
76
73
74
74
75
72
73
75
73
75
73
76
Probability
than average
Wetter than
avg. days
.63
.67
.66
.67
.65
.67
.63
.66
.60
.60
.68
.60
.64
of greater
DO deficit
Dryer than
avg. days
.49
.45
.40
.37
.43
.45
.47
.41
.38
.44
.40
.46
.46
when deficit
is greater
than average
Wetter than
avg. days
1.23
1.44
1.61
1.56
1.33
1.18
1.25
1.37
7.64
1.43
1.76
1.58
1.34
Dryer than
avg. days
1.21
1.40
2.01
1.42
1.18
1.14
1.13
1.29
1.77
1.75
1.72
1.86
1.27
Average
percentage of
saturation when deficit-
is worse
Wetter than
avg. days
.64
.84
.83
.84
.73
.72
.71
.72
.77
.56
.77
.83
.71
than average
Dryer than
avg. days
.72
.86
.85
.86
.74
.76
.71
.67
.76
.62
.83
.76
.74
Scioto R. at Chillicothe 72
MASSACHUSETTS (USGS)
North Nashau R. near 70
Lancaster (Leominster)
Westfield R. at Westfield 75
.64
.63
.62
.48
.50
.48
1.21
1.17
1.69
1.15
1.17
2.65
.41
.53
.78
.51
.62
.88
*Stations are grouped by state, states are in alphabetical order, stations are alphabetized within groups.
(continued)
-------
TABLE 10 (continued).
Ln
Strength of DO deficit
Monitor site
by state and agency
MISSOURI (USGS)
Wilsons Cr. near
Springfield
NEW JERSEY (USGS)
Delaware R. at Trenton
Manasquan R. at Squankum
(This station later aban-
Water
Year
(19--)
73
75
76
73
75
71
72
Probability
than average
Wetter than
avg. days
.70
.81
.77
.64
.61
.61
.66
of greater
DO deficit
Dryer than
avg. days
.42
.39
.44
.45
.40
.58
.52
when deficit
is greater
than average
Wetter than
avg. days
1.27
1.27
1.28
1.83
2.79
1.12
1.22
Dryer than
avg. days
1.20
1.16
1.14
6.66
3.73
1.13
1.11
Average
percentage of
saturation when deficit
is worse
Wetter than
avg. days
.60
.56
.58
.85
.91
.51
.50
than average
Dryer than
avg. days
.64
.69
.65
.93
.96
.55
.56
doned for lack of urban
area; it is near a swamp.)
Raritan R. near South 75
Bound Brook (below
Callo Dam at Boundbrook)
OREGON (USGS)
Umpqua R. near 76
Brockway
PENNSYLVANIA (USGS)
Delaware R. at Bristol 75
Delaware R. at Chester 73
TEXAS (USGS)
Trinity R. below Dallas 77
.60
.63
.71
.67
.60
.37
1.60
.55
.43
.65
.52
3.82
1.33
1.13
1.13
1.65
2.30
1.24
1.61
1.14
.71
.82
.84
.36
.28
.80
.83
.84
.42
.26
(continued)
-------
TABLE 10 (continued).
Monitor site
by state and agency
Water
Year
(19-)
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
Strength of DO deficit
when deficit is greater
than average
Wetter than Dryer than
avg. days avg. days
Average percentage of
saturation when deficit
is worse than average
Wetter than Dryer than
avg. days avg. days
COLORADO (EPA)
S. Platte R. at Denver, 71
Burl. Ditch at York St.
MDSDD #1
S. Platte at Denver, S. 76
Platte at #60 Ave,
MDSDD #1
WISCONSIN (WDNR)
Wisconsin R. at Biron 74
Wisconsin R. at Dubay 74
i Dam 76
Wisconsin R. at Petenwell 73
.60
.66
.64
.60
.64
.60
.44
.44
1.35
1.21
.46
.52
.50
.40
1.48
1.08
1.08
1.20
1.30
1.28
1.13
1.11
1.13
1.16
.72
.61
.60
.36
.36
.56
.72
.70
.62
.44
.44
.62
-------
TABLE 11. USGS MONITOR SITES WITH 60 PERCENT OR GREATER PROBABILITY
OF LOW DO ON DAYS WITH RAINFALL
Monitor site
MASSACHUSETTS
Connecticut R. at W.
Springfield
(Thompsonvillo)
Westfiekl R. at Westfield
MISSOURI
Center Cr. near Carterville
James R. near Boa?
Wilsons Cr. near
Battlefield
Wilsons Cr. near
Springfield
NEW JERSEY
Delaware R. at Trenton
OHIO
Ash tabula R. at Ashtabula
Water
Year
(19--)
73
74
75
75
74
75
73
74
75
76
74
75
76
75
76
72
75
76
Probability of greater
than average DO deficit
Wetter than
avg. days
.60
.66
.63
.63
.64
.65
.64
.64
.65
.63
.63
.62
.63
.67
.77
.62
.63
.67
Dryer than
avg. days
.52
.47
.47
.45
.49
.50
.47
.54
.53
.52
.56
.53
.52
.38
.43
.46
.45
.42
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.67
1.32
1.26
2.13
1.14
1.14
1.14
1.13
1.14
1.11
1.08
1.12
1.06
1.26
1.24
1.79
1,24
1.22
Dryer than
avg. days
2.98
1.21
1.23
2.46
1.10
1.15
1.12
1.13
1.09
1.11
1.10
1.08
1.05
1.15
1.16
5.35
1.18
1.24
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.80
.78
.81
.81
.59
.54
.57
.49
.44
.30
.09
.14
.03
.61
.61
.86
.70
.69
Dryer than
avg. days
.76
.81
.83
.87
.62
.54
.58
.48
.45
.28
.10
.14
.04
.68
.64
.87
.68
.68
"Stations are grouped by state, states are in alphabetical order, stations are alphabetized within groups.
(continued)
-------
TABLE 11 (continued).
Monitor site
Water
Year
(19--)
Probability of greater
than average DO deficit
Strength of DO deficit
when deficit is greater
than average
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
Dryer than
avg. days
Wetter than
avg. days
Dryer than
avg. days
Wetter than
avg. days
Dryer than
avg. days
OHIO (continued)
Blanchard R. near Findlay
Cuyahoga R. at
Independence
Cuyahoga P.. at Old
Portage
Grand R. near
Painesviiie
Little Miami R. near
Spring Valley
Mad R. near Dayton
Mahoning R. at OH.-PA.
State Line below
Loweliville
Sandusky R. near Upper
Sandusky
Scioto R. at Chillicothe
73
75
76
73
76
75
74
75
72
73
74
75
76
76
74
75
76
76
.65
.62
.64
.60
.66
.60
.60
.61
.66
.64
.60
.63
.65
.60
.62
.62
.69
.60
.48
.44
.49
.41
.44
.59
.43
.43
.43
.34
.45
.41
.45
.44
.45
.41
.42
.39
1.29
1.17
1.22
1.25
1.27
1.19
1.27
1.17
1.20
1.26
1.17
1.20
1.21
1.18
1.77
1.38
1.40
2.15
1.20
1.16
1.19
1.22
1.19
1.16
1.21
1.14
1.14
1.14
1.13
1.20
1.15
1.14
2.62
1.21
1.31
.65
.50
.57
.72
.67
.72
.73
.74
.72
.78
.73
.74
.68
.54
.71
.70
.71
.54
.58
.55
.59
.70
.75
.66
.74
.75
.70
.78
.72
.75
.70
,52
.70
.74
.75
.58
(continued)
-------
TABLE 11 (continued).
Strength of DO deficit
Monitor site
PENNSYLVANIA
Lehigh R, at Easton
(Glendon)
Water
Year
(19-)
72
73
Probability
than average
Wetter than
avg. days
.67
.62
of greater
DO deficit
Dryer than
avg. days
.51
.45
when deficit
is greater
than average
Wetter than
avg. days
1.35
1.50
Dryer than
avg. days
1.56
1.60
Average
percentage of
saturation when deficit
is worse
Wetter than
avg. days
.77
.83
than average
Dryer than
avg. days
.76
.82
L71
Ln
-------
• If the DO deficit on a particular dryer-than-
average day is worse than average, how much
worse is it?
• On wetter-than-average days when the DO deficit
is worse than average, what percentage of satu-
ration is present?
• On dryer-than-average days when the DO deficit
is worse than average, what percentage of satur-
ation is present?
The first and third questions were of primary concern to
this study. If the answer to the first was "yes" and the answer
to the third was "less than 75 percent saturation," then the
site in question was felt to be worth closer examination.
The results of the analysis to examine the DO deficit levels
for days on which the flow equaled various percentages of the
mean annual value are not presented here. In general, they were
used to augment or confirm judgments made from information in
Appendix C.
The following section describes the more-detailed analyses
that were performed at the sites that exhibited strong correla-
tions. The detailed analysis was designed to determine which
of the sites with strong correlations also had water quality
standard violations.
56
-------
SECTION 7
DETAILED SITE ANALYSIS
GENERAL CONSIDERATIONS
Water quality monitor sites where correlations existed be-
tween rainfall and/or flow and low DO were identified in Sec-
tion 6. The daily data used to identify these sites did not
allow a determination of specific water quality standard viola-
tions. Hourly data analysis was required at sites that showed
strong deficits (DO levels less than 75 percent saturation).
For these reasons and because of other contract requirements,
it was necessary to examine those sites with correlations in
greater detail.
The contract requirement that required the most additional
information was the Streeter-Phelps analysis. This is a pro-
cedure used to calculate the location of the maximum DO deficit
downstream of point or distributed sources of oxygen-demanding
material. The purposes of such analysis were to determine if a
given water quality monitor was in a position to sense the maxi-
mum deficit from urban runoff and if observed deficits appeared
reasonable based on the best estimates of input loads. Urban
population, drainage area, streamflow, temperature, BOD loads,
and other data were required.
In addition to selecting sites for hourly and Streeter-
Phelps analysis, it was desired to form some opinion on why a
correlation might exist between flow/rainfall and low DO at a
given site. Information was needed on travel time from the
urban area to the monitor, distance from the rain gage to the
urban area, and other factors that might influence the DO that
a monitor "sees."
The overall process of the detailed site analysis studies
will now be described. Specific aspects of the analysis, such
as Streeter-Phelps, will be singled out and discussed in detail.
These detailed discussions occupy the remainder of the section.
The first step in the detailed site analysis was the gath-
ering of available maps. The monitor and flow gage and the
weather station, if any, were then located precisely. An in-
formation sheet that summarized what could be learned from USGS
57
-------
records, the map, or other sources was next filled out. Infor-
mation sheet data were used to select sites with the simplest
hydraulic conditions for analysis by the Streeter-Phelps tech-
nique. Finally, the plots of daily average flow, DO, and rain-
fall values were examined and periods for hourly analysis were
selected. From 10 to 30 days or more of hourly data were
examined at each site where the hourly data could be obtained.
The total of all this information was then studied carefully to
see what could be concluded. The results form the substance of
Section 8, which follows.
GENERALIZED INFORMATION ANALYSES
The gathering of. information; at each site was systematized
by developing a standard form. The two-page form, filled out
for,the- Scioto River at Chillicothe, Ohio, monitor is illustrated
in-Exhibits 2 and 3.
It was originally intended that completed forms for each
of the sites in TABLES 9, 10, and 11 would be included as an
appendix to this report.. However, experience indicated that
much of the information was unavailable. For this reason, each
site will be discussed individually in the appendix and as much
general information as possible tabulated there.
In order to keep the main body of the text to a reasonable
size, the remainder of the discussion^on the completion and use
of the site .analysis forms will be structured around the Scioto
River at Chillicothe. This will serve as an example of the
method used. 'The reader is referred to Appendix D, where the
remaining sites, as well as the Scioto, are presented.
The first step in filling out the analysis forms was usually
to assemble USGS 7.5-minute topographic sheets to form a complete
picture of the particular urban area in question. These ranged
in size from a single sheet to as many as nine sheets for the
Philadelphia-Trenton area. Next, the water quality monitor, the
stream gage, and the precipitation station were located. This
was done by plotting the reported latitude and longitude of the
facilities. This had been done in the early phases of the study
to select monitors for inclusion in the daily analysis. Many of
the maps used, however, were in the USGS library. More maps
were purchased and brought to Sutron for the detailed analysis
work.
The assembled maps were large and cumbersome. Therefore, a
sketch of relevant features was transferred to the site form.
This aided in remembering each one later. Any outstanding fea-
tures such as industry or sewage treatment plants were noted.
Other information gathered from the maps included the urban
drainage area, stream width, the distance from monitor to stream
gage, and distances from the monitor to sewage outfalls. The
58
-------
Exhibit 2. Site analysis form - Streeter-Phelps.
MONITOR NAME: Scioto R.at Chillicoths, OH.
AGENCY: US6S
MONITOR LOCATION: Lat.: 392029 Long.: 825816
AGENCY ID NO.: 03231500
SKETCH OF SITE:
POP. OF COLUMBUS
566,600 11,069,400]
PHYSICAL DESCRIPTION: 43 milas south of Columbus, OH.
HILLICOTHE FED.
REFORMATORY
OUTSTANDING FEATURES: Sewage disposal downstream of monitor, no obvious problems
STREAM GAGE NAME: Scioto R.at Chitlicothe, OH.
AGENCY: USGS
STREAM GAGE LOCATION: Lat.: 392029 Long.: S25S16
AGENCY ID NO.: 03231500
PRECIPITATION GAGE NAME: Chillicoths - Mound City
NWS ID NO.: 1528
DISTANCE-MONITOR TO STREAM GAGE: 0
DISTANCE - RAIN GAGE TO MONITOR: ~2 mi
QUALITY OF RECORDS: Looks good most of the time
DESCRIPTION OF STREAM:
APROX. WIDTH: 250ft AVG. DISCHG.: ^3400 cfs
RANGE OF DISCHG.: 500-50,000 cfs
APROX. DEPTH AT AVG. Q.:
APROX. MEAN VEL. AT AVG. Q.:
OTHER INFORMATION:
59
-------
Exhibit 3. Site analysis form.
MONITOR NAME: Scioto R. at Chillicothe, OH.
TOTAL DRAINAGE AREA AT MONITOR: 3850 TOTAL URBAN DRAINAGE: 3-4 mi2
% URBAN CONTRIBUTING AT MONITOR: 50% % CONT. URB./TOTAL: <1%
APROX. POPULATION OF URBAN AREA: ___^_
RESULTS OF CORRELATION ANALYSIS:
DAILY FLOW - COMMENT ON PROBABILITY OF OCCURRENCE: .64-1972
COMMENT ON STRENGTH OF SAG: 1.21 wet, 1.15 dry -1972 / .41% wet, .51 dry -1972
DAILY RAIN - COMMENT ON PROBABILITY OF OCCURRENCE: .6-1976
COMMENT ON STRENGTH OF SAG: 2.15 wet, 1.31 dry -1976 / 54% wet, 58% dry - 1976
STREETER-PHELPS ANALYSIS PARAMETERS
STREAM Q, cfs: WASTE DISCHG., cfs: STM. VEL., ft/sec.:
DEPTH,ft.: INITIAL DEFICIT, mg/l:
DEFICIT INFLOW,, no./day: BOD INFLOW,no^day:
DIST. BOD SOURCE, mg/l/day: UPSTRM. BOD IN RIVER, mg/l: _
DIST. BENTHAL DMND., g/m2/day: REAERATiON COEF., Ka, I/day:
BOD DECOY COEF., Kr, I/day: DEOXYGENATION COEF., Kd, I/day:
STREETER-PHELPS RESULTS
COMMENT ON MONITOR LOCATION:
COMMENT ON STREAM MILES AFFECTED:
CONFIDENCE IN RESULTS:
ANALYSIS OF PROBLEM AT THIS SITE:
Chillicothe is 44 mi downstream of Columbus. This could be near the Columbus sag point - bad water quality
probably originates at Columbus.
60
-------
urban area was estimated as accurately as possible from the rose-
tinted areas on the maps. Distances were measured by map scale
and dividers.
After tabulating the physical features of the site, the
stream flow was examined. The flow and DO records were gath-
ered and the range of values entered. The plots of DO and dis-
charge were examined to determine the percentage of missing
records. This examination was combined with the quality esti-
mates in the agency reports to rate the record. Records were
"excellent" if only one or two weeks a year were missing. A
"good" rating went to stations with one or two months missing.
A "poor" rating went to all others.
While examining the flow records, typical July-August low
flow values were selected. These were for later use in the
Streeter-Phelps analysis (if performed). If rating curves were
available, they were used to determine the depth and velocity,
which were also required for Streeter-Phelps.
The results of the daily correlation analysis were entered
on the form in a summarized fashion. This was done to give at
a glance the magnitude of the "problem" at each site. This was
omitted in cases where there was too much information to fit.
A note was entered referring to a separate summary table.
The population figures were obtained from a World Almanac
from the 1970 census. No attempt was made to account for growth
increases. The maps were often more out of date than the census.
For consistancy, no adjustments were made to "urban" areas
either.
The Streeter-Phelps portion of the site forms was completed
only if the analysis was actually performed. Development of
complex water quality models was not within the Scope of Work.
For this reason, sites with fairly simple hydraulic conditions
were selected for analysis. Meaningful analysis using the
Streeter-Phelps method required streams with few tributaries
and not too many waste sources. Five sites were required by the
contract. Thirteen were done because the writers felt that it
contributed a worthwhile insight into the location of monitors.
The actual Streeter-Phelps procedure is complex enough to warrant
separate discussions. It is the subject of the following sec-
tion.
Streeter-Phelps Analysis
The bulk of the "Streeter-Phelps" theory used in this study
was derived from Thomann's (2) excellent book on water quality
management. Virtually, all of Section 5 of the book is directed
toward one-dimens.ional modeling of stream quality., Streeter-
Phelps is a subset of the methods covered,
61
-------
LQ
The term "Streeter-Phelps" is somewhat misleading. The
original work of Streeter and Phelps in 1925 was directed specif-
ically at determining the DO deficit versus distance caused by a
point source of carbonaceous BOD (CBOD) , LO , and some initial DO
deficit, D0. In the intervening years, other researchers have
added additional capabilities to the original model. These in-
clude nitrogenous BOD (NBOD) , photosynthesis , respiration, ben-
thai oxygen demand, and distributed sources of CBOD and NBOD.
The "Streeter-Phelps" equations used in this study were
selected primarily on the basis of data available to set the
various parameters. In all cases, the data were very limited.
It was necessary to work from good estimates of the urban area,
the population, and the data obtained for the study, that is,
flow, temperature and DO levels. No CBOD, NBOD, photosynthesis,
respiration, or rate constant information was available. In the
absence of these key numbers, it was decided to be consistent
with the technique, even though accuracy was out of the question.
The equation chosen for use is as follows:
D = [(wd/Q) + DQ] exp [-Ka(X/V>]
V(Kd - Kr) |eXP [-VX/V>] - exp [-Ka(X/V)]})
(Kd/(KaKr) j1 - exP [-Ka(X/V)]})Lrel
(Ka - Kr) HP -VX/V) -exp -K& (X/V)
Ka
where D = DO deficit (milligrams per liter) ,
Wd/Q = weight of initial DO deficit divided by river
discharge (Wd is expressed in pounds, Q in
cubic feet per second, and the quotient con-
verted to milligrams per liter) ,
exp = exponential function,
,_ •]
Ka = reaeration rate constant (days ) ,
X = distance from initial source (miles) ,
V = mean stream velocity (miles per day) ,
K^ = CBOD oxidation rate coefficient (days " ) ,
Kr = CBOD settling plus oxidation rate coefficient
(days -1) ,
L0 = initial point source CBOD (milligrams per liter) ,
Lrd = distributed CBOD source (milligrams per liter
per day) , and
Sb = oxygen uptake of benthic deposits (milligrams
per liter per day) .
62
-------
The initial value of CBOD, LQ, is found by a mass balance
at X = 0 as
L0 = (W + LUQ)/Q + Qw
where W = weight of CBOD input (pounds),
Lu = upstream CBOD (pounds), and
Qw = waste discharge rate (cubic ft per sec).
The quotient is converted to milligrams per liter.
Benthal demand, S]-,, is normally reported as grams oxygen per
meter squared per day. It is converted to milligrams per liter
per day by dividing by the hydraulic radius of the channel.
Photosynthesis/respiration effects were not considered.
Likewise NBOD, both point source and distributed, were not con-
sidered .
Before discussing the means used to evaluate the many co-
efficients and parameters in the equations, it is worth discus-
sing the method in general. There are a number of reasons to be
skeptical of the outcome from the start. First, the equation is
one-dimensional in nature. All waste sources and other param-
eter values are assumed to be uniformly distributed across the
length and breadth of the stream. Experience has shown this to
be a fairly reasonable assumption for the case in which lateral
mixing is rapid or in which waste sources are deliberately in-
troduced through diffusers placed across the channel. Second,
the equations were designed for steady state conditions. That
is, the waste discharges, the stream flow, and all the rate
controls are time invariant. Assuming such condition to exist
under the influence of storm runoff is highly suspect. Dis-
charge increases of factors of ten are common in small streams
under summer storm runoff conditions. Peak flows seldom last
an entire day- Often storm hydrographs last less than one day.
Finally, as mentioned earlier, virtually no data are available
on the various rate constants. Empirical formulas are available,
but seldom specific measurements.
The writers would have preferred to see the "Streeter-
Phelps" analysis replaced with a coupled unsteady-flow/water-
quality model. This would have been much more realistic and is
still strongly recommended. However, this would require a con-
centrated data collection effort at one site. Recommendations
on suitable places will be made later.
Since application of the Streeter-Phelps methodology was a
requirement in the Scope of Work, every effort was made to
implement it in a consistent and reasonable way. Although it
was not within the Scope of Work to develop water quality models,
the first step was to develop a flexible Streeter-Phelps solu-
tion algorithm. Direct solution of the equation confined the
63
-------
answer to a single point source of CBOD as well as single iden-
tifiable point sources. In order to handle these situations, a
model was developed capable of superimposing solutions at vari-
ous distances. A starting position, X=0, was specified for each
point source desired, and the program kept track of each and
added them all together to form an overall DO deficit profile
for the reach in question. Take for example a situation in which
three sewage treatment plants are present in a city seven miles
long and a reach length of 70 miles was desired. The procedure
is as follows: First, assign X = 0 to the upstream edge of the
city and solve the equation for a distributed demand over the
seven miles of city. Store this result. Next, solve the equation
for a point source of CBOD at the location of the first treat-
ment plant. Add this to the first solution with the origin off-
set appropriately to reflect the plant location. Finally, do the
same for the other two plants.
In practice, the scheme was implemented on. a large desktop
calculator with a plotter. The "Streeter-Phelps" equation was
solved at one-quarter-mile intervals for whatever total reach
length was specified. A large one-dimensional matrix saved the
solutions and accumulated them for plotting. The calculator
handled all the offset calculations automatically. Examples of
the results will be shown after discussing parameter estimation.
The following is a list of the parameters that had to be
estimated for the Streeter-Phelps technique:
® Flow parameters: stream distance
discharge
depth
flow velocity
® Rate constants: Ka, Kr, Kd
• Demand and loading constants: D.L.Q.L,, S,,W.
o o w rd b
As mentioned previously, the hope for accuracy of the tech-
nique was not great. Consistency was sought, however, so that
relative results could be compared. Thus, each site considered
for "Streeter-Phelps" analysis was treated in exactly the same
way.
The flow parameters were the easiest to estimate. Even
here, however, detail was lacking. Stream distances were meas-
ured by picking a starting point for X = 0. This was usually
the upstream edge of the urban area as indicated by the rose-
colored area on the 7.5-minute quad sheets. In some instances,
it was taken as the outfall of the farthest upstream treatment
plant. Stationing in miles was then established by using
64
-------
dividers set at one-quarter-mile spacing proceeding down the
channel centerline.
When very large distances (10 miles or more) were involved,
atlases or 1:25,000 scale maps were used. In the absence of
suitable maps, channel length was estimated using a standard
geomorphology formula (Leopold et al. (9)). This states that
channel length, L, is proportional to drainage area to the 1.4
power. The USGS publishes drainage areas at all gages and the
distance between each could be found as
1.4 1.4
AL = A, -A,
dl d2
Values obtained by this method were always checked for reason-
ableness by measuring the straight-line distance and assuring
that AL was neither shorter nor greater than a factor of five
longer.
The magnitude of discharge to use was an important consid-
eration. The final decision was to use two separate values for
each site. First, a typical July-August low flow was picked by
eye from a year when a correlation existed between discharge or
rainfall and low DO. This was done by looking at the plots of
mean daily discharge and selecting the low point. Next, a
typical July-August storm flow was selected. This was also
chosen of the mean daily value plots.
The reasoning behind using two flows was based on the
Streeter-Phelps limitation of steady flow. It seemed unwarrant-
ed to merely change the loading into the low flow conditions to
simulate storm events when it was known that storm flows were
often 10 times or more greater than low flow. By using realis-
tic low flow loadings in a low flow and realistic storm loadings
in a storm flow, it was hoped that reasonable results would be
obtained.
After selecting the two discharges, it was necessary to
estimate the depth and flow velocity. Readers familiar with
river mechanics can appreciate the difficulty of assigning a
single depth and velocity to reaches as long as 60 miles. As
stated earlier, the objective was not to reinvent the water
quality model, so reach-by-reach depths were not entered. If a
rating curve was available, the depth and area were taken from
that. If no rating curve was available, the widely used Manning's
equation for discharge,
Q = (1.486/n) AR2/3S01/2 ,
was used to estimate depth. Here, n = channel roughness param-
eter, taken to be 0.035 (typical of many open channels); A =
cross-sectional area, in square feet; R = hydraulic radius
(approximately equal to depth for wide channels); and So = the
65
-------
energy slope (bed slope for steady flow in a uniform rectangular
channel). By assuming a rectangular channel, R = depth, and
taking So = bed slope,
Q = (1.486/n) WY(Y2/3)Sb1/2
= (1.486/n) WY5/3Sb1/2 .
By estimating the width, W, and the slope, Sb, from maps, the
depth was easily computed. Once the depth was known, the veloc-
ity was also easily computed.
The rate coefficients were all computed from standard for-
mulas or typical values reported in the literature. The reaera-
tion rate coefficient, Ka, was extensively studied by Rathbun
and Bennett of the USGS (10). They recommend the formula
K = 2.59 (V0
cl
where Y is mean stream depth. This was based on a very large
data base in a variety of rivers collected by many other re-
searchers .
The CBOD oxidation and settling rate constants, Kr and Kd,
are normally determined in the laboratory from stream samples.
In the absence of such samples, rate constants typical of streams
in urban areas were used. Thomann (2) gives a table of values
of typical rate constant, Kr (values at 20°C):
Kr = 1.0+3.0 in shallow streams, high oxidation,
rapid settling
= 0.6+0.8 some settling
= 0.1+0.6 "normal"
Kd * 0.5 Kr "normal."
For this study, both K^ and Kr were considered to be 0.3. This
was approximately the middle of the "normal" range. Rate con-
stants were adjusted for temperature by the formula:
K(T) =K (20°C) (1.024(T~20)) .
It was particularly difficult to make good estimates of the
CBOD loads and the upstream DO levels. Again, the goal was
one of consistency rather than accuracy. The first loading con-
sidered was usually the initial point source of CBOD. Two cal-
culations were performed. The first was for "normal" sewage
flow. The second was for "storm" flow. If more than one treat-
ment plant was involved, more calculations were made and the
multiple solution capability used.
66
-------
The "normal" sewage flow was readily computed from the pop-
ulation of the urban area. The population figures were obtained
from a World Almanac, if possible. If not possible, a population
density of 7000 per square mile was assumed. This figure was
obtained by averaging 25 cities ranging in population from
50,000 to 5,000,000. Raw-sewage flow is characteristically
taken to be 125 gallons per capita per day. This gave Qw. The
raw weight per day, W, of the CBOD load was computed by multi-
plying the population by 0.5 pounds per capita per day (typical
according to Thomann (2)). This was multiplied by 0.15 to give
an estimate of the treated load (i.e., we assumed 85 percent
treatment efficiency). The "Streeter-Phelps" calculator program
used the information to compute Lo.
The "storm" loadings were a somewhat more difficult matter.
Two steps were necessary. First, an assumption was required
regarding the quantity of rainfall and the percentage of runoff.
Second, the CBOD load associated with this runoff was required.
Thomann (2) again provided an answer. A study of combined
sewer overflows in North Philadelphia, Pennsylvania, produced a
graph of CBOD load in pounds per day per acre versus storm
duration in hours at various intensities. Arbitrarily a rain-
fall rate of 0.15 inch per hour and a storm duration of 10 hours
was selected. This gave 1.5 inches total rainfall, which seemed
somewhat reasonable. The graph indicated a 10 pounds per acre
per day load for these conditions. Twenty percent of the water
that fell was assumed to run off. This again was based on the
Philadelphia study as reported by Thomann. The procedure then
was: (a) multiply the urban area by 10 pounds per acre per day
to get W, (b) multiply the urban area by 1.5 inches and convert
to cubic feet, (c) divide the cubic feet by 10 hours and convert
to cubic feet per second, (d) multiply cubic feet per second by
the 20 percent runoff factor to get Qw. If the urban area had a
single treatment plant, this "storm" load was entered as a point
source there. If more than one plant was present, the load was
entered at the midpoint of the urban area. Typical "storm"
loadings computed by this method roughly equaled the estimated
raw sewage yield. This seemed reasonable as the Durham, North
Carolina, studies (1,3) and a study of San Francisco CSO's in-
dicate that such discharges are roughly equal to raw sewage in
quality.
Three parameters remain. These are Do, Lrcj, and Sb. The
value of D0, the initial deficit, was assumed to be zero. That
is, the stream was assumed saturated at X = 0. This was done
simply for lack of a way to make an intelligent guess. Similar-
ly, the distributed CBOD source was set to zero. The solution
was extremely sensitive to this value and again no good way was
available to estimate it. The distributed benthal demand, Sj-,,
was taken as 1.5 grams per meter squared per day. According to
Thomann, this is typical of aged municipal sewage. The S^ term
67
-------
was needed only when a concentration of industry or treatment
plants were present.
TABLE 12 identifies the 13 sites where the Streeter-Phelps
analysis was applied. As mentioned previously, these are sites
on rivers with few tributaries and fairly well-defined waste
loads. These were felt to be sites where analysis would be most
meaningful. The results for the Scioto River at Chillicothe,
Ohio, are illustrated in Figure 4. Figure 4 is fairly typical
TABLE 12. STREETER-PHELPS ANALYSIS SITES
No. Monitor name
1 Lehigh R. at Easton, PA.
2 Scioto R. at Chillicothe, OH.
3 Sandusky R. near Upper Sandusky, OH.
4 Portage R. at Railroad Bridge at Woodville, OH.
5 Maumee R. at Defiance, OH.
6 Mahoning R. at OH.-PA. State Line below
Lowellville, OH.
7 Mad R. near Dayton, OH.
8 Hocking R. below Athens, OH.
9 Cuyahoga R. at Old Portage, OH.
10 Cuyahoga R. at Independence, OH.
11 Blanchard R. near Findlay, OH.
12 Raritan R. near South Bound Brook, NJ.
13 Connecticut R. at West Springfield, MA.
of all the Streeter-Phelps analysis results. Plotted on the
graph are DO deficit versus distance below Columbus, Ohio.
Columbus is the only significant urban area upstream of Chilli-
cothe. Two DO deficit curves are illustrated. The lower curve
represents the theoretical DO deficit for August low-flow con-
ditions. The upper curve represents the theoretical deficit for
an August "storm flow with urban runoff contributing oxygen-de-
manding material at Columbus. The location of the Chillicothe
monitors is shown at 80 miles downstream of Columbus. The DO
sag point, or maximum deficit would theoretically occur 26 miles
below Columbus. The monitor is, thus, poorly located. Water
quality violations could occur near the sag point and not be
observed at Chillicothe. The analysis indicates that a deficit
of slightly over 4 mg/1 could be observed at Chillicothe during
storm flow conditions at Columbus. A deficit of 7 mg/1 could
theoretically be observed at the sag point. Also illustrated in
the figure are the saturation DO levels for 20°C and 27°C. The
68
-------
15.0
1
I
10.0 --
SATURATION AT 20 C = 9.07 mg/J
SCIOTO R. AT CHILLICOTHE, OH.
O)
E
N
O
U,
UJ
Q
O
Q
SATURATION AT 27°C = 8.07 mg/l
5.0 --
STORM FLOW
•COLUMBUS
t
90
CHILLICOTHE
MONITOR
100
DISTANCE, mi
Figure 4. Streeter-Phelps analysis results for Scioto River at Chillicothe, OH.
-------
difference between these lines and the deficit curves is the
absolute DO level. In the worst case (27°C), absolute DO levels
at Chillicothe could theoretically fall below 4.0 mg/1 (8.07 -
4.1) under storm conditions. At the sag point, DO levels of 1.0
mg/1 (8.07 - 7.0) could theoretically be observed. These obser-
vations are typical of the information available from the
Streeter-Phelps analysis. Discussion of the Streeter-Phelps
analysis for all 13 sites is included in Appendix D. Reference
will be made to the results in Section 8.
Hourly Data Analysis
The last and most important of the detailed site analysis
techniques was the hourly data analysis. This was ultimately
completed at 22 of the 30 USGS sites that indicated a strong
correlation between flow and/or rainfall and periods of low DO.
These sites are listed in TABLE 13. No decision process was in-
volved in reducing the number of sites for hourly analysis from
30 to 22. Those sites that are included had hourly data avail-
able. The eight missing sites did not.
TABLE 13. SITES AT WHICH HOURLY DATA
WERE PROCESSED
No. Monitor name
1 Connecticut R. at West Springfield, MA.
2 North Nashua R. near Lancaster, MA.
3 Wilsons Cr. near Springfield, MO.
4 Ashtabula R. at Astabula, OH.
5 Blanchard R. near Findlay, OH.
6 Cuyahoga R. at Independence, OH.
7 Cuyahoga R. at Old Portage, OH.
8 Grand R. at Painesville, OH.
9 Hocking R. below Athens, OH.
10 Little Miami R. at Miamiville, OH.
11 Little Miami R. near Spring Valley, OH.
12 Mad R. near Dayton, OH.
13 Portage R. at Railroad Bridge at Woodville, OH.
14 Sandusky R. near Upper Sandusky, OH.
15 Scioto R. at Chillicothe, OH.
16 South Umpqua R. near Roseburg, OR.
17 Delaware R. at Bristol, PA.
18 Delaware R. at Chester, PA.
19 Lehigh R. at Easton, PA.
20 Schuylkill R. at Philadelphia, PA.
21 Trinity R. below Dallas, TX.
22 Trinity R. near Rosser, TX.
70
-------
Looking at every piece of hourly data available was out of
the question. Five years of hourly data for flow, DO, and tem-
perature plus the computation of an hourly DO deficit would be
5.25 million numbers. This would not be infeasible if the
numbers were available in digital form. Unfortunately, they are
not. All hourly data had to be transcribed by hand from computer
listings provided by USGS. Further discussion of this follows
later.
Instead of examining all available hourly data for each
site, a "windowing" technique was used with daily data. For a
given station the plots of mean daily discharge, DO and DO defi-
cit were located for years with a 60 percent or greater proba-
bility of a low DO event coinciding with a flow or rainfall
event. These were searched to find typical periods ("windows")
when such events occurred. Hourly data were then processed for
these periods and plotted.
Advantages of USGS Data—
Some explanation of why only USGS sites were used in the
hourly analysis and how hourly data are obtained is in order.
Several federal and state agencies gather and file hourly data.
Most notable of these are USGS, ORSANCO, EPA (through the STORET
system), the Metropolitan Sanitation District of Greater Chicago,
and WDNR.
Recall from earlier discussions that ORSANCO stores hourly
values at some 20 places along and tributary to the Ohio
River. There are two major drawbacks to its data. First, no
flow data are collected. The COE collects those. Second, the
data are given to outside users in a very awkward format for a
project such as this. Each file on an ORSANCO digital tape con-
tains data from all the stations reporting on a given hour.
Thus, it is necessary to execute 8760 tape reads to recover a
single year of data at one station. The computational expense
is prohibitive. (This is not an indictment of the method used
but simply a warning to other potential users.)
EPA's STORET data base contains several types of data mixed
together. There is a good deal of usable daily average data,
as mentioned in the section on site screening. The writers were
unable to locate hourly data in significant amounts at any sites
of interest. The Chattahoochee River near Atlanta, Georgia,
monitor is typical. Well over five years of daily data were
included in the STORET records. Scattered through this were
random days when one or two diurnal oxygen cycles had been
sampled on an hourly basis. There were not enough data to con-
tribute to this study.
The Metropolitan Sanitation District of Greater Chicago
keeps track of hourly data on a number of sites. However, it is
not sufficiently staffed to copy these data onto a digital tape
71
-------
for outside users. It will, however, allow access to its data
for hands-on copying of numbers. This represents a sizable man-
power cost for useful amounts of data.
The WDNR stores both hourly and daily data on 14 stream
sites. It was highly cooperative at providing a tape of the
daily average values. No attempt was made to obtain the hourly
data because only three of the 11 showed any degree of correla-
tion between rainfall/flow and low DO. Neither of the correla-
tions was exceptionally strong. There is no reason to believe
that hourly data could not be readily obtained from WDNR.
Hourly Rainfall Data--
For reasons of economy the hourly rainfall data for the
detailed analysis was obtained from a National Climatic Center
publication called "Local Climatologic Data." This is a monthly
publication that summarizes various data at primary weather
stations throughout the country. Each state contains three to
five such stations. Each monthly summary contains the hourly
rainfall. These data were obtained free from NOAA at no charge.
The cost of obtaining hourly data on one year from the National
Climatic Center in Asheville, North Carolina, on digital tape is
$50.
The rainfall data were entered by hand into files on digi-
tal magnetic-tape cartridges. The program for creating the
files and plotting the data will be discussed in the next sec-
tion on USGS hourly data.
Obtaining and Processing USGS Data—
Most USGS data are recorded at monitor or stream gage sites
on digital punched tape. In rare instances, the stream gages
still record stage on strip chart recorders. The digital tape
punches used by USGS are capable of punching at intervals from
five minutes to one hour. The tapes are collected once a month
or more. The monitors are serviced and brought to working order
if necessary when the tapes are changed. Any clock descrepan-
cies or monitor malfunctions are noted on the tape.
The next step in recovering data from the digital punched
paper tape is to copy it to magnetic tape. This is done by
reading the paper tape via phone connection to the USGS central
computer system in Reston, Virginia. There, the digital magnetic
tape is created. Some state offices have facilities for making
their own magnetic tapes.
The next step in the data recovery process is to examine
and edit the numbers on the magnetic tape. It is at this stage
that hourly data become available. Standard processor programs
are run to produce hourly or bihourly listings of the data on
the digital magnetic tape. These listings are called primary
computation sheets or "primaries" by the USGS. They are the
72
-------
only record of hourly or bihourly data available. The digital
tapes from which the primaries are made are reused, thus destroy-
ing the only digital form of the data. Each primary sheet is
carefully examined by a hydrologist and periods of erroneous,
missing, or shifted data are identified. A second series of
programs is then run to create a file of corrected data. The
corrected files are then daily averaged and the results stored
in the USGS's backfile library of daily values. The daily
values are published each year on a state-by-state basis.
Obtaining hourly data from USGS thus is a matter of obtain-
ing copies of the primary computation sheets. No software is
maintained by USGS to provide the information in magnetic-tape
form. Again, this is because the digital tapes on which the
information from the paper tapes is copied are not saved. In
general, it is not too difficult to obtain copies of the sheets.
A district office is maintained in almost every state. The
district chief will usually direct an interested user to the
correct person to talk to. Occasionally reluctance is shown
because the data are in raw form and must be interpreted with
caution.
For this study, the primary sheets were obtained for 22 of
the 30 USGS sites that showed positive correlation with rain or
flow. For various reasons hourly data could not be obtained at
the remaining eight sites. In most cases, the problem was in-
ability to obtain the primary sheets. A number of older sta-
tions still record data on strip charts. These are not converted
to digital form. In a very small number of instances the
District Chief of a particular state would not release the pri-
maries because of alleged deficiencies in the data. Emphasis
was placed on obtaining periods of hourly data during June,
July, August, and September. In addition to the primary sheets,
appropriate rating curves were obtained for use in converting
the stage data to discharge. Primary sheets report hourly stage
instead of hourly flow.
Recovering useful data from the primary sheets requires a
great deal of hand work. This was minimized for this study by
developing a flexible plotting and analysis sequence for a desk-
top calculator. First, a program was developed which created
digital magnetic-tape cartridge files. Each day of data of a
particular kind (stage, DO, rainfall, or temperature) was
entered as a separate tape file. Files were created only for
periods when all four types of data were available simultan-
eously. Next, a processor and plotting program was developed.
This program reads the files of data, applies the rating curve
to the stage data to obtain flow, calculates the saturation DO
level, and plots the results, including rainfall. The program
contains shift table capability so that the stage datum can be
corrected just as is done by USGS. The program allows periods
from a single day up to one month to be studied.
73
-------
A one-month period of hourly data for the Scioto River at
Chillicothe, Ohio, is illustrated in Figure 5. A portion of the
plot of 1972 daily data which contains the month is illustrated
in Figure 6. These two plots are fairly typical of the hourly
analysis procedure. The correspondence between descreasing DO
O)
E
O
Q
20
10
SCIOTO R. AT CHILLICOTHE, OH.
8/1772 TO 8/30/72
SATURATION DO LEVEL
DO LEVEL—
3.0 T
I
0
t/5
Q
d
> 0
< 3^
-- H-
tt E
< ^
X
0 O
W Q
O
2.5 :
2.0 •
1.5 \
1.0 ;
0.5 ;
0.0 3
DIS
U I O
: r— DO
: PRE
^y\f\
LOffi
DISCHARGE / AVG. DISCHARGE
DO DEFICIT/10, mg/l
PRECIPITATION, inches
8 10 12 14 16 18 20 22 24
TIME FROM START OF PERIOD, days
26 28 30
Figure 5. One month of hourly data for the Scioto River at Chillicothe, OH.
and increased flow is fairly evident in Figure 6. The upper line
in the figure is the observed daily minimum DO level, the lower
line is the mean daily discharge. A decrease in the DO level
can be seen in almost all instances when the flow increases.
The period from day 215 to day 245 was selected as typical.
Figure 5 provides a highly detailed hourly look at this same
time period. The upper graph in Figure 5 illustrates the behavior
of the absolute DO level. Daily fluctuations of approximately
4.0 mg/l were common before the rainfall and subsequent runoff
event. The upper line in the DO graph is the saturation level
based on the hourly water temperature. Note periods of super-
74
-------
60 -,
40 -
SCIOTO RIVER AT CHILLICOTHE, OH.
I PERIOD OF HOURLY
ANALYSIS
DO
- 15
- 10
en
o"
Q
„ 5
180
300
I
TIME, day of year, 1972
Figure 6. Hourly data for one-month period,
Scioto River at Chillicothe, OH.
saturation occurring on days 3 and 11. Also note that the DO
level falls below 2.0 mg/1 and remains there for three days
beginning at day 18. This precisely coincides with the peak of
the flow hydrograph in the lower graph. The lower graph in Figure
5 illustrates the behavior of the flow, rainfall, and DO deficit.
The flow has been plotted as discharge divided by averge dis-
cahrge for the period. This was simply a convenience to avoid
rescaling every graph. The shape of the hydrograph is preserved
and no useful information is lost. The DO deficit level is the
difference between the upper and lower lines in the upper graph
of Figure 5. This was included in the lower graph to aid in see-
ing the coincidence between a change in the behavior of the DO
level and the change in flow. Rainfall amounts are plotted as
histograms. The period of greater than average DO deficit that
occurs during the time of higher-than-average flow can clearly
be seen extending from day 18 to day 30. There is a clear cor-
respondence between the rainfall events and the increases in
flow.
Further discussion of the hourly data plots is included in
Section 8. Additionally daily and hourly plots are included in
Appendix D.
75
-------
SECTION 8
ASSESSMENT OF EXTENT AND CAUSES OF PROBLEM
GENERAL CONSIDERATIONS
This section of the report presents an analysis of the
findings. A very large number of data have been examined in
various ways. The purpose of this final section is to determine
what has been learned.
This section is divided into five subsections. Each sub-
section is designed to answer a question or several related
questions concerning the correlation between DO deficit and
urban runoff. These questions are as follows:
• Is there a correlation between the presence of
urban runoff and DO deficits? If so, what is
the nature of the correlation and are water
quality standards violated when the correlation
is present?
• Is the problem national in scope?
• What are the causes of the problem?
• Where is the problem worth studying further?
• What should be studied and how?
THE EXISTENCE OF A CORRELATION, ITS NATURE AND SEVERITY
Daily Correlation Analysis
The answer to the first part of the first question (Is
there a correlation between periods of higher-than-average DO
deficit and storm runoff?) is an unqualified "yes." Some general
statistics and results will now be presented to support this
conclusion.
Attempts were made to run daily correlation analysis using
either flow or rainfall or both on 104 stations. These included
USGS, STORET, and WDNR data. Of these, 83 had sufficient data
or data that could be economically analyzed to produce results.
76
-------
These included 55 USGS monitors, 17 STORET monitors, and 11 WDNR
monitors.
Of the USGS monitors, 19 exhibited a 60 percent or greater
probability of low DO at times of high flow. Eighteen exhibited
a 60 percent or greater probability of low DO on days with rain-
fall. Eight of the stations correlated with both flow and rain-
fall. The total number of stations exhibiting either type cor-
relation was 30. Of the 17 STORET monitors, three exhibited a
60 percent or greater probability of low DO at times of high
flow. Of the 11 WDNR monitors, three exhibited the 60 percent
or greater probability-
Out of 104 candidates for daily analysis, 35 (19 USGS flow
+18 USGS rain -8 USGS overlap +3 STORET +3 WDNR) gave positive
results in the correlation analysis. Twenty-one of the monitors
could not be correlated because of problems in obtaining or using
data. Thus, 42 percent of the monitors successfully examined,
34 percent of the likely candidates, gave positive correlation
results. For discussion purposes, it will be assumed that four
monitors in 10 placed near an urban area might indicate low DO
at times of storm runoff. That is, a correspondence between
lower-than-average DO and higher than average flow or rainfall
can be identified. This does not mean that water quality stan-
dards are always violated.
An interesting question can be raised concerning the pres-
ence of a correlation at a given station: What is the implica-
tion of finding a correlation in only one or two of the years
examined? Only general comments can be offered in answer. One
problem comes from the statistical problem of defining a corre-
lation in the first place. Recall that this was defined as a
60 percent probability of DO lower than the seven-day average on
a day with flow or rainfall higher than the seven-day average. One
can easily question the 60 percent cutoff level even though this
has statistical significance. For instance, the Little Miami
River near Spring Valley, Ohio, monitor showed strong correlation
(65 and 67 percent) with flow in two of the five years examined.
Two additional years had 58 and 59 percent probabilities. Should
these years be counted as "correlated"? Examination of the daily
DO and flow plots indicate that there are certainly times when
the DO decreases on days with increased flow. A second problem
in determining the presence or absence of a correlation comes
from physical reality. Since the cause of the DO depressions is
not known, it is not possible to say why a correlation is present
one year and not the next. If the cause was sewage treatment
plant overflow, perhaps a new plant was added. If the cause
depends on time between storms for pollutant accumulation, then
perhaps one year was wetter than the next. The answer to these
questions will have to come from further research.
77
-------
Nature of the Correlation
The second part of the first question concerned the nature
of the correlation. When hourly data from sites with strong
correlations were examined, similar DO records were found at
sites with widely different physical settings. Data from the
Scioto River at Chillicothe, Ohio, were presented earlier. Data
from the Little Miami River near Spring Valley, Ohio, are repro-
duced from Appendix A in Figure 7.
The three-day period prior to the storm event (Figure 7) is
characterized by fairly large diurnal cycles in the DO level,
4 to 5 mg/1 is not unusual. Periods of supersaturation are often
indicated. These are clearly visible at the Little Miami sta-
tion. There is some question, of course, whether these periods
are real or a monitor malfunction. Conversations with the USGS
office in Columbus, Ohio, indicate that periods of supersatura-
tion, due to algal growth, are common. The precipitation re-
sponsible for the rise in flow is visible just prior to the
hydrograph along the bottom axis. Approximately six hours elapsed
from the time of the rainfall to the beginning of the flow change.
Approximately two inches of rainfall fell at Dayton for the
several small events shown. As the flow increases, the diurnal
cycles in the DO record disappear. This may be due to increased
turbidity and depth cutting off the sunlight to the aquatic
plant life at times of high flow. At the time of peak flow a
deficit level 40 percent higher than the maximum achieved during
a nonhigh-flow diurnal cycle is reached. This percentage in-
crease at other sites with correlations varied from zero to 50
percent. In the case of the Little Miami, the peak deficit
during diurnal cycles was 2.5 to 2.6 mg/1. At peak flow, the
deficit reached 3.5 mg/1 and the absolute DO level reached 5 mg/1.
The correlation between the DO deficit and the flow is very
obvious in the lower graph. Beginning at the seventh day, seven
small events follow in which the deficit curve mimics the flow
hydrograph almost exactly. There is roughly a six-hour lag be-
tween the time the flow begins to increase and the time the
deficit begins to increase.
The effect of the storm flow on the DO level is quite long
lasting. Note, that on the Little Miami the DO has not returned
to the diurnal cyclic behavior in 17 days. The DO depression
caused by the large flow hydrograph that began on day 3 would
probably have only lasted four to six days if it had not been
followed by the smaller events on days 7, 10, and 13. This be-
havior is typical of all the sites with pronounced correlation.
Sites at which a single hydrograph peak was examined usually re-
covered in three to five days.
This long-term effect gives added validity to the daily
analysis procedure. If the DO depressions only lasted a few
78
-------
o>
£
20
"
R.
7/1/75 TO 7/20/75
SATURATION DO LEVEL
DO LEVEL
VALLEY, OH.
3.0 T
2.5
•"-DISCHARGE / AVT. DISCHARGE
DO DEFICIT / 10, mg/l —
-PRECIPITATION, inches
10 12 14
TfME FROM START OF PERIOD, days
16
18
20
Figure 7. DO/fiow correlation on the Little Miami River near Spring Va!!ey, OH.
-------
hours after storm events, it would not be possible to detect
them using daily average data.
Severity of Problem
The final part of the first question concerned the severity
of the DO deficits associated with urban runoff. The fact that
the DO decreases at times of storm runoff is scientifically of
interest. However, if the level does not consistently decrease
below prescribed standards, no "problem" exists.
Water quality standards were mentioned in the introduction.
There is no uniform standard for the entire country. Generally,
for a warm-water system containing fish, other aquatic life, and
wildlife, a level of 5 mg/1 is recommended (Thomann (2)). The
ORSANCO recommends that DO levels not drop below 5 mg/1 for more
than eight hours a day or below 3.0 mg/1 at any time. The EPA
standards recommend that DO not drop below 2.0 mg/1 for more
than four hours or below 3.0 mg/1 for three days. The hourly
DO data in Appendix D were examined carefully and compared to
these standards.
The first standard used to examine the hourly data was sim-
ply 5.0 mg/1. This standard appears so frequently in the liter-
ature that the writers felt it was worthwhile to identify those
sites that would not meet it. While 5.0 mg/1 is not a severe
water quality problem, it is certainly poor water quality.
Eleven monitor sites on nine rivers had hourly DO levels that
fell to 5.0 mg/1 or below. These 11 monitor sites are identi-
fied in TABLE 14. Also listed in the table are the minimum
observed DO level and the time (hours) that the DO remained be-
low 5.0 mg/1 because of the change in flow.
The next standard used to examine the hourly data was the
EPA recommended standard. Six monitor sites on five rivers had
hourly DO levels that violated the EPA standard. These six
sites are listed in TABLE 15. Also listed in the table are the
minimum observed DO level, the approximate DO level before the
runoff event, and the time the DO remained below 2.0 mg/1.
The final examination of the data with respect to standards
was to determine how many additional monitor sites would show
EPA standard violations if the monitor were more appropriately
located. Minimum observed values were reduced by the difference
between the Streeter-Phelps predicted minimum value and the
Streeter-Phelps predicted value at the monitor site. Four addi-
tional monitor sites would probably show EPA standard violations
if properly located. These are listed in TABLE 16. Also listed
for each site are the minimum observed DO level, the Streeter-
Phelps correction, and the predicted minimum for a properly
located monitor. Hourly data were not available at two of the
four sites listed. The observed values were taken from plots of
80
-------
TABLE 14. MONITOR SITES AT WHICH DO LEVELS
BELOW 5.0 mg/l WERE OBSERVED
No.
1
2
3
4
5
6
7
8
g
10
11
Monitor name
N. Nashua R. near Lancaster, MA.
Wilsons Cr. near Springfield, MO.
Cuyahoga R. at Independence, OH.
Cuyahoga R. at Old Portage, OH.
Little Miami R. at Spring Valley, OH.
Mad R. near Dayton, OH.
Sandusky R. near Upper Sandusky, OH.
Scioto R. at Chillicothe, OH.
Schuylkill R. at Philadelphia, PA.
Trinity R. below Dallas, TX.
Trinity R. near Rosser, TX.
Minimum
observed
DO level.
mg/l
1.8
0.0
4.9
1.9
5.0
4.9
1.5
0.9
3.0
0.9
0.0
Length of time
below 5.0 mg/l.
hours
8
84
3
4
1 or less
1 or less
240+
168
72
All the time
All the time
TABLE 15. MONITOR SITES AT WHICH EPA DO STANDARDS
WERE NOT MET DURING RUNOFF EVENTS
No.
1
2
3
4
5
6
Monitor name
IM. Nashua R. near Lancaster, MA.
Wilsons Cr. near Springfield, MO.
Sandusky R. near Upper Sandusky, OH.
Scioto R. at Chillicothe, OH.
Trinity R. below Dallas, TX.
Trinity R. below Rosser, TX.
Minimum DO
level
mg/l
1.8
0.0
1.5
0.9
0.0
0.0
DO level
before
runoff.
mg/l
5-6
6-10
5-9
4-10
2-3
1-2
Time DO
remained below
2.0 mg/l, hours
4
9
18
16
Much of the time
(6-8 hr/event)
Much of the time
(6-8 hr/event)
the daily minimum for years with positive correlation. One
additional monitor site, the Little Miami River near Spring
Valley, Ohio, monitor, would probably be included with those in
TABLE 16 if the Streeter-Phelps analysis had been used there.
The monitor is near the outskirts of Dayton. Experience with
other sites indicates that if the monitor had been 30 to 40
miles further downstream, it might have sensed a greater deficit,
Absolute DO levels of 5.0 mg/l were observed at the monitor's
present location.
81
-------
TABLE 16. MONITOR SITES WITH POTENTIAL FOR EPA STANDARD
VIOLATIONS IF PROPERLY LOCATED
No. Monitor name
1 Cuyahoga R. at Independence,
OH.
2 Cuyahoga R. at Old Portage,
OH.
3 Mahoning R. at OH.-PA.
State Line below
Lowellville, OH.
4 Maumee R. at Defiance, OH.
Problem with
location
Too close to
urban area
Too close to
urban area
Too close to
urban area
Too far from
urban area
Minimum
observed DO
level, mg/l
4.9
1.9
0-1.5
(daily)
2.5-4.0
(daily)
Streeter/Phelps
correction,
mg/l
4.9
1.9
1.5
4.0
Possible
minimum DO
level, rng/l
0.0
0.0
0.0
0.0
Several other general observations were made from the hourly
data concerning the severity of the problem. Three of the USGS
sites examined were unusual. All showed strong correlation be-
tween flow and low DO but none were truly urban in character.
Center Creek near Carterville, Missouri, for instance, is near
a relatively small town in an isolated area. There is a good
deal of strip mining near the monitor. The site on the Manasquan
River near Squankum, New Jersey, was near a large federal muni-
tions reservation and a few suburbs, but nothing truly urban.
Much of the drainage appeared swampy. The site on the Little
Miami River at Miamiville, Ohio, is in the far outskirts of
Cincinnati and has no direct drainage from the city. Five of
the sites examined on an hourly basis did not have water quality
problems associated with urban runoff, at least concerning DO.
However, some depression during storm events could usually be
seen. These five sites are on the
& Ashtabula R, at Ashtabula, OH;
@ Grand R. at Painsville, OH;
• Hocking R. at Athens, OH;
© Delaware R. at Chester, PA; and
o Lehigh R. at Easton, PA.
In summary, the question "Do severe (EPA standard) DO defi-
cits occur downstream of urban areas due to urban runoff?" can-
not be answered conclusively. There is definite evidence that
82
-------
decreases in DO levels occur downstream of a number of urban
areas after rainfall or flow events. However, only six of the
monitor sites examined for this study showed clear evidence of
EPA water quality standard violations. For purposes of discus-
sion, the following facts might be considered. A population of
104 monitor sites were considered for analysis. Data could only
be obtained for 83 of these. Out of the 83, only some of the 55
USGS sites could be examined on an hourly basis. Of the 55 USGS
sites, 29 showed positive correlation between either rainfall or
flow events and low DO. Of these 29, only 22 could be examined
on an hourly basis. Eleven of the 22 could not meet a 5.0 mg/1
DO standard and six could not meet the lesser EPA standard of
2.0 mg/1 for four hours. When poor monitor location is accounted
for using the Streeter-Phelps technique, a total of 10 sites
might not meet the EPA standard. If it is assumed that the
group of monitors actually examined closely (22) contains the
same percentage of monitors with water quality violations as the
candidate group (104), the following is obtained. Including
Streeter-Phelps sites, (13/29) of the 42 percent ((29 + 3+3)/83)
of all monitors that might be correlated, or 19 percent of the
candidate monitors, might not meet a 5.0-mg/1 standard. (Eight/
22) of the 42 percent, or 15 percent, might not meet the EPA
standard of 2.0 mg/1 for four hours. The frequency of these
violations would appear to be fairly low. No exact figures can
be given because programs were not developed to address this
question. Three to five violations per year at a site with a
strong correlation might be a reasonable guess. This question
of frequency should probably be the subject of further investi-
gation.
In fairness, it must be pointed out that the water quality
at sites with EPA standard violations is marginal at all times.
Storm events merely push the level down further. General im-
provements in water quality at all the critical (EPA standard)
sites would help alleviate the problem. For example, the main-
stream of the Delaware River below Trenton, New Jersey, consis-
tently shows a correlation between DO deficit and flow events.
The DO levels are high enough and the river big enough to absorb
the effect without standard violations. Note that it is not
possible from this study to define how the water quality might
be improved at critical sites. Further investigation would be
required to determine why quality is poor.
SCOPE OF THE PROBLEM
The question of whether urban runoff caused DO deficits is
a national problem is difficult to answer. One item to consider
is the geographic coverage of this study.
The USGS data base maintained at the USGS National Head-
quarters contains records for 150 water quality monitors. These
monitors are located in 30 of the 48 conterminous states, i.e.,
83
-------
63 percent coverage. The distribution of states containing moni-
tors is highly nonuniform. Thirteen states are east of the
Mississippi River and 17 are west. Only 47 of the 150 monitors
are in western states. If the dividing line between east and
west is considered as the western boundaries of Louisiana,
Arkansas, Iowa, and Minnesota, the distribution becomes even
more unbalanced. Only 19 monitors are then located in the west.
The distribution of monitor by state is highly nonuniform
also. The top five monitoring states and the numbers of moni-
tors for USGS are:
• Ohio, 32;
• New Jersey, 14;
• Pennsylvania, 13;
• Louisiana, 11; and
• South Carolina, 10.
These five states contain over half of all USGS monitors. Ohio
alone contains over 20 percent.
The STORET data base contains over 1000 monitor records.
It added, however, only three additional states to the geographic
coverage. Most of the STORET monitors applicable to this study
are located in North Carolina along the Neuse River.
The major monitoring networks that contributed to this
study are not designed to provide uniform geographic coverage.
They are heavily biased toward the northeast, with a fair amount
of coverage in the south. Some investigation of whether the
networks are adequate to detect a national problem is in order.
It is reasonable to assume that the more water quality
monitors are present in a state, the more likely any existing
DO problems would be found. The data used in this study were
examined by state to determine the number of sites with water
quality standards violations versus the number of monitors pres-
ent. The number of correlations found versus the number of
monitors was also considered. The results are given in TABLE 17.
Listed are states with violations, the rank of the state in
terms of number of monitors, and the type of violation (5.0 mg/1
or EPA, 2.0 mg/1 for four hours). Streeter-Phelps sites are in-
cluded in the totals. The sites counted in the visible correla-
tion column are those where DO depressions of any size could be
seen at the time of storm events. No consideration was given to
standards.
-------
TABLE 17. PROBLEM SITES PER STATE VERSUS
NUMBER OF MONITORS PER STATE
State
Ohio
Massachusetts
Missouri
Texas
New Jersey
Pennsylvania
Oregon
No. of USGS
monitors in
state
32
9
10
3
14
13
1
Rank in
terms of
monitors
1
6
5
10
2
3
Tie for 12
No. of sites
with visible
correlations
10
3
3
2
2
1
1
No. of sites
with 5.0 mg/l
standard violations
8
1
1
2
0
1
0
No. of sites
with EPA
standard violations
6
1
1
2
0
0
0
The probability of finding visible evidence of a correlation
between low DO and runoff events appears to be one in three (num-
ber of correlations divided by number of monitors in state).
The probability of finding a site where the 5.0-mg/l standard
is violated appears to be between one in 10 and one in four.
The probability of finding an EPA standard violation is between
one in 10 and one in five.
The writers feel that the geographic distribution of the
nation's water quality monitoring network is not adequate to
conclude anything on national scope of urban runoff related DO
problems. The odds seem to indicate that if monitors were
located in a Streeter-Phelps correct way, then one in three would
show some evidence of a correlation between urban runoff and DO
depression. Of those, one in 10 to one in four might indicate
standards violations.
There is little other information on which to base an esti-
mate of the national impact of urban runoff. Considerable time
was spent trying to determine if any government agency kept
statistics on the distribution of the nation's population along
rivers. The U.S. Water Resources Council is considering such an
effort but to date has not implemented a project. If results
from such a project were available, it might be possible to
estimate the number of potential stream miles affected using the
Streeter-Phelps technique. Even this would not be conclusive
because there is no well-defined relationship between size of
the urban area and the presence of correlations or standards
violations. This is discussed in the following subsection. The
monitoring network is also not adequate to extrapolate the prob-
lem from one area of the country to another.
-------
POSSIBLE CAUSES OF PROBLEM
The Scope of Work did not call for a determination of the
cause of specific problems. It did call for considerations of
the percentage of contributing area when selecting monitors for
inclusion in the study. The percentage of urban area was also
to be considered when stating that a low-DO condition was re-
lated to the presence of urban area. This will be discussed
along with some other possible contributing factors.
It is intuitively inviting to assume that the larger the
percentage of a drainage basin that is urban, the more likely it
should be that a low-DO runoff correlation would exist. This
did not prove to be the case for the data examined here. Fig-
ure 8 illustrates the nonrelation between the probability of a
greater-than-average DO deficit during higher-than-average flow
and the percentage of urban drainage. Data for this figure were
taken from Appendix D for most of the 29 USGS sites that showed
positive correlation between flow or rainfall and low DO. The
percentage of urban drainage ranges from less than one to more
than 80.
Several other relationships were investigated in an effort
to determine what effect the percentage of urban area has on the
presence of a correlation. Equally discouraging results were
obtained. For instance, a regression was run using the mean
probability of low DO at times of high flow as one parameter and
the estimate of urban storm flow from the Streeter-Phelps analy-
sis divided by the mean low flow at the station as the second
variable. The idea was that if the volume of urban storm runoff
was high compared to the normal stream flow, then the influence
of the runoff would be greater and, hence, the correlation would
be greater. Many streams are regulated and the percentage of
drainage is no longer a measure of the percentage of flow that
may be contributed. The results looked encouraging but statis-
tically the correlation coefficient was 0.002. Regressions were
also run using estimated urban flow divided by average flow as
one parameter and both maximum and mean probability of low DO
at times of high flow as the second parameter. These were also
unsuccessful. The problem is too complex to study using simple
relations.
As stated earlier, it was not part of the Scope of Work to
precisely identify the cause of each problem. However, it is
worthwhile to examine some other possible causes since the ob-
vious urban area percentage proved discouraging. These are the
types of things that should be studied in order to better under-
stand the problem.
The urban areas with strong correlations are of many sizes,
covering a fairly large range. The population of Leominster,
Massachusetts, is approximately 33,000, while the population of
86
-------
CO
o
Q
O
O
>
CD
O
oc
Q-
Y = 50.33 + .4006 log X
Correlation Coefficient = .0018
10
100
RATIO OF URBAN FLOW TO AVERAGE OR LOW FLOW
Figure 8. Probability of low DO at high flow versus percentage of urban area.
-------
Akron, Ohio (Cuyahoga River at Independence monitor), is nearly
three quarters of a million. Several of the monitor sites with
poor monitor location in a Streeter-Phelps sense were near urban
areas of 20,000.
Close study of the sites with strong correlations did not
reveal any notable common characteristics. For instance, the
Mahoning River at the Ohio-Pennsylvania state line, below Lowe11-
ville monitor is just downstream of Young's town, Ohio. Youngs town
is in a very hilly country. A large concentration of steel mills
and several sewage treatment plants are located along the river.
The monitor is located on the fringe of the urban development.
The monitor on the Sandusky River near Upper Sandusky, Ohio, is
in a fairly rural setting some 30 miles downstream of Bucyrus,
Ohio. Bucyrus is a city of 30,000.
Proximity to sewage treatment facilities is a matter of some
interest. In most cases, at least one treatment plant was
located upstream of the monitor site. Distances varied from
less than one mile to 8 to 10 miles. The monitor on the Raritan
River near South Bound Brook, New Jersey, had six treatment
plants within five miles upstream. Reintrainment of oxygen-
demanding materials in solids deposits downstream of outfalls
could certainly be a factor at some locations. Combined sewer
overflows could also be a factor. Quality and capacity of treat-
ment facilities should also be considered. If significant quan-
tities of sewage are bypassed during times of high flow, the
impact on a receiving water could be significant. A high per-
centage of BOD removal still leaves significant quantities of
BOD for high-volume facilities. The remainder must be oxidized
by the stream in some way.
The hourly records examined did not show any "first flush"
effect. The DO deficit increased in a smooth way and remained
high for lengthy periods. This might indicate that for the
cases examined the problem is not related to a buildup of mate-
rial that is flushed rapidly away. The length of the deficit
period points to sources of oxygen-demanding material that would
continue throughout a storm event.
The type of activity carried out along the stream in the
urban environment should certainly be investigated as a cause of
wet weather DO deficits. Sewage treatment has already been
mentioned. Certain types of industry continue to discharge
treated waste. Gravel dredging and concrete mixing activities
often concentrate along rivers. Each site has its own peculiar-
ities .
SPECIFIC STUDY SITES
The data examined here indicate that water-quality viola-
tions associated with flow events can occur downstream of urban
-------
areas. The 5.0 mg/1 and EPA (2.0 mg/1 for four hours) standards
are both violated. The frequency of such violations was not
determined exactly, but no need for panic treatment appears to
exist. There is, however, a need to understand the causes of
the DO deficits observed here . An exact understanding of the
causes can be used to prevent the frequency of occurrences from
increasing. The purpose of this subsection is to identify and
discuss those monitor sites that the writers feel are worthy of
further investigation.
TABLE 18 identifies those monitor sites that are most likely
to yield useful results from further study. Listed along with
each site are the minimum observed DO level, the standard that
was violated, and a short description of why the site was chosen.
The sites are listed in the order of preference for study. Two
monitor groups are identified. The first group is hydraulically
simple sites; single streams with major urban areas present.
The second group is complex sites with branched streams, several
towns, and other factors--sites where more sophisticated analy-
sis techniques and data collection will be required.
The first group contains five monitor sites. These are the
Scioto River, near Chillicothe, Ohio; the Cuyahoga River at Old
Portage and Independence, Ohio; the Mahoning River at the Ohio-
Pennsylvania state line near Lowellville, Ohio; the Sandusky
River near Upper Sandusky, Ohio; and the Little Miami River near
Spring Valley, Ohio.
The Scioto River site is highly attractive. It has one of
the most distinct correlations of any station examined on an
hourly basis. The monitor is located fairly well in a Streeter-
Phelps sense to detect the sag from Columbus, Ohio. The USGS is
establishing a monitor at Circleville, which is halfway between
Chillicothe and Columbus. An additional monitor called Scioto
River at Shadeville is situated in Columbus. The Ohio USGS
personnel are extremely cooperative and would probably be in-
terested in collecting sediment or other supplementary data to
examine this site. This is also an excellent choice for further
study.
The Cuyahoga River flows southwestward, generally parallel-
ing the southern shore of Lake Erie. It turns north at Akron,
Ohio, joining the Little Cuyahoga, and flows through Cleveland
into the lake. All of the flow from Akron is sensed by both
monitor sites. The monitor at Independence is below the mouth
of Tinkers Creek, which drains part of Cleveland and its suburbs.
The Akron sewage treatment facilities lie between the two moni-
tors. They are separated by roughly 13 miles. Sediment records
are available at both sites. The Cuyahoga is famous for having
caught fire in the early 1960. This, however, was in the in-
dustrial part of Cleveland, which is below both monitors. There
should be considerable other data available here.
39
-------
TABLE 18. SITES RECOMMENDED FOR FURTHER STUDY
Site Monitor Min. DO
classification normal level, mg/l
Simple -^
Complex -^m
Scioto R. near 0.9
Chillicothe, OH.
Icuyahoga R. at 1.9,4.9
Independence and
Old Portage, OH.
Mahoning R. at 0-1.5
OH.-PA. State
Line below
Lowellville, OH.
Sandusky R. 1.5
near Upper
Sandusky, OH.
Little Miami R. 5.0
near Spring
Valley, OH.
r North Nashua R. 1.8
near Lancaster,
MA.
Schuylkill R. at 3.0
Philadelphia,
PA.
Trinity R. below 0.0
Dallas, TX.
Wilsons Cr. near 0.0
Springfield, MO.
Wilsons Cr. near —
Battlefield, MO.
James R. near —
Boaz, MO.
Standard
violated
EPA
5.0 mg/l,
potential
EPA
Potential
EPA based
on daily
EPA
5.0 mg/l,
potential
EPA
EPA
5.0 mg/l
EPA
EPA
Reason for recommendation
Clear correlation in hourly data; clearly
defined urban area and hydraulics;
potential for cooperation with USGS
Sediment data available; clearly defined
urban area; national reputation for
poor quality.
Large, clearly defined urban area draining
to single stream; good example of heavily
industralized flood plain.
Well located to sense runoff from Bucyrus;
straightforward hydraulics;.
Close to outskirts of Dayton; simple hydraulics;
clear DO depression with storm events .
Significant problem; two distinct
urban areas.
Major urban area; empties into Delaware R.
Biggest urban area of any site; worst
water quality.
Clear correlation; second worst water quality
of any monitor site.
-------
The remaining three sites in the first group are all equally
attractive. The Mahoning River site is just below Youngstown,
Ohio, an industrial city of half a million. The Sandusky River
monitor is in a excellent position to study runoff from Bucyrus,
Ohio, a city of 20,000. The Little Miami River site receives
the runoff from Dayton and Xenia, Ohio. All three are single
streams draining well-defined urban areas with no other signi-
ficant urban area upstream.
The second group of sites contains six monitor sites and
four stream systems. These are given in roughly the order of
preference for study. All are interesting, but they would re-
quire extensive investment in complex models and supportive data
collection. They are the North Nashua River near Lancaster,
Massachusetts; the Schuylkill River at Philadelphia, Pennsylvania;
the Trinity River below Dallas, Texas; Wilsons Creek near Spring-
field and Battlefield, Missouri; and the James River near Boaz,
Missouri.
The North Nashua River site is complicated by two distinct
urban areas, Fitchburg and Leominster. Each are cities of
roughly 35,000. The monitor and flow gages are separated by
several miles and a lake dilutes the stream below the monitor.
The Schuylkill River site is in the heart of industrial
Philadelphia, just upstream from the junction with the Delaware
River. It may be partially tidally affected. There is no
question of it being an urban site, but only a small reach of
river is available for study before reaching the Delaware.
The Trinity River below Dallas, Texas, has the worst water
quality observed anywhere during the study. Six lakes regulate
the flow and in the summer the stream is mostly treated sewage
effluent. The DO falls to zero at a monitor 40 miles downstream
of Dallas during storm events. The site is discouraging just
from the physical size of the contributing urban area, over 700
square miles.
The Wilsons Creek and James River sites in Missouri also
have very bad water quality. These sites have the highest per-
centage urban contributing area in the study, 85 percent. The
two Wilsons Creek monitors would be interesting to study, but
the flow in the James is regulated by a lake that catches some
of the runoff from Springfield. A number of small creeks feed
the runoff from Springfield into Wilsons Creek.
RECOMMENDATIONS FOR STUDY PROCEDURES
Two types of further study would be worthwhile. The first
is a general site survey. The second is a specific study of
detailed site(s).
91
-------
The general survey study would not require a great deal of
resources. The study procedure would be to visit each of the
monitor sites where a strong correlation has been identified.
Characteristics of the stream and the land use along it would
be identified. Local agencies would be visited to determine
sewage treatment practices, industrial practices, and other
activities impacting water quality. The study objective would
be to develop a general picture of the places where the problem
has been observed. A study of this type could more accurately
locate sewage outfalls, combined sewer overflow locations, in-
dustrial waste deposits and other things not part of the present
study.
The detailed site-specific study (studies) should be di^-
rected toward understanding exactly what happens at one err several
locations where strong correlations have been observed. This
will require considerably more effort. Ideally, a well-designed
data collection program should be undertaken to support a mathe-
matical model. The study should be undertaken in three parts.
First, the necessary information to build an accurate model of
the receiving water flow should be obtained. This will include
stage records, cross-sections, and tributary and mainstream dis-
charge data. The second step contains two substeps. First, an
accurate unsteady flow model should be set up and calibrated.
Steady-flow modeling should not even be considered. The problem
is associated with unsteady flow and should be studied with the
appropriate models. The technology is available and not overly
expensive. Second, the unsteady flow model should be coupled to
a conservative mass transport model. If at all possible, this
should be verified by conducting a dye study during a storm
event. This procedure positively guarantees a model that will
produce meaningful transport predictions.
The final step is to model the transport of nonconservative
constituents such as BOD and DO. No data should be collected
until the conservative transport model has been verified. It
should then be modified to handle the appropriate constituents
and used to simulate expected conditions during several storm
events. Best guesses are used for all unknown coefficients.
All data sampling should be geared to define those receiving
water reaches and at those times the model indicates as critical
or of most interest. This minimizes sample cost and maximizes
model accuracy when calibration is undertaken. Transport in
transient flow is not easy to second guess. This has been amply
demonstrated by Jobson and Reefer (11) on the Chattanoochee River.
Only an accurate process model can tell the investigator when
and where to look.
After the necessary data are collected for the constituents
of most interest, the model can be calibrated for these. If
this proves impossible, further data collection and model modi-
fication can be undertaken. Since it will be known that the
92
-------
model accurately predicts conservative transport, any errors in
the nonconservative transport have to be due to inadequately
defined coefficients or processes that were not considered. By
carefully proceeding to add capability to the model (e.g., ben-
thic sediment, chemical oxygen demand, and others), the exact
cause of the DO-deficit/high-flow correlation can be identified.
It is highly unlikely that adequate data for a modeling
effort of the type exists. Attempting such a study with exist-
ing monitor records and randomly collected water quality samples
is probably a waste of time.
The writers would strongly recommend the Scioto River be-
tween Columbus and Chillicothe, Ohio, for a study of this type.
This would provide an excellent site and the maximum opportunity
for cooperation with USGS. As a second choice, the Cuyahoga
River between Akron and Cleveland, Ohio, would be good. There
are some sediment records that may be helpful and the hydraulic
conditions are simple to model. This is also true of the Scioto
River.
93
-------
REFERENCES
1. Rimer, Alan E., and James A. Nissen. Characterization and
Impact of Stormwater Runoff from Various Land Cover Types.
J. Water Pollution Control Federation, Feb. 1978. pp. 252-
264.
2. Thomann, Robert V. Systems Analysis and Water Quality
Management. Environmental Research Applications, Inc.,
New York, New York, 1972. p. 13.
3. Triangle J Council of Governments. Areawide Water Quality
Management Plan. 208 Pollution Source Analysis. Research
Triangle Park, North Carolina, Feb. 1976. Section V.E.
4. City of Dallas, Texas. Urban Storm Runoff Sampling Program.
Storm of February 11, 1977. DWV Operations Division, Nov.
1977.
5. Unpublished data. U.S. Geological Survey , Missouri District,
Water Resources Division, Rolla, Missouri.
6. Committee on Water Quality Criteria. Water Quality Criteria,
1972. Environmental Protection Agency, Report R3-73-033,
March 1973. p. 134.
7. Office of Water Programs Operations, U.S. Environmental
Protection Agency. 1978 Needs Survey. Report to Congress
on Cost Estimates for Control of Combined Sewer Overflow
and Stormwater Discharge.
8. Bendat, Julius S., and Allan G. Piersol. Measurement and
Analysis of Random Data. John Wiley & Sons, New York,
New York, 1958.
9. Leopold, Luna B., Gordon M. Wolman, and John P. Miller.
Fluvial Processes in Geomorphology. W. H. Freeman and Co.,
San Francisco, California, 1964. p. 251.
10. Bennett, James P., and R. E. Rathbun. Reaeration in Open-
Channel Flow. U.S. Geological Survey, open file report,
Fort Collins, Colorado, April 1971.
94
-------
11. Jobson, H. E., and T. N. Keefer. Thermal Modeling of Highly
Transient Flows in the Chattahoochee River Near Atlanta,
Georgia. Proceedings, American Water Resources Association
Symposium on River Quality Assessments, Tucson, Arizona,
Nov. 2-3, 1977.
95
-------
APPENDIX A
WATER QUALITY DATA COLLECTION AGENCIES
This appendix contains a complete listing of all the agen-
cies that collect water quality data as reported by the USGS
Office of Water Data Coordination (OWDC). The agencies and their
identification codes are listed according to the regions de-
scribed in Section 5 of the report.
96
-------
REGION 1
FEDERAL AGENCIES
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
ERD Energy Research and Development Administration
FS Forest Service (Agriculture)
GS Geological Survey (Interior)
NFE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
CANADA
WQB Environment Canada, Water Quality Branch
WSC Environment Canada, Water Resources Branch
NON-FEDERAL AGENCIES
Connecticut
DOO Environmental Health Service Division,
State Department of Health
D01 The Water Bureau of the Metropolitan District,
Hartford
D02 Bridgeport Hydraulic Company
DOS Connecticut Department of Environmental Protection
New York
P50 New York State Department of Environmental
Conservation
97
-------
REGION 2
FEDERAL AGENCIES
AHS Army Health Services Command (Army)
CE Corps of Engineers (Army)
ERD Energy Research and Development Administration
EPA Environmental Protection Agency
GS Geological Survey (Interior)
MC Marine Corps (Navy)
MFS National Marine Fisheries Service (Commerce)
NpE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
CANADA
WQB Environment Canada, Water Quality Branch
WSC Environment Canada, Water Resources Branch
NON-FEDERAL AGENCIES
Delaware
050 Delaware Geological Survey
Maryland
D51 Baltimore County Health Department
D52 City of Baltimore, Water Supply Treatment and
Pumping Division
District of Columbia
D53 Department of Sanitary Engineering
D54 Department of Environmental Health Administration
New Jersey
050 Passaic Valley Water Commission
051 New Jersey State Department of Environmental
Protection
052 North Jersey District Water Supply Commission
054 Delaware River Joint Toll Bridge Commission
New York
P50 New York State Department of Environmental
Conservation
Pennsylvania
S50 Pennsylvania Department of Environmental Resources
98
-------
Virginia
WOO State Water Control Board
West Virginia
X50 West Virginia Department of Natural Resources
X51 West Virginia Department of Health
REGIONS
FEDERAL AGENCIES
AHS Army Health Services Command (Army)
CE Corps of Engineers (Army)
ERD Energy Research and Development Administration
El'A Environmental Protection Agency
FS Forest Service (Agriculture)
GS Geological Survey (Interior)
MC Marine Corps (Navy)
NFE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
NON-FEDERAL AGENCIES
Alabama
A04 Alabama Water Improvement Commission
Florida
EDO Hollywood Reclamation District
EO1 Hillsborough County Health Department
E02 Manatee County Health Department
£03 Central and Southern Florida Flood Control District
E17 State of Florida Department of Pollution Control
E21 Hillsborough County Environmental Protection
Commission
Georgia
E50 Savannah Department of Water and Sewage
E51 Thomasville Water and Light Department
E52 Valdosta Water and Sewer Department
E5 3 Water Works, City of Gainesville
E54 Rome City Manager
E5 5 Water Works, City of Griffin
E56 Board of Water Commissioners, City of Macon
E57 Atlanta Water Works
E58 Columbus Water Works
E60 Georgia Department of Natural Resources
99
-------
Louisiana
151 Louisiana State Department of Health
159 Louisiana Wildlife and Fisheries Commission
Mississippi
L51 City of Jackson Water Works
L52 Pearl River Valley Water Supply District
L53 City of Meridian Water and Sewer Dept.
L54 City of Columbus Light and Water Dept.
L55 Mississippi State Board of Health
North Carolina
QOO North Carolina Department of Human Resources
Q01 North Carolina Department of Natural and Economic
Resources
South Carolina
T50 Agricultural Engineering Department, Clemson University
T51 Greenville Water System
T52 Spartanburg Water Works
T53 South Carolina Department of Health and Environmental
Control
Tennessee
U52 Cleveland Water System
Virginia
WOO State Water Control Board
REGION 4
FEDERAL AGENCIES
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
ERD Energy Research and Development Administration
FS Forest Service (Agriculture)
FWS Fish and Wildlife Service (Interior)
GLB Great Lakes Basin Commission
GS Geological Survey (Interior)
NFE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
CANADA
WQB Environment Canada, Water Quality Branch
WSC Environment Canada, Water Resources Branch
100
-------
NON-FEDERAL AGENCIES
Illinois
GOO Illinois Department of Public Health
G01 Metropolitan Sanitary District of Greater Chicago
G03 Illinois Department of Public Works and Buildings
Indiana
G50 Indiana State Board of Health
Michigan
K50 Michigan Department of Natural Resources
Minnesota
L02 Minnesota Department of Natural Resources
L06 Water, Gas and Sewage Treatment Department,
City of Duluth
L07 Minnesota Ore Operations, USS Corp.
L09 Minnesota Power and Light Company
LI 1 Minnesota Pollution Control Agency
New York
P50 New York State Department of Environmental
Conservation
Ohio
R03 Ohio Environmental Protection Agency
Pennsylvania
S50 Pennsylvania Department of Environmental Resources
Wisconsin
YOO Wisconsin Department of Natural Resources
Y05 Wisconsin Michigan Power Company
REGIONS
FEDERAL AGENCIES
AHS Army Health Services Command (Army)
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
ERD Energy Research and Development Administration
FS Forest Service (Agriculture)
FWS Fish and Wildlife Service (Interior)
GS Geological Survey (Interior)
NWS National Weather Service (Commerce)
TVA Tennessee Valley Authority
101
-------
NON^FEDERAL AGENCIES
Illinois
GOO Illinois Department of Public Health
GO 2 Illinois Department of Registration and Education
G03 Illinois Department of Public Works and Buildings
Indiana
G50 Indiana State Board of Health
Kentucky
100 Kentucky Department for Human Resources
101 Kentucky Department for Natural Resources and
Environmental Protection
New York
P50 New York State Department of Environmental
Conservation
North Carolina
Q01 North Carolina Department of Natural and Economic
Resources
Ohio
ROO Ohio Department of Natural Resources
RO1 The Miami Conservancy District
R02 Ohio River Valley Water Sanitation Commission
R03 Ohio Environmental Protection Agency
Pennsylvania
S50 Pennsylvania Department of Environmental Resources
South Carolina
T53 South Carolina Department of Health and Environmental
Control
Tennessee
U50 Tennessee Wildlife Resources Agency
U51 Tennessee Department of Public Health
Virginia
WOO State Water Control Board
West Virginia
X50 West Virginia Department of Natural Resources
X51 West Virginia Department of Health
102
-------
REGION 6
FEDERAL AGENCIES
EPA Environmental Protection Agency
ERD Energy Research and Development Administration
FS Forest Service (Agriculture)
GS Geological Survey (Interior)
NWS National Weather Service (Commerce)
TVA Tennessee Valley Authority
NON-FEDERAL AGENCIES
Alabama
A01 Geological Survey of Alabama
Georgia
E60 Georgia Department of Natural Resources
Kentucky
100 Kentucky Department for Human Resources
101 Kentucky Department for Natural Resources and
Environmental Protection
Mississippi
L55 Mississippi State Board of Health
North Carolina
QOO North Carolina Department of Human Resources
QO1 North Carolina Department of Natural and Economic
Resources
Tennessee
U50 Tennessee Wildlife Resources Agency
US 1 Tennessee Department of Public Health
U52 Cleveland Water System
U54 Bristol Water Plant
U55 University of Tennessee Experiment Station
U57 Water Resources Research Center, University of
Tennessee
REGION?
FEDERAL AGENCIES
CE Corps of Engineers (Army)
103
-------
EPA Environmental Protection Agency
ERD Hnergy Research and Development Administration
FS Forest Service (Agriculture)
FWS Fish and Wildlife Service (Interior)
GS Geological Survey (Interior)
NWS National Weather Service (Commerce)
NON-FEDERAL AGENCIES
Illinois
GOO Illinois Department of Public Health
G01 Metropolitan Sanitary District of Greater Chicago
G02 Illinois Department of Registration and Education
G03 Illinois Department of Public Works and Buildings
Indiana
C50 Indiana State Board of Health
Iowa
HOO Iowa State Hygenic Laboratory
H02 Des Moines Water Works
H03 Ottumwa Water Works
H05 Iowa Department of Preventive Medicine and
Environmental Health
H06 Agricultural Engineering Department, Iowa State
University
H07 Fort Dodge Department of Municipal Utilities
H09 Des Moines County Drainage District No. 7
H10 Green Bay Levee and Drainage District No. 2
Minnesota
LOO Hennepin County Highway Department
L02 Minnesota Department of Natural Resources
LOS Otter Tail Power Company
L04 Ramsey County Engineer's Department
LOS Northern State Power Company
LOS Blandin Paper Company
L09 Minnesota Power and Light Company
L10 Metropolitan Sewer Board
LI 1 Minnesota Pollution Control Agency
LI 2 Washington County Highway Department
Missouri
MOO Division of Health of Missouri
M02 University of Missouri at Rolla
M03 Metropolitan St. Louis Sewer District
M04 Little River Drainage District
M06 Missouri Clean Water Commission
104
-------
Wisconsin
YOO Wisconsin Department of Natural Resources
Y02 Dairyland Power Cooperative
Y04 Northern States Power Company
REGION 8
FEDERAL AGENCIES
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
GS Geological Survey (Interior)
MFS National Marine Fisheries Service (Commerce)
NFE Naval Facilities Engineering Command
NWS National Weather Service (Commerce)
TVA Tennessee Valley Authority
NON-FEDERAL AGENCIES
Arkansas
B50 Bureau of Environmental Engineering, Arkansas State
Department of Health
B5 1 Arkansas Game and Fish Commission
B52 Arkansas Pollution Control and Ecology
Kentucky
100 Kentucky Department for Human Resources
101 Kentucky Department for Natural Resources and
Environmental Protection
Louisiana
151 Louisiana State Department of Health
152 Houma Light and Water Plant
153 Jefferson Water Works District No. 2
154 Lafourche Water Works District No. 1
155 East Jefferson Water Works District No. 1
156 New Orleans Sewerage and Water Board
158 Utilities Commission Water Treatment Plant, City of
Monroe
159 Louisiana Wildlife and Fisheries Commission, Division of
Water Pollution Control
Mississippi
L50 City of Vicksburg Water Treatment Plant
L55 Mississippi State Board of Health
105
-------
Missouri
MOO Division of Health of Missouri
M04 Little River Drainage District
M06 Missouri Clean Water Commission
Tennessee
U50 Tennessee Wildlife Resources Agency
US 1 Tennessee Department of Public Health
U55 University of Tennessee Experiment Station
REGION 9
FEDERAL AGENCIES
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
FS Forest Service (Agriculture)
GS Geological Survey (Interior)
NWS National Weather Service (Commerce)
CANADA
WQB Environment Canada, Water Quality Branch
WSC Environment Canada, Water Resources Branch
NON-FEDERAL AGENCIES
Minnesota
L02 Minnesota Department of Natural Resources
L03 Otter Tail Power Company
L07 Minnesota Ore Operations, USS Corp.
L09 Minnesota Power and Light Company
LI 1 Minnesota Pollution Control Agency
North Dakota
Q50 North Dakota Game and Fish Department
Q51 North Dakota State Department of Health
Q52 Minot City Water Treatment Plant
Q55 Grand Forks Water Treatment Plant
106
-------
REGION 10
FEDERAL AGENCIES
BR Bureau of Reclamation (Interior)
CE Corps of Engineers (Army)
F.PA Environmental Protection Agency
FS Forest Service (Agriculture)
FWS Fish and Wildlife Service (Interior)
GS Geological Survey (Interior)
NFE Naval Facilities Engineering Command (Navy)
NWS National Weather Service (Commerce)
SCS Soil Conservation Service (Agriculture)
CANADA
WQB Environment Canada, Water Quality Branch
WSC Environment Canada, Water Resources Branch
NON-FEDERAL AGENCIES
Colorado
C50 Board of Water Commissioners, City and County of
Denver
C51 Division of Water Resources, Office of Colorado State
Engineer
C53 Boulder City County Health Department
Iowa
HOO Iowa State Hygenic Laboratory
HOI Director of Lakeside Laboratory, University of Iowa
H06 Agricultural Engineering Department, Iowa State
University
H08 Council Bluffs Water Works
Kansas
H50 Kansas State Department of Health
H51 Board of Public Utilities, Kansas City
H52 Kansas State Board of Agriculture
H53 Topeka Water Department
HS4 Kansas Forestry, Fish, and Game Commission
Minnesota
L02 Minnesota Department of Natural Resources
LI 1 Minnesota Pollution Control Agency
107
-------
Missouri
MOO Division of Health of Missouri
M06 Missouri Clean Water Commission
MOV Union Electric Company, Bagnell Dam
Montana
M50 Montana Fish and Game Department
M51 Montana University Joint Water Resources Research
Center
M52 Montana State Department of Health and Environmental
Sciences
M53 Montana Department of Natural Resources and
Conservation
Nebraska
NOO Nebraska Game and Parks Commission
N01 State of Nebraska Department of Environmental Control
N02 Metropolitan Utilities District, City of Omaha
N03 Soil and Water Testing Laboratory, University of
Nebraska
North Dakota
Q50 North Dakota Game and Fish Department
Q51 North Dakota State Department of Health
Q52 Minot City Water Treatment Plant
Q53 City of Bismarck Water Department
Q54 City of Dickinson Water Treatment
South Dakota
UOO Water Resources Research Institute, South Dakota State
University
U01 East Dakota Conservancy Sub-District
Wyoming
Y50 City of Casper Board of Public Utilities
Y51 Sheridan Water Department
Y52 Wyoming State Engineer
Y53 Water Resources Research Institute, University of
Wyoming
REGION 11
FEDERAL AGENCIES
BR Bureau of Reclamation (Interior)
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
108
-------
FS Forest Service (Agriculture)
FWS Fish and Wildlife Service (Interior)
CiS Geological Survey (Interior)
NFE Naval Facilities Engineering Command
NW'S National Weather Service (Commerce)
NON-FEDERAL AGENCIES
Arkansas
B-50 Arkansas State Department of Health
QS] Arkansas Game and Fish Commission
B5Z Arkansas Pollution Control and Ecology
Colorado
C51 Division of Water Resources, Office of State Engineer
CS2. City of Colorado Springs, Water Division
C54 Pueblo Board of Water Works
Kansas
H50 Kansas State Department of Health
H52 Kansas State Board of Agriculture
H54 Kansas Forestry, Fish and Game Commission
Louisiana
151 Louisiana State Department of Health
157 Bossier City Water Plant
159 Louisiana Wildlife and Fisheries Commission
160 City of Shreveport Department of Water Utilities
Missouri
MOO Division of Health of Missouri
M06 Missouri Clean Water Commission
Oklahoma
R50 Oklahoma State Department of Health
Texas
VOO Texas Water Development Board
REGION 12
FEDERAL AGENCIES
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
GS Geological Survey (Interior)
109
-------
IBW International Boundary and Water Commission
NFE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
NON-FEDERAL AGENCIES
Louisiana
159 Louisiana Wildlife and Fisheries Commission
Texas
VOO Texas Water Development Board
REGION 13
FEDERAL AGENCIES
BR Bureau of Reclamation (Interior)
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
FS Forest Service (Agriculture)
GS Geological Survey (Interior)
IBW International Boundary and Water Commission
NWS National Weather Service (Commerce)
NON-FEDERAL AGENCIES
Colorado
C51 Division of Water Resources, Office of Colorado State
Engineer
New Mexico
POO New Mexico State Engineer's Office
REGION 14
FEDERAL AGENCIES
BR Bureau of Reclamation (Interior)
EPA Environmental Protection Agency
FS Forest Service (Agriculture)
FWS Fish and Wildlife Service (Interior)
110
-------
GS Geological Survey (Interior)
NWS National Weather Service (Commerce)
NON-FEDERAL AGENCIES
Colorado
C50 Board of Water Commissioners, City and County of
Denver
C51 Division of Water Resources, Office of Colorado State
Engineer
C52 City of Colorado Springs, Water Division
Utah
V50 Utah State Health Department
V52 Utah Department of Natural Resources
Wyoming
Y52 Wyoming State Engineer
REGION 15
FEDERAL AGENCIES
BLM Bureau of Land Management (Interior)
BR Bureau of Reclamation (Interior)
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
FS Forest Service (Agriculture)
GS Geological Survey (Interior)
IBW International Boundary and Water Commission
NWS National Weather Service (Commerce)
SCS Soil Conservation Service (Agriculture)
NON-FEDERAL AGENCIES
Arizona
BOO Salt River Project
B01 Water Resources Research Center, University of Arizona
B02 Roosevelt Irrigation District
BOS Arizona Game and Fish Department
B04 Maricopa County Municipal Water Conservation
BOS Gila Water Commissioner
111
-------
California
COO California Department of Water Resources
Nevada
N50 Nevada State Health Division
Utah
V50 Utah State Health Department
REGION 16
FEDERAL AGENCIES
BLM Bureau of Land Management (Interior)
BR Bureau of Reclamation (Interior)
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
ERD Energy Research and Development Administration
FS Forest Sendee (Agriculture)
GS Geological Survey (Interior)
NFE Naval Facilities Engineering Command (Navy)
NWS National Weather Service (Commerce)
SCS Soil Conservation Service (Agriculture)
NON-FEDERAL AGENCIES
California
COO California Department of Water Resources
Idaho
F50 Idaho State Fish Hatchery
F52 Idaho Department of Environmental and Community
Services
Nevada
N50 Nevada State Health Division
N51 Walker River Irrigation District
Utah
V50 Utah State Health Department
V51 Metropolitan Water District of Salt Lake City
V53 Salt Lake County Water Conservancy District
V54 Salt Lake City Water Supply and Waterworks
V55 Ogden Bay Waterfowl Management Area
V56 Clear Lake Waterfowl Management Area
V58 Utah Geological and Mineralogical Survey
V59 Ogden River Water Users
V60 Weber Distribution System
112
-------
REGION 17
FEDERAL AGENCIES
HPA Bonnevillc Power Administration (Interior)
BR Bureau of Reclamation (Interior)
CF Corps of Engineers (Army)
HPA Environmental Protection Agency
I- KD Energy Research and Development Administration
I;S Forest Service (Agriculture)
I-'WS Fish and Wildlife Service (Interior)
CiS Geological Survey (Interior)
NIC Marine Corps (Navy)
MFS National Marine Fisheries Service (Navy)
N'FH Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
SCS Soil Conservation Service (Agriculture)
CANADA
WQB Environment Canada, Water Quality Branch
WSC Environment Canada, Water Resources Branch
NON-FEDERAL AGENCIES
Idaho
F5l Water Resources Research Institute, University of Idaho
F52 Idaho Department of Environmental and Community
Services
Montana
M50 Montana Fish and Game Department
M53 Montana Department of Natural Resources and
Conservation
Nevada
N50 Nevada State Health Division
Oregon
SOO Department of Forest Engineering, Oregon State
University
SOI Oregon Wildlife Commission
S02 Douglas County Water Resource Survey
SOS Oregon State Engineer
S04 Fish Commission of Oregon
S13 Portland General Electric
113
-------
Washington
XOO State of Washington, Department of Ecology, Water
Resources Division
X01 Public Utility District No. 1, Skagit County
X02 Chelan County PUD No. 1
X03 College of Fisheries, University of Washington
X04 College of Engineering Research, Washington State
University
X05 Department of Zoology, University of Washington
X06 City of Bremerton Water Department
X07 City of Everett Department of Water
X08 Seattle Water Department
X09 City of Tacoma, Department of Public Utilities
XI2 Municipality of Metropolitan Seattle
XI6 King County, Washington, Department of Public Works
X19 The Washington Water Power Company
X20 Douglas County Public Utility District
X21 Public Utility District of Grant County
X24 Puget Sound Power and Light Company
X25 City of Seattle
REGION 18
FEDERAL AGENCIES
BLM Bureau of Land Management (Interior)
BR Bureau of Reclamation (Interior)
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
FS Forest Service (Agriculture)
GS Geological Survey (Interior)
IBW International Boundary and Water Commission
MC Marine Corps (Navy)
MFS National Marine Fisheries Service (Commerce)
NFE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
SCS Soil Conservation Service (Agriculture)
NON-FEDERAL AGENCIES
California
COO California Department of Water Resources
CO I Los Angeles County Flood Control District
C03 Alameda County Water District
C04 County of Sacramento, Water Resources Division
COS Whitewater Mutual Water Company
114
-------
C06 California Water Quality Control Board
C09 Ventura County Flood Control District
CIO San Diego Department of Sanitation and Flood Control
Cl 1 Orange County Flood Control District
Cl 2 Merced Irrigation District
C13 Turlock Irrigation District
C.14 Tridam Irrigation District
C15 Pacific Gas and Electric
C16 Oroville-Wyandotte Irrigation District
C17 Mosquito Irrigation District
CIS Contracting Entities
Cl9 East Bay Municipal Utility District
C20 Modesto Irrigation District
C21 El Nido Irrigation District
C22 Madera Irrigation District
C23 Hetch Hetchy Water Supply, City and County of
San Francisco
C24 Southern California Edison Company
C25 Pacific Power and Light
C26 Kings River Water Association
C27 Fresno Irrigation District
C28 Kaweah and St. Johns Water Association
C29 Tulare Irrigation District
C30 Delano-Earlimart Irrigation District
C31 Kern County Land Company
C32 Buena Vista Water Storage District
C33 Terra Bella Irrigation District
C34 Sausalito Irrigation District
C35 Monterey County Flood Control and Water Conservation
District
C36 San Luis Obispo County Flood Control and Water
Conservation District
C37 Montecito County Water District
C38 Santa Barbara County Flood Control and Water
Conservation District
C39 Metropolitan Water District of Southern California
C40 Marin Municipal Water District
C41 Marin, North, County Water District
C42 Sonoma County Flood Control and Water Conservation
District
C43 Alameda County Flood Control and Water Conservation
District
C44 Santa Clara Valley Water District
C45 Tule Irrigation District
C46 Montague Water Conservation District
C47 City of Los Angeles, Dept. of Water and Power
C48 Palm Springs Water Company
C49 Escondido Mutual Water Company
115
-------
ZOO San Bernardino County Flood Control District
Z01 San Antonio Water Company
Z02 Temescal Water Company
Z03 Riverside County Flood Control and Water Conservation
District
Z04 Ventura River Municipal Water District
Z05 Ventura County Water Resources Division
Z06 United Water Conservation District
Z07 Kings River Water Conservation Board
Z08 La Canada Irrigation District
Z09 San Gabriel Electric Company
Nevada
N50 Nevada State Health Division
N54 Nevada Irrigation District
Oregon
SO 1 Oregon Wildlife Commission
SOS Oregon State Engineer
REGION 19
FEDERAL AGENCIES
EPA Environmental Protection Agency
FS Forest Service (Agriculture)
GS Geological Survey (Interior)
MC Marine Corps (Navy)
MFS National Marine Fisheries Service (Commerce)
NFE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
SCS Soil Conservation Service (Agriculture)
CANADA
WSC Environment Canada, Water Resources Branch
NON-FEDERAL AGENCIES
Alaska
A50 Chugach Electrc Association
A51 Alaska Department of Highways
ASS Alaska Department of Environmental Conservation
116
-------
Washington
X03 College of Fisheries, University of Washington
REGION 20
FEDERAL AGENCIES
CE Corps of Engineers (Army)
EPA Environmental Protection Agency
GS Geological Survey (Interior)
MC Marine Corps (Navy)
NFE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Interior)
NON-FEDERAL AGENCIES
Hawaii (and other Pacific Islands)
FOO Board of Water Supply, City and County of Honolulu
F01 Department of Water, County of Kauai
F02 Board of Water Supply, County of Maui
F03 Board of Water Supply, County of Hawaii
F04 Department of Hawaiian Home Lands, State of Hawaii
F05 Department of Land and Natural Resources, State of
Hawaii, Division of Fish and Game
F06 Department of Land and Natural Resources, State of
Hawaii, Division of Water and Land Development
F07 Public Utility Agency Water Division, Government
of Guam
F08 Ryukyu Industrial Research Institute, Government of
Ryukyu Islands
F09 Ryukyu Meteorological Agency, Government of
Ryukyu Islands
REGION 21
FEDERAL AGENCIES
GS Geological Survey (Interior)
NFE Naval Facilities Engineering Command (Navy)
NOS National Ocean Survey (Commerce)
NWS National Weather Service (Commerce)
117
-------
NON-FEDERAL AGENCIES
Puerto Rico
TO-1 Puerto Rico Water Resources Authority
118
-------
APPENDIX B
MONITOR SITES CONSIDERED FOR ANALYSIS
This appendix contains a listing of the 104 monitor sites
that were considered for analysis in this study- All the sites
listed were considered to be sufficiently close to some populat-
ed area to warrant analysis. Further investigation sometimes
revealed that data were not available or some other factor was
present and prevented analysis. Those sites that could not be
used and the reasons why are identified in Table B-l.
119
-------
TABLE B-1. LIST OF ALL SITES CONSIDERED FOR DAILY OR HOURLY ANALYSIS
State
Alabama
Colorado
Georgia
Illinois
Louisiana
Station name
Coosa R. at Verbena
Coosa R. at Gadsen
S. Platte R., 60 ave.
S. Platte R., 88 ave.
Burlington Ditch at
York St.
Sand Cr. at
Burlington Ditch
Chattahoochee R. at
Atlanta
Peachtree Cr. near
Atlanta1'
Ocmulgee R. near
Warner-Robins1
Calumet R. STW1
Chicago R. Bridge*
Chicago Sanitation
and Ship Canal
at Lockport*
Bayou Teche at
Olivier
Houma Nav. Canal
near Dulac*
Agency* Lat.
GS
EPA
EPA
EPA
EPA
EPA
EPA
GS
GS
G01
G01
G01
GS
GS
324756
340057
394826
395115
394802
394837
335132
335133
324017
413946
415333
413408
295718
292306
Long.
862602
855843
1045730
1045615
1045730
1045659
842716
842716
833611
873940
873832
880441
914254
904347
Water discharge . Distance to
Site Established , ram . . ..
available? ram gage (mt)
gage
Stream
Stream
Stream
Stream
Canal
Stream
Stream
Stream
Stream
Canal
Stream
Canal
Stream
Stream
1974
1971
1968
1968
1967
1967
1960
_
1970
1969
1969
1969
1972
1973
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
Yes
No
Montgomery —
Gadsen —
Denver <10
Denver <10
Denver <10
Denver <10
Atlanta 10
Atlanta 10
Macon 20
Chicago 20
Chicago 5
Chicago 35
_ _
_ _
Applicability
to study
code
4
4
1
1
1
1
1
1
4
2
1
2
_
_
•The agency codes, taken from OWDC, identify the agency that was in charge of the data; specifically, GS * Geological Survey, EPA = Environmental Protection Agency, GDI = Metn
politan Sanitation District of Greater Chicago, R02 = ORSANCO, WDNR <= Wisconsin Department of Natural Resources.
tShe was later dropped.
(continued)
-------
TABLE B-1 (continued).
State
Louisiana
(continued)
Massachusetts
Mississippi
Maine
Station name
Ouachita R. at
Monroe
Blackstone R. at
Millville
Connecticut R. at
W. Springfield
Connecticut R. at
Agawamt
Hoosic R. below
Williamston
Merrimack R. at W.
Newbury
Merrimac R. above
Concord R. at
Lowell*
N. Nashua R. near
Leominster
Quinebaug R. near
Dudley
Westfield R. at
Westfield
Chicopee R. at
Chicopee Falls
Yellow Cr. near
Doskie
Vermillion R. near
Empire City
Agency
GS
GS
GS
GS
GS
GS
GS
GS
GS
GS
GS
GS
GS
Lat.
323019
420116
420546
420257
424428
424838
423857
423006
420140
420559
420937
345402
444000
Long.
920732
713404
723543
723635
731247
710002
711950
714323
715722
723828
723452
881735
930317
Primary „
.. . Water discharge Distance to
Site Established .. . . , raln i -\
available? ram gage (mi)
gage
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
1954
1969
1972
1968
1968
1969
1967
1968
1968
1972
1973
1973
1974
Yes Shreveport 90
Yes
Yes Hartford 25
Yes Hartford 25
Yes
Yes
No Boston 25
Yes
Yes
Yes Hartford 25
Yes Hartford 30
Yes
Yes
Applicability
to study
code
4
4
2
2
4
4
4
4
4
2
4
4
4
(continued)
-------
TABLE B-1 (continued).
State
Missouri
New Jersey
North
Carolina
Station name Agency
Center Cr. near
Carterville
James R. near
Boaz
Wilsons Cr. near
Battlefield
Wilsons Cr. near
Springfield
Delaware R. at
Trenton
Manasquan R. at
Squankum
Pompton R. at
Two Bridges
Passaic R, at
Little Falls*
Raritan R. at S.
Bound Brook
Cape Fear R. at
Channel Marker 50
Muddy Cr. 500' below
effluent
N. Fork Muddy Cr.
Muddy Cr. atS.R.
1485
Muddy Cr. at S.R.
2991
GS
GS
GS
GS
GS
GS
GS
GS
GS
EPA
EPA
EPA
EPA
EPA
Lat.
370826
370025
370702
370706
401318
400947
405352
405305
403305
341533
360312
360220
355690
360030
Long.
942257
932150
932414
932414
744642
740921
741622
741335
743254
775624
794424
801810
802100
802000
Primary r.
Water discharge Distance to
Site Established ...... rain . . ..
available? ram gage (mi)
gage
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
1962
1972
1972
1967
1923
1970
1969
1963
-
1974
1974
1974
1974
1974
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Springfield 70
Springfield 25
Springfield 20
Springfield 20
Trenton 10
-
Trenton 10
Newark 15
- -
Wilmington <10
Greensboro —
Winston-Salem —
Winston-Salem —
Winston-Salem —
Applicability
to study
code
4
4
2
2
1
4
1
1
4
1
2
2
2
2
(continued)
-------
TABLE B-1 (continued).
[\J
State Station name Agency
Roanoke, R. 1 mi.
upstream of Welch
Cr.
Neuse R. at U.S. 70
Neuse R. at U.S. 42
Neuse R. at U.S. 98
Neuse R. at U.S. 50
Ohio Ashtabula R. at
Ashtabula
Black R. at Elyria
Black Fork at
Loudonville
Blanchard R. near
Findlay
Cuyahoga R. at
Independence
Cuyahoga R. at Old
Portage
Cuyahoga R. at
Superior St. Bridget
Cuyahoga R. at W.
3rd St. Bridge
Grand R. near
Painsville
Hocking R. below
Athens
Huron R. below
Milan
EPA
EPA
EPA
EPA
EPA
GS
GS
GS
GS
GS
GS
GS
GS
GS
GS
GS
Lat.
355200
353041
353815
355840
360052
415400
412442
403809
410321
412343
410808
412939
412917
414409
391939
412006
Long.
764700
782100
782347
78380"
789130
804744
820545
821422
834117
813748
813250
814212
814107
811559
820018
823438
Water discharge Distance to
Site Established .. . . , rain . . ..
available? ram gage (mi)
gage
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
1974
1975
1975
1975
1975
1968
1966
1968
1965
1948
1965
1950
1966
1950
1966
1970
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yej
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
-
_
_
-
Cleveland
Cleveland
Mansfield
Toledo
Cleveland
Cleveland
Cleveland
Cleveland
Cleveland
Parkersburg,
WV.
_
-
—
—
-
60
20
20
50
10
25
10
10
25
30
_
Applicability
to study
code
4
4
4
4
4
2
4
4
4
1
1
1
1
4
4
3
(continued)
-------
TABLE B-1 (continued).
State Station name
Ohio Little Miami R. at
(continued) Miamiville
Little Miami R.
near Spring Valley
Mad R. near Dayton
Mahoning R.atOH.-
PA. State Line
below Lowellville
Maumee R. at
Defiance
Maumee R. at
Mouth at Toledo
Maumee R. at
Waterville
Muskingham R. at
McConnelsville
Ohio R. at West End
(Cincinnati)^
Ohio R. at
Andersons Ferryt
Portage R. at
Woodville
Sandusky R. near
Freerrtont
Sandusky R. near
Upper Sandusky
Scioto R. at
Chillicothe
Scioto R. below
Shadeville
Agency
GS
GS
GS
GS
GS
GS
GS
GS
R02
R02
GS
GS
GS
GS
GS
Lat.
391238
393500
394750
410153
411643
414136
413000
393842
Long.
841733
840149
840519
803110
842307
832820
834246
815100
None Published
None Published
412658
412212
405102
392029
394737
832129
830610
831523
82581,6
830040
. , Primary
_. _ „ ... . . Water discharge . Distance to
Site Established ., ., , raln . / ,
available? ram gage (mi)
gage
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
1970
1968
1961
1967
1966
1967
1950
1950
1961
1961
1971
1970
1965
1950
1965
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Cincinnati
Dayton
Dayton
Youngstown
Toledo
Toledo
Toledo
Parkersburg,
WV
Cincinnati
Cincinnati
— .
—
Mansfield
Columbus
Columbus
15
15
10
10
50
5
25
35
10
10
_
_
40
40
5
Applicability
to study
code
3
3
3
1
4
1
3
4
1
1
3
3
4
1
1
(continued)
-------
TABLE B-1 (continued).
N>
Ul
State
Ohio
(continued)
Oregon
Pennsylvania
Station name
Tuscarawas R. at
Navarre
S. Umpqua R. near
Brockway
Willamette R. at
Portland*
Willamette R, at
Oregon City*
Willamette R. above
Oregon City*
Allegheny R. at
Oakmont*
Beaver R. at
Beaver Falls*
Delaware R. at
Bristol
Delaware R. at
Chester
Delaware R. at
Easton
Delaware R. at
Torresdale Ints1--
Kiskiminetas R. at
Vandergrift*
Monongahela R. at
S. Pittsburgh*
Lehigh R. at Easton
Agency
GS
GS
EPA
EPA
EPA
R02
R02
GS
GS
GS
GS
R02
R02
GS
Lat.
404336
431320
453349
452154
452030
Long.
813147
1232445
1224317
1223603
1223733
None Published
4048
400455
395012
404243
400157
403620
402436
404112
8019
745158
752200
751148
745946
793315
795715
751232
Site Established
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Estuary
Esutary
Stream
Stream
Stream
Stream
Stream
1970
1970
1962
1968
1968
1962
1961
1949
1961
1947
1949
_
-
1961
Primary
Water discharge . Distance to
available? rain gage (mi)
gage
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
No
Yes
Yes
_
Eugene
Portland
Portland
Portland
Pittsburgh
Pittsburgh
Trenton
Philadelphia
Allentown
Philadelphia
Pittsburgh
Pittsburgh
Allentown
_
60
5
20
20
15
40
10
20
15
10
30
10
15
Applicability
to study
code
3
4
1
4
4
3
4
3
2
3
2
4
1
3
(continued)
-------
TABLE B-1 (continued).
State Station name
Pennsylvania Schuylkill R. at
(Continued) Philadelphia
W. Br. Brandywine
Cr. at Modena
Tennessee W. Fork Stones R.
at Manson Park
Texas Trinity R. below
Dallas
Virginia N. Fork Holston R.
near Gate Cityt
[ i Washington Green-Duwamish R.
I\J Station 307
O1
West Ohio R. at
Virginia Huntington^
Wisconsin Fox R. at Menasha
Fox R. at Appleton
Fox R. at Rapide
Croche
Fox R. at DePere
Fox R. at Green
Bay
Wisconsin R. at
Wasau
Agency
GS
GS
GS
GS
EPA
EPA
R02
WDNR
WDNR
WDNR
WDNR
WDNR
WDNR
Lat.
395800
395742
355125
324227
363631
Long.
751120
754806
862443
964408
823405
None Published
3825
441151
441517
442653
443129
445725
8225
882653
882443
880402
880026
893808
Site
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Stream
Established
1945
1974
1973
1967
1969
1969
1961
1971
1971
1971
1971
1971
1971
Water discharge
available?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Primary
rain
gage
Philadelphia
_
Nashville
Dallas
Bristol, TN.
_
Huntington
Green Bay
Green Bay
-
Green Bay
Green Bay
Green' Bay
Distance to
rain gage (mi)
10
_
25
20
25
5
30
25
-
5
5
90
Applicability
to study
code
1
4
4
1
4
_
1
4
4
4
3
1
4
(continued)
-------
TABLE B-1 (continued).
State
Wisconsin
(continued)
Station name
Wisconsin at
Moisinee
Wisconsin at
Dubay Dam
Wisconsin at
Biron
Wisconsin at Port
Edwards
Wisconsin at
Petenwell
Agency Lat.
WDNR
WDNR 442450
WDNR
WDNR 442552
WDNR
Long. Site
— Stream
895121 Stream
— Stream
894647 Stream
— Stream
Established
1971
1971
1971
1971
1971
Water discharge
available?
Yes
Yes
Yes
Yes
Yes
Primary _.
Distance to
ram . , ..
ram gogc (mi)
gage
£treen Bay —
Green Bay 90
Green Bay —
Green Bay 90
Green Bay —
Applicability
to study
code
4
-
4
_
-------
APE-ENDIX C
RESULTS OF DAILY CORRELATION ANALYSIS
This appendix contains the results of all the daily corre-
lation analyses for both flow and rainfall. Table C-l contains
the results of flow correlation analysis at USGS monitor sites.
Table C-2 contains the results of flow correlation analyses for
STORET data base monitors. Table C-3 contains the results of the
flow correlation analyses for the WDNR monitor sites. Table C-4
contains the results of daily rainfall correlation analysis at
the USGS monitor sites.
In the following tables the stations are grouped by state.
The states are in alphabetical order and the stations in a group
are alphabetized. Three types of statistics are included in each
table. These are the probability of a worse than average dissolv-
ed oxygen deficit, the strength of the deficit, and the average
percent saturation which occurs when the dissolved oxygen def-
icit is worse than average. The three types of statistics answer
the following six questions about a given monitor site:
• What is the probability at this station that a worse-
than-average (seven-day moving) DO deficit will occur
on a wetter-than-average (seven-day moving) day?
• What is the probability at this station that a worse-
than-average (seven-day moving) DO deficit will occur
on a dryer-than-average (seven-day moving) day?
» If the DO deficit on a particular wetter-than-average
day is worse than average, how much worse is it?
* If the DO deficit on a particular dryer-than-average
day is worse than average, how much worse it it?
• On wetter-than-average days when the DO deficit is
worse than average, what percentage of saturation is
present?
• On dryer-than-average days when the DO deficit is worse
than average, what percentage of saturation is present?
128
-------
TABLE C-1. USGS MONITOR SITE SUMMARY OF DAILY FLOW CORRELATION ANALYSIS RESULTS
to
Monitor site*
ALABAMA
Coosa R. near Verbena
(at Jordan Dam near
Wetumbwa)
LOUISIANA
Bayou Teche at Olivier
(at Keystone Lock near
St. Martinsville)
Houma Nav. Canal at
Dulac
Quachita R. at Monroe
MASSACHUSETTS
Blackstone R. at MUlville
Water
Year
(19-)
74
75
72-
74
75
74<
75
69
70
71
72
69
70
71
72
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
- ABORTED
.44
- ABORTED
.52
- ABORTED
.45
.50
.51
.38
- ABORTED
.43
.48
.48
—
.60
—
.53
—
.60
.53
.46
.46
—
.65
.65
.63
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
—
1.16
—
1.15
—
1.39
1.19
1.28
1.28
—
1.09
1.06
1.15
Dryer than
avg. days
—
1.46
—
1.18
—
1.35
1.21
1.33
1.36
_
1.11
1.11
1.09
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
—
.62
—
.24
—
.29
.42
.35
.33
_
.45
.44
.62
Dryer than
avg. days
—
.72
-
.31
—
.47
.33
.38
.43
—
.49
.45
.55
"Stations are grouped by state, states are in aiphadettcal order, stations are alphabetized within group; ABORTED indicates that less than 180 days of data were available that
year.
(continued)
-------
TABLE C-1 (CONTINUED).
h-
OJ
O
Monitor site
MASSACHUSETTS (continued)
Connecticut R. at W.
Springfield (Thompsonville)
Hoosic R. below
Williamstown
Merrimack R. at W.
IMewbury
North Nashau R. near
Lancaster (Leominster)
Quinebaug R. near Dudley
(at Quinebaug CN.)
Westfield R. at Westfield
Water
Year
(19-)
72
73
74
75
73
74
75
72
73
74
75
76
68
69
70
71
72
69
70
71
72
72
73
74
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
- ABORTED
.46
.43
.41
- ABORTED
.46
- ABORTED
.46
- ABORTED
.44
.46
.47
ABORTED
.55
.63
.47
.32
.38
.33
.47
.39
- ABORTED
.46
.52
—
.60
.60
.61
—
.58
-
.57
—
.65
.52
.45
—
.58
.50
.52
.69
.61
.52
.67
.66
—
.51
.47
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
—
1.76
1.27
1.21
—
1.37
-
1.63
-
1.52
1.50
1.09
—
1.14
1.17
1.14
1.20
1.38
2.14
1.68
1.28
—
1.42
1.30
Dryer than
avg. days
—
2.86
1.25
1.25
—
1.26
-
1.54
—
1.49
2.48
1.13
—
1.13
1.17
1.12
1.13
1.48
1.79
1.50
1.24
—
2.46
1.26
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
—
.72
.81
.81
_
.64
-
.74
—
.55
.63
.63
_
.48
.53
.48
.66
.83
.88
.74
.70
—
.76
.79
Dryer than
avg. days
—
.80
.79
.82
_
.60
-
.74
—
.67
.72
.61
—
.53
.62
.57
.59
.74
.83
.67
.71
—
.75
.80
(continued)
-------
TABLE C-1 (CONTINUED).
Monitor site
Water
Year
(19-)
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
Dryer than
avg. days
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
Dryer than
avg. days
MASSACHUSETTS (continued)
Westfield R. at Westfield
MISSISSIPPI
Yellow Cr. near Doskie
MAINE
Vermillion R. near Empire
MISSOURI
Center Cr. near Carterville
James R. near Boaz
Wilsons Cr. near
Battlefield
75
76
73
74
75
74
75
72
73
74
75
76
72
73
74
75
76
73
74
.62
.30
- ABORTED
.58
- ABORTED
- ABORTED
- ABORTED
.53
- ABORTED
.49
.43
- ABORTED
- ABORTED
.44
.40
.43
.44
.36
.51
.48
.48
—
.50
-
-
—
.52
—
.56
.58
-
—
.58
.64
.62
.57
.63
.61
1.69
1.68
-
2.16
—
-
—
1.68
—
1.12
1.10
-
—
1.14
1.11
1.11
1.09
1.06
1.06
2.65
1.79
—
2.80
—
-
—
1.71
—
1.11
1.16
-
_
1.13
1.14
1.11
1.11
1.10
1.10
.78
.89
-
.74
—
-
—
.59
—
.58
.46
-
—
.53
.42
.32
.18
.09
.06
.88
.82
—
.76
—
-
—
.57
—
.61
.56
-
—
.59
.50
.48
.31
.13
.10
(continued)
-------
TABLE C-1 (CONTINUED).
CO
N>
Monitor site
MISSOURI (CONTINUED)
Wilsons Cr. near
Battlefield (continued)
Wilsons Cr. near
Springfield
NEW JERSEY
Delaware R. at
Trenton
Manasquan R. at Squankum
(This station later abandoned
for lack of urban area. It
is near a swamp.)
Passaic R. at Two Bridges
Raritan R. near South Bound
Brook (below Callo Dam at
Water
Year
(19-)
75
76
72
73
74
75
76
72
73
74
75
76
70
71
72
73
74
72
73
74
74
75
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.53
.57
- ABORTED
.70
.54
.81
.77
.58
.64
.44
.61
.51
.52
.61
.66
.55
.48
.48
.34
- ABORTED
- ABORTED
.60
.57
.54
_
.42
.48
.39
.44
.49
.45
.47
.40
.51
.48
.58
.52
.44
.52
.69
.66
-
—
.37
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.10
1.04
_
1.27
1.30
1.27
1.28
2.12
1.83
2.75
2.79
4.71
1.20
1.12
1.22
1.16
1.26
1.07
1.12
-
—
1.60
Dryer than
avg. days
1.09
1.05
1.20
1.18
1.16
1.14
4.60
6.66
2.74
3.73
3.66
1.16
1.13
1.11
1.13
1.22
1.09
1.21
-
1.65
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.10
.02
.60
.58
.56
.58
.85
.85
.89
.91
.91
.52
.51
.50
.68
.67
.52
.48
-
.72
Dryer than
avg. days
.16
.04
.64
.64
.69
.65
.87
.93
.91
.96
.97
.56
.55
.56
.71
.70
.42
.50
—
_
.80
Boundbrook)
(continued)
-------
TABLE C-1 (CONTINUED).
OL)
LO
Monitor site
OHIO
Ashtabula R. at Ashtabula
Black R. at Elyria
Black Fork at
Loudonville
Blanchard R. near Findlay
Cuyahoga R. at
Independence
Water
Year
(19--)
72
73
74
75
76
70
71
72
73
74
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.44
.52
.42
.63
.58
- ABORTED
- ABORTED
- ABORTED
.54
.49
.57
.48
.52
.47
.56
.55
.53
.47
.49
.41
.39
.54
.52
.42
.56
.54
.47
.57
.49
.50
—
—
—
.52
.53
.45
.51
.46
.50
.45
.39
.55
.59
.52
.60
.58
.48
.56
.53
.47
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.29
1.40
1.50
1.23
1.25
—
—
—
1.81
1.43
2.59
?
1.24
1.21
1.22
2.96
1.19
1.14
1.18
1.23
1.23
1.24
1.15
1.22
1.24
Dryer than
avg. days
1.35
2.02
1.32
1.21
1.21
—
=-
—
1.47
1.61
1.27
1.49
1.12
1.12
1.12
.67
.26
.11
.16
.19
1.16
1.24
1.15
1.21
1.14
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.72
.76
.76
.64
.68
—
—
—
.57
.57
.60
.71
.67
.67
.67
.61
.63
.42
.47
.50
.69
.70
.62
.69
.71
Dryer than
avg. days
.67
.78
.72
.72
.69
—
—
—
.63
.54
.65
.73
.68
.74
.72
.69
.60
.44
.55
.61
.62
.72
.65
.74
.75
(continued)
-------
TABLE C-1 (CONTINUED).
Monitor site
OHIO (continued)
Cuyahoga R. at Old Portage
Cuyahoga R. at W. 3rd St.
Bridge
Grand R. near Painesville
Hocking R. below Athens
Huron R. below Milan
Water
Year
(19-)
72
73
74
75
76
72
73
74
75
76
75
76
72
73
74
75
76
70
71
72
73
74
75
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.45
.37
.39
.47
.41
.47
.44
.43
.41
.45
.45
.34
.46
.50
.42
.57
.67
- ABORTED
.49
.43
.41
.57
.50
.51
.63
.58
.49
.61
.54
.54
.59
.51
.52
.67
.63
.55
.46
.55
.60
.45
—
.52
.37
.54
.45
.49
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
2.27
1.47
1.24
1.36
1.25
1.14
1.10
1.09
1.13
1.17
1.16
1.26
2.02
1.74
1.39
1.33
1.44
—
1.54
3.33
2.21
1.49
1.76
Dryer than
avg. days
1.18
3.33
1.35
1.28
1.22
1.21
1.11
1.09
1.12
1.13
1.18
1.56
1.33
1.60
1.34
1.31
1.40
—
1.51
3.14
5.60
1.50
2.24
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.63
.72
.75
.58
.63
.41
.45
.38
.43
.38
.72
.63
.65
.82
.84
.79
.84
__
.70
.81
.87
.77
.67
Dryer than
avg. days
.67
.76
.75
.74
.74
.35
.37
.37
.41
.39
.68
.67
.64
.82
.84
.82
.86
.78
.83
.80
.79
.72
(continued)
-------
TABLE C-1 (CONTINUED).
Monitor site
OHIO (continued)
Little Miami R. at
Miamiville (Milford)
Little Miami R. near
Spring Valley
Mad R. near Dayton
Mahoning R. at OH.-PA.
State Line (Lowellville)
Maumee R. at Defiance
Water
Year
(19-)
71
72
73
74
72
73
74
75
76
72
73
74
75
76
73
74
75
76
72
73
74
75
76
Probability of greater
than average DO deficit
Wetter than
avg. days
.59
.57
.66
.67
.58
.49
.65
.67
.59
.63
.50
.49
.56
.50
.50
.38
.48
.39
.51
.66
.56
.60
.55
Dryer than
avg. days
.47
.37
.40
.37
.42
.40
.43
.45
.39
.47
.45
.52
.48
.54
.54
.58
.54
.56
.42
.41
.44
.38
.42
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.51
1.85
1.61
1.56
1.17
1.49
1.33
1.18
1.23
1.25
1.31
1.19
1.23
1.16
1.16
1.15
1.10
1.22
1.94
1.37
2.38
7.64
4.92
Dryer than
avg. days
2.07
2.37
2.01
1.42
.15
.30
.18
.14
.13
.13
.19
1.13
1.19
1.19
1.29
1.13
1.09
1.14
1.44
1.29
3.22
1.77
11.23
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.68
.84
.83
.84
.62
.68
.73
.72
.63
.71
.76
.73
.73
.65
.53
.54
.51
.47
.74
.72
.63
.77
.78
Dryer than
avg. days
.71
.81
.85
.86
.64
.71
.74
.76
.67
.71
.79
.72
.75
.71
.53
.52
.57
.54
.77
.67
.71
.76
.83
(continued)
-------
TABLE C-1 (CONTINUED).
Monitor site
OHIO (continued)
Maume R. at Mouth
at Toledo
Maumee R. at
Waterville
Muskingum R. at
McConnelsville
Portage R. at
Woodville
Sandusky R. near
Fremont
Water
Year
(19-)
69-
70
71
72
69
70
71
72
73
74
75
76
71
72
73
74
75
70
71
72
73
74
75
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
- ABORTED
.53
.49
.42
- ABORTED
- ABORTED
.58
- ABORTED
.57
.54
.45
- ABORTED
.60
.54
.68
.52
.52
.53
.45
.42
.47
.52
.52
.61
—
—
.49
—
.43
.43
.44
-
.44
.46
.40
.64
.54
.47
.68
.53
.48
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
6.65
1.34
11.76
—
—
1.45
—
1.56
1.69
1.87
-
1.43
1.32
1.76
1.33
1.19
5.14
1.45
1.65
1.34
Dryer than
avg. days
1.41
1.65
3.25
—
—
1.63
—
1.62
2.57
3.50
—
1.75
2.25
1.72
1.28
1.22
2.48
9.71
2.61
1.26
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.38
.46
.79
—
—
.55
—
.80
.85
.81
—
.56
.64
.77
.52
.44
.65
.68
.63
.59
Dryer than
avg. days
.44
.59
.70
—
—
.64
—
.77
.85
.81
—
.62
.63
.83
.53
.60
.67
.65
.60
.75
(continued)
-------
TABLE C-1 (CONTINUED).
U)
-j
Monitor site
OHIO (continued)
Sandusky R. near
Upper Sandusky
Scioto R. at
Chill icothe
Scioto R. below
Shadeville
Tuscarawas R. at
Navarre (Massillion)
Water
Year
(19-)
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
70
71
72
73
74
75
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.52
.60
.57
.54
.64
.64
.52
.52
.54
.58
.39
.31
.40
.56
.49
- ABORTED
.54
.50
.43
.40
.40
.47
.46
.49
.46
.46
.48
.59
.50
.48
.40
.60
.67
.63
.60
.55
—
.55
.54
.61
.64
.56
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
2.03
1.58
—
2.67
1.34
1.21
1.19
1.19
1.21
2.19
1.23
1.16
1.12
1.16
1.14
_
1.09
1.12
1.12
1.06
1.08
Dryer than
avg. days
4.47
1.87
1.81
1.57
1.27
1.15
1.15
1.11
1.16
1.23
1.25
1.20
1.13
1.12
1.16
1.08
1.10
1.11
1.09
1.11
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.78
.83
.71
.69
.71
.41
.59
.47
.43
.51
.42
.58
.48
.43
.51
.17
.29
.41
.40
.45
Dryer than
avg. days
.80
.76
.70
.75
.74
.51
.65
.52
.60
.61
.54
.58
.46
.48
.59
_
.28
.26
.41
.32
.39
(continued)
-------
TABLE C-1 (CONTINUED).
UJ
CO
Monitor site
OREGON
South Umpqua near
Brockway
PENNSYLVANIA
Delaware R. at
Bristol
Delaware R. at
Chester
Delaware R. at
Easton
Water
Year
(19--)
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.57
.56
.33
.49
.63
.51
.56
.50
.71
- ABORTED
.49
.67
.52
.35
.57
- ABORTED
- ABORTED
- ABORTED
.50
.44
.42
.47
.48
.49
.55
.53
.45
.61
.43
-
.62
.65
.73
.77
.54
—
—
—
.60
.49
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
3.25
1.79
3.57
2.98
3.82
1.21
1.95
1.69
1.33
-
.05
.13
.04
.07
.07
—
—
—
1.53
2.33
Dryer than
avg. days
1.91
12.68
1.94
1.69
2.30
1.22
1.67
2.00
1.24
-
1.07
1.61
1.06
1.07
1.07
_
—
—
2.62
1.83
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.84
.83
.96
.87
.82
.72
.77
.69
.84
-
.31
.36
.31
.37
.36
—
—
—
.85
.86
Dryer than
avg. days
.88
.87
.82
.81
.83
.68
.79
.65
.84
-
.28
.42
.33
.42
.42
—
—
.86
.92
(continued)
-------
TABLE C-1 (CONTINUED).
U)
Monitor site
PENNSYLVANIA
(continued)
Delaware R. at Torresdale
Intake, Philadelphia
Lehigh R. at Easton
(Glendon)
Schuylkill R. at
Philadelphia
West Br. Brandywine
Cr. at Modena
TENNESSEE
W.F. Stones R. at Manson
Pk., at Murpheesboro
TEXAS
Trinity R. below Dallas
Water
Year
(19-)
72
73
74
75
76
72
73
74
75
76
69
70
71
72
71
72
73
74
73
74
75
76
77
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.46
.49
.50
.44
.41
.53
.46
.39
.37
.50
.43
.48
.39
.61
- ABORTED
.44
.58
.53
- ABORTED
- ABORTED
.52
- ABORTED
.60
.52
.53
.49
.55
.62
.58
.55
.63
.60
.55
.54
.55
.58
.55
—
.49
.53
.49
—
-
.53
—
.52
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.24
1.63
1.17
1.25
1.19
1.28
1.71
1.44
1.50
2.85
1.25
1.42
1.45
1.24
—
1.58
1.32
1.40
—
—
1.71
-
1.13
Dryer than
avg. days
1.30
1.30
1.18
1.37
1.58
1.55
1.48
4.43
1.30
1.24
1.47
2.26
1.60
1.53
—
2.41
1.76
1.46
—
—
2.29
-
1.14
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.72
.58
.54
.59
.67
.71
.83
.77
.80
.79
.66
.66
.68
.76
—
.83
.84
.69
—
—
.85
-
.28
Dryer than
avg. days
.70
.58
.52
.67
.74
.78
.82
.75
.77
.83
.72
.68
.71
.79
—
.83
.81
.62
_
—
.86
-
.26
-------
TABLE C-2. STORE! DATA BASE MONITOR SITES
SUMMARY OF DAILY FLOW CORRELATION ANALYSIS RESULTS
>£>
O
Monitor site*
ALABAMA
Coosa R. at Gadson
Coosa R. Left Side
COLORADO
S. Platte R. at Denver
Sand C. Birlington
Ditch MDSDD # 1
S. Platte R. at Denver
S. Platte R. at 88th Ave.
MDSDD # 1
Water
Year
(19-)
71
72
73
74
75
76
77
67
70
71
72
73
74
75
76
77
69
70
71
72
73
74
75
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.58
.49
.47
.46
.56
.41
- ABORTED
- ABORTED
.53
.55
.52
.51
.47
.50
.50
.41
- ABORTED
.46
.51
.57
.51
.35
.52
.51
.41
.51
.52
.40
.43
—
—
.48
.45
.50
.47
.53
.42
.56
.40
—
.51
.47
.55
.52
.65
.54
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.82
1.72
1.63
1.34
1.27
1.40
—
—
1.17
1.25
1.63
1.51
1.22
1.52
1.61
1.93
—
.09
.19
.17
.20
.09
.15
Dryer than
avg. days
1.56
2.00
1.70
1.31
1.34
1.49
—
—
1.13
1.21
1.41
1.40
1.26
1.30
1.79
1.85
—
1.11
1.16
1.14
1.14
1.12
1.18
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.88
.86
.82
.77
.73
.78
-
—
.64
.68
.77
.79
.70
.84
.78
.89
_
.57
.68
.68
.70
.69
.67
Dryer than
avg. days
.83
.88
.85
.78
.74
.79
-
—
.66
.70
.79
.76
.72
.81
.84
.84
.57
.68
.67
.69
.79
.68
•StBtlora are grouped by state, states are in alphabetical order, stations are alphabetized within group; ABORTED indicates that less than 180 days of data were available that
year.
(continued)
-------
TABLE C-2 (CONTINUED).
Monitor site
COLORADO (continued)
S. Platte, 88 Ave. (cont.)
S. Platte R. at Denver
Burlington Ditch at York
St. MDSDD #1
Flow-S. Plane at Denver
S. Platte #60 Ave.
MDSDD #1
GEORGIA
Chattahoochee R. at
Atlanta
Water
Year
(19--)
76
77
67
68
69
70
71
72
73
74
75
68
69
70
71
72
73
74
75
76
77
67-
68
Probability of greater
than average DO deficit
Wetter than
avg. days
.51
.50
.48
.51
.48
.49
.60
.52
.53
.37
.40
.56
Dryer than
avg. days
.40
.52
.46
.46
.57
.44
.44
.55
.56
.60
.46
.58
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.29
1.26
1.17
1.47
1.09
1.14
1.35
1.28
1.14
1.09
1.13
1.29
Dryer than
avg. days
1.21
1.24
1.21
1.11
1.08
1.13
1.30
1.20
1.15
1.10
1.10
1.09
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.63
.62
.35
.65
.56
.67
.72
.67
.66
.68
.66
.60
Dryer than
avg. days
.67
.61
.44
.64
.55
.69
.72
.66
.67
.63
.68
.65
ABORTED -----
.49
.57
.51
.47
.38
.42
.66
.53
- ABORTED
.48
.47
.54
.53
.56
.52
.44
.51
— highly
1.15
1.30
1.26
.17
.14
.28
.21
.31
intermittent data
1.14
1.28
1.23
1.20
1.19
1.14
1.28
1.28
-
.65
.75
.72
.73
.74
.76
.61
.60
—
.68
.74
.69
.73
.71
.77
.70
.60
—
(continued)
-------
TABLE C-2 (CONTINUED).
Monitor site
NORTH CAROLINA
Cape Fear R. at Lock #1
near Kelly, Cape Fear R. at
Channel Marker #50
Muddy Cr. near Muddy Cr.
500 below efficient discharge
Muddy Cr. near Muddy Cr.
Muddy Cr. at SR 1485
i_, Muddy Cr. near Muddy Cr.
*• Muddy Cr. at SR 2991
NJ
Muddy Cr. near Muddy Cr.
NFK, Muddy Cr,
Roanoke R. at Roanoke Rapid
Roanoke R. 1 mi. upstream
of Welch Cr.
Neuse R. near Northside
Neuse R. at U.S. Hwy. 42
Neuse R. near Northside
Neuse R. at U.S. Hwy. 50
Neuse R. near Northside
Neuse R. at U.S. Hwy. 70
Neuse R. near Northside
Neuse R. at U.S. Hwy. 98
Water
Year
(19-)
74
75
74
75
74
75
74
75
74
75
74
75
75
76
75
76
75
76
75
76
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
—
.52
.69
.64
—
.49
.50
.53
—
.51
.55
.49
—
-
—
-
—
-
—
—
ABORTED
ABORTED
ABORTED
ABORTED
ABORTED
ABORTED
ABORTED
ABORTED
ABORTED
ABORTED
ABORTED
—
.55
.45
.44
—
.45
.52
.54
—
.36
.74
.55
_
-
—
-
—
-
—
—
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
-
1.20
1.24
1.46
—
1.64
1.13
1.20
—
2.01
1.27
1.27
_
-
—
-
—
-
—
—
Dryer than
avg. days
—
1.20
1.38
1.38
—
1.75
1.24
1.14
—
1.86
1.26
1.19
—
-
—
-
—
-
—
—
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
—
.44
.73
.71
_
.75
.75
.73
—
.81
.68
.75
-
—
-
—
-
_
_
Dryer than
avg. days
—
.40
.68
.74
_
.79
.75
.64
—
.86
.68
.65
__
_
__ .
-
_
-
_
_
(continued)
-------
TABLE C-2 (CONTINUED).
Strength of DO deficit Average percentage of
Probability of greater when deficit is greater saturation when deficit
Monitor site Water than average DO deficit than average is worse than average
Year Wetter than Dryer than Wetter than Dryer than Wetter than Dryer than
(19--) avg. days avg. days avg. days avg. days avg. days avg. days
WASHINGTON
Green R. at Tuckwilla 69- - ABORTED - - - - -
Green Duwamish R. 73
Sta. 307
-------
TABLE C-3. WDNR MONITOR SITES SUMMARY OF DAILY FLOW CORRELATION ANALYSIS RESULTS
Monitor site
Fox R. at Menasha
Fox R. at Appleton
Fox R. at Rapide Croche
Fox R. at Depere
Fox R. at Green Bay
Water
Year
(19~)
73
74
75
76
77
73
74
75
76
77
73
74
75
76
77
73
74
75
76
77
73
74
Probability of greater
than average DO deficit
Wetter than
avg. days
.57
.47
.50
.46
.58
.54
.48
.42
.43
.52
.40
.41
.45
.46
.48
.53
.39
.47
.47
.46
.48
.41
Dryer than
avg. days
.39
.50
.48
.62
.47
.56
.45
.50
.50
.57
.60
.59
.59
.63
.59
.54
.56
.54
.59
.52
.58
.55
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
3.15
2.12
1.75
1.29
1.58
1.34
1.92
1.34
1.39
1.55
.69
.38
.28
.31
.29
1.80
1.51
2.51
1.28
1.39
1.26
1.34
Dryer than
avg. days
2.31
2.68
1.73
1.50
2.32
1.42
1.37
1.33
1.52
1.25
1.65
1.58
2.74
.14
.20
.45
.59
.67
.29
.48
1.59
1.31
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.84
.87
.85
.82
.83
.73
.80
.81
.77
.78
.70
.71
.65
.62
.60
.75
.82
.75
.76
.69
.66
.74
Dryer than
avg. days
.87
.88
.87
.80
.86
.69
.74
.74
.77
.78
.61
.54
.60
.56
.60
.81
.78
.73
.72
.71
.56
.61
(continued)
-------
TABLE C-3 (CONTINUED).
Monitor site
Fox R. at Green Bay
(continued)
Wisconsin R. at Wausau
Wisconsin R. at Mosinee
Wisconsin R. at Dubay Dam
Wisconsin R. at Biron
Water
Year
(19-)
75
76
77
73
74
75
76
77
73
74
75
76
77
72
73
74
75
76
72
73
74
75
76
Probability
than average
Wetter than
avg. days
.44
.44
.56
.40
.56
.57
.45
.49
.40
.38
.47
.53
.44
.55
.52
.60
.58
.64
.48
.42
.64
.49
.50
of greater
DO deficit
Dryer than
avg. days
.59
.43
.49
.58
.63
.55
.57
.54
.66
.57
.58
.62
.61
.59
.60
.52
.55
.50
.68
.59
.46
.52
.54
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.29
1.14
1.17
1.12
1.18
1.14
1.15
1.19
1.20
1.06
1.12
1.06
1.09
1.10
1.11
1.08
1.13
1.08
1.13
1.24
1.48
1.18
1.14
Dryer than
avg. days
1.40
1.24
1.16
1.13
1.32
1.14
1.16
1.10
1.16
1.14
1.25
1.12
1.13
1.13
1.15
1.11
1.18
1.13
1.47
1.38
1.13
2.04
1.20
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.66
.72
.57
.70
.71
.62
.54
.56
.51
.46
.40
.33
.40
.27
.52
.36
.47
.36
.46
.72
.60
.61
.50
Dryer than
avg. days
.56
.58
.60
.76
.71
.59
.55
.60
.54
.44
.38
.36
.52
.36
.56
.44
.49
.44
.56
.79
.62
.71
.64
(continued)
-------
TABLE C-3 (CONTINUED).
Strength of DO deficit
Monitor site
Wisconsin R. at Port
Edwards
Wisconsin R. at Petenwell
Water
Year
(19-)
73
74
75
76
73
74
75
76
Probability
than average
Wetter than
avg. days
.42
.54
.44
.50
.60
.53
.54
.56
of greater
DO deficit
Dryer than
avg. days
.59
.59
.70
.54
.40
.45
.56
.61
when deficit is greater
than
Wetter than
avg. days
1.24
1.14
1.11
1.14
1.20
1.10
1.11
1.13
average
Dryer than
avg. days
1.38
1.42
1.24
1.20
1.16
1.21
1.12
1.14
Average percentage of
saturation when deficit
is worse than
Wetter than
avg. days
.72
.56
.49
.50
.56
.53
.34
.56
average
Dryer than
avg. days
.79
.61
.58
.64
.62
.43
.47
.60
en
-------
TABLE C-4. USGS MONITOR SITES SUMMARY OF DAILY RAINFALL CORRELATION ANALYSIS RESULTS
Monitor site*
MASSACHUSETTS
Connecticut R.at W.
Springfield (Thompsonville)
Chicopee R. at Chicopee
Falls
Westfield R. at Westfield
MISSOURI
Center Cr. near Carterville
James R. near Boaz
Water
Year
(19--)
72
73
74
75
73
74
75
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
- ABORTED
.60
.66
.63
- ABORTED
.59
.56
- ABORTED
.56
.59
.63
.38
.58
- ABORTED
.64
.65
- ABORTED
- ABORTED
.64
.64
.65
.63
—
.52
.47
.47
_
.60
.58
—
.46
.45
.45
.45
.50
—
.49
.50
-
_
.47
.54
.53
.52
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
—
1.67
1.32
1.26
—
1.30
1.36
—
1.41
1.27
2.13
1.64
1.67
—
1.14
1.14
-
—
1.14
1.13
1.14
1.11
Dryer than
avg. days
—
2.98
1.21
1.23
—
1.16
1.23
—
2.49
1.28
2.46
1.84
1.71
—
1.10
1.15
-
1.12
1.13
1.09
1.11
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
—
.80
.78
.81
—
.56
.41
.74
.78
.81
.86
.62
_
.59
.54
-
.57
.49
.44
.30
Dryer than
avg. days
—
.76
.81
.83
—
.45
.39
.76
.80
.87
.82
.55
—
.62
.54
—
__
.58
.48
.45
.28
'Stations are grouped by state, states are in alphabetical order, stations are alphabetized within group; ABORTED indicates that
year.
; than 180 days of data were available that
(continued)
-------
TABLE C-4 (CONTINUED).
Monitor site
MISSOURI
(continued)
Wilsons Cr. near Battlefield
Wilsons Cr. near
Springfield
NEW JERSEY
Delaware R. at Trenton
OHIO
Ashtabula R. at Ashtabula
Water
Year
(19-)
73
74
75
76
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.55
.63
.62
.63
- ABORTED
.58
.56
.67
.77
.62
.57
.47
.52
.51
.51
.58
.56
.63
.67
.55
.56
.53
.52
—
.45
.46
.38
.43
.46
.50
.46
.44
.51
.51
.41
.47
.45
-42
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.11
1.08
1.12
1.06
—
1,21
1.24
1.26
1.24
1.79
2.14
2.05
4.66
2.91
1.29
1.51
1.41
1.24
1.22
Dryer than
avg. days
1.09
1.10
1.08
1.05
—
1.23
1.18
1.15
1.16
5.35
5.76
3.07
2.53
4.58
1.35
2.13
1.34
1.18
1.24
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.12
.09
.14
.03
—
.61
.59
•§1
.61
,86
.87
.90
.92
.96
.71
.77
.75
.70
.69
Dryer than
avg. days
.13
.10
.14
.04
—
.64
.64
.68
.64
.87
.90
.90
.95
.95
.67
•77
.72
.68
-68
(continued)
-------
TABLE C-4 (CONTINUED).
Monitor site
OHIO (continued)
Black R. at Elyria
Blanchard R. near Findlay
Cuyahoga R. at
Independence
Cuyahoga R. at Old
Portage
Cuyahoga R. at W. 3rd St.
Bridge
Water
Year
(19-)
70
71
72
73
74
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
Probability of greater
than average DO deficit
Wetter than
avg. days
Dryer than
avg. days
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
Dryer than
avg. days
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
Dryer than
avg. days
- ABORTED _____
- ABORTED -
- ABORTED _____
.55
.52
.58
.65
.59
.62
.64
.50
.60
.52
.57
.56
.55
.57
.56
.58
.66
.59
.57
.55
.51
.53
.50
.52
.57
.48
.52
.44
.49
.54
.41
.49
.43
.39
.42
.48
.43
.38
.44
.45
.46
.52
.45
.47
1.55
1.51
1.79
1.29
1.14
1.17
1.22
1.20
1.25
1.15
1.24
1.20
1.71
1.63
1.30
1.31
1.27
1.20
1.12
1.10
1.14
1.15
1.64
1.58
2.65
1.20
1.10
1.16
1.19
1.16
1.22
1.15
1.17
1.14
1.40
4.14
1.34
1.31
1.19
1.17
1.09
1.09
1.12
1.14
.68
.59
.65
.65
.44
.50
.57
.66
.72
.67
.74
.74
.66
.74
.75
.68
.67
.41
.42
.43
.48
.42
.53
.51
.67
.58
.43
.55
.59
.61
.70
.62
.70
.74
.66
.76
.76
.68
.75
.31
.35
.31
.34
.36
(continued)
-------
TABLE C-4 (CONTINUED).
Monitor site
OHIO (continued)
Grand R. near Painesville
Grand R. at Painesville
Little Miami R. at
Miamiville (Milford)
Little Miami R. near
Spring Valley
Mad R. near Dayton
Mahoning R. at OH.-PA.
State Line below
Lowellville
Water
Year
(19-)
75
76
72
73
74
71
72
73
74
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.60
.54
.46
.54
.46
- ABORTED
.41
.56
.53
.59
.53
.60
.61
.58
.66
.64
.60
.63
.65
.49
.56
.51
.59
.60
.59
.49
.51
.49
.48
—
.44
.42
.43
.37
.36
.43
.43
.38
.43
.34
.45
.41
.45
.50
.50
.49
.45
.44
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.19
1.70
1.45
1.45
1.23
—
2.32
1.50
1.47
1.17
1.31
1.27
1.17
1.19
1.20
1.26
1.17
1.20
1.21
1.15
1.17
1.16
1.09
1.18
Dryer than
avg. days
1.16
1.31
2.10
1.37
1.23
2.08
2.22
1.49
1.13
1.43
1.21
1.14
1.16
1.14
1.14
1.13
1.20
1.15
1.16
1.32
1.12
1.10
1.14
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.72
.69
.79
.81
.71
.85
.85
.83
.65
.70
.73
.74
.65
.72
.78
.73
.74
.68
.42
.53
.56
.57
.54
Dryer than
avg. days
.66
.64
.74
.68
.57
.81
.84
.86
.61
.70
.74
.75
.66
.70
.78
.72
.75
.70
.37
.52
.49
.52
.52
(continued)
-------
TABLE C-4 (CONTINUED).
Ln
Monitor site
OHIO (continued)
Sandusky R. near Upper
Sandusky
Scioto R. at Chillicothe
OREGON
South Umpqua near
Brockway
PENNSYLVANIA
Lehigh R. at Easton
(Glendon)
Water
Year
(19-)
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
72
73
74
75
76
Probability of greater
than average DO deficit
Wetter than Dryer than
avg. days avg. days
.58
.56
.62
.62
.69
- ABORTED
- ABORTED
.53
.56
.60
.49
.51
.40
- ABORTED
- ABORTED
.67
.62
.57
.61
.65
.43
.47
.45
.41
.42
—
—
.49
.47
.39
.46
.42
.47
—
-
.51
.45
.54
.47
.47
Strength of DO deficit
when deficit is greater
than average
Wetter than
avg. days
1.89
1.84
1.77
1.38
1.40
—
—
1.14
1.20
2.15
1.85
15.62
3.07
—
-
1.35
1.50
1.52
1.37
1.42
Dryer than
avg. days
5.08
1.68
—
2.62
1.21
—
—
1.13
1.17
1.31
2.88
5.49
1.86
—
-
1.56
1.60
4.98
1.34
1.96
Average percentage of
saturation when deficit
is worse than average
Wetter than
avg. days
.74
.79
.71
.70
.71
—
—
.51
.48
.54
.90
.93
.95
—
-
.77
.83
.77
.77
.81
Dryer than
avg. days
.83
.78
.70
.74
.75
—
—
.52
.57
.58
.84
.84
.80
—
-
.76
.82
.75
.77
.83
-------
APPENDIX D
RESULTS OF DETAILED SITE ANALYSES
This appendix contains the results of the detailed site
analyses. These analyses were performed at 30 sites where a 60
percent or greater probability existed of a high flow or rain-
fall event occurring at the same time as a low dissolved oxygen
event. The sites are grouped by state. The states are in alpha-
betical order within each state group.
Each site analysis contains some or all of the following:
9 water quality monitor site name;
• monitor location, latitude and longitude, agency
I.D. number;
• flow gage site name;
• flow location, latitude and longitude, agency
I.D. number;
• rain gage location, I.D. number;
• short summary of daily correlation results;
• assessment of record quality;
• discharge, drainage area and population description;
• Streeter-Phelps analysis results;
« conclusions and comments; and
9 daily or hourly plots of data from years with
positive correlation.
Three types of illustrations are included in this appendix.
Streeter-Phelps analysis results are presented as plots of DO
deficit (mg/1) versus distance (miles) downstream of the urban
area. Two lines generally appear on each illustration. The
lower line depicts the deficit due to "normal" flows and the
second line depicts the deficit due to "storm" flows. The
152
-------
saturation DO level for typical conditions is also illustrated.
Important features such as the stream gage and DO monitor loca-
tion as well as sewage treatment facilities are identified. The
second type of illustration is the hourly data analysis. Two
separate graphs are included in each illustration. The upper
graph illustrates absolute DO level and saturation DO level
(both in mg/1) versus time in days. The lower graph presents
the stream flow, DO deficit, and rainfall. The flow line has
been normalized by dividing the actual flow values by the average
value for the entire period illustrated. This preserves the
hydrograph shape and avoids rescaling the graph for every new
station. The rainfall is shown as hourly histograms (vertical
rectangular bars). The format of the illustrations allows fairly
easy visualization of the relationship between the several vari-
ables. The final illustration type is the daily average value
plots. These generally illustrate either flow in cubic feet per
second versus time in days or DO and DO deficit in mg/1 versus
time in days.
153
-------
STATE: MASSACHUSETTS
Monitor Name; Connecticut River at West Springfield
USGS I.D.:01 177 200 Latitude: 42 05 46 Longitude: 72 35 43
Stream Gage Name; Connecticut River at Thompsonville, Conn.
USGS I.DTI01 184 OOP Latitude: 41 59 14 Longitude 72 36 21
Rain Gage Name; Springfield
Weather Bureau I.P.: 8046
Daily Data Analysis Results: Water years 72-75 were examined.
The probability of low dissolved oxygen with high flow
averaged 40 percent. The probability of low dissolved
oxygen on days with rainfall was greater than 60 percent
in 1973, 1974, and 1975. Dissolved oxygen levels remained
around the 75-80 percent of saturation level on days with
rain.
Quality of Records; The monitor is visited once a month or
more. Record is reliable except for two to three weeks
each summer.
Discharge Characteristics: Average flow 16,250 cfs, approx.
range 1,000 to 100,000 cfs
Drainage Area at Monitor: 9,623 square miles
Urban Area(s) Contributing at Monitor: Turners Falls,
Northampton, Holyoke, Chicopee
approximate population (Almanac) 87,500
Approximate Urban Area Contributing at Monitor: 60 sq mi or
less than 1 percent of total
Results of Streeter-Phelps Analysis: Monitor is located
satisfactorily for normal flows. At high flows a location
20 to 30 miles downstream would be better. Storm flow
could produce D.O. deficits of 4.5 mg/1 resulting in D.O.
levels below 5 mg/1 during storm events. There is some
evidence of this on daily plots. Simulation results
follow in Figure D-l.
Hourly Data Analysis Results: The period examined illustrates
a mild tendency for the deficit to decrease slowly during
the flow event. Water quality is never very bad at this
site. The period of hourly data is illustrated in Figure
D-2.
154
-------
Ln
01
15.0 _
10.0 .
O
u.
LU
Q
O
Q
5.0
0.0
SAJURATK3NjAT^O c =_9.17jn&n_
SATURATION AT_27^C_=_8.07_mg/l_
STORM
NORMAL
-T"
i
I
t
-*—
0
I
w
VIC
-1 — 1 — 1 1 1 • 1 — 1 — 1 — t — 1 — 1 — 1 — 1 — 1 — » — 1 — 1 — 1 — 1 — t — 1 — 1— — t — 1— 1 — 1
' | T20 30 40 50 6(
| DISTANCE, mi
1 FLOW GAGE
LONG MEADOW AND SUBURBS
WESTFIELD R.
.SPRINGFIELD
)NITOR
SPRINGFIELD N. CHICOPEE R.
HOLEYOKE
SOUTH HADLEY
Figure D-1. Streeter-Phelps analysis results for Connecticut R. at West Springfield, MA.
-------
20 _
O"
o
10 ..
SATURATION DO LEVEL
U1
o
Q
HI
i
z
5
3.0 -
2.5 :,
2.0 ::
DISCHARGE / AVG. DISCHARGE
4 6
TIME FROM START OF PERIOD, days
10
Flgurt D-2. Hourly data for Connecticut R. at W. Springfield, MA.
(4/1774 to 4/10/74).
-------
Conclusions and Comments: During June 1974, there appears
to have been a period when the D.O. dropped drastically
during a flow event. The monitor quit during the most
interesting time. This site does not appear to have a
water quality problem attributable to urban runoff. It
would be interesting, however, to put a monitor at the
theoretical D.O. sag point and see what was observed
then.
157
-------
STATE: MASSACHUSETTS
Monitor Name: North Nashua River near Lancaster
USGS I.P.; 01 094 700 Latitude: 42 28 47 Longitude: 71 41 04
Stream Gage Name: North Nashua River near Leominster
USGS I.P.; 01 094 500 Latitude: 42 30 06 Longitude: 71 43 23
Rain Gage: Fitchburg 4 SE
Weather Bureau I.P..:
Daily Data Analysis Results; Water years 1968-72 were examined.
The probability of low dissolved oxygen with high flow
exceeded 60 percent in 1970 and the summer of 1971. Paily
rainfall correlation was not run at this station. Hourly
rainfall data were used in the detailed analysis, however.
Dissolved oxygen levels are definitely low, averaging
less than 60 percent of saturation most of the time and
falling to 40-50 percent during wet weather.
Quality of Records; Monitor is visited once a month or more.
Record is highly reliable for flow and good for the monitor
except for short periods scattered throughout the year.
Discharge Characteristics: Average flow approximately 175
cfs, approx. range 50-1,000 cfs
Drainage Area at Monitor: 128 square miles
Urban Area(s) Contributing at Monitor: Fitchburg, Leominster,
approximate population: 43,300 and 33 ,000, respectively
. 2
Approximate Urban Area Contributing at Monitor: 11 mi
(based on 7100/mi^)or 9 percent of total
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Data Analysis Results: The period examined had a
definite tendency for the D.O. deficit to increase at
the beginning of a storm event. The D.O. level dropped
from an average of over 6 mg/1 to less than 5 and re-
mained there for one day. The average D.O. level was
definitely worse during the flow events examined. The
period of hourly data is illustrated in Figure D^3.
The EPA suggested standard of 2.0 mg/1 for 4.0 hours
was violated on the third and fourth days of the period
examined.
158
-------
20
o"
Q
10 .
SATURATION DO LEVEL
DO LEVEL
•o
O
Q
C3
I
Q
>
UJ
c
I
O
DISCHARGE /AVG. DISCHARGE
PRECIPITATION, inches
DO DEFICIT/ 10, mg/l
6 8 10 12 14 16
TIME FROM START OF PERIOD, days
18
20
22 24
Figure D-3. Hourly data for N. Nashua R. near Leominster, MA.
(6/1770 to 6/25/70).
-------
Conclusions and Comments; A definite problem exists here.
The monitor is 5 miles downstream of the Leominster
Sewage Treatment Plant. It is 10 miles below the town.
This probably puts it in an excellent position to sense
a D.O. sag. It is not possible without detailed exami-
nation of the site to tell whether the problem is urban
runoff, combined sewer overflows, sediment entrainment,
or some other problem.
160
-------
STATE: MASSACHUSETTS
Monitor Name: Westfield River at West Springfield
USGS I.D.; 01 183 600 Latitude: 42 05 59 Longitude: 72 38 28
Stream Gage Name: Westfield River near Westfield
USGS I.D.; 01 183 500 Latitude: 42 06 24 Longitude: 72 41 58
Rain Gage: Springfield
Weather Bureau I.P.; 8046
Daily Data Analysis Results: Water years 1972-76 were examined.
The probability of low dissolved oxygen with high flow
reached 62 percent in 1975. The same year the probability
of low D.O. on days with rainfall was 63 percent. Dis-
solved oxygen levels are not bad, running approximately
75 percent saturation most of the time.
Quality of Records: The monitor is visited once a month or
more. Flow records are excellent. Monitor records are
poor in the summer with entire months missing.
Discharge Characteristics: Average flow approx. 940 cfs,
approx. range 100-13,000 cfs
Drainage Area at Monitor: 497 square miles
Urban Area(s) Contributing to Monitor: Pittsfield, Westfield,
West Springfield
approximate populations: estimate 35-50,000
Approximate Urban Area Contributing at Monitor; 5 to 7 square
miles or approximately 1 percent of total
Results of Streeter-Phelps Analysis: Not performed at this
site*
Hourly Data Analysis Results; Hourly data could not be ob-
tained for this site.
Conclusions and Comments; Examination of the daily record
indicates a number of periods where the D.O. level appears
to drop at the time of flow events. The monitor is poorly
located. The "problem" may originate at Pittsfield which
is 40 miles upstream. A monitor closer to there might
have detected more. The water quality is generally good
at this site.
161
-------
STATE: MISSOURI
Monitor Name: Center Creek near Carterville
USGS I.D.; 07 186 400 Latitude: 37 08 26 Longitude: 94 22 57
Stream Gage Name: Same
USGS I.D.: Same Latitude: Same Longitude: Same
Rain Gage: Springfield
Weather Bureau I.P.; 8046
Daily Data Analysis Results: Water years 1972-76 were examined.
The probability of low dissolved oxygen with high flow
never exceeded 53 percent. The probability of low dis-
solved oxygen on days with rainfall reached 64 percent in
1974 and 65 percent in 1975. Pissolved oxygen levels in
general are low ranging from 45 to 60 percent of satura-
tion.
Quality of Records: The monitor was visited once a month or
more. Monitor records are excellent with only a few short
periods per year missing.
Discharge Characteristics: Average flow 200 cfs, a.pprox.
range 10 to 37,000 cfs.
Drainage Area at Monitor: 232 square miles
Urban Area(s) Contributing at Monitor: Perhaps a little of
Carthage and the fringes of Joplin, approx. population -
no t de te rmi ned .
Approximate Urban Area Contributing at Monitor; 2 or 3 square
miles at most or less than 1 percent of total.
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Data Analysis Results: No hourly data were obtained
at this site.
Conclusions and Comments: Very little urban area seems to
contribute at monitor. Correlation is with rain gage
60 miles away. Examination of daily record seems to
indicate isolated events where D.O. decreases with flow.
This does not seem to be related to urban runoff. There
is a great deal of strip mining between the monitor and
Joplin.
162
-------
STATE: MISSOURI
Monitor Name; James River near Boaz, Missouri
USGS I.D.; 07 052 250 Latitude: 37 00 25 Longitude: 93 21 50
Stream Gage Name: Same
USGG I.^.: Same Latitude: Same Longitude: Same
Rain Gage: Springfield
Weather Bureau I.P.: 7976
Daily Data Analysis Results: Water years 1972-76 were examined.
The probability of low dissolved oxygen with high flow never
exceeded 44 percent. The probability of low dissolved
oxygen on days with rainfall exceeded 60 percent in 1973, 74,
75, and 76. The correlation could be clearly seen by eye.
Dissolved oxygen levels were extremely low, ranging from
23 to 60 percent saturation.
Quality of Records: Rated fair by USGS, monitor is visited
once a month. Flow record is continuous, monitor looses
from 1 to 2 months a year, usually at summer low flow.
Discharge Characteristics: Average flow approx. 300 cfs,
approx. range 55 to 31,500 cfs,
Drainage Area at Monitor: 462 square miles
Urban Area(s) Contributing at Monitor: Springfield, approx.
population 180,000.
Approximate Urban Area Contributing at Monitor: 25 square
miles or 5 percent of the total
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Data Analysis Results: Hourly data were not available
for this site.
Conclusions and Comments: The James River receives treated
sewage from the Springfield Southwest sewage treatment
plant. It also receives the flow from Wilsons Creek
which is very low in D.O. There is a definite D.O.
problem here, but whether it is direct urban runoff or
something else is hard to say. The dissolved oxygen
level is typically below 5 mg/1 4 months of every year.
163
-------
STATE: MISSOURI
Monitor Name: Wilsons Creek near Battlefield
USGS I.D.1 07 052 160 Latitude: 37 07 02 Longitude: 93 24 14
Stream Gage Name: Same
USGS I.D.: Same Latitude: Same _ Longitude: Same
Rain Gage: Springfield
Weather Bureau I.P.: 7976
Daily Data Analysis Results: Water years 1973-76 were examined,
The probability of low dissolved oxygen at the time of a
flow event never exceeded 60 percent but did reach 57 per-
cent in 1976. The probability of low D.O. on days with
rainfall exceeded 60 percent in 1974, 75, and 76. Pissolved
oxygen levels are very low, ranging from 2 to 16 percent
of saturation.
Quality of Records: Monitor is visited once a month or more.
Flow record is generally continuous. Monitor record is
intermittent during the summer months.
Discharge Characteristics; Average flow approx. 100 cfs,
approx. range 25-4,000 cfs,
Drainage Area at Monitor: 55 square miles.-
Urban Area(s) Contributing at Monitor: Springfield, approx.
population 180,000-
Approximate Urban Area Contributing at Monitor; 25 square
miles or 45 percent of total.
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Data Analysis Results: Hourly data could not be
obtained for this site.
Conclusions and Comments; The problem at this site is con-
nected to the problem at Wilsons Creek near Springfield,
Missouri. Hourly data were examined there and may be
seen by referring to that station. The problem is no
doubt related to the fact that Springfield is nearly
1/2 of the drainage basin.
164
-------
STATE: MISSOURI
Monitor Name: Wilsons Creek near Springfield
USGS I.D.: 07 052 100 Latitude: 37 10 06 Longitude: 93 22 14
Stream Gage Name: Same
USGS I.D.; Same Latitude: Same Longitude: Same
Rain Gage; Springfield
Weather Bureau I.P.: 7976
Daily Data Analysis Results: Water years 1972-76 were examined,
The probability of low dissolved oxygen with high flow ex-
ceeded 70 percent in 1973, 75, and 76. It hit a peak of
81 percent in 1975. This is the highest probability en-
countered at any site examined. The probability of low
D.O. on days with rainfall exceeded 60 percent in 1975 and
1976. D.O. levels are characteristically low, averaging
50 to 60 percent saturation.
Quality of Records: The monitor is visited once a month or
more. Flow records are virtually continuous. The monitor
quits from one to two months a year, usually at times of
low flow.
Discharge Characteristics: Average flow approximately 20 cfs,
range from 0 to 2800 cfs.
Drainage Area at Monitor: 31.4 square miles-
Urban Area(s) Contributing at Monitor: Springfield, approx.
population 180,000.
Approximate Urban Area Contributing at Monitor: 25 square
miles or 80 percent of total.
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Data Analysis Results: The hourly data periods ex-
amined are shown in Figures D*-4, D-5 , and D-6. Water
quality is very poor at all times, but definitely decreases
more at times of high flow. The EPA suggested standard
of 2.0 mg/1 for 4.0 hours is violated several times in the
periods examined.
Conclusions and Comments: This site had the highest percent-
age of urban drainage of any in the entire study. There
can be no doubt that the large percentage of urban area
contributes to the bad quality. It is not possible, how-
ever, to tell the exact cause of the low D.O. - high flow
without physically examining the site.
165
-------
20 T
cr
a
10 . r
SATURATION DO LEVEL
3.0 ,.
CTi
0.0
DISCHARGE / AVG. DISCHARGE
4 6 8 10
TIME FROM START OF PERIOD, days
12
14
Figure D-4. Hourly data for Wilsons Cr. near Springfield, MO.
(4/15/73 to 4/29/73).
-------
cr
o
20
10
SATURATION DO LEVEL
CTi
•o
O
Q
.-
O
I
O
C/J
Q
CD
UJ
C3
CC
<
O
VI
3.0
2-5
2.0
1.5
1.0
0.5 •
0.0
DISCHARGE / AVG. DISCHARGE
n PRECIPITATION, inches
4 6
TIME FROM START OF PERIOD, days
10
Figure D-5. Hourly data for Wilsons Cr. near Springfield, MO.
(8/10/75 to 8/19/75).
-------
O"
a
20 T
10 ..
SATURATION DO LEVEL
DO LEVEL
CTl
03
•o
O
3.0 ,.
2.5 ::
J 2.0 X
Q 1.5 ±
•«. 1.0 ::
UJ
o
tr ••.
< 0.5 ::
0.0
DISCHARGE / AVG. DISCHARGE
DO DEFICIT / 10, mg/l
TIME FROM START OF PERIOD, days
10
Figure D-6. Hourly data for Wilsons Cr. near Springfield, MO.
(2/15/76 to 2/24/76).
-------
STATE: NEW JERSEY
Monitor Name: Delaware River at Trenton
USGS I.D.: 01 463 500 Latitude: 40 13 18 Longitude: 74 46 42
Same Longitude: Same
8883
Stream Gage Name: Same
USGS I.D.: Same Latitude:
Rain Gage: Trenton
Weather Bureau I.D.:
Daily Data Analysis Results: Water years 1972-76 were examined,
The probability of low D.O. with high flow exceeded 60 per-
cent in 1973 and 1975. It reached 58 percent in 1972.
The probability of low D.O. on days with rainfall exceeded
60 percent in 1972. It reached 57 percent in 1973 and
1976. Water quality here is generally good with D.O.
levels from 85 to 97 percent saturation. The correlation
between flow and low D.O. is so obvious that it can easily
be seen by eye. This is illustrated in Figure D-7.
Quality of Records: The monitor is visited once a month or
more. The flow record is virtually continuous. The moni-
tor record is good overall but from two weeks to a month
are missing each year.
Discharge Characteristics: Average flow 12,000 cfs, approx.
range 1,200 to 325,000 cfs.
Drainage Area at Monitor: 6,780 square miles
Urban Area(s) Contributing at Monitor: Trenton and suburbs,
population estimated at greater than 1/2 million.
Approximate Urban Area Contributing at Monitor: 60 to 70
square miles or 1 percent of total.
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Data Analysis Results:
tained at this site.
Hourly data could not be ob-
Conclusions and Comments; Trenton, New Jersey marks the up-
stream edge of an urban megalopolis. The water quality
in the Delaware deteriorates rapidly downstream. See
for example Delaware at Chester and Bristol, Pennslyvania.
While the water quality is good at Trenton, it is defi-
nitely not good as one proceeds downstream. The fact that
such a strong correlation exists at Trenton makes one
wonder what would have been measured on a smaller stream.
169
-------
WATER YEAR = 1972
20.00n
i- 80.00
15.00-
10.00-
z
LU
a
x
o
Q
HI
5.00-
0.00
1SOOO
-5.00 -I
0.00
Figure D-7. Daily DO and flow for Delaware R. at Trenton, NJ.
-------
The monitor is downstream of Trenton on the New Jersey
shore. It is in a good place to sense urban runoff
before it mixes laterally across the stream. It is
much too close to Trenton to sense a fully developed
D.O. sag. No water quality standards violations were
seen here.
171
-------
STATE: NEW JERSEY
Monitor Name: Manasquan River at Squankum
USGS I.D.:01 408 OOP Latitude: 40 09 47 Longitude: 74 09 21
Stream Gage Name: Same
USGS I.D.: Same Latitude:
Same Longitude: Same
Rain Gage:
Weather Bureau I.D.:
Daily Data Analysis Results: Water years 1970-74 were examined.
The probability of low D.O. at high flow exceeded 60 percent
in 1971 and 1972. No daily rainfall correlation was per-
formed at this site. D.O. levels are generally low ranging
from 50 to 70 percent saturation. The daily record for
1972 is illustrated in Figure D-8.
Quality of Records: USGS rates the flow record as excellent.
The monitor is visited at least once a month. Monitor
data is fairly typical with from two to four weeks missing
each year.
Discharge Characteristics:
13 to 3,000 cfs.
Average flow 75 cfs, approx. range
Drainage Area at Monitor; 43.4 square miles
Urban Areas Contributing at Monitor: Farmingdale, Naval
Reserve Ammunition Dump, several unnamed suburbs, popu-
lation unknown.
Approximate Urban Area Contributing at Monitor: 1-2 square
miles or 2-4 percent of total.
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Data Analysis Results: Not performed at this site.
Conclusions and Comments; Very little urban area could be
found here. There is a good deal of swamp upstream of the
monitor. This may lend some additional evidence to the
swamp flushing theory advanced by the Triangle J208 Study
(3) . The water quality is definitely bad and there is an
obvious correlation with flow events. The problem, how-
ever, does not appear to be related to the objectives of
this study.
172
-------
WATER YEAR = 1972
U)
10.00 -I
-10.00 -"
210.00
TIME, days
DO DEFICIT
60.00
0.00
240.00 270.00
Figure D-8. Daily DO and flow for Manasquan R. at Squankum, NJ.
-------
STATE: NEW JERSEY
Monitor Name: Raritan River near South Bound Brook
USGS I.D.: 01 404 100 Latitude: 40 30 47 Longitude: 74 32 24
Stream Gage Name; Raritan River Below Calco Dam at Bound Brook
USGS I.D.: 01 403 060 Latitude: 40 33 05 Longitude: 74 32 54
Rain Gage: Bound Brook
Weather Bureau I.P.: 0927
Daily Data Analysis Results: Water years 1974 and 1975 were
examined. The probability of low D.O. at high flow reached
60 percent in 1975. D.O. levels were generally high,
ranging from 75 to 80 percent saturation. The daily data
for 1975 is illustrated in Figure D-9. Several obvious
instances where the D.O. decreased with flow events were
found. Just as many times can be found when the opposite
is true.
Quality of Records: Much of the 1974 D.O. record was missing.
~~ The flow record is excellent. The 1975 monitor record is
good except for the winter months. The monitor is visited
once a month or more.
Discharge Characteristics: Average flow 1250 cfs, approx.
"range 1,000 to 17,000 cfs.
Drainage Area at Monitor: 862 square miles
Urban Area(s) Contributing at Monitor: Somerville, Bound
Brook, Middlesex, Manville, South Bound Brook, approximate
population 47,000.
Approximate Urban Area Contributing at Monitor: 6.63 square
miles or 1 percent of total.
Results of Streeter-Phelps Analysis; The results of the
analysis are illustrated in Figure D-10. The monitor
location is too close to the urban area to sense
maximum, sag. A location 10 miles or more downstream
would be better.
Hourly Data Analysis Results: Hourly data could not be
obtained for this site.
Conclusions and Comments: Water quality is not good here.
There are 6 sewage treatment plants along the reach
immediately above the monitor. The entire area is neavily
industrialized. The monitor is too close to the urban
area to sense the maximum sag in a Streeter-Phelps sense.
174
-------
WATER YEAR = 1975
15.00 -i
10.00 -
Z
LU
C3
X
o
Q
in
o
5
5.00-
0.00
150.00
-5.00 -
-10.00 -1
180.00 210.00
TIME, days /*"
r eo.oo
4O.OO
O
cc
- 20.00 0
270.00
DO DEFICIT
Figure D-9. Daily DO and flow for Raritan R. near South Bound Brook, NJ.
-------
01
15.0 T
10.0 1
| SATURATION ATJ00C_=_7.63_rng/l_
LU
a
o
a
5.01
SATURATION AT20°<^=J3.17jTig/l
0
A
1
10
MONITOR
MIDDLESEX
BOUND BROOK
LOW GAGE
MANMILLE
FINDERINE
SOMERVILLE
STORM
20 30
DISTANCE, mi
Figure D-10. Streeter-Phelps analyses results for Raritan R. near South Bound Brook, NJ.
-------
STATE: OHIO
Monitor Name: Ashtubula River at Ashtabula
USGS I.D.; 04 212 700 Latitude: 41 54 00 Longitude: 80 47 44
Stream Gage Name: Ashtabula River near Ashtabula
USGS I.D.: 04 212 500 Latitude: 41 51 20 Longitude: 80 45 44
Rain Gage: Ashtabula
Weather Bureau I.P.: 0264
Daily Data Analysis Results: Water years 1972-76 were examined,
The probability of low D.O. with high flow reached 60 per-
cent in 1975. The probability of low D.O. on days with
rainfall exceeded 60 percent in 1975 and 1976. Dissolved
oxygen levels are in the 65 to 70 percent saturation range.
Quality of Records: The monitor is visited once a month or
more. Both the flow and monitor records are excellent
with little missing data.
Discharge Characteristics: Average flow 150 cfs, approx.
range 0-10,000 cf s.
Drainage Area of Monitor: 136 square miles
Urban Area(s) Contributing at Monitor: Ashtabula, Harbor,
East Ashtabula, approximate population 64,000
Approximate Urban Area Contributing at Monitor: 9 square
miles or 7 percent of total
Results of Streeter-Phelps Analysis: Not performed at this
site - river empties into Lake Erie less than one mile
downstream.
Hourly Data Analysis Results: The period examined shows a
definite tendency for the D.O. to improve with increased
flow. Water quality is poor here with D.O. less than
5 mg/1 during much of the summer. The hourly data are
illustrated in Figure D-ll.
Conclusions and Comments: The lower reaches of the Ashtabula
are probably more like an estuary than a river. There is
a heavy industrial concentration near the monitor. Poor
water quality does not appear directly related to urban
surface runoff.
177
-------
o
Q
SATURATION DO LEVEL
00
T3
O
C3
I
CD
111
CD
CJ
in
3.0 r
2.5 ..
2.0 ::
DISCHARGE / AVG. DISCHARGE
CIPITATIQN. inches -i i I
DO DEFICIT/10, mg/l
6 8 10 12
TIME FROM START OF PERIOD, days
Figure D-11. Hourly data for Ashtabula R. near Ashtabula, OH.
(6/1/75 to 6/20/75).
-------
STATE: OHIO
Monitor Name; Blanchard River near Findlay
USGS I.P.; 04 189 OOP Latitude: 41 03 21 Longitude: 83 41 17
Stream Gage Name: Same
USGS I .D. ; Same Latitude: Same Longitude: Same
Rain Gage; Findlay STP
Weather Bureau I.P.; 2791
Daily Data Analysis Results: Water years 1972-1976 were examined.
The probability of low P.O. with high flow never exceeded
55 percent. The probability of low P.O. on days with rain-
fall exceeded 60 percent in 1973, 1975, and 1976. Water
quality levels are generally poor, ranging from 40 to 70
percent P.O. saturation.
Quality of Records: Both the monitor and flow records are
excellent. Virtually, no data are missing from either.
The monitor is visited once a month or more.
Pischarge Characteristics; Average flow 250 cfs, approx.
range 10 to 15,000 cfs
Drainage Area at Monitor: 346 square miles
Urban Area(s) Contributing at Monitor: Findlay, approximate
~ population 38,200
Approximate Urban Area Contributing at Monitor; 7 square
miles oar 2 percent of total
Results of Streeter-Phelps Analysis: Results of the analysis
indicates that the monitor is too close to Findlay to
sense the maximum deficit. A location 8 to 10 miles down-
stream would be better. The results of the analysis are
illustrated in Figure P-12.
Hourly Pata Analysis Results: The period examined seemed to
indicate a correlation between the presence of rainfall
and higher than average P.O. deficit. High flow appears
to decrease the deficit. Water quality is poor with many
days below 5 mg/1 P.O. Given the monitor location this
may actually be a case of urban runoff caused P.O. defi-
cit. The hourly data are illustrated in Figure p-13.
179
-------
co
o
15.0 _
10.0 ..
O
il
ui
O
8
5.0 ..
STORM
ft
10
20
MONITOR, FLOW
FINDLAY STP
30
DISTANCE, mi
40
50
60
Figure D-12. Streeter-Phelps analysis results for Blanchard R. near Findlay, OH.
-------
o
D
20 ,
10 .
SATURATION DO LEVEL
JDO LEVEL
00
•o
O
O
X
$
o
o
in
3.0 ,.
2.5 :
2.0 :
DISCHARGE / AVG. DISHCARGE
^ PRECIPITATION, inches i
6 8 10 12 14
TIME FROM START OF PERIOD, days
16
18
20
Figure D-13. Hourly data for Blanchard R. near Findlay, OH.
(7/1776 to 7/20/76).
-------
Conclusions and Comments: This would be an interesting site
to study in detail. It is hydraulically simple. Findlay
is the only urban area. The monitor is 1.5 miles down-
stream of the sewage treatment plant. Sediment records
are available. There is some strip mining and consider-
able Oil activity in the area. Ohio State University
examined this site in an independent effort and concluded
that poor treatment facilities were part of the problem.
They concluded that it would be difficult to separate the
CSO effects from just plain bad treatment effluent.
182
-------
STATE: OHIO
Monitor Name: Cuyahoga River at Independence
USGS I.D.: 04 208 OOP Latitude: 41 23 43 Longitude: 81 37 48
Stream Gage Name: Same
USGS I.P. : Same Latitude: Same Longitude: Same
Rain Gage: Cleveland WB APT
Weather Bureau I.P.; 1657
Daily Data Analysis Results: Water year 1972-1976 were ex-
amined. The probability of low P.O. with high flow never
exceeded 60 percent. The probability of low P.O. on
days with rainfall exceeded 60 percent in 1973 and 1976.
Water quality levels are in the range 60 to 70 percent
of saturation.
Quality of Records: The monitor is visited once a month or
more. Both the flow and monitor records are virtually
continuous with little or no missing values.
Discharge Characteristics: Average flow approx. 800 cf s ,
range 55 to 25,000 cfs
Prainage Area at Monitor: 707 square miles
Urban Area(s) Contributing at Monitor: Cleveland suburbs,
Akron, approximate population 2,500,000+
Approximate Urban Area Contributing at Monitor: 60 square
miles counting Akron or 8 percent of total, 110 square
miles counting Cleveland suburbs or 16 percent of total.
Results of Streeter-Phelps Analysis: The analysis indicates
that the monitor is in a good place to sense a sag due
to influences at Akron. The analysis indicates that storm
flow at Akron could drive the P.O. at Independence to
zero. The results are shown in Figure D-14.
Hourly Pata Analysis Results: The period examined showed a
definite tendency for the P.O. deficit to increase at
high flow and on days with rain. The problem was not
overly severe for the period investigated with P.O. levels
averaging 6 mg/1. This is shown in Figure D-15.
Conclusions and Comments: This would be a reasonably good
site to study in detail. The lower reaches of the Cuyahoga
are famous in water quality annuals for catching fire in
the early 1960s. Sediment records are available here. The
urban percentage is large and the monitor is well located.
183
-------
15.0 _
10.0 ..
li.
O
Q
5.0 ..
0.0
SATURATION AT 26°C = 8.22 mg/l
* j
J
, i
, f. f . y i y i • T ' f- If f f —91
'10 20 i
| DISTANCE, mi
AKRON STP
CUYAHOGA AT OLD PORTAGE MONITOR
STORM FLOW AKRON
SiPPROX. EDGE OF AKRON
•r — r "
30 '
i
i
1— -1 —
k '
— —r r r * • r "t-"""
40
t
CENTER OF CLEVELAND
CLEVELAND STP
CUYAHOGA AT INDEPENDENCE MONITOR
TINKERS CREEK
SMALL STP
Figure D-14. Streeter-Phelps analysis results for Cuyahoga R. from Akron to Cleveland, OH.
-------
20 ^
o-
Q
10 ..
SATURATION DO LEVEL
DO LEVEL
I—
00
U1
O
'3
O
Q
(3
O
UJ
CD
ec
I
5
3.0 „
2.5 ::
2.0
Q 1.5
6
>
1.0
0.5
0.0
DISCHARGE / AVG. DISCHARGE
O DEFICIT / 10, mg/l
PRECIPITATION, inches
46 8 10 12 14 16 18 20
TIME FROM START OF PERiOD, days
22 24 26 28 30
Figure D-15. Hourly data for Cuyahoga R. at independence, OH.
(6/19/76 to 7/18/76).
-------
The hydraulics of the Akron sewer system are quite com-
plex, but the reach from the Cuyahoga at Old Portage
gage to the Independence gage would be good because the
input and output D.O. levels are monitored. The Akron
treatment plant lies between the two gages.
186
-------
STATE: OHIO
Monitor Name: Cuyahoga River at Old Portage
USGS I.D.: 04 206 OOP Latitude: 41 08 08 Longitude: 81 32 50
Stream Gage Name: Same
USGS I .D. : Same Latitude: Same Longitude: Same
Rain Gage: Cleveland WB APT
Weather Bureau I.P.: 1657
Daily Data Analysis Results: Water years 1972-1976 were ex-
amined. The probability of low D.O. with high flow never
exceeded 47 percent. The probability of low D.O. or days
with rain exceeded 60 percent in 1976 and averaged 57
percent for the period examined. Water quality is mar-
ginal at 60 to 75 percent of saturation.
Quality of Records: Monitor is visited once a month or more.
Both flow and monitor records are very good with little
missing data.
Discharge Characteristics: Average flow40Qcfs, range 25-
8,000 cfs
Drainage Area at Monitor: 404 square miles
Urban Area(s) Contributing at Monitor: Akron, approx. popu-
lation 690,000, including suburbs
Approximate Urban Area Contributing at Monitor: 50 square
miles or 12 percent of total
Results of Streeter-Phelps Analysis: The analysis indicates
that.the monitor is too close to Akron to sense the maxi-
mum sag in a Streeter-Phelps sense. The Cuyahoga at
Independence monitor is better situated. The analysis
is illustrated in Figure D-14 .
Hourly Data Analysis Results: The period examined is illus-
trated in Figure D-16. This is the same time period
examined for the Cuyahoga at Independence monitor. A
slight increase in the D.O. deficit and a definite damp-
ening of the diurnal D.O. cycles can be seen for the
first two flow events. The effect is not as pronounced
as at Independence. Water quality is marginal, averaging
6 mg/1. It is difficult to conclude much from this hourly
data. Examination of 1976 daily data indicates several
places where D.O. levels well below 5 mg/1 occur with
increased flow.
187
-------
o
Q
20 ,
10 ..
SATURATION DO LEVEL
oo
oo
x
u
Q
(5
LU
<5
1
3.0
2.5 ::
o 2.0 ::
x
1.5 ::
DISCHARGE / AVG. DISCHARGE
8 10 12 14 16 18 20
TIME FROM START OF PERIOD, days
DO DEFICIT/ 10, mg/l
PRECIPITATION, inches
22 24 26 28 30
Figure D-16. Hourly data for Cuyahoga R. at Old Portage, OH.
(6/19/76 to 7/18/76).
-------
Conclusions and Comments: A problem can be identified here,
but not as well as at Independence. It would be an inter-
esting place to study in detail because sufficient evidence
exists to show a correlation between both rain and flow
and low D.O. The availability of monitor records here and
at Independence would be very good for model calibration.
189
-------
STATE: OHIO
Monitor Name: Grand River at Painsville
USGS I.D.: 04 212 200 Latitude: 41 44 09 Longitude: 81 15 59
Stream Gage Name; Grand River near Painsville
USGS I.D.; 04 212 100 Latitude: 41 43 08 Longitude 81 13 41
Rain Gage: Cleveland
Weather Bureau I.P.: 1657
Daily Data Analysis Results: Water years 1975 and 1976 were
examined. The maximum probability of low D.O. with high
flow was 45 percent. The probability of low D.O. on days
with rain was 60 percent in 1975. D.O. levels fall in
the 60 to 70 percent saturation range.
Quality of Records: The monitor is visited once a month or
more.Flow records are good but the monitor record is
very poor with nearly 1/3 of the year missing.
Discharge Characteristics; Average flow 1,000 cfs, range
50-13,000 cfs
Drainage Area at Monitor; 701 square miles
Urban Area(s) Contributing at Monitor: Painsville, approx.
population 30,000
Approximate Urban Area Contributing at Monitor: 4 square
miles or less than 1 percent of total
Results of Streeter-Phelps Analysis: This was not considered
here. River empties into Lake Erie several miles down-
stream.
Hourly Data Analysis Results: The hourly data are illustrated
in Figure D-17. There is some indication of increased
deficit for the small flow event but the D.O. improves with
the larger event. The evidence is inconclusive.
Conclusions and Comments: This site has occasional water
quality problems with the D.O. dropping to less than
5 mg/1. There is no real evidence to connect this to
direct urban runoff. There is no obvious correlation
with flow and the rainfall correlation is suspect because
of the distance (60 miles) to the rain gage from the
monitor.
190
-------
o
o
20 ,.
10 ..
SATURATION DO LEVEL
DC LEVEL
•o
O
Q
6
I
HI
O
K
<
O
3.0 ^
2.5 ::
2.0 ::
1.5 ::
1.0
o.5 ::
o.o
DISCHARGE / AVG. DISCHARGE
6 8 10 12
TIME FROM START OF PERIOD, days;
14
16
18
20
Figure D-17. Hourly data for Grand R. at Painesville, OH.
(8/25/75 to 9/13/75).
-------
STATE: OHIO
Monitor Name: Hocking River Below Athens
USGS I.D.: 03 159 510 Latitude: 39 19 39 Longitude: 82 00 18
Stream Gage Name: Hocking River at Athens
USGS. I.P.; 03 159 500 Latitude: 39 19 44 Longitude: 82 05 16
Rain Gage: Athens
Weather Bureau I.P. ;
0279
Paily Pata Analysis Results; Water years 1972-1976 were ex-
~ amined. The probability of low D.O. and high flow reached
67 percent in 1976. The dissolved oxygen levels are
generally high, averaging 65 to 86 percent of saturation.
Quality of Records: The monitor is visited once a month or
~~ more. Both flow and monitor records are good with little
missing data.
Discharge Characteristics;
range 25-30,000 cfs
Average flow approx. 1,000 cfs,
Drainage Area at Monitor: 943 square miles
Urban Area(s) Contributing at Monitor: Athens, approx. popu-
lation 15,000
Approximate Urban Area Contributing at Monitor; 2 square
miles or less than 1 percent of total
Results of Streeter-Phelps Analysis; Analysis indicates
that the monitor is too close to town. A location 15 to
20 miles downstream would have been better. Even storm
flow seems incapable of doing much harm to water quality
here. The analysis results are illustrated in Figure D-18.
Hourly Data Analysis Results: The period examined is illus-
trated in Figure D-19.It is highly inconclusive evidence
of anything. The D.O. level decreases slightly as the
flow events continue to occur, but that is all. P.O.
levels are high, running quite near saturation. Some
evidence of supersaturation is apparent for the first
three days of the period examined.
Conclusions and Comments: Examination of the daily record
indicates that the P.O. does drop below 5 mg/1 in the
summer, but no evidence is present in the hourly to say
that it is an urban runoff problem. Possibly the level
of sewage treatment is inadequate. There is little else
shown on the map to suspect.
192
-------
U)
15.0 T
10.0 .
H
LU
Q
5.0
SATURATION AT 20°C = 9.17 mg/l
AT 28°C =
DISTANCE, mi
ATHENS STP MONITOR
STORM
30
Figure D-18. Streeter-Phelps analysis results for Hocking R. below Athens, OH.
-------
20 ,.
O
Q
10 .
SATURATION DO LEVEL
DO LEVEL
u
'i
•O
O
Q
C3
O
UJ
O
oc
<
I
8
O
3.0 ,,
2.5 ::
DISCHARGE / AVG. DISCHARGE
PRECIPITATION, inches
DO DEFICIT/10, mg/l
8 10 12 14 16 18 20
TIME FROM START OF PERIOD, days
22
24 26 28
30
Figure D-19. Hourly data for Hocking R. at Athens, OH.
(6/19/76 to 7/18/76).
-------
STATE: OHIO
Monitor Name: Little Miami River at Miamiville
USGS I.D.; 03 245 300 Latitude: 39 12 38 Longitude: 84 17 33
Stream Gage Name: Little Miami River at Milford
USGS I.D.; 03 245 500 Latitude: 39 10 17 Longitude: 84 17 53
Rain Gage: Cincinnati ABBE OBS
Weather Bureau I.P.: 1561
Daily Analysis Results: Water years 1971 to 1974 were ex-
amined. The probability of low D.O. and high flow reached
67 percent in 1974 and 66 percent in 1973. The probability
of low D.O. on days with rain never exceeded 60 percent.
Water quality is generally good at the site with D.O.
levels averaging 80-85 percent saturation.
Quality of Records; The monitor is visited once a month or
more. Both flow and D.O. records are generally continuous.
Discharge Characteristics: Average flow 1,200 cfs, range
100-60,000 cfs
Drainage Area at Monitor: 1200 square miles
Urban Area(s) Contributing at Monitor: Suburbs of Cincinnati,
population difficult to estimate
Approximate Urban Area Contributing at Monitor: unknown -
probably less than 1 percent
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Data Analysis Results: The period of hourly data
examined is illustrated in Figure D-20. There is clear
evidence of an increased D.O. deficit at the time of
the flow events shown. The dissolved oxygen level fell
near the 5 mg/1 level during the events.
Conclusions and Comments: There is clearly a correlation here,
but the cause is not clear. The writers visited the
site and it is not obviously urban. The monitor is near
several small towns in the fringe suburbs of Cincinnati.
There is a good deal of gravel mining and cement pro-
cessing upstream. Deficit levels are not severe enough
to warrant further study here.
195
-------
20 T
O
Q
SATURATION DO LEVEL
•a
O
Q
CJ
X
o
I
5
DISCHARGE / AVG. DISCHARGE
6 8 10 12
TIME FROM START OF PERIOD, days
14
16
20
Figure D-20. Hourly data for L. Miami R. at Miamiville, OH.
(8/1/73 to 8/30/73).
-------
STATE: OHIO
Monitor Name: Little Miami River near Spring Valley
USGS I.D.: 03 242 050 Latitude: 39 35 00 Longitude: 84 01 49
Stream Gage Name: Same
USGS I.D.: Same Latitude: Same Longitude: Same
Rain Gage: Dayton WB APT
Weather Bureau I.P.; 2075
Daily Data Analysis Results: Water years 1972-1976 were ex-
amined. The probability of low P.O. with high flow reached
58 percent in 1972, 59 percent in 1976, and exceeded 60
percent in 1974 and 1975. The probability of low P.O. on
days with rainfall exceeded 60 percent in 1974 and 1975.
Pissolved oxygen levels ranged from 60 to 75 percent satu-
ration.
Quality of Records: The monitor is visited once a month or
more. The flow record is continuous. The monitor record
is generally continuous with an occasional two to three
weeks missing.
Pischarge Characteristics: Average flow 386 cfs, range 23-
18,400 cfs
Prainage Area at Monitor: 366 square miles
Urban Area(s) Contributing at Monitor: Payton, Payton sub-
urbs, Xenia, approximate populations 60-70,000 and 25,000
respectively
Approximate Urban Area Contributing at Monitor: 11 to 13
square miles or 3 percent of total
Results of Streeter-Phelps Analysis: Not performed at this
site.
Hourly Pata Analysis Results: A clear correlation exists
between the presence of storm event flow and a lowering
of the dissolved oxygen level for the period examined.
This is illustrated in Figure P-21. There is a fairly
clear relationship between the rainfall events and
small increases in the P.O, deficit,, The P.O. level
is driven close to the 5 mg/1 level by the flow event.
Conclusions and Comments: Sutron contacted the cities of
Payton and Xenia concerning the correlation observed
here. The Assistant City Manager at Xenia indicates
that the discharge of raw sewage at times of high flow
from the city of Spring Valley is probably the reason
197
-------
20
o
Q
10
SATURATION DO LEVEL
DO LEVEL
3.0 ,
00
DISCHARGE / AVG. DISCHARGE
6 8 10 12 14
TIME FROM START OF PERIOD, days
16
18
20
Figure D-21. Hourly data for L. Miami R. near Spring Valley, OH.
(7/1/75 to 7/20/65).
-------
for poor water quality here. The correlation is unmis
takeable here and a limited further investigation may
be warranted.
199
-------
STATE: OHIO
Monitor Name: Mad River near Dayton
USGS I.D.: 03 270 OOP Latitude: 84 05 19 Longitude: 84 05 19
Stream Gage Name: Same
USGS I.D.: Same Latitude: Same Longitude: Same
Rain Gage; Dayton WB APT
Weather Bureau I.D.: 2075
Daily Data Analysis Results: Water years 1972-1976 were ex-
amined. The probability of low D.O. at times of high flow
reached 56 percent in 1975 and exceeded 60 percent in 1972.
The probability of low D.O. days with rainfall exceeded
60 percent all 5 years. Dissolved oxygen levels are gen-
erally in the 60 to 75 percent of saturation range.
Quality of Records; The monitor is visited once a month or
more.Records are generally continuous for both flow and
the monitor. Occasional periods of one to two weeks are
missing.
Discharge Characteristics: Average flow 620 cfs, range 100-
20,000 cfs
Drainage Area at Monitor: 635 square miles
Urban Area(s) Contributing at Monitor; Springfield, small
portion of Dayton suburbs, Wright Patterson Air Force
Base, approximate population 90-100,000
Approximate Urban Area Contributing at Monitor: 11-12 square
miles or 2 percent of total
Results of Streeter-Phelps Analysis: Streeter-Phelps analysis
indicates that a storm flow at Springfield could produce
a deficit of 4-5 mg/1 at the monitor. The monitor is prob-
ably 10 to 12 miles too far downstream to best sense
effects from Springfield. The results are illustrated
in Figure D-22.
Hourly 'Data Analysis Results; The hourly data examined are
illustrated in Figure D-23. There is a fairly clear cor-
relation between periods of high flow and periods when the
D.O. deficit remains high. The minimum value at times of
high flow is no worse than on normal days but no recovery
to "better" conditions takes place. The correlation be-
tween the rainfall and flow events is fairly clear. D.O.
levels fall to roughly 5-6 mg/1. While the correlation
is clear, no real problem seems to exist here as far as
standards go.
200
-------
15.0 ,.
10.0
SATUftATJON AT_17_°C_=_9.74_ma/l_
SATURATION AT_27^C_=_8.07jng/l_
m
O
O
Q
5.0
STORM
SPRINGFIELD
MONITOR
HUBER HEIGHTS
Figure D-22. Streeter-Phelps analysis results for Mad R. near Dayton, OH.
-------
20 „
O
D
SATURATION DO LEVEL
3.0 T
K)
O
NJ
.e
u
•o
8
I
i
I
Q
DISCHARGE / AVG. DISCHARGE
8 10 12 14 16 18 20 22 24 26 28 30
TIME FROM START OF PERIOD, days
Figure D-23. Hourly data for Mad R. near Dayton, OH.
(6/26/72 to 7/25/72).
-------
Conclusions and Comments; This is a fairly good example
of a correlation. It would require further study to
determine if the monitor is actually located in a poor
place and if the problem is worse upstream as indicated
by Streeter-Phelps. The Streeter-Phelps results indi-
cate that D.O. levels as low as 3.0 mg/1 might be found,
203
-------
STATE: OHIO
Monitor Name: Mahoning River at Ohio-Pennsylvania State Line
below Lowellville
USGS I.D.: 03 099 510 Latitude: 41 01 53 Longitude: 80 31 10
Stream Gage Name: Mahoning River at Lowellville
USGS I.D.: 03 099 500 Latitude: 41 02 12 Longitude: 80 32 11
Rain Gage: Youngstown
Weather Bureau I.D.: 9406
Daily Data Analysis Results: Water years 1973-1976 were
examined. The probability of low D.O. at times of high
flow days with rainfall reached 60 percent in 1976.
Water quality is fairly poor with D.O. levels averaging
from 45 to 55 percent saturation. The 1976 daily data
are illustrated in Figure D-24. There are clearly
identifiable periods when flow events coincide with
periods of near zero dissolved oxygen. The rainfall
correlation is also clear.
Quality of Records: The monitor is visited once a month or
more.Both flow and monitor records are excellent with
only isolated periods of 3-5 days missing.
Discharge Characteristics: Average flow 1400 cfs, range
700-7,000 cfs
Drainage Area at Monitor: 1073 square miles
Urban Area(s) Contributing at Monitor; Niles, Girard,
Youngstown, approx. population 550,000, including
metropolitan area.
Approximate Urban Area Contributing at Monitor: 25 square
miles or 2 percent of total
Results of Streeter-Phelps Analysis; The results of the
analysis are shown in Figure D-25. Storm flow from Youngs-
town is theoretically capable of driving the D.O. to zero.
The monitor should ideally be 10 to 15 miles further down-
stream to sense this. It seems to be getting the full
effect where it's at now, however.
Hourly Data Analysis Results; Hourly data could not be
obtained at this site. The daily analysis clearly in-
dicates the correlation between both rainfall, flow and
low D.O.
204
-------
10.00-1
to
O
en
r40.00
o
^
1
-20.00
0.00
270.00
-10.00 -I
Figure D-24. Daily DO and flow for Mahoning R. at
OH.-PA. state line below Lowellville, OH.
-------
15.0 „.
o
CTi
10.0 ..
h-
O
iZ
HI
Q
O
Q
SAJURAT1QN AT_31°C ^_6.5 mg/j
5.0 ..
NORMAI
30 40 50
DISTANCE, mi
60
70 80
MONITOR
FLOW
YOUNGSTOWN BELOW R. STP
YOUNGSTOWN ABOVE R. STP
GIRARDSTP
NILESSTP
NILES
Figure D-25. Streeter-Phelps analysis results for Mahoning R. at
OH.-PA. state line near Lowellville, OH.
-------
Conclusions and Comments: Youngstown is very heavily in-
dustrialized. The entire river valley is a mess of
rail yards and steel plants. The entire urban area
drains into the river. Sutron contacted the Youngs-
town treatment facility and learned that no facilities
above primary level exist in the Mahoning Valley. A
secondary treatment facility for Youngstown has
reached final design stages.
207
-------
STATE: OHIO
Monitor Name: Maumee River at Defiance
USGS I.P.: 04 184 100 Latitude: 41 16 43 Longitude: 84 23 07
Stream Gage Name: Maumee River at Antwerp
USGS I.D.: 04 183 500 Latitude: 41 11 56 Longitude: 84 44 40
Rain Gage: Defiance
Weather Bureau I.P.: 2098
Paily Pata Analysis Results: Water years 1972-1976 were ex-
amined.The probability of low P.O. at times of high flow
exceeded 60 percent in 1973 and 1975. The P.O. levels are
generally in the 60 to 70 percent of saturation range.
Quality of Records; The monitor is visited once a month or
more.The record quality, however, is poor. One to two
months of record are missing each year. Much of the record
missing is in the summer when it would have been of most
interest.
Pischarge Characteristics: Average flow 2680 cfs, range 26-
26,200 cfs
Drainage Area at Monitor: 2300 square miles
Urban Area(s) Contributing at Monitor: Fort Wayne, Indiana
(none of Defiance), population 182,000
Approximate Urban Area Contributing at Monitor: 25 square
miles or 1 percent of total
Results of Streeter-Phelps Analysis: The Streeter-Phelps tech-
nique was used to determine if the problem observed could
originate at Fort Wayne, Indiana. This was done because
the monitor is upstream of all of Defiance. The results,
illustrated in Figure D-26, indicate that storm flow from
Fort Wayne could cause a 4-5 mg/1 deficit (2-3 mg/1 absolute
DO level) at Defiance. The monitor would have been better
located at Antwerp, Indiana, to sense the maximum deficit.
Hourly Data Analysis Results; Hourly data were not examined
at this site because of the large distance between the
monitor and flow gages (20+ miles).
Conclusions and Comments; It seems quite probable that 5.0
mg/1 standard violations would be detected by a properly
placed monitor. EPA standards violations might also be
found. The daily data are of poor quality, however, and
the monitor is not in the best location. The reach from
Fort Wayne to Antwerp would definitely be worth investi-
gating.
208
-------
t-o
o
15.0 ,.
10.0 ..
Q
5.0 ..
'•
JLAJU_RATJC3W AT_20.5?C_=_9.g7 jpg/l
0.0
^--"~ ~ 7~Z~TZT~7~~^:T:::ri — i — ; — ; — ; — ; — ; — : — :
10 20 30 4 40 50 '
!
:ORT WAYNE, IN. FLOW
i i
'60
JEF
DISTANCE, mi
DEFIANCE, OH.
MONITOR
Fioure D-26. Streeter-Phelps analysis results for Maumee R. at Defiance, OH.
-------
STATE: OHIO
Monitor Name: Portage River at Railroad Bridge at Woodvilie
USGS I.D7;04 195 600 Latitude: 41 26 58 Longitude: 83 21 29
Stream Gage Name; Portage River at Woodville
USGS I.D.; 04 195 500 Latitude: 41 26 58 Longitude: 83 21 41
Rain Gage: Freemont
Weather Bureau I.P.; 2974
Daily Data Analysis Results: Water years 1971-75 were ex-
amined. The probability of low D.O. with high flow
exceeded 60 percent in 1973 and reached 68 percent in
1975. Dissolved oxygen levels generally fall in the
50 to 70 percent saturation range.
Quality of Records: The monitor is visited once a month or
more.The flow record is continuous. The monitor record
usually contains periods of from two weeks to a month or
more missing each year.
Discharge Characteristics: Average flow 300 cfs, range 0-
17,000 cfs
Drainage Area at Monitor: 428 square miles
Urban Area(s) Contributing at Monitor-: Bowling Green, Portage,
approximate population 22,000 and less than 5,000 respec-
tively
Approximate Urban Area Contributing at Monitor: 3 square
miles or 1 percent of total
Results of Streeter-Phelps Analysis: Results of the analysis
are shown in Figure p-27. They indicate that storm flow at
Bowling Green might create a D.O. deficit of 3.5 to 4 mg/1
at Woodville. The monitor location is too far downstream
for ideal study of a problem at Bowling Green.
Hourly Data Analysis Results: The period examined is illus-
trated in Figure D-28. There is a clear tendency for the
D.O. deficit to increase and hold steady at times of
high flow. The correlation of the flow events with rain-
fall is also clear. D.O. levels fall to roughly 6 mg/1
at times of high flow.
Conclusions and Comments: There is clear evidence of a corre-
lation here. Water quality in the river may be worse than
indicated because of poor monitor location. The urban area
involved is relatively small, but would probably be worth
investigating.
210
-------
15.0
10.0 L - SAJURATJON AT_15.5fC_=JOjp5_mg/l_
y
LL
UJ
D
O
a
5.0
SATURATION AT 30°C = 7.63 mg/l
10
BOWLING GRtEN
50
60
WOODVILLE
DISTANCE, mi
Figure D-27. Streeter-Phelps analysis results for Portage R. at
Railroad Bridge at Woodville, OH.
-------
20 „
r
o
Q
SATURATION DO LEVEL
3.0 „
DISCHARGE / AVG. DISCHARGE
6 8 10 12 14
TIME FROM START OF PERIOD, days
Figure D-28. Hourly data for Portage R. at Woodviile, OH.
(8/14/75 to 9/2/75).
-------
STATE: OHIO
Monitor Name: Sandusky River near Upper Sandusky
USGS I.D. ; 04 196 500 Latitude: 40 51 02 Longitude: 83 15 23
Stream Gage Name: Same
USGS I.D. : Same Latitude: Same Longitude: Same
Rain Gage: Upper Sandusky
Weather Bureau I.P.: 8534
Daily Data Analysis Results: Water Years 1972-76 were examined.
The probability of low D.O. with high flow reached 60 per-
cent in 1973 and exceeded 60 percent in 1976. The proba-
bility of low D.O. on days with rainfall exceeded 60 percent
in 1974, 75 and 76. D.O. levels were generally in the 70-
75 percent of saturation range. Periods when low D.O.
occurs at times of both flow and rainfall can be clearly
identified on the daily plots.
Quality of Records: The monitor is visited once a month or
more. The flow record is continuous. The monitor record
has minor interruptions occasionally but these are in-
frequent and of short duration.
Discharge Characteristics: Average flow 250 cfs, range 10-
10,000 cfs
Drainage Area at Monitor: 298 square miles
Urban Area(s) Contributing at Monitor: Bucyrus, Upper San-
dusky, approximate population 13,000 and 7500 respectively
Approximate Urban Area Contributing at Monitor: 3 square
miles or 1 percent of total
Results of Streeter-Phelps Analysis; The results of the analy-
sis are shown in Figure D-29. The storm flow from Bucyrus
could conceivably cause a 3-4 mg/1 deficit at Upper Sandusky.
The monitor is not in a good position to sense the effect of
either town.
Hourly Data Analysis Results; The period analyzed is illus-
trated in Figure D-30. There is clear evidence that a
rainfall event followed by a flow increase is accompanied
by an increase in D.O. deficit. The period examined
contained one severe deficit where the D.O. level fell
to less than 1 mg/1.
Conclusions and Comments: There is clear evidence of a problem
here"The exact cause is not obvious, although it probably
originates at Bucyrus. The monitor is rather poorly located
213
-------
15.0
10.0 ..
H
ui
Q
O
a
5.0 ..
=L9X>7_mg/l_
SATURATJON
BUCYRUS
30
FLOW, MON.
STP
UPPER SANDUSKY
STORn/|
DISTANCE, mi
Figure D-29. Streeter-Phelps analysis for Sandusky R. near Upper Sandusky, OH.
-------
o
a
20 ,.
10
SATURATION DO LEVEL
Ul
DISCHARGE / AVG. DISCHARGE
6 8 10 12 14
TIME FROM START OF PERIOD, days
16
18
20
Figure D-30. Hourly data for Sandusky R. near Upper Sandusky, OH.
(8/6/73 to 8/25/73).
-------
in a Streeter-Phelps sense but still picks up the problem.
This site would also be worth examining more closely. In
an independent study, Ohio State University indicated that
poor treatment facilities at Bucyrus may cause problems in
the Sandusky.
216
-------
STATE: OHIO
Monitor Name: Scioto River at Chillicothe
USGS I.D.; 03 '231 500 Latitude: 39 20 29 Longitude: 82 58 16
Stream Gage Name: Same
USGS I.D.: Same Latitude: Same Longitude: Same
Rain Gage: Chillicothe-Mound City
Weather Bureau I.P.; 1528
Daily Data Analysis Results: Water years 1972-1976 were ex-
amined. The probability of low D.O. with high flow ex-
ceeded 50 percent all five years and exceeded 60 percent
in 1972. The probability of low D.O. or days with rainfall
exceeded 60 percent in 1976. D.O. levels are generally
in the 40 to 60 percent of saturation range. Periods of
low D.O. which coincide with flow and rainfall events can
be clearly identified on the daily data plots.
Quality of Records; The monitor is visited once a month or
more. The flow record is continuous and the monitor
record is very good. Only isolated periods of a few
days are missing.
Discharge Characteristics: Average flow 35,000 cfs, range
500-50,000 cfs
Drainage Area at Monitor: 3850 square miles
Urban Area(s) Contributing at Monitor; Columbus, approximate
population 1,069,000
Approximate Urban Area Contributing at Monitor; 22 square
miles or 1 percent of total
Results of Streeter-Phelps Analysis; The results of the analy-
sis are shown in FigureD^Sl.ft is theoretically possible
for storm flow from Columbus to create a 3-4 mg/1 deficit
at' Chillicothe. 'The monitor is poorly located. It would
probably see nearly twice the deficit if it were at Circle-
ville, halfway between Columbus and Chillicothe.
Hourly Data Analysis Results; The period examined is illus-
trated in Figure D-32.There is clear evidence of a corre-
lation between both rainfall and flow and increased D.O.
deficit. In one storm event, the D.O. drops very near zero.
This violates the EPA suggested 2.0 mg/1 for 4.0 hours
standard.
217
-------
15.0 T
to
M
00
10.0 ..
LU
Q
O
Q
SATURATION AT 27°C = 8.07 mg/l
20 30 40 50 60 70 8(
COLUMBUS
CHILLICOTHE
MONITOR
Figure D-31. Streeter-Phelps analysis for Scioto R. at Chillicothe, OH.
-------
o
Q
20 ,.
10 ..
SATURATION DO LEVEL
3.0 ,
DISCHARGE / AVG. DISCHARGE
8 10 12 14 16 18 20
TIME FROM START OF PERIOD, days
22 24 26 28 30
Figure D-32. Hourly data for Scioto R. at Chillicothe, OH.
(8/1772 to 8/30/72).
-------
Conclusions and Comments: This is a good example of a site
with a definite storm runoff related problem. It would
definitely merit further study. One Sutron engineer
worked with USGS in Columbus and is familiar with the
Chillicothe monitor. He states that the diurnal oxygen
variations are due to heavy algal growth in the channel.
He also indicates that Columbus often flushes the
trickling filters at the treatment plant during storm
flows. Conversations with Columbus treatment facility
operators indicate that storm flow bypassing does occur.
220
-------
STATE: OREGON
Monitor Name: South Umpqua River near Roseburg
USGS I.D.: 14 312 260 Latitude: 43 13 20 Longitude: 123 24 45
Stream Gage Name: South Umpqua River near Brockway
USGS I.P.; 14 312 OOP Latitude: 43 08 00 Longitude: 123 23 50
Rain Gage: Roseburg KQEN
Weather Bureau I.P.: 7331
Daily Pata Analysis Results; Water years 1972-1976 were ex--
amined. The probability of low P.O. with high flow reached
63 percent in 1976 and reached 57 and 56 percent in 1972
and 1973 respectively. The probability of low P.O. on
days with rainfall never exceeded 50 percent. P.O. levels
are generally high, averaging 80 to 87 percent saturation.
Quality of Records: The monitor is visited once a month or
more. The flow and monitor records are both intermittent
in the winter but are generally good.
Pischarge Characteristics; Average flow 3,000 cfs, range 50
to 40,000 cfs
Prainage Area at Monitor: 1800 square miles
Urban Area(s) Contributing at Monitor: Winston, Roseburg,
approximate combined population of 65,000 or more
Approximate Urban Area Contributing at Monitor: 9 square
miles or 1 percent of total
Results of Streeter-Phelps Analysis; Not performed at this
site.
Hourly Pata Analysis Results; The period examined is illus-
trated in Figure P-33. There is clear evidence of an
increased P.O. deficit accompanying an increase in flow.
There is no real problem in a quality sense, as the P.O.
remains above 8 mg/1 at all times.
Conclusions and Comments: There is no doubt that something
happens here. Periods of low P.O. at times of high flow
can be clearly identified on the daily data plots. Based
on other sites, the monitor would probably see a greater
deficit were it located 20 miles further downstream. The
low magnitude of the deficits do not appear to merit fur-
ther study.
221
-------
20 ,.
O
Q
10 .«
SATURATION DO LEVEL
_—• -
-fT— ~_
DO LEVEL
O
t—
5
O
i
•O
O
Q
13
<
cc
<
x
3.0 T
2.5 ::
2.0 ::
1.5 ::
1.0 ::
0.5 ::
0.0
DISCHARGE / AVG. DISCHARGE
DO DEFICIT/10, mg/l
6 8 10 12
TIME FROM START OF PERIOD, days
14
16
18
20
Figure D-33. Hourly data for South Umpqua R. near Brockway, OR.
(3/1776 to 4/6/76)
-------
STATE: PENNS YLVANIA-
Monitor Name; Delaware River at Bristol
USGS I.D.: 01 464 &00 Latitude: 40 05 55 Longitude: 74 51 58
Stream Gage Name; Delaware River at Trenton, New Jersey
USGS I.D.: 01 463 500 Latitude: 40 13 18 Longitude: 74 46 42
Rain Gage; Trenton
Weather Bureau I.P.: 8883
Daily Data Analysis Results; Water years 1972-1976 were ex-
amined. The probability of low D.O. at times of high flow
exceeded 50 percent all years. It reached 71 percent in
1975. Average D.O. levels ranged from 65 to 85 percent
of saturation.
Quality of Records; The monitor is visited once a month or
more. Records in the summer for the monitor are very bad
with as much as two months missing. The flow record is
continuous.
Discharge Characteristics: Average flow 12,000 cfs, range
200-32,500 cfs
Drainage Area at Monitor: 7163 square miles
Urban Area(s) Contributing at Monitor: Trenton, Levitown,
Bristol, Bordentown, PA, and suburbs, population greater
than 1/2 million
Approximate Urban Area Contributing at Monitor: 65 square
miles or 1 percent of total
Results of Streeter-Phelps Analysis; Not performed at this
site.
Hourly Data Analysis Results: Hourly flow data were not
available. The Trenton daily flows are plotted with the
hourly D.O. and D.O. saturation data in Figure D-34.
There is clearly an increase in D.O. deficit with in-
creased flow. There is no real water quality problem,
however, as the D.O. never falls below 8 mg/1.
Conclusions and Comments; The Trenton area forms the up-
stream end of a large urban megalopolis. Water quality
deteriorates steadily from here to Philadelphia and
below. The problem here is not bad enough to warrant
further study.
223
-------
O
D
20 -r
SATURATION DO LEVEL
DO LEVEL
to
tsJ
a
i
Q
CJ
UJ
O
tt
<
I
CJ
3.0
2.5 ::
2.0 ::
1.5 ::
1.0 ::
o.s ::
0.0
DISCHARGE / AVG. DISCHARGE
DO DEFICIT / 10, mg/l
6 8 10 12 14
TIME FROM START OF PERIOD, days
16
18
20
F/aure D-34. Hourly data for Delaware R, at Bristol, PA.
(4/20/75 to 5/9/75)
-------
STATE: PENNSYLVANIA
Monitor Name: Delaware River at Chester
USGS I.p.: 01 464 600 Latitude: 40 05 55 Longitude: 74 51 58
Stream Gage Name: Delaware River at Trenton, New Jersey
USGS I.D. .- 01 463 500 Latitude: 40 13 18 Longitude: 74 46 42
Rain Gage: Philadelphia WB APT
Weather Bureau I.P.; 6889
Daily Data Analysis Results: Water years 1972-1976 were ex-
amined. The probability of low D.O. with high flow ex-
ceeded 50 percent in 1973, 1974, and 1976, reaching a
peak of 67 percent in 1973. D.O. levels average 65 to
75 percent of saturation.
Quality of Records: The monitor is visited once a month or
more. Flow records are continuous. Monitor records are
highly intermittent with roughly 1/3 of each month missing.
Discharge Characteristics: Average flow 12,000 cfs, range
1200-32,500 cfs
Drainage Area at Monitor: 10,300 square miles
Urban Area(s) Contributing at Monitor: Trenton, New Jersey,
Philadelphia metropolitan area
Approximate Urban Area Contributing at Monitor: Greater than
100 square miles but less than 2 percent of total
Results of Streeter-Phelps Analysis: Not performed at this
site
Hourly Data Analysis Results: Hourly flow data were not
available here.The hourly D.O. and D.O. saturation
values are plotted with the Trenton, New Jersey, daily
flows in Figure D-35. The D.O. level drops to less than
5 mg/1 as the flow increases for the period examined.
The relationship is not as pronounced as at other sites.
Conclusions and Comments: There is evidence of a correla-
tion here.The fact that the D.O. can be driven down
4 mg/1 by a drainage area which represents less than
2 percent of the total speaks for the potency of the
runoff here. This is a very large river and the changes
are much slower in a time sense than at other locations.
This is probably too large and complex a site to study
effectively at reasonable cost.
225
-------
20 ,.
O
Q
10 ..
SATURATION DO LEVEL
DO LEVEL
CTi
•s
T3
O
Q
a
O
LU
C3
cc
3.0 ^
2.5 ::
2.0 ::
DISCHARGE / AVG. DISCHARGE
6 8 10 12 14
TIME FROM START OF PERIOD, days
16
20
Figure D-35. Hourly data for Delaware R. at Chester, PA.
(2/26/73 to 3/17/73).
-------
STATE: PENNSYLVANIA
Monitor Name; Lehigh River at Easton
USGS I.D.; 01 454 720 Latitude: 40 41 12 Longitude: 75 12 32
Stream Gage Name: Lehigh River at Glendon
USGS I.P.; 01 454 700 Latitude: 40 40 09 Longitude: 75 14 12
Rain Gage: Allentown WB APT
Weather Bureau I.P.; 0106
Paily Data Analysis Results: Water years 1972-1976 were ex-
amined. The probability of low D.O. with high flow reached
53 percent in 1972 but averaged 35 to 40 percent. The
probability of low D.O. on days with rainfall exceeded
60 percent every year but 1974. That year it was 57 per-
cent. D.O. levels in general are good, averaging 75 to
85 percent saturation. Periods of low D.O. at times of
rianfall are discernable on the daily data plots.
Quality of Records: The monitor is visited once a month or
more. Flow records are continuous. The monitor record
exhibits intermittent missing periods of a week or less.
Discharge Characteristics: Average flow 3,000 cfs, range
1500-16,000 cfs
Drainage Area at Monitor: 1360 square miles
Urban Area(s) Contributing at Monitor: Allentown, Bethlehem,
population approximately 610,000 and 71,500 respectively.
Approximate Urban Area Contributing at Monitor: 25 square
miles or 2 percent of total
Results of Streeter-Phelps Analysis: The results of the
analysis are shown in Figure D-36. Theoretically, storm
flow from Allentown-Bethlehem could produce a deficit of
3-4 mg/1 at Easton. The monitor could not be located at
the sag point under any circumstances because the Lehigh
joins the Delaware at Easton.
Hourly Data Analysis Records: The period examined is illus-
trated in Figure D-37. The results do not clearly show
anything. The D.O. deficit appears to reach a maximum
between the first and second rainfall events. As the
flow increases, the D.O. level improves, nearly reaching
saturation at one point. This type of behavior can also
be seen on the daily plots. The deficit nears the 5 mg/1
level at its worst.
227
-------
M
15.0 ,.
10.0 ..
I
O
H
Ul
Q
O
Q
5.0 ..
_ SAJURATION AT_27°C =^8.07jng/l
" ~ ^CsZJiamlS/r
EDGE
OF
ALLENTOWN
2.0 4.0
ALLENTOWN
8.0 10.0
DISTANCE, mi
BETHLEHEM
12.0
14.0
16.0
EDGE
OF
EASTON
T 18.0
MONITOR
Figure D-36. Streeter-Phelps analysis for Lehigh R. at Easton, PA.
-------
o
Q
20 „
10
SATURATION DO LEVEL
NO
to
O
Q
C3
O
CO
Q
O
LLJ
C3
cc
<
o
CO
3.0 T
2.5 ::
2.0 ::
DISCHARGE / AVG. DISCHARGE
2 4 6 8 10 12 14 16
TIME FROM START OF PERIOD, days
18 20
Figure D-37. Hourly data for Lehigh R. at Easton, PA.
(8/1776 to 8/20/76).
-------
Conclusions and Comments; There does not appear to be a
real problem at this site, even though Streeter-Phelps
indicates one could theoretically exist. Water quality
is generally good. The D.O. generally increases with
increased flow. It would be interesting to see why the
D.O. seems to consistently be low at times of rainfall.
The cause is not obvious.
230
-------
STATE: PENNSYLVANIA
Monitor Name: Schuylkill River at Philadelphia
USGS I.P.; 01 474 500 Latitude: 39 58 00 Longitude: 75 11 20
Stream Gage Name: Same
USGS I.P.: Same Latitude: Same Longitude: Same
Rain Gage; Western Philadelphia, Drexel Institude of Technology
Weather Bureau I.P.; 6879
Daily Data Analysis Results: Water years 1969-1972 were ex-
amined. The probability of low D.O. with high flow reached
61 percent in 1972. D.O. levels averaged 60 to 70 percent
of saturation.
Quality of Records: The monitor is visited once a month or
more. The flow record is continuous. The monitor record
is fair with periods of up to one month missing, generally
in the summer.
Discharge Characteristics: Average flow 1900 cfs, range 0-
100,000 cfs
Drainage Area at Monitor: 1893 square miles
Urban Area(s) Contributing at Monitor: Philadelphia and
suburbs
Approximate Urban Area Contributing at Monitor; 40 square
miles or 2 percent of total
Results of Streeter-Phelps Analysis: Not performed at this
site
Hourly Data Analysis Results; The period examined is illus-
trated in Figure D-38.There is clear evidence of a D.O.
deficit increase at the time of the flow event. The
correlation between the rainfall and increase in flow is
also clear. The D.O. drops below 5 mg/1 during the event
shown.
Conclusions and Comments: There is clear evidence of a cor-
relation here. The monitor is probably not in the best
location based on similar sites elsewhere. Unfortunately,
it could only be moved 4 or 5 miles downstream, as the
Schuylkill joins the Delaware there. The junction is
just above the Delaware at Chester monitor which is dis-
cussed elsewhere. There is no question of urban area
here, but other sites would be simpler hydraulically for
detailed studies.
231
-------
20 T
NJ
CO
NJ
O
Q
.S
u
•o
O
Q
O
ui
O
K
<
U
CO
SATURATION DO LEVEL
DISCHARGE / AVG. DISCHARGE
6 8 10 12
TIME FROM START OF PERIOD, days
Figure D-38. Hourly data for Schuylkil! R. at Philadelphia, PA.
(8/19/72 to 9/7/72).
-------
STATE: TEXAS
Monitor Name: Trinity River below Dallas
USGS I.D.: 08 057 410 Latitude: 32 42 27 Longitude: 96 44 Qi
Same
Longitude: Same
Stream Gage Name: Same
USGS I.P.; Same Latitude:
Rain Gage: Dallas FAA APT
Weather Bureau I.P.: 2244
Daily Data Analysis Results: Only water year 1977 was avail-
able, as the stream has not been monitored long. A 60
percent probability of low D.O. at high flow was computed.
The dissolved oxygen level averaged an extremely low 26
percent of saturation.
Quality of Records: The monitor was serviced once a month or
more. The flow record is continuous for the entire year.
The monitor record was highly intermittent in the summer.
Pischarge Characteristics:
1,000-8,000 cfs
Average flow 2,000 cfs, range
Drainage Area at Monitor: 6278 square miles
Urban Area(s) Contributing at Monitor: Dallas-Fort Worth
Metropolitan Area
Approximate Urban Area Contributing at Monitor; 724 square
miles or 12 percent of total
Results of Streeter-Phelps Analysis; Not performed at this
site
Hourly Data Analysis Results: Two sets of hourly data were
examined for the Dallas gage as well as one for the
Trinity River at Rosser, Texas. Rosser is some 40 miles
south of Dallas and seemed to be in a better position to
sense a deficit in the Streeter-Phelps sense. All three
data sets are illustrated in Figures D-39, D-40, and D-41.
There is no doubt that a problem exists here. The flow
of the Trinity is almost all treated sewage under low
flow conditions. The D.O. is often zero during flow
events. Note that even at Rosser the D.O. goes to zero
with a flow event.
Conclusions and Comments: The Trinity River below Dallas
has some of the worst water quality encountered anywhere
in this study. During summer low flow periods, the river
is primarily treated sewage effluent. It would be par-
233
-------
20 T
O
Q
10 ..
SATURATION DO LEVEL
DO LEVEL
3.0 T
to
Ul
I
O
Q
(D
O
(S
til
O
oc
<
8
a
DISCHARGE / AVG. DISCHARGE
8 10 12 14 16 18 20
TIME FROM START OF PERIOD, days
22 24 26 28
30
Fiaure D-39. Hourly data for Trinity R. below Dallas, TX.
(5/21/77 to 6/19/77).
-------
20 ,.
O
Q
10 ..
SATURATION DO LEVEL
to
CO
O
Q
i
<
CD
CC
<
O
CO
DISCHARGE / AVG. DISCHARGE
4 6
TIME FROM START OF PERIOD, days
Figure D-40. Hourly data for Trinity R. near Rosser, TX.
(9/4/77 to 9/13/77).
-------
20 „
o
a
10 .
SATURATION DO LEVEL
?t'%>*^^
CO
CTl
o
£
•o
O
O
C3
O
la
Q
CD
m
O
oc
<
o
GO
DISCHARGE / AVG. DISCHARGE
8 10 12 14 16 18 20
TIME FROM START OF PERIOD, days
22 24 26 28 30
Figure D-41. Hourly data for Trinity R. below Dallas, TX.
(9/1777 to 9/30/77).
-------
ticularly interesting to see exactly what causes the
D.O. to go to zero when flow increases. This is perhaps
a reintrainment of oxygen demanding material. The urban
area involved is very large and a rather expensive samp-
ling program might be required to adequately define the
problem.
237
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-600/2-79-156
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
November 1979
(Issuing Date)
DISSOLVED OXYGEN IMPACT FROM URBAN STORM RUNOFF
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Thomas N. Keefer, Robert K. Simons, and Raul S,
McQuivey
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
The Sutron Corporation
1925 North Lynn Street Suite 700
Arlington, Virginia 22209
10. PROGRAM ELEMENT NO.
1BC822
11. CONTRACT/GRANT NO.
No. 68-03-2630
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Gin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final 11-18-77 thru 5-1-79
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer:
John N. English 513/684-7613
16. ABSTRACT The primary objective of the research reported nere is to determine if
on a national basis a correlation exists between strength of dissolved oxygen
(DO) deficits and the presence of rainfall and/or storm runoff downstream of
urban areas. A secondary objective is to estimate the magnitude and extent of
the problem.
One hundred and four water quality monitoring sites in and downstream of urban
areas throughout the country were considered for inclusion in the study. These
were screened from over 1000 monitors maintained by federal and state agencies
such as the U.S. Geological Survey, Environmental Protection Agency (EPA), Ohio
River Valley Sanitation Commission and Wisconsin Department of Natural Resources.
Daily data were obtained and processed for 83 of the 104 candidate sites. Of the
83 monitors considered, 42 percent or roughly four monitors in ten of the 104
candidates demonstrated a 60 percent or greater probability of a higher than
average DO deficit occurring at times of higher-than-average stream flow or on
days with rainfall. In general, the data examined here indicate that 19 percent
of the 104 candidate monitors might not meet a 5.0-mg/l standard and 15 percent
might not meet a 2.0-mg/l standard. Frequency of violations was not tabulated
exactly but appears to be zero to five times per year at sites with
correlations.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Rainfall
•-Surface Water Run off
Combined Sewers
Water Pollution
*Water Quality
^Dissolved Oxygen
Urban Runoff
"Streeter-Phelps"
13B
EMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
250
2O. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
238
SUSGPO: 1980-657-146//5513
------- |