Region 111 Library
Environmental Protection Agency ^ £n,lronmr,{a! Pr3tec«on Ageacj
Region 1H information Resource.
Center (-PM52)
841 Chestnut Street
Pbite&lpiiia, PA 1910Z - -" -*
WATER QUALITY MONITORING PROGRAMS FOR SELECTED
SUBESTUARIES OF THE CHEASAPEAKE BAY
to
Environmental Protection Agency,
Environmental Research Laboratory
Athens, GA
Environmental Protection Agency, Region III,
Chesapeake Say Program
Annapolis, MD
Maryland Department of Natural Resources,
:,teter Resources Administration
Annapolis, MD
'-'irginia State idater Control Board,
/'ureau of Water Control Management
rlichmond, VA
SM Brown
RM Ecker
June 27, 1973
3ATTELLE,
Pacific'Northwest Laboratories
Rich land, UA 99352
EPA Report Collection
Information Resource Center
US EPA Region 3
Philadelphia, PA 19107
-------
-------
s ;. - : i'.^..judi'i Source
SUMMARY C,. ;•;;,„/:]
L:; :/, ., :.::„,A
Fi,;:dJ,;.,,:u;hi 19107 .
In order to obtain the data required to tast and compare existing, state-
of-the-act estuary water quality models, comprehensive sampling programs have
been designed for the following subestuaries of Chesapeake Bay: Chester River,
Maryland; Patuxent River, Maryland; Poquoson River, Virginia; and Ware River,
Virginia. Water quality data collected under these programs include those
parameters required to characterize eutrophication, carbon cycling and pollu-
tant and gross sediment transport. While data on sediment-bound contaminants,
such as pesticides and trace metals, and detailed ecosystem dynamics were not
included in the program design, the programs have been structured so that
these types of sampling can be added in the future. In addition to providing
data for model testing, the recommended sampling programs have also been
designed to provide the States of Maryland and Virginia with the information
they require for water quality planning and management.
The physical, chemical and biological parameters which have to be
sampled to characterize the dominant water processes occurring in each
subestuary are identified. In addition, the EXPLORE model is used to illus-
trate the relationship between the sampled parameters and model input and
calibration/evaluation data requirements. EXPLORE was recommended as the basis
for sampling program design and parameter selection by the Environmental
Protection Agency's Environmental Research Lab in Athens, Georgia.
A generalized sampling program consisting of two components was formulated.
The first component involves conducting two intensive sampling programs each
year to identify short-tami (i.e., tidal cycle) variations in water quality
conditions. The second component consists of monthly trend monitoring to
identify long-term (i.e., seasonal) variations. Both types of data are
required for model testing and evaluation. Also included in the generalized
program are sampling procedures, site selection criteria, parameter sampling
frequencies, ana sample preservation and handling guidelines.
Specific sampling programs were developed by appling the generalized
program to each subestuary. Sampling transect and station locations and
recommended sampling depths are identified along with the number of samples
-------
which have to be taken of each parameter. The number of samples are given
on a per intensive sampling program basis or per year of trend monitoring
basis for use in estimating the costs of subestuary sampling.
-------
t »
-------
CONTENTS
SUMMARY -Hi
LIST OF FIGURES ix
LIST OF TABLES xi
INTRODUCTION 1
STATEMENT OF THE PROBLEM 1
PURPOSE Z
SCOPE 2
PARAMETER SAMPLING AND USE IN MODELING 5
PHYSICAL DATA REQUIREMENTS 6
Hydrodynanri c Data 7
Meteorological Data 9
Salinity 9
Water Temperature. . 10
pH Data 10
Suspended Sediment Load and Bed Sediment Characteristics. . 10
CHEMICAL DATA REQUIREMENTS 10
Carbonaceous Biochemical Oxygen Demand Data 11
Carbon Data 1Z
Nitrogen- Data 12.
Phorphorus Da.ta 13
Dissolved Oxygen Data 13
BIOLOGICAL DATA REQUIREMENTS 13
Phytoplankton Data H
-------
* *
-------
Bacteria Data 14
ESTUARY MODELING 15
Carbonaceous Biochemical Oxygen Demand. ....... 16
Benthic Biochemical Oxygen Demand 16
Total Organic Carbon 17
Nitrogen 17
Phosphorus 18
Dissolved Oxygen 20
Carbon Cycle "... 22
GENERALIZED SAMPLING PROGRAM 25
SAMPLING METHODOLOGY 25
Intensive Sampling Program 25
Trend Monitoring 25
SAMPLING PROCEDURES 27
Sita Selection 27
Site Location in the- Field 28
Sampling Frequencies 29
Bathymetric Surveys 29
Data Collection 30
Sampling Preservation and Handling 31
SPECIFIC SAMPLING PROGRAMS 35
CHESTER RIVER, MARYLAND 36
Site Description 36
Sampling Site Locations 38
PATUXENT RIVER, MARYLAND 42
-------
Site Description ..... 42
Sampling Site Locations 42
POQUOSON RIVER, VIRGINIA 47
Site Description 47
Sampling Site Locations . 47
WARE RIVER, VIRGINIA 49
Site Description 49
Sampling Site Locations 53
REFERENCES 57
-------
-------
FIGURES
1 Cheasapeake Bay -. . 3
2 Example Field Log 33
3 Chester River 37
4 Recommended Sampling Locations - Chester River 40
5 Patuxent River 43
6 Recommended Sampling Locations - Patuxent River 45
7 Poquoson River 48
8 Recommended Sampling Locations - Poquoson River 50
9 Ware River. 52
10 Recommended Sampling Locations - Ware River 54
-------
TABLES
1 Samoling Frequencies for Physical Parameters ....... 29
2 Sampling Frequencies for Chemical Parameters . . 30
3 Sampling Frequencies for Biological Parameters 31
4 Recommended Preservation Methods 32
5 Chester River Sampling Locations . . 39
6 Total Number of Samples Required on the Chester River Per
Intensive Program or Per Year of Trend Monitoring 41
7 Patuxent River Sampling Locations 44
8 Total Number of Samples Required on the Patuxent River Per
Intensive Program or Per Year of Trend Monitoring 46
9 Poquoson River Sampling Locations 49
10 Total Number of Samples Required on the Poquoson River Per
Intensive Program or Per Year of Trend Monitoring 51
11 Ware River Sampling. Locations 53
12 Total Number of Samples Required on the Ware River Per
Intensive Program of Per Year of Trend Monitoring 55
-------
INTRODUCTION
Numerous mathematical computer models have been developed to simulate the
contribution of diffuse or nonpoint source pollutants to waterbodies as well as
the transformations which occur to both point and nonpoint source pollutants
in the waterbodies themselves. Nonpoint source pollution models generally
simulate the processes of rainfall-runoff, soil erosion and pollutant transport.
Water quality models simulate the dominant physical, chemical, and biological
processes which influence water quality conditions in various types of water-
bodies, including rivers, lakes, estuaries and oceans. The capabilities of
these models range from simple, steady-state models which simulate the long-
term, (i.e. , annual or seasonal) variation of selected parameters to complex,
dynamic (i.e., time-varying) models which simulate parameter changes in
several spatial dimensions. The time scale of the water quality condition
or process being evaluated usually determines which type of model is applied.
Mathematical computer models also provide both diagnostic and predictive
capabilities—diagnostic in that the proper application of a model requires
the user to identify and understand the specific factors affecting water
quality changes and predictive in that the model can be used to simulate the
effects of proposed changes in a watershed or waterbody. These capabilities
render mathematical models extremely powerful tools in water quality planning
and management.
STATEMENT OF THE PROBLEM
To date there has been no coordinated effort to collect the water quality
data required: to adequately apply, test and compare various types of nonpoint
and water quality models in order to assess their relative- effectiveness as
management tools. This condition has prompted the Environmental Protection
Agency (EPA) to implement a set of comprehensive water quality sampling pro-
grams in several basins and subestuaries. of the Chesapeake Bay. The intent
of these sampling programs is to obtain complete and valid data sets which can
be used to calibrate and evaluate existing state-of-the-art nonpoint source
and river and estuary water quality models. The data collected under these
-------
programs are to be sufficiently detailed for use with the more sophisticated
models in each class as well as the simplified models.
PURPOSE
The purpose of the model testing is twofold. First, the model testing
will help identify the most economical models and combination of models which
provide the information required to make water quality planning and management
decisions. Secondly, the testing of various models against a comprehensive
set of data will make it possible to determine: whether existing models ade-
quately simulate the processes which control water quality changes. This will
aid in the identification of research which must be completed in order to
further the state-of-the-art of water quality and nonpoint source modeling.
BatteHe-Northwest and Hydrocomp, Inc. have been given the responsibility
to design sampling programs for the following tributaries of the Chesapeake
Bay: Patuxent River, Maryland; Chester River, Maryland; Ware River, Virginia;
and Poquoson River, Virginia (see Figure 1). Battelle-Northwest is respon-
sible for developing the sampling programs for the subestuaries formed by each
of the above tributaries. For the purpose of this program, a subestuary is
defined as the body of water bounded by the head-of-tide (i.e., the location
where the tributary is no longer tidally influenced) to the intersection of
the subestuary with Chesapeake Bay proper. Hydrocomp, Inc. is developing the
river water quality sampling program for locations upstream' of the head-of-
tide in addition to the nonpoint source sampling programs. The States of
Maryland and Virginia will be responsible for implementing the programs in
their respective basins. Care has been taken to ensure that the water quality
parameters collected irr the freshwater systems are compatible with those
collected in- the- subestuaries.
SCOPE
Data collection under the- programs being designed by Eattelle-Northwest
includes those parameters which are required to characterize eutrophication
and pollutant transport in the subestuaries. These sampling programs should
-------
-------
BALTIMORE*^
':CHESTER R
WASHINGTON-
./&/PATIJXENTR.
POTOMAC R.^r
RAPPOHANNOCK Rh
POQUOSON
NEWPORT NEWS
FIGURE 1. Chesapeake Bay
-------
generate the data required to calibrate and apply one of the more sophisti-
cated estuary water quality models currently available for use. The EPA's
Environmental Research Lab in Athens, Georgia recommended using the EXPLORE
model [Baca et a!., 1973] as the basis for sampling program design and para-
meter selection. In addition, the sampling programs are to generate data on
the carbon cycle and gross sediment transport. While data on sediment-bound
contaminants, such as pesticides and trace metals, and. on detailed ecosystems
dynamics are not to be collected initially, the programs are structured so
that this type of sampling can be added in the future.
The next three sections of this report comprise Battelle's design of the
sampling programs for eutrophication and pollutant transport in the subestuaries.
The first section discusses the parameters to be sampled and their relation-
ship to estuarine water quality modeling with EXPLORE. The second section
presents the components of a generalized estuary sampling program including
sampling methodologies and procedures, site selection, sampling frequencies
and sample handling. This generalized program is applied to each of the four
subestuaries in the final section.
Due to time and funding limitations, it was difficult to incorporate many
of the specific characteristics of each subestuary into the program design.
In addition, the major consideration in designing these programs was to obtain
the data required to meet the model testing objectives. Sampling costs have
not been considered in detail since this task was left up to the individual
states. Therefore, it is possible that the cost of the recommended programs
may exceed the funds allocated to each state for sampling. As a result, the
States of Maryland and Virginia may have to refine and possibly condense the
recommended programs based on: 1) their knowledge of factors which control
water quality conditions in each subestuary, 2) past,, present or planned data
collection efforts which could possibly supplement elements of the proposed
programs, 3) available sampling and water quality analysis resources, and
4) the State's priorities with respect to water quality management in each
subestuary.
-------
PARAMETER SAMPLING AND USE IN MODELING
In the application, calibration and evaluation of water quality models
three categories of data are required: interpretive data, model input data,
and comparative data. Interpretive data include various information sources
such as maps, charts, past water quality studies and modeling efforts, and
basin plans which are not directly used as input to the model but are necessary
background information required to familiarize the model user with the water
body. This report will not address this category of data since it is not a
normal element of a sampling program.
Input data are the data required to drive the model and have it simulate a
specific sequence of events or water quality condition. In estuarine modeling,
these data generally consist of point and nonpoint inflow quantities and quali-
ties, tidal fluctuations and meteorological conditions for the period of simula-
tion. Data specifying the water quality and hydrodynamic conditions at the
beginning of the simulation period (i.e., initial and boundary conditions) are
also required as input.
The final category, comparative data, are required in the model calibra-
tion and evaluation processes. It is difficult to specify the values of cer-
tain system parameters such as chemical reaction rates and biological uptake
rates a priori. Although the order of magnitude of these parameters, often
called calibration coefficients, is usually known, the means for more precise
independent specification for many of the coefficients are presently not avail-
able. Consequently, model predictions are compared to field data and the
calibration coefficients are varied until simulated values match measured
values to an acceptable level.
Following this coefficient adjustment process, models are- then evaluated.
Water quality model evaluation or verification involves additional model test-
ing to determine how closely the model predictions can match different sets of
measured conditions without a change in coefficient values. It is best if the
measured conditions used for model evaluation represent substantially different
water Quality and flow conditions than those used for calibration. This pro-
vides added confidence in the calibration coefficients and in the model's
-------
ability to simulate widely varying conditions. It is often recommended that
at least three sets of data be used in the model calibration/evaluation phase,
one for gross calibration, the second for refined calibration and the third
for verification.
Existing estuary water quality models simulate the concentration or distri-
bution of a constituent, property or parameter in an estuary. Accordingly, the
processes represented by the model are generally classified as either transport
processes or reaction processes. Transport processes are basically hydrodyna-
mic and include advection, turbulent diffusion, and, when spatial averaging is
involved, dispersion. Reaction processes encompass the sources and sinks to
which the parameter is subjected and may be physical, chemical or biological,
(e.g., sedimentation and flocculation of organics, uptake of oxygen in bio-
chemical degradation of wastes, algae loss by zooplankton grazing, etc.). The
relative importance of transport processes and reaction processes obviously
varies from estuary to estuary, and depends upon the parameter modeled as well
as the required spatial and temporal refinement. The following describes the
physical, chemical and biological parameters which will be sampled for estuary
model testing. The collection of these data for the purpose of model input,
calibration/evaluation or both will also be discussed.
PHYSICAL DATA REQUIREMENTS
Physical data which will be collected during the sampling programs for
input to and for use in the calibration of water quality models include the
following:
* hydrodynamic data (estuary geometry, water surface elevation variations,
currant velocity and direction, freshwater discharge, and point source
discharge)
• meteorological data (solar radiation" and/or air temperature, precipita-
tion, wind speed and direction-, and wave height, period and direction)
* salinity data
-------
-------
» water temperature data
• pH data
•• suspended sediment load and bed sediment characteristics
Hydrodynamic Data
Estuary Geometry
Estuary geometry data in the form- of bathymetric data are required input
to most estuary water quality models for use in calculating the flow field.
These data consist of either cross-section profiles for calculation of cross-
sectional area at designated transects along the estuary or 'discrete point
depths such as would be developed by superimposing a "grid" or "cell" pattern
over the estuary. These data should be referenced to some datum such as mean
sea level or mean low water. Bathymetric data are usually available in the-form
of U.S. Coastal and Geodetic Navigation Charts and Boat Sheets (National Ocean
Survey), U.S. Army Corps of Engineers Hydrographic Surveys, and U.S. Geological
Survey surveys. The bathymetric data obtained from these sources should be
supplemented and verified in the field, at least to the extent of conducting
longitudinal profiles and cross-section profiles at designated transects.
Water Surface Elevation
Water surface elevation changes due to tidal and nontidal effects are
also needed for the calculation of the flow field. Depending on the water
quality node! employed, the time varying water surface elevations can be input
at each computation point or only at the open boundaries. In the latter case,
water surface elevations at computation points other than the open boundaries
ara computed internally within the computer program. Even with the latter
case models,, it is necessary to have the water surface elevation data at other
points for calibration and verification purposes. Water surface elevation data
for the Chesapeake Bay area are available through the U.S. Department of Commerce,
National Ocean Survey '''High and Low Water Predictions, East Coast of North and
South America" [Tide Tables, 1978]. These annually published tide tables do
not include day to day variations in sea level due to changes in winds and
barometric conditions, nor do they account for unusual changes in freshwater
7
-------
discharge conditions, all of which will cause the tide to be higher or lower
than predicted. Time of high and low water can also differ from the predicted
times due to these factors. To correlate the predicted tides to the actual
tides, tide gages should be established at strategic locations in the sub-
estuaries. The National Ocean Survey should be contacted for the locations of
existing gages and for help in selecting the locations and number of new gages.
Current Velocity and Direction
Current velocity and direction are not normally required as input to the
hydrodynamic portion of a water quality model. They are calculated based on
the time varying water surface elevations and estuary geometry. However,
current velocity and direction data are required for the calibration and veri-
fication of the hydrodynamic submodels. Lateral and longitudinal time
varying current velocity data should be collected at various depths in the
flow field, especially in the lower saline reaches of the estuary regardless of
whether the models are one- or multi-dimensional. Current velocity data are
required for two reasons: 1) to evaluate the dimensional effectiveness of a
particular model, i.e., whether a model of less than three dimensions can
adequately simulate the hydrodynamics of a three-dimensional system, and 2) to
obtain lateral, vertical or time-averaged velocity fields for calibration of
the models. Current velocity data are especially important in the lower
reaches of the estuary because of the possibility of multi-layered circulation
patterns from vertical variations in water density. Predicted tidal current
tables for various locations within the Chesapeake Bay area are available
through the U.S. Department of Commerce, National Ocean Survey [Tidal Current
Tables, 1978], but these data are not published in sufficient detail to be used
for calibration- purposes. The predicted values may also vary from the actual
values for the same reasons the actual tidal elevations differ from the pre-
dicted tidal elevations.
Freshwater and Point Source Discharges
Freshwater and point source pollution discharges into estuaries are used
as input parameters in water quality models. Daily discharges of the major
river and stream tributaries to each subestuary should be obtained from
-------
the U.S. Geological Survey and cooperating Stata and other Federal agencies.
Flow data on point sources of pollution, such as municipal sewage treatment
plants and power plants, should be obtained from the dischargers if possible. If
such data are not available, estimates will have to be made from other sources
of information (e.g., discharge permits). Data on freshwater discharges from
rainfall-runoff can be obtained from the Hydrocomp, Inc. program discussed-
previously.
Meteorological Data
Meteorological data may or may not be used as input to an estuary water
quality model. Models which calculate water temperature instead of requiring
it as an input need the following input data: daily solar radiation, average
air temperature, average dewpoint, average relative humidity and wind speed.
All of these parameters are normally available through the National Weather Ser-
vice. Precipitation data are often required as input to water quality models.
These data can also be obtained from the National Weather Service. A few models
calculate wave height, period and direction from wind speed and direction data.
Wave data for use in calibrating these models may be obtained from direct observa-
tions during sampling or from wave gaging stations.-
Salinity
Salinity (or specific conductance) data should be collected for water
density computations, the interpretation of seasonal and diurnal characteris-
tics and the evaluation of vertical and longitudinal circulation patterns.
Salinity distributions in estuaries have a profound effect on suspended sedi-
ment loadings in the water column, and sediment deposition and erosion. The
sedimentation patterns and loadings in turn have an effect on nutrient and con-
taminant transport and loadings. Salinity distributions are accounted for in
many water quality models only in terms of its effect on other water quality
parameters. Although-, salinity data may not be used directly in many of these
models, the data should be collected for evaluating the processes responsible
for the distribution of other water quality parameters.
-------
Water Temperature
Water temperature is an input parameter on many water quality models,
requiring the collection of water temperature during the field sampling pro-
gram. Where water temperatures are computed from meteorological parameters,
water temperature data are required for model calibration and evaluation.
Vertical and lateral temperature data are needed for averaging purposes on one-
and two-dimensional models. The water temperature of all major freshwater
inflows and point sources of pollution should also be sampled.
pH Data
Data on the hydrogen ion activity or pH of a water body are required as
input to some models, but are generally used more as indicators of the chemi-
cal composition of a water sample. For example, pH data are required to cal-
culate the carbonate, bicarbonate and carbon dioxide concentrations from
alkalinity. In addition, pH has been found to be a useful indicator of the
potential for ammonia oxidation to nitrate. This use of pH data is discussed
in greater detail in the Estuary Modeling Section. pH data should be collected
at all subestuary sampling locations and in the major inflows to each sub-
estuary.
Suspended Sediment Load and Bed Sediment Characteristics
Suspended sediment loadings and bed sediment characteristics data should
be collected because of the important role sediments play in the availability
of nutrients and transport of contaminants in estuaries. Time varying sus-
pended sediment concentration data should be- obtained at aach sampling station
and depth in order to estimate gross suspended sediment transport and for
the analysis of the nutrient contents, described in the following section on
chemical data requirements. Bed. sediment samples should also be collected for
grain size determinations and analyses of chemical parameters discussed in the
following section.
CHEMICAL DATA REQUIREMENTS
The primary categories of chemical data which should be collected for
model input, calibration and evaluation include the following:
10
-------
-------
« carbonaceous biochemical oxygen demand data
* carbon data
• nitrogen data
• phosphorus data
• dissolved oxygen data
The specific parameters to be collected under each of the above categories of
chemical data are discussed below. Later in this section the relationship
between these parameters and estuarine water quality modeling with EXPLORE
is discussed. The sampling frequencies for each parameter are discussed
in the Generalized Sampling Program Section.
Carbonaceous Biochemical Oxygen Demand Data
Carbonaceous biochemical oxygen demand (CBOD) is a measure of the amount
of oxygen required by micro-organisms to decompose aerobically the carbonaceous
fraction of the organic matter present in a water sample. The "ultimate" CBOD
is the amount of oxygen required for complete oxidation. For convenience in
laboratory analysis, the 28-day CBOD of a sample can be used as an estimate
of ultimate CBOD. A chemical inhibitor should be used to insure, that the
— • . *\ j^
nitrogenous portion of the 800 is suppressed. -r/^''^, a- ' --a?
•joe, o : oy*rt .-» , A1 ~£,
Ultimate CBOD data are required for both model input and model calibration/
evaluation. All point sources of pollution (e.g., municipal sewage treatment
plants), tributary inflows and the watsrs entering a subestuary from Chesapeake
Bay should be sampled for model input. During the characterization of bed
sediments, the area! extent (i.e., mg per unit area) of benthic CBOD should
also be sampled at each sampling location for input purposes. This will
require the use of a. core sampler or other sampling device- which collects
bottom- samples fronr a known area. Nonpoint source inputs of CBOD do not
have to be sampled under this program since they will be accounted for in the
Hydrocomp, Inc. program.
For model calibration and evaluation, ultimate CBOD samples should be
taken in the water column and from the bottom at each subestuary sampling
location. These data will be used to calibrate and verify the model rate
coefficients for CBOD degradation, settling and scour.
11
-------
Carbon Data '- --'"•--•*
Two types of carbon data should be coll acted during the sampling programs:
data required for modeling total organic carbon (TOC) variations and data for
characterizing the carbon cycle in each subestuary. The data which should be
collected in order to model TOC include the TOC and chemical oxygen demand
(COD) of all inputs to the subestuary. These inputs include tributary and
waste source inflows and waters from the Chesapeake Bay. TOC and COD deter-
minations should also be made on water samples collected at each sampling
station. These data are required to calibrate and evaluate model reaction
rates and scour coefficients. The use of COD in modeling TOC is described
later in this section.
The characterization of the carbon cycle in each subestuary will involve
the sampling of total alkalinity in the inflows as well as in the water
column at each sampling location. Alkalinity can be used in conjunction with
pH and water temperature to determine the concentrations of individual inorgan-
ic carbon components (i.e., CO-, HCQZ, COZ) [Standard Methods, 1975]. In
addition to taking these data in the water column, bed sediments should be
collected at each location for the analysis of organic and inorganic carbon
content. This sampling can be conducted in conjunction with the bed sediment
characterization sampling. Changes in- TOC and the inorganic carbon components
throughout the duration of the sampling programs will provide a basis for
characterizing both the short-term (i.e., diurnal) and long-term (i.e., sea-
sonal) carbon cycling in each subestuary.
Nitrogen Data
In order to obtain data suitable for modeling nitrogen cycling in the-
subestuaries, the following nitrogen species' should be analyzed in all the-
inflows to each subestuary and at each sampling location:
• ammonia-nitrogen
» nitrate-nitrogen
» nitrite-nitrogen
* Kjeldahl-nitrogen
12
-------
These species should be measured in both filtered and unfiltered samples.
The inflow nitrogen data are required for model input. The subestuary sampling
station data are required for the calibration and evaluation of the various
rate constants used in modeling nitrogen cycling.
Phorphorus Data
Filtered and unfiltered water samples should be collected in each- inflow
to the suitestuaries and at each sampling station location. These samples
wilV'be analyzed for the following phosphorus components to obtain data for
model input and calibration/evaluation:
\ \ ^__
f? total phosphate r~-Jv+4<>*
C»_ total orthophosphate
^ total filtrable phosphate
\» filtrable orthophosphate
In addition to these data taken in the water column, bed sediment samples
should also be taken at each subestuary sampling location. These sediment
samples should be analyzed for orthophosphate and organic phosphate.
Dissolved Oxygen Data
The dissolved oxygen (DO) input data which have to be collected during
the sampling programs are- the DO concentrations in all inflows to each
subestuary. These inflows include freshwater tributaries, point source waste
loads and Chesapeake Bay waters entering at the mouth of the subestuary.
For model calibration, dissolved oxygen data should be collected at each
sampling station in the subestuary. These data are required to establish the
coefficients in the expressions for dissolved oxygen reaeration. Following
model calibration, dissolved oxygen data from each sampling station are also
required to evaluate- the model's capabilities to simulate dissolved oxygen
variations under a range of hydrologic and water quality conditions.
BIOLOGICAL DATA REQUIREMENTS
Estuary water quality changes are controlled not only by numerous physi-
cal and chemical factors but also by the growth and death of various biological
13
-------
species. This section will briefly discuss the biological parameters which
should be sampled in order to generate the required model input and calibration/
evaluation data. Two categories of biological parameters should be sampled,
those related to phytoplankton and those related to bacteria. The use of
these parameters in modeling water quality Variations with EXPLORE is dis-
cussed later in this section.
Phytoplankton Data
The input data required for modeling phytoplankton dynamics includes
various physical and biological parameters. The physical parameters include
depth, water temperature and water clarity. The sampling requirements for
the first two parameters were discussed previously in this section. Water
clarity data are obtained by using a Secchi Disk. The depth at which this
device is barely visible when submerged is called the Secchi Disk depth.
The inverse of this measurement, called the extinction coefficient, is used
as input to the model to represent the degree of water clarity. Secchi Disk
readings should be taken at each sampling location in the subestuary.
The biological parameters which are required as input include phytoplankton
and zooplankton concentrations (expressed as mg-C/i) for each tributary inflow
and across the mouth of the subestuary. Any of several collection methods,
including plankton and zooplankton towing nets or a Juday trap, can be used
to sample these locations. Sufficient quantities of water must be sampled
in order to obtain- enough organisms for dry and ash-free weight determinations
[Standard Methods, 1975]. Since biomass is composed of approximately 50*
carbon by weight, the dry and__ash-free weight data can de used to determine
the milligrams of carbon per unit volume for both plankton and zooplankton
species.
Similar data should also be- collected at each subestuary sampling loca-
tion. These data are- required for the calibration and evaluation of phyto-
plankton growth- and death rates and zooplankton grazing rates.
Bacteria Data
Currently, there are no reliable techniques for independently estimating
the values of the rates of ammonia oxidation to nitrite and nitrite oxidation
-------
to nitrate other than through the model calibration process. Bacteria counts,
however, are one biological parameter which can be used as an indicator in
determining if nitrification is occurring. Ammonia, nitrite and nitrate
concentrations can be continually changing in an aquatic system but if the
correct bacteria (nitrosomonas and nitrobacter) are not available, the nitrifi-
cation process and the resultant uptake of dissolved oxygen will not proceed.
Tuffey, Hunter, and. Matulewich [1974] have found that established communities
of nitrifying organisms only predominate under certain hydraulic conditions.
They noted, that shallow, surface-active areas (i.e., water bodies with rocky
bottoms that offer a high surface area to waters flowing nearby) and deep,
slow moving areas with long residence times show^ the highest degree of nitrifi-
cation. It is recommended that enumerations be made of water samples and
sediment samples taken at each sampling station to determine the concentration j -• -4
ESTUARY MODELING
In order to illustrate how the physical, chemical, and biological para-
meters which have been recommended for sampling are used in water quality
modeling, the EXPLORE model is used as an example. EXPLORE is composed of a
set of" submodels which describe the behavior and interactions of various water
quality parameters. The primary water quality submodels in EXPLORE include:
» carbanaceous biochemical oxygen demand (CBOD)
» benthic biochemical oxygen demand
* total organic carbon
» nitrogen
* phosphorus
*• dissolved oxygen-
Various physical, chemical,, and biological data are required as input to each
of these submodels as well as for use in selecting the appropriate values for
various reaction rates and model coefficients. Each of the above submodels
are briefly discussed below along with the water quality parameters which are
used in each submodel. Detailed descriptions of the mathematical formulations
15
-------
used in each submodel are included in the Appendix. Following the submodel
discussion a brief description of the carbon cycle is presented.
Carbonaceous Biochemical Oxygen Demand
In general, 800 has been observed to occur in two stages, the carbonaceous
stage in which heterotrophic organisms oxidize the carbonaceous fraction of the
organic material and the- nitrogenous stage in which autotrophic bacteria
oxidize the nitrogenous fraction of the organic matter. In EXPLORE, nitrogenous
BOD is accounted for in the nitrogen submodel. The behavior of carbonaceous
BOD is modeled as a first-order reaction in which the rate of change of CBOD,
sometimes called the deoxygenation rate, is proportional to the amount of
CBOD present.
In order to obtain an estimate of this rate independent of the calibration
procedure, CSOD time-series data should be obtained in the laboratory. CBOD
•
time-series consist of measuring the oxygen depletion due to CBOD decay in a
series of samples after different periods of time (i.e., 1, 3, 5, 10, 15, 20,
and 28 days). The rate of CBOD decay can then be calculated using one of
several methods [Hewitt and Hunter, 1975]. . Wh i 1 e _th ls_ pcacedur e_. doesjig t_
provide exact rates of decay for use in modelcalibration, it does establish
the probable lower limit for this model coefficient.
Benthic Biochemical Oxygen Demand
In EXPLORE,oxygerr demand of the benthic layer is considered to be composed
of two separate parts:; the demand of the organic sludge undergoing decomposi-
tion on the bottom of the water body and the photosynthetic and respiratory
contributions of attached algae and rooted aquatic plants. The modeling of
attached algae is discussed later in this section. Benthic oxygen demand is
modeled in much the same way as CSOD in EXPLORE.
^--Ai» ill Ltiy- cyuunmended-CBOD analysis procedures, time-series laboratory
analyses should also be performed on the benthic samples to obtain an estimate
"of the~benthic deoxygenation coefficient. If benthic BOD is a major sink of
dissolved oxygen in a subestuary, this sampling procedure should greatly
enhance the model calibration process since this component is often either
ignored or treated as an unknown.
t •
^
^a^i^ . .
\ ' i —
1 z-<*'~* o. • -**-£. I3w<, .
''1
.- e. c a ^ w^ -, ^
. -
-------
-------
Total Organic Carbon
Many attempts have been made to develop a general relationship between
total organic carbon (TOC) data and BOD or COD data. However, two problems
have been encountered in attempting to correlate the results of these data.
First, the ratios of TOC/BOD and TOC/COD change from one type of waste to
another. This variation is attributed to the fact that while the TOC test
measures the amount of organic carbon present, it yields no information about
the chemical form of the carbon or its ability to be degraded biologically.
Second, the ratios of TOC/BOD and TOC/COD change as the waste undergoes oxida-
tion so that ratios determined at one point in time would not be the same at a
later time.
Organic carbon can generally be classed into one of two groups: carbon
that is part of a compound which will be degraded biologically and carbon that
is refractory with respect to biological decomposition. The behavior of\TOC is
described in EXPLORE by a first-order equation in a manner similar to
equation used to model CBOD behavior. J
The input data required by the TOC submodel include total organic and
refractory organic carbon concentrations in all the inflows to the estuary.
Refractory organic carbon is that fraction of the TOC which is not biologically
degradable. Since this fraction is not directly, attainable from established
laboratory procedures, it is possible to calculate the refractory fraction
using COD data. The procedure by which the refractory fraction of TOC is
calculated from COD determinations is presented by Baca et al . [1973b].
Nitrogen
The relationships between the various forms of nitrogen in natural
waters is highly complex. Elemental nitrogen can be reduced (nitrogen- fixa-
tion) for use in- building cells by various bacteria and blue green algae.
Certain autotrophic bacteria will oxidize the ammonia (NFU) present in water
to nitrite (NO^-) and nitrite can be further oxidized to nitrate (NO^-). This
process, nitrification, represents the major portion of the nitrogenous com-
ponent of BOD as discussed earlier. In addition, the ammonia can be directly
incorporated into an organic form by either autotrophic or heterotrophic
17
-------
organisms. Ammonia can also escape from (or enter) the water in a gaseous
form. The nitrate produced is available as a nutrient to many organisms which
can convert it into an organic form. In some cases the nitrate will undergo
denitrification through biological action and be converted either to nitrite or
elemental nitrogen. The nitrogen in organic form can be present either as part
of a living organism or as part of a nonliving organic substance (such as a
soluble organic compound or part of a decaying organism). Nitrogen released
from the nonliving organic substances will in general return to the water in
the form of ammonia-. A significant portion of the nitrogen in the organic form
is refractory and will not readily return to solution or re-enter the nitrogen
cycle.
The EXPLORE model has two nitrogen submodels which represent the dominant
transformations of various nitrogen species in aquatic systems. One submodel
considers the affects of algae (Figure 5A in the Appendix) while the second
includes only ammonia-nitrate dynamics (Figure 58 in the Appendix). The
ammonia, nitrate, nitrite, and Kjeldahl nitrogen data obtained in the sampling
programs will be used to establish the reaction rates for these nitrogen
submodels.
In addition to the concentration of nitrifying bacteria, pH has been
shown to be a good indicator of the occurrence of nitrification. The oxida-
tion of ammonia to nitrite and. nitrate has an optimum pH range of 8.0 to
8.5 [Wild et al., 1971]. Nitrification can- occur, however between- 7.0 and
8.5. If the pH is much lower than 7.0, the process will probably be negli-
gible. Since pH will be measured at all locations, this parameter and the
concentration of nitrifying bacteria will be useful in model calibration.
Phosphorus
Phosphorus occurs in natural waters and; in wastewaters almost solely in
the form of various types of phosphate. These forms are commonly classified
into orthophosphates, condensed phosphates (pyro-, meta-, and polyphosphates)
and organically bound phosphates. As with other nutrients, phosphorus cycling
is mediated by chemical, biological and even geochemical processes. Inorganic
phosphorus is most commonly present in minerals or dissolved in water as
18
-------
orthophosphate (PO*). Particulate inorganic phosphates are in general only
slightly soluble in water and are not used by aquatic life. Complex or con-
densed phosphates are unstable in water and will hydrolyze slowly to ortho-
phosphate. Plants and other aquatic organisms can assimilate either dissolved
orthophosphate or soluble organic phosphate. The biota will excrete either
soluble organic phosphate or orthophosphate during both growth and death. If
the soluble organic phosphate is not reassimilated it will hydrolyze to ortho-
phosphate. Upon death of the aquatic organisms, a certain fraction of the
phosphorus in the cells is retained in a refractory state. Phosphorus also
interacts readily with sediment materials which can act as either a source or
sink for both organic and inorganic phosphorus compounds.
Three phosphorus submodels are included in EXPLORE. The first submodel
(Figure 6A in the Appendix) includes the first order reactions of soluble
phosphorus with algae and sediments. The second (Figure 6B in the Appendix)
is a first-order reaction model between soluble and sediment phosphorus and the
third submodel (Figure 6C in the Appendix) is a simple second-order decay model
for soluble phosphorus. The final submodel was incorporated into EXPLORE for
situations where data are not available to evaluate the first-order model
coefficients or where the available data indicate a second-order reaction is
occurring.
The input data required for the algae phosphorus model include the follow-
ing parameters:
•> soluble inorganic phosphorus (orthophosphate)
* sediment phosphorus
* organic phosphorus
The laboratory procedures used in analyzing phosphorus compounds have to be
carefully selected to insure the proper separation of total phosphorus into its
various components for use in modeling. Soluble inorganic phosphorus is the
"filtrable" fraction of orthosphosphate. Standard Methods recommends the use
of a 0.45-um membrane filter to separate filtrable from particulate forms.
Sediment phosphorus in EXPLORE refers to the inorganic phosphate bound on both
suspended and bed sediment materials (i.e., total particulate orthophosphate).
The organic phosphorus component consists of filtrable and particulate (i.e.,
total) organic phosphate.
19
-------
As in the case of the nitrogen reaction coefficients, the rate of ortho-
phosphate uptake by algae can not be readily determined independently of model
calibration. Phosphorus data are also required at each of the subestuary
sampling locations in order to calibrate the soluble inorganic phosphorus-
sediment phosphorus reaction rates and the organic phosphorus-soluble phos-
phorus reaction rate.
Dissolved Oxygen
Dissolved Oxygen concentrations in estuaries are influenced by a variety
interrelated factors. The most important factors include:
• water temperature
• bacterial oxidation of organic matter
* nitrification
• algal photosynthesis and respiration
• reaeration
The first factor, water temperature, is an input to the EXPLORE model which
controls the maximum amount of oxygen which can be contained in water under
equilibrium conditions (i.e., dissolved oxygen saturation). Water temperature
also influences the rates at which the next three factors deplete oxygen from
water. The submodels related to the bacterial oxidation of organic matter and
nitrification are coupled to the DO submodel and were discussed earlier in
this section.
The modeling of algae photosynthesis and respiration is inherently com-
plex because of the heterogeneity of algal populations and the myriad of
interacting environmental factors which influence algal dynamics. For example,
algae growth rates are complex functions of a large number of environmental
parameters such as temperature, pH, nutrient conconcentrations, irradiance,
and the presence of trace minerals. In addition, the effects of these para-
meters on the algae growth rates vary widely from one type of algae to another.
Algae are also part of a complicated food web which is difficult in itself
to model.
20
-------
In EXPLORE, algae have been separated into two categories, 1) planktonic
(unattached) algae, and 2) sessile (attached) algae. Sessile algae, attached
plants, and other micro- and macro-organisms that reside on the bottom of the
water course have been collectively classified as benthic organisms.
The diurnal changes in dissolved oxygen caused by algal photosynthesis
and respiration are simulated two different ways in EXPLORE. The first approach
involves modeling oxygen production from both attached algae and aquatic plants
and unattached algae-independent of algal growth. Two sinusoidally varying
terms are used to represent diurnal variations in photosynthesis and respira-
tion. The maximum rates of photosynthesis and respiration are calibrated by
comparing predicted and measured dissolved oxygen concentrations. This approach
has been found to be fairly reasonable as long as all the other factors affect-
ing 00 variations have been determined. In using this approach, diurnal varia-
tions in photosynthesis and respiration are assumed to be constant from day
to day.
In order to incorporate algal growth directly into the modeling of photo-
synthesis and respiration, a second approach can be used. This approach couples
phytoplankton (unattached algae) density to the time varying equation described
above. Phytoplankton growth using this approach is a function of nutrient
(nitrogen and phosphorus) availability, sunlight, temperature, zooplankton
grazing and death due to other causes such as toxic materials. Water depth
and temperature, the extinction coefficient and phytoplankton and zooplankton
concentrations in the inflows are input to EXPLORE and the concentration of
these species in the water column are- used to calibrate growth and death \
) _e-)Cln- order to cali-
brate the coefficients of" the reaeration equations, dissolved, oxygen data
at each sampling location in the subestuary are required. Following the
calibration of these coefficients, DO data are also used to evaluate the model's
capabilities to simulate dissolved oxygen variations under different hydrologic
and water quality conditions.
vrc*.«^T.> * !T \ 1
ce_ cctfk^
-------
Carbon Cycle
The cycling of organic and inorganic carbon compounds in natural waters
is similar to the cycling of other nutrients in that inorganic carbon is incor-
porated into organic carbon which is subsequently mineralized into inorganic
carbon. The majority of the carbon in aquatic systems occurs as equilibrium
products of carbonic acid. Atmospheric carbon dioxide (002) dissolves in
water and forms carbonic acid (H-CO-j) which dissociates into various
inorganic species (i.e., C02, HCO," and CO-"). The complex dissociation
reactions of carbonic acid are well understood and tend to be highly dependent
upon pH.
Synthesis of organic matter by photosynthesis of algae and submersed
aquatic macrophytes requires an abundant and readily available source of
inorganic carbon. Abundant physiological evidence indicates that free C02
is most readily utilized by nearly all algae and larger aquatic plants. Many
algae and aquatic vascular plants are capable of assimilating bicarbonate
(HCO^") ions when free COo is in very low supply. There is no clear evidence
that algae or higher aquatic plants assimilate C0o~ directly as a carbon
source.
Microbial decomposition of organic materials replenishes the inorganic
carbon pool. The dominant sources of organic carbon, other than the dissolved
and organic matter which enters estuaries from various inflows, are cellular
constituents of organisms at death and secretions from actively growing algae
and large aquatic plants. The rate of" decomposition of these substances is
a function of their concentration and of the activity level of the bacteria.
The soluble organic matter synthesized by planktonic production and
released by secretion and. death decompose fairly rapidly at the site of
generation and release. Particulate organic materialsr however,, undergo
much slower degradation. As a result, particulate organic carbon can reach
the sediment interface prior to complete degradation. The amount of this
material which reaches the bottom is a function of the depth of the estuary
and the time required for decomposition. Once particulate organic materials
are incorporated into the bed sediments they are transformed into inorganic
materials through bacterial activity.
-------
The characterization of the carbon cycle in each subestuary requires
the sampling of various inorganic species of carbon and organic carbon in
the inflows to each subestuary and within each subestuary. These data will
provide a basis for characterizing both the short-term (i.e., diurnal) and
long-term (i.e., seasonal) fluctuations in carbon.
23
-------
GENERALIZED SAMPLING PROGRAM
SAMPLING METHODOLOGY
The overall objective of the sampling program is to collect data on the
physical, chemical, and biological characteristics of the four aforementioned
subestuaries for the calibration and verification of various water quality
models. Many of these models will be intertidal, that is,, they will be capable
of simulating real time variations in water quality parameters. Thus, for
calibration purposes, both seasonal and tidal cycle variations of the selected
parameters are of importance. In developing the generalized program it was
deemed necessary to conduct a minimum of two intensive sampling programs
annually, accompanied by less intensive trend monitoring on a monthly basis.
Intensive Sampling Programs
The intensive sampling programs are designed to provide longitudinal,
lateral and vertical variations of the previously discussed parameters over two
consecutive tidal cycles of approximately 25 hours duration. Sampling over two
tidal cycles should be conducted because of the mixed-semidiurnal nature of the
tides in Chesapeake Bay, i.e., two high waters and two low waters occur during
a tidal day where an inequality exists between the two high and two low waters.
A minimum of two intensive sampling programs should be conducted in each
subestuary per year to represent high and low freshwater discharge conditions
in Chesapeake Bay. High discharge conditions generally occur during the months
of February through April and low discharge conditions occur from June to
October. If possible, the high discharge sampling program should coincide
and preferably just follow a major storm event so that the nonpoint source
sampling programs carr be integrated, with- the subestuary program.
As will be-discussed subsequently the intensive sampling programs will be
comprised of a number of sampling transects along the estuarial reaches of the
subestuaries. The transects will consist of one to three stations in the
lateral direction and from one to three sampling depths. Generally, in the
broader lower reaches of the subestuaries three stations will be required to
represent conditions occurring in the main flow channel and adjacent channel
S> c
«£^JK C-S.-0
-------
-------
margins and subtidal flats. Multiple depth sampling to represent near-bottom,
mid-depth and near surface conditions will also be necessary in much of the
lower saline portions of the subestuaries because of the possibility of the
layered flow characteristics. In the upper freshwater tidal portions of the
subestuaries the lateral and vertical dimensions become less important because
of the general lack of extensive channel margins and subtidal flats, and due
to the lack of layered flow characteristics. Thus, in the freshwater tidal
portions of the subestuaries one sampling location in the cross-section and
one sampling depth is often sufficient to represent conditions occurring along
the transect.
Ideally all transects within a subestuary should be sampled simultaneously
during each intensive sampling program to represent conditions occurring
during a set time frame of two tidal cycles. However, this is often not
possible because the large number of transects and stations require too much
equipment and funding limitations are generally present. As a substitute it
is often appropriate to use a limited number of vessels to sample individual
transects during succeeding tidal cycles, provided that conditions such as
freshwater discharge, water surface elevation variations, and meteorological
conditions, vary only within allowable limits during the entire sampling
period.
Trend Monitoring
Trend monitoring of selected sampling parameters should be conducted on a
monthly basis along the- longitudinal axis of the subestuaries. Trend monitor-
ing should be- conducted- on either high or low water slack following the same
tidal cycle during the duration of the trend monitoring program for each sub-
estuary. As long- as: the general hydrodynamic and meteorological conditions
remain relatively constant, the data obtained from each longitudinal station
will be comparable irr terms of tidal cycle variations.
Trend monitoring data will be used to track changes in the sampling
parameters between the intensive sampling programs and also to correlate
changes between adjacent transects. Except in the lower deep portions of the
subestuaries one sampling depth in mid-channel should be sufficient for
correlation- of trend monitoring data. However, where two or three layered
25
-------
-------
flow regimes exist, two or three depths will have to be sampled due to the
large vertical and longitudinal variations that will be encountered throughout
the year.
SAMPLING PROCEDURES
The intensive sampling and trend monitoring programs as discussed above
are designed to provide comprehensive tidal cycle conditions on a seasonal
basis and slack water conditions on a monthly basis, respectively. Sampling
procedures for the intensive sampling programs will consist of the selection
of specific transects, stations, and depths; locating the sites in the field;
selection of sampling frequencies for collection of water quality parameters;
conducting a bathymetric survey; collection of water quality data; and sample
preservation and handling. The following is a brief discussion of these pro-
cedures.
Site Selection
A number of considerations should be evaluated in selecting the transects
to be sampled during the intensive sampling program. Foremost is that the
number of transects be great enough and the transects be so located as to give
complete coverage of the estuary. Modeling considerations are also important
in the selection of the number and location of transects. The distance between
transects should not be shorter than the distance between model computation
points. However, it is not necessary in most cases to sample at every computa-
tion point. Two mandatory areas to establish transects are the upper and
lower open boundaries, i.e. r the mouth of the estuary and the head-of-tide.
Transects should also be established at the junctions of major tributaries
even if they are not to be modeled. Another important area to establish
transects is the null zone (turbidity maximum), if such an- area exists,
because this is the transition area between the tidal salt water and tidal
freshwater portion of the estuary where the most radical physical and chemical
changes occur. Other considerations for the selection of transect locations
are: 1) the transect should be located in a fairly straight length of chan-
nel, 2) locate transects in known areas of large fluctuations in water quality
parameters, and 3) areas of high contaminant inout such as industrial and
municipal discharges. A "rule of thumb" for determining distance between
27
-------
-------
transects is that the distance be equal to the length of one tidal excursion.
However, where the estuary is very long this method of selecting distances is
often not possible.
Station locations along a transect and selection of sampling depths are
important in order to distinguish lateral and vertical changes in water quality
parameters- Along the main stem in the lower portion of the estuary multiple
cross-channel stations and depths should be established because of the vast
differences in the cross-channel bathymetry and because of the possible pre-
sence of a multi-layered flow regime. Three cross-channel stations are nor-
mally sufficient to characterize the lateral variations' in water quality para-
meters; one station located in the deeper main flow channel and one each on the
adjacent channel margins or subtidal flats. Three sampling depths in the main
flow channel and one or two depths on the channel margins or subtidal flats
should be sufficient to characterize ttie vertical variations. In the tidal
freshwater portion of the estuary and secondary estuary stems, one station in
the lateral and one depth will normally be sufficient for characterizing physi-
cal, chemical and biological conditions.
Site Location in the Field
Locating sampling stations in the field and properly documenting the
locations is of prime importance so that the stations can be re-established
with a great deal of accuracy during subsequent sampling periods. Where
possible, the transects should be located between prominent shore features
such as a promontorie or where abrupt changes irr topography occur, or where
identifiable man-made structures exist. Locating stations near navigation aids
is very helpful in relocating these stations at later times. In the absence of
navigation- aids horizontal control methods will have to be used in locating
stations„ The best method; for horizontal control is the use of electronic
positioning equipment such as a range-range instrument. In lieu of this,
horizontal control in locating sampling stations can be established using
three-point sextant triangulation, Loran, or radar. The minimum accuracy for
station relocations should be about ±15 m.
23'
-------
Sampling Frequencies
No consideration need be given to sampling frequencies of individual
parameters during each of the trend monitoring sampling periods since these
samples will represent only one point in time, i.e., slack water conditions.
Sampling frequencies for each physical, chemical, and biological parameter are
given in Tables 1, 2 and 3. The frequencies presented in these tables were
selected on the basis of possible temporal variations in each parameter over
a tidal cycle, the length of time required to collect the sample and the total
number of samples.
TABLE 1. Sampling Frequencies for Physical Parameters
Description of Parameters
Bathymetric Survey
Tidal Stage
Current Velocity and
Direction
Freshwater and Point Source
Di scharge
Wave Height, Period, and
Direction
Salinity
Water Temperature
oH
Suspended Sediment Load.
Bed; Sediment Characteristics
Particle Size Distribution-
Sampling Frequencies
for Each Depth or Station
Intensive
Sampling
Annually
Continuous
30 Minutes
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Trend
Monitoring
___
—
—
Daily Average
Monthly
Monthly
Monthly
Monthly
Monthly
Onca
Monthly
Sathymetric Surveys
Bathymetric surveys using a high frequency, continuous recording fatho-
meter should be conducted at each transect at least on an annual basis. This
depth data will be used for comparison with the hydroqraphic charts of the area
Z9
-------
TABLE 2. Sampling Frequencies for Chemical Parameters
Sampling Frequencies for
Each Depth or Station
Description of Parameters
Ultimate Carbonaceous BOD
Benthic Carbonaceous BOD
Total Organic Carbon
Chemical Oxygen Demand
Alkalinity
Bed Sediment Carbon Content
Ammonia-Nitrogen^ '
Nitrite-Nitrogen
Nitrate-Nitrogen
Kjeldahl-Nitrogen'
Total Phosphate
Total Orthophosphate
Total Filtrable Phosphate
Filtrable Orthophosphate
Bed Sediment Phosphate Content
Dissolved Oxygen
0)
(1)
,0)
Intensive
Sampling
Hourly
Once
Hourly
Hourly
Hourly
Once
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Once
Hourly
Trend
Monitoring
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
NOTE:
1. Fi 1tered and unfi 1 tered
and for computations of cross-sectional areas. Horizontal control in these
bathymetrfc surveys is an- important aspect because of their use in determining
the cross-sectional areas. These depth- soundings should be referenced to some
datum such as mean low water or mean sea level, requiring- the strategic place-
ment of tide gages within the estuaries.
Data Collection
Detailed data collection procedures need not be developed herein. All
stations on a particular transact should be manned during the same tidal cycle,
30
-------
TABLE 3. Sampling Frequencies for Biological Parameters
Sampling Frequencies for
Each Depth or Station
IntensiveTrend
Description of Parameters Sampling Monitoring
Phytoplankton Concentration
Zooplankton Concentration
Secchi Disk Depth
Nitrifying Bacteria Enumeration
Twi ce
Twi ce
Hourly
Once
Monthly
Monthly
Monthly
Monthly
requiring the use of at least three fully equipped vessels during most of the
intensive sampling program. Multiple casts of water samples at the designated
depths should be taken simultaneously and multiple depth current sensors should
be used if possible. If it is not possible to collect data at all depths
simultaneously, consecutive samples should be collected beginning at the near
surface or near bottom depth and progressively working downward or upward.
Data from each depth should be collected within a set time frame with no
deviations.
Sample Preservation and Handling
Instant analyses of the water quality parameters previously discussed is a
practical impossibility, therefore preservation methods should be employed
to retard any reactions which tend to alter the sample. Preservation methods
are generally limited, to pH control, chemical addition, refrigeration, and
freezing. The U.S. Environmental Protection Agency, Environmental Monitoring
Support Laboratory (EMSL) [Methods for Chemical Analysis of Water and Waste,
1974] has compiled, a list of recommendations for preservation of samples
according to the measurement analysis to be performed. Table 4 is a summary of
applicable water quality parameters.
Proper sample handling includes correct sample container labeling and
documentation. Each sample container should have a designation that uniquely
distinguishes it from all other samples. Water-proof labels and indelible ink
should be used in marking sample containers. Field sampling logs, similar to
the example field log shown in Figure 2 should be prepared as part of the
documentation of sample collection.
31
-------
TABLE 4. Recommended Preservation Methods
(1)
Volume
Measurement
Required
(ml)
1000
50
300
300
400
500
500
100
50
25
25
50
50
50
50
100
100
1000
Container
P, G<2>
P, G
G only
G only
P, G
P, G
P, G
P, G
P, G
P, G
P, G
P, G
P, G
P, G
Pi*
Ij
P, G
P, G
P, G
Preservative
Cool, 4°c
H2S04 to pH <2
Determine Onsite
Fix on Site
Cool, 4°C
Cool, 4°C
H2S04 to pH <2
Cool, 4°C
H2S04 to pH <2
Cool, 4°C
H2S04 to pH <2
Cool, 4°C
Cool, 4°C
H2S04 to pH <2
Cool, 4°C
Determine Onsite
Filter on Site
Cool, 4°C
Cool, 4°C
H2S04 to pH <2
Cool, 4°C
Filter on Site
Cool, 4°C
Cool, 4°C
Cool, 4°C
Determine Onsite
Holding
Time(S)
o hr
7 days
No Holding
4-8 hr
24 hr
7 days
24 hr
24 hr
5 hr^
24 hr
24 hr
7 days
24 hr
7 days
1 A '«*•*'
£,*» fir
No Holding
300
COD
Dissolved Oxygen
Probe-
Winkler
Ni troqen
Ammonia
Kjeldahl
Total
Nitrate
Nitrite
Organic Carbon
PH
Phosphorous
Ortnophospnate,
Dissolved
Hydrolyzable
Total
Total, Dissolved
Suspended Solids
Specific Conductance
Temperature-
NOTES:
1. Taken- from- ENSL (1974).
2. Plastic or glass
3. If samples can not be returned to the laboratory in less than 5 hr
and holding time exceeds this limit, the final reported data should
indicate the holding time.
4. If the sample is stabilized by cooling, it should be warmed to 25°C
for reading, or temperature correction made and results reported at
25°C.
5. It has been- shewn that samples orooerly oreserved may be held for
extended periods Deyond the recommended holding time.
32
-------
.
3*
"? i
a -j
w >•
£ — •
c <
3 2
Ml
41
O.
41
1 1
oi Sa
5 2
« .. i i -
jui £ ^1 f| X .- z
>• ^ il 3 5 £ S 2 X
— v? vi § v< vi ac a |! a
i , | * a 1 S ! - J I i
I i
•• ^ -J 'oj
LU i. O CL
a! a; * S
5 trt O i—
S3 22 w
* i i 1 s
< 2 O .^
I ,
2 S
si
i/i
isSSllS s I = 1 S IIHS"11'10'"31 1
-
ae
a*
01
§
01
0
— J
2
ul
u4
OC
UJ
§
t-J
3C
U4
3.
i " i
—
1 -
I a
I
u u —
I § 2 5
0.
3
DATA (tSIIHAlfU)
II:
C.
1
*^ ^
S ^
"
LU
<^ O
i u!
g
1
i.
S S z- .... "a
^i*» ui«5
S'_j«^i 3» — SS a outsat*— 30 i — *•*! £
Sc»~
-------
SPECIFIC SAMPLING PROGRAMS
• The sampling programs for the four subestuaries of Chesapeake Bay described
in this section have been designed without specific regard to available fund-
ing limitations. However, certain constraints were adhered to in the design
which ultimately would decrease the costs. The principal constraint is that
of time. The sampling programs have been developed on the basis of a maximum
of 10 days field effort for each intensive sampling program, using three
sampling vessels. These sampling vessels must be large enough to operate
undermost meteorological conditions, but be of shallow enough draft to navi-
gate in some of the.shallower areas. Four intensive sampling periods have
been recommended for each subestuary in order to obtain data from two high
and two low freshwater discharge conditions over a two year period. For
model calibration and evaluation purposes it may be possible to decrease the
number of intensive sampling periods to three, but this is not recommended
because of the possibility of required changes in the sampling program as
a result of experience gained during the first intensive sampling program.
Transect locations and number and locations of sampling stations and
depths were based on the geography of the subestuary, variations in cross-
channel bathymetry, and width of the estuary at a particular transect. Cross-
channel stations have been limited to a maximum of three on the basis that
three vessels may man one transect simultaneously. Generally, it will be
found that in the lower portions of the subestuaries three stations and three
depths have been established along each transect. In the upper saline regions
of the estuary where the channel is narrower and there is a lack of extensive
channel margins and subtidal flats, the number of stations and depths per
transect will generally decrease. In the tidal freshwater region and secon-
dary stems the number of stations and depths are generally limited to one each
because of the narrowness of the waterway and uniform shallow conditions.
The location of specific sampling stations for point sources of pollution
and freshwater discharges to each subestuary have not been identified. The
States of Maryland and Virginia will be responsible for selecting the waste
35
-------
-------
loads and tributaries to be sampled based on their local knowledge of the
relative impact these inflows have on subestuary water quality.
The specific sampling programs for each of the four subestuaries are
described below. Included in each description are the recommended locations
for sampling transects and stations as well as the number of depths to be
sampled at each station. Also included are tables summarizing the total num-
ber of samples which have to be taken of each physical, chemical, and biologi-
cal parameter during the intensive and trend monitoring programs. The num-
bers shown represent either samples per intensive program or samples per year
of trend monitoring. This information will help the States of Maryland and
Virginia in estimating the costs of subestuary sampling.
CHESTER RIVER, MARYLAND
Site Description
The Chester River, shown in Figure 3, is located in northeastern
Chesapeake Bay in the State of Maryland. The main course of the Chester River
is approximately 51 miles long with a watershed covering about 440 sq_mi. The
boundary between tidal and nontidal water lies at Cypress Point near Millington,
Maryland some 42 mi upstream from the mouth of the river. The mouth of Chester
River opens between Eastern Neck and the north end of Kent Island (Love Point).
In addition to this main connection to Chesapeake Bay, the Chester River also
communicates with Eastern Bay to the south through Kent Narrows.
The Chester River has a rather sinuous course and is quite broad in its
lower reaches. The main river channel is fairly deep, averaging well over
20 ft in the lower reaches. The main river channel consists of a series of
deep basin-liker depressions on- the order of 50 to 60 ft. in depth and are
separated by shallower areas 20 to 40 ft deep. The river channel is flanked
by extensive subtidal flats, extending to a depth of about 6 ft, and channel
margins extending from 6 to 20 ft in depth-. In the lower reaches of the
Chester River these subtidal flats and channel margins encompass from 40% to
50% of the water surface area. In the upper reaches they are much more exten-
sive and cover up to 80% of the surface water area.
36
-------
-------
i.
-------
The lower wide portion of the Chester River results from the junction of
three major tidal streams, the West Fork and East Fork of Langford Creek, and
the Corsica River. Langford Creek is a large watershed of approximately 40 sq
mi. About 2 mi above its mouth it divides into two stems, the West Fork and
East Fork. The Corsica River is a large tidal estuary which drains about
36.3 sq mi. Southeast Creek, another large- tributary of the Chester River
lies on the eastern bank between the headwaters area and the Corsica River.
Its drainage area is about 55 sq mi.
The volume of the Chester River has been calculated to be approximately
19.6 billion ft [Chester River Study, 1972] and the surface area is about
2
1.55 billion ft . The mean annual discharge of the Chester River into
Chesapeake Bay is estimated to be about 450 ft /sec. The highest flow rates
generally occur from February to April and the lowest from June to October.
The Chester River, like most of the major tributaries of Chesapeake Bay,
is classified as partly-mixed. In this type of estuary the vertical salinity
distribution is neither completely uniform nor segregated sharply into a
highly saline bottom water overlain by freshwater. The Chester River estua-
rine system generally has a multi-layered circulation pattern, tending toward
two-layered flow with upstream flow predominance of bottom waters and down-
stream flow predominance of surface waters.
Sampling Site Locations
Recommended, transect locations and number of sampling stations and depths
for the intensive sampling program are given in Table 5 and shown in Figure 4.
Eleven transects are located on the main stem of the Chester River, two
transects on Grays Inn Creekr four transects on Langford Crsek, two transects
on Corsica River and one transect is located on Southeast Creek. There is a
total of 31 sampling stations and 68 discrete sampling points. The trend
monitoring station locations should consist of collecting one slack water
sample in mid-channel and at mid-depth at aach of the- 20 transects. Table 6
gives the total number of samples required for each intensive program and
year of trend monitoring on the- Chester River.
38
-------
TABLE 5. Chester River Sampling Locations
Location of Transect
Main Stem Chester River
Love Pt. to Eastern Nk.
Long Pt. to Narrows Pt.
Belts Bar Pt. to
Piney Pt.
Nichols Pt. to
Hoi ton Pt.
Quaker Neck Ldg.
Wi liner Pt.
Chestertown
Travilla Whf.
C romp ton
Millington
Grays Inn Creek
Little Gum Pt. to
Grays Inn Pt.
Browns Pt.
Lanqford Creek
Grays Inn Pt. to
Nichols Pt.
Island Pt.
Hawbush Pt.
Longmarsh Pt.
Corsica River
Town Pt.
Fort Pt.
Southeast Creek
Deep Pt.
TOTAL
Approximate
River Mile
(N.M.)
0.0
4.0
7.2
11.0
15.0
18.5
22.4
26.9
30.0
T3 n
JO . U
36.5
9.5
0.0
1.4
10.5
0.0
2.0
2.3
3.4
0.5
1.7
18.8
0.0
Number of
Stations
3
3
3
3
2
2
2
1
1
i
1
1
1
1
1
1
1
1
1
1
J_
31
Number of
Sampling
Depths
3,3,3
3,3,3
3,3,3
3,3,3
3,2
3,2
3,2
1
1
i
1
1
1
1
3
1
1
1
3
1
1
68
9
5
5
5
1
1
1
1
3
1
68
39
-------
01
-------
TABLE 6. Total Number of Samples Required on the Chester River
Per Intensive Program or Per Year of Trend Monitoring
Description of Parameters
Physical Parameters
Sathymetric Survey
Tidal Stage
Current Velocity and Direction
Freshwater and Point Source
Discharge
Wave Height, Period and
Direction
Salinity
fat&r Temperature
pH
Suspended Sediment Load
Bed Sediment Characteristics
Chemical Parameters
Ultimate Carbonaceous 300
Sentnic Carbonaceous 300
Total Organic Carbon
Chemical Oxygen Demand
Alkalinity
3ed Sediment Carbon Content
^4'0'
Ammonia-Nitrogen
Nitrite-nitrogen
'4'5'
Total Phospnate1
Total Orthopnosphate
(5)
Total Filtrable Phosohate^ '
(4.)
Filtrable Orthophospnate
Bed Sediment Phosonate Content
Dissolved Oxygen-
31ological Parameters
Phytoplankton Concentration
Zooolanktorr Concentration
Secchi Disk Depth
Vitrifying 3acteria
Enumeration(S)
."umLar Per
intensive
Prcaram
(D
(2)
3400
(3)
500
1700
1700
1700
1700
31
1700
31
1700
1700
1700
31
3400
3400
3400
3400
1700
1700
1700
1700
31
1700
52
52
775
'lumbar Per
Year of Trend
Von i tori no
—
53
—
ZQ
S3
53
5S
53
31
63
31
53
S3
53
31
136
136
136
136
53
53
53
53
31
53
31
31
31
52
31
.NOTTS:
1. -sr^rm once /early
2. Continuous same ling recommended
3. To be Identified by the State of Maryland
i. Filtered
5. Jn filtered
5. in trie *atar column and bed sediment
41
-------
PATUXENT RIVER, MARYLAND
Site Description
The Patuxent River, shown in Figure 5, is located along the western shore
of Chesapeake Bay in the State of Maryland. The overall length of the Patuxent
River is 110 miles. The Patuxent River watershed is approximately 930 sq mi
and encompasses the metropolitan regions of Baltimore, Maryland and Washington,
O.C. Tidal influence of the Patuxent River extends to Hardesty, Maryland near
Route 214, about 55 mi above the mouth. Saltwater occasionally reaches Lyons
Creek, about 35 mi above the mouth.
The average discharge of the Patuxent River at its mouth is about 1078 cfs.
The discharge varies from 55 cfs to 6930 cfs.
The tidal portion of the Patuxent River can be characterized into three
hydrodynamic systems. Between Hardesty and Lyons Creek, the river is tidal
freshwater. Between Lyons Creek and Sheridan Point the system is estuarine,
being subject to tidal action and salinity gradients. Below Sheridan Point
the Patuxent River has been described as an embayment of Chesapeake Bay, i.e.,
it is chemically similar to and hydrodynamically driven by the bay [Patuxent
River Program, 1978].
Sampling Site Locations
Recommended transect locations and number of sampling stations and depths
for the intensive sampling program are given in Table 7 and shown in Figure 6.
Twelve transects are located on the main stem of'-the Patuxent River and two
each are located on Saint Leonard Creek, Battle Creek and Western Branch.
There are a total of 30 sampling stations and 60 discrete sampling points.
The trend monitoring station- locations should consist of collecting one
slack water sample in mid-channel and at mid-depth at each of the 18 transects.
Table- 8 gives: the* total number of samples required for each intensive program
and each year of trend monitoring on- the Patuxent River.
-------
FIGURE 5. Patuxent River
43
-------
TABLE 7. Patuxent River Sampling Locations
Location of Transect
Main Stem Patuxent River
Fishing Pt. to Drum Pt.
Half Pone Pt. to
Hoopers Neck
Captains Pt. to
Broomes Island
Sheridan Pt.
Chalk Pt. to Gods
Grace Pt.
Cocktown Cr.
Jones Pt.
Lyons Whf.
Queen Anne Bridge
Route 214 Bridge
St. Leonard Creek
Rodney Pt.
Johns Creek
Battle Creek
Prison Pt.
Long Cove
Western Branch
TOTAL
Approximate
River Mile
(N.M.)
0.0
5.2
9.8
15.8
21.0
26.5
31.0
35.0
&/) £
*rC • O
47.5
m n
J 1 . U
55.0
8.0
0.5
2.5
13.3
0.0
0.5
n n
u • u
7 n
U . U
Number of
Stations
3
3
3
3
3
1
3
1
1
i
i
1
1
1
1
1
i
1
1
I
30
Number of
Sampling
Depths
3,3,3
3,3,2
3,3,3
2,2,3
3,2,1
3
1,3,1
1
i
i
1
1
i
1
3
1
1
1
1
1
(
50
Total Number
of Sampling
Points
9
8
9
7
6
3
5
1
» 1
1
i
i
1
3
1
1
1
1
i
1
60
44
-------
FIGURE 6. Recommended Sampling Locations
Patuxent River
-------
TABLE 8. Total Number of Samples Required on the Patuxent River
Per Intensive Program or Per Year of Trend Monitoring
Description of Parameters
Physical Parameters
Satnymetric Survey
Tidal Stage
Current Velocity and Direction
Freshwater and Point Source
01scharge
Wave Height, Period, and
Direction
Salinity
Water Temperature
pH
Suspended Sediment Load
3ed Sediment Characteristics
Chemical Parameters
Ultimate Carbonaceous 300
Sentnic Carbonaceous SOD
Total Organic Carton
Chemical Oxygen Demand
Alkalinity
Sed Sediment Carbon Content
Ammonia-Nitrogen'4'3'
Nitrate-Nitrogen'*'0'
N1tr1te-Ni trogen'*'3'
Kj el dan 1-Nitrogen'
Total Phosphate'
/ e\
Total Orthophosphatav
Total riltrable Phosphate'
i 4)
riltraole Ortnoonosofiats'
3ed Sediment Phosphate Content
Dissolved Oxygen
Biological Parameters
.'.*)
Phytoplanfcton Concentration-
Zooplanlcton Concentration-
Secctii Disk Death
Nitrifying Sactaria.
Enumeration! 5)
Number Per
Intensive
P^oaram
(1)
(2)
3000
(3)
450
1500
1500
1500
1500
30
1500
30
1500
1500
1500
30
3000
3000
3000
3000
1500
1500
1500
1500
30
1500
60
60
750
Numoer Par
Year of Trend
tfonitorina
—
60
—
18
60
50
50
50
30
50
30
50
50
50
30
50
30
50
SO
50
50
50
50
30
50
30
30
30
50
30
.NOTES:
Perform onca yearly
Continuous sampling recommended
To be determined 3y :tie State of '
Filtered
Unfiltered
In the water column and sed saaiment
-------
POQUQSON RIVER, VIRGINIA
Site Description
The Poquoson River shown in Figure 7, is located on the southwestern
shore of Chesapeake Bay in the State of Virginia. The Poquoson River is a
small coastal basin between two large estuarine systems, the James River to
the south and the York River to the north. The drainage area of the Proquoson
River is small and the topographic relief is slight, resulting in the lack of
free flowing circulation such as is typical with the larger estuaries. The
Poquoson River has a dendritic pattern with many smaller distributaries enter-
ing it. The distributaries are drowned or flooded by Chesapeake Bay waters,
as is the Poquoson River and, therefore, are greatly influenced'by the tidal
circulation of Chesapeake Bay. The Poquoson River is dammed approximately
5 mi upstream of the mouth. Much of this water is diverted for water supply
purposes, so during parts of the year freshwater flow is nonexistent. Because
of the low freshwater inflow, tidal mixing will generally be the primary cir-
culation process, the longitudinal salinity gradient will be mild and the
vertical stratification is often nearly eliminated.
The Poquoson River is very shallow, generally less than 12 ft in depth.
The two principle distributaries are Chisman Creek on the north side of the
basin and Bennett Creek on the south side.
Sampling Site Locations
Recommended transect locations and number of sampling stations and depths
for the intensive sampling program are given in Table 9 and shown in Figure 8.
Eight transects are located on the main stem of the Poquoson River, and three
transects each are located on Bennett and Chisman Creeks. There are a total
of 19 sampling stations and 39 discrete sampling points. The trend monitor-
ing station locations should consist of collecting one slack water sample in
mid-channel and at mid-depth at each of the 14 transects. Table 10 gives
the total number of samples which have to be collected for each intensive
program and year of trend monitoring on the Poquoson River.
47
-------
s.
O>
c
O
(/I
O
3
CT
O
Q_
LU
Of
48
-------
TABLE 9. Poquoson River Sampling Locations
Location of Transect
Main Stem Poquoson
York Pt. to Cow 1
Hunts Pt.
Hunts Whf.
Lambs Creek
Hunters Creek
Quarter March Creek
Moores Creek
Mill Farms
Bennett Creek
Bay Pt. Buoy #4
Buoy #8
Floyds Bay
Chisman Creek
Ship Point
Buoy #6
TOTAL
WARE RIVER, VIRGINIA
Site Description
Approximate
River Mile
(N.M.)
0.0
1.3
1.8
2.2
2.6
3.1
3.4
4.4
0.2
0.6
1.0
1.0
0.0
n 5
U. 3
1.4
Number of
Stations
3
2
2
2
1
1
1
1
1
1
1
1
1
i
J_
19
Number of
Sampling
Depths
3,3,3
3,3
2,2
2,2
1
1
1
1
3
2
1
3
?
c~
]
39
Total Number
of Sampling
Points
9
6
4
4
1
1
1
1
3
2
1
3
2
c*
J_
39
The Ware River shown in Figure 9 is a small coastal basin within Mobjack
Bay on the southwestern shore of Chesapeake Bay in the State of Virginia. The
Ware River from the mouth to Deacons Neck, a distance of about 6 mi, is gener-
ally under the influence of Chesapeake Bay tidal waters. The freshwater
discharge of the Ware River is small, the average discharge at the Beaverdam
Swamp gaging station being 7 cfs. During high freshwater discharge conditions
it is likely that a two-layered circulation pattern exists in the tidal por-
tions of the Ware River.
49
-------
j.
-------
TABLE 10. Total Number of Samples Required on the Poquosson River
Per Intensive Program or Per Year of Trend Monitoring
Description of Parameters
Physical Parameters
Sathymetric Surrey
Tidal Stage
Current Velocity and Direction
Freshwater and Point Source
01 scharge
Wave Height, Period, and
Direction
Salinity
Water Temperature
pH
Suspended Sediment Load
3ed Sediment Characteristics
Chemical Parameters
U1 timate Carbonaceous 300
3enthic Carbonaceous 300
Total Organic Carbon
Chemical Oxygen Demand
Alkalinity
3ed Sediment Carbon Content
.(4,5)
Ammonia-iMi trogenl
Nitrata-Nitrogen1-
(1,5)
Nitrite-Mitrogen^*'5'
Kjeldahl-Nitrogen^ >3
Total Phosphate^5'
Total Orthophosphate'
Total "iltra&le Phosphate^ '
"iltrable Orthopnosphate•
3ed Sediment Phosphate Content
Dissolved Oxygen-
Biological Parameters
Phytoplankton Concentration
Zooplankton Concentration
Seccni Qisk Oepth
Nitrifying Sacteria
inumerarloniS)
Number Par
Intensive
Proaram
nj
(2)
1950
(3)
350
975
375
975
975
19
975
975
975
975
975
19
1950
1950
1950
1950
975
975
975
975
19
975
33
28
47?
?lumoer Per
Year of Trend
Monitorinq
—
39
—
14
39
39
39
39
19
39
39
39
39
39
19
39
39
39
39
39
39
39
39
19
39
19
19
19
33
19
NOTES:
°erfnrm ones- yearly
Continuous samoling recammenaed
To be determined by the State of Virginia
Un filtered
In the *ater column and oed sediment
51
-------
WILSON CR. WINDMILL PT.
FIGURE 9. Ware River
The main channel of the tidal Ware River is generally broad and shallow,
varying from about 25 ft deep near the- mouth to less than 10 ft in the upper
tidal reaches. The channel margins and subtidal flats are generally narrow,
making up less than 20% of the water surface area.
52
-------
Sampling Site Locations
Recommended transect locations and number of sampling stations and depths
for the intensive sampling program are given in Table 11 and shown on Figure 10.
Six transects are located on the main stem of the Ware River and two transects
are located on Wilsons Creek. There are a total of 12 sampling stations and
27 discrete sampling points. The trend monitoring station locations should
consist of collecting one slack water sample in mid-channel and at mid-depth
at each of the eight transects. Table 12 gives the total number of samples
to be collected for each intensive program and year of trend monitoring on
the Ware River.
TABLE 11. Ware River Sampling Locations
Location of Transect
Approximate Number of
River Mile Number of Sampling
(N.M.) Stations Depths
Total Number
of Sampling
Points
Main Stem Ware River
Ware River Pt. to
Ware Neck Pt.
Windmill Pt.
Baileys
Cottage Pt.
Landing Pt.
Pig Pt.
Wilsons Creek
Roanes Whf .
TOTAL
•
0.0
1.8
3.7
4.5
5.4
6.1
3.1
0.0
0.0
3
3
1
1
1
1
1
J_
12
3,3,3
3,3,3
3
1
1
1
2
1
27
9
9
3
1
1
1
2
J_
27
53
-------
-------
FIGURE 10. Recommended Sampling Locations - Ware River
54
-------
TABLE 12.
Total Number of Samples Required on the Ware River Per
Intensive Program or Per Year of Trend Monitoring
Description of
Physical Parameters
Sathymetric Survey
Tidal Stage
Current Velocity and Direction
Freshwater and Point Source
Discharge-
Wave Height, Period, and
Direction
Salinity
Water Temperature
PH
Suspended Sediment Load
3ed Sediment Characteristics
Chemical Parameters
Ul timate Carbonaceous 300
3enthic Carbonaceous 300
Total Organic Caroon
Chemical Oxygen Demand
Alkalinity
3ed Sediment Carbon Content
Arnmoni a-N1trogen ^
Nitrate-Nitrogen^
(4,5)
Nitrite-Nitrogen'4'51
Kjeldahl-Nitrogen'4'51
Total Phosphate'3'
(cl
Total Orthophosphatev
(4)
Total Filtrable Phosohata;
Filtrable Orthoonosphate'
3ed Sediment Phospnate Content
Dissolved Oxygen
31oioq7cal Parameter?
Phytopiankton Concentration-
Zaoplankton Concentration-
Seccht Oisic Oeptfr
Nitrifying Sactaria
£numeration(5)
Number ?sr
Intensive
Prsaram
(D
(2}
1350
(3)
200
575
575
575
575
12
575
575
575
575
675
12
1350
1350
13SO
1350
575
575
575
57S
12
575
24
24-
300
Number Per
year of Trend
.'"omtorina
—
27
—
8
27
27
27
27
12
27
27
27
27
27
12
27
27
27
27
27
27
27
27
12
27
12
12
12
NOTES:
Perform onca yearly
Continuous samollnq recommended
To 5e aetarmined by the Stats- of Virginia
Fi1terea
Unf i 1 tared
in -he water column and bea sediment
55
-------
CONCLUSIONS AND RECOMMENDATIONS
The application and evaluation of mathematical models developed to pre-
dict estuary water quality dynamics require detailed, valid data on both the
short- and long-term variations of many physical, chemical and biological
parameters. Such data are currently not available and, therefore, comprehen-
sive water quality sampling programs should be implemented.
A generalized sampling program designed to obtain the required data should
consist of two components: a set of intensive programs designed to collect data
on water quality variations throughout complete tidal cycles and a trend moni-
toring program to track seasonal and annual variations. This generalized pro-
gram should also provide sampling procedures which aid in the selection of
sampling locations and frequencies as well as ensuring the collection of
quality data.
The generalized sampling program as applied to the Chester, Patuxent,
Poquoson and Ware Rivers should be considered as recommended sampling pro-
grams. Even though the specific sampling programs developed for each sub-
estuary may exceed the states' water quality planning and management needs, such
data are required to adequately test and compare state-of-the-art models.
The States of Maryland and Virginia should refine the proposed, programs
based on 1) local knowledge of the factors which control water quality in
their respective subestuaries, 2) past, present or planned water quality
sampling efforts, 3) available sampling and analysis resources, and 4) their
priorities with respect to water quality management in each subestuary.
57
-------
REFERENCES
• Baca, R. G., W. W. Waddel, C. R. Cole, A. Brandstatter, and 0. B. Cearlock,
EXPLORE-I; A River Basin Water Quality Model. Battelle, Pacific Northwest
Laboratories, Richland, WA to U.S. Environmental Protection Agency, August
1973a.
Baca, R. G., W. W. Waddel, C. R. Cole, A. Brandstetter, and D. B. Cearlock,
Literature Review for EXPLORE-I: A River Basin Water Quality Model. Battel1e,
Pacific Northwest Laboratories, Rich!and, WA to U.S. Environmental Protect!on
Agency, August 19735.
Chester River Study, Volume II. A Joint Investigation by the State of Maryland
A Department of Natural Resources and Westinghouse Electric Corporation, November
9 1972.
Hewitt, J. P. and J. V. Hunter, "A Comparison of the Methods Used to Calculate
First Order BOD Equation Constants." Water Research, 9_: 683-687, 1975.
»
£ Methods for Chemical Analysis and Waste. A Technology Transfer Publica-
tion, U.S. Environmental Protection Agency, Environmental Monitoring Support
Laboratory, Cincinnati, OH, 1974.
Patuxent River Program. Maryland Department of Natural Resources, Water
Resources Administration, Water Quality Services, January 1978.
Standard Methods for the Examination of Water and Wastewater, Fourteenth
Edition. American Public Health Association - American Water Works Associa-
tion - Water Pollution Control Federation, 1975.
Tidal Current Tables 1978, Atlantic Coast of North America. U.S. Depart-
• ment of Commerce, National Oceanic and Atmospheric Administration, National
Ocean Survey, 1977.
Tide Tables 1978, High and Low Water Predictions, East Coast of North and
South America, Including Greenland. U.S. Department of Commerce, National
Oceanic and Atmospheric Administration, National Ocean Survey, 1977.
•
Tuffey, T. J., J. V. Hunter and V. A. Matulewich, "Zones of Nitrification.11
Water Resources Bulletin, JOJ3):555-563, 1974.
Wild, H. W., C. N. Sawyer and T. C. McMahon, "Factors Affecting Nitrifica-
tion Kinetics." Journal of the Water" Pollution Control Federation,
• 43:1845, 1971.
-------
APPENDIX
WATER QUALITY AND HYDRODYNAMIC
SUBMODELS OF EXPLORED
[]} This Appendix is composed of Sections IV and V of the EXPLORE model
documentation report [Baca at al., 1973a]
-------
SECTION -V
Sa QUALITY SUBMODELS
The water quality submodels used to describe the reactions of
various quality parameters in bc-ch the river and estuary and
dee? reservoir models are discussed in this section. A review
of all quality models identified during an extensive literature
survey is contained in Appendix A.
la choosing the models for each quality parameter, a number of
guidelines were considered. Past successful use of a model was
an important criterion, because the purpose of this study was
to identify and use the bes-c existing models rather than to
develop new ones. Ln addition, the simplest models which would
provide acceptably accurate results were selected. Consider-
ation was also given to the model's ease of calibration with
existing water quality data.
The program consists of eleven integrated water quality sub-
models which predict the behavior of sixraen water quality
parameters. A diagram of the relationship between these sub-
models is shown in Figure 4 and Table 1 provides a list of the
sixteen water quality parameters .
BIOCHEMICAL OXYGEN DEMAND
A number of models reviewed treat the biochemical oxygen demand
(BCD) reaction in a river system as other than a first order
reaction. These models have two things in common: first,
chey do not predict 30D concentrations with any more accuracy
than the first order models. Second, considerably more data
are required, to determine the large number of coefficients in
these models . Hence , a model using first order 2GD raacricns
was
The carbonaceous and nitrogenous ccmpcr.en.ts of 3CD were sepa-
rated to allow individual, treatment of each, reaction. 'The,
3CD equations ara written as
(2)
Cf I/Tr—"7 f 1
< = v rip1* a (. j- ^^ ^ / ±\
*••, '^T I «•« I' •- _ ' V ** ^
I L CT
A-l
-------
-------
0
s
H-
<
OS
'-3
A-2
-------
-------
TA3L2 1. WATZE QUALITY ?A?JU
-------
-------
(5)
, n • r..,n, .(T-20)
>
-------
1 dLd Ld c
H at" = " k4 H?c * k3 Lc (
:<4 =* :<4(20) s^2a) (3)
where
L, = araal benthic 30D, mc/a
!<4 * deoxygenation coefficient for benthic 3CD
k4(2Q) =» deoxygenation coefficient at 2QaC: 0.0-0.3/day
3^ * teaiperature coefficient: l.QS
S = average river death/ m
?,!<-, L are defined above
c j c
This equation is coupled to the carbonaceous 3CD equation
through the constants ? / k? anc* ^c-
^OTAL QRGA^IIC C\P^CN
}To iBodels were identified for siaulating- the reactions of
cotal organic carbon (TOO during the literature review. The
following- first order niodel was developed (sae iccendi^c A)
and is used. \n the c.
dt
where
Cm = cotal organic carhcn,
C_ = refractory organic carbon, nic/1
A-5
-------
-------
M-, = deoxygenation coefficient: 0.1-0.8/day
M, (20) = deoxygenation coefficient at 20°C
M., = sedimentation coefficient: Q.O-3.5/day
A^ - proportionality cons-ant: 10.0
L^ = carbonaceous 30D, mg/1
?3 = scour of degradabie organic carbon: 0.0-0.3 mg/l-day
?P = scour of refractory organic carbon: 0.0-0.3 mg/l-day
9_ = temperature coefficient: 1.05
This rodel is rela-ced -o the 30D model since -she rate of reaction
depends upon the concentration of carbonaceous 300.
NIT30G5N MODSL
Two nitrogen models have been included in zhe program. One con-
siders the effects of algae while che second includes ammonia-nitrate
dynamics. The algal nicrogen model (Figure 3a) assumes that each reac-
tion can be described, by a first order equation. The concentrations
FIGURE 5. NITROGEN
MODELS
A-S
-------
-------
of the various nitrogen f orris ara given by C^ and the ratas of
reaction are given by ji_. The various nitrogen components wera
formulated as a first order n:cdel for two reasons. First, the
relative simplicity of the first order equations with only one
constant to be evaluated (along with the boundary conditions)
is attractive. Detailed data for regression on a large nuaber
of cons-tants will generally not be available. Second, the'
fi.rt order model has been shown to work satisfactorily, at
leasr under certain conditions. The codel is described by the
ecuaticns:
dC
J3C4 -
ffl
dt
dC3
dt"
dC.
2C2
'IT?
*
(12)
(13)
(14)
(15)
where
C.J =» asctcnia concentra-tion, ag/7!
C7 * nitrite concentration, rag/1
C- =* nitrate, concentration, ac/1
C = organic nitrogen concenrrarion , -Tig/1
^ , ^ , 3
rats ccns-canrs : 0 . 1-0 . 5 , 5 . 0— 1Q . a , a . 1-0 . 4/day
* chyiroplanlc-cn and soaplankton: ccncen-rations , aig-C/1
^A.^-. = raric of nitrogen to carbon in ohytopianxton
and in zocplanJcton : 0.17, 0.15 ag— N/ag-C
/G^ — growth rates of phy-opiarJcton and scop lanktcn, /day
/D = death rates of phytoplanktcn and =ocplan:ctcn,/cay
D-, and D_ ara f emulated, in, the algae -ccai.
s a.
A-7
-------
-------
A simple first ordsr ammonia-nitrate mods! (Figure 3b) has been
included in the coda for use when algal affects are not. to be
considered:
•*• _ _ T r
~ ' U2C1
dC,
at
(15)
(17)
PHOSPHORUS MODEL.
Three phosphorus models are included in the quality model pro-
gram. One includes the first.order reactions of soluble
phosphorus with algae and with the sediments. The second is a
first order reaction model between solvable and sediment phos-
phorus . The zhird model is a second order decay model for
soluble phosphorus.
•The algal phosphorus model selected (Figure 6a) is similar to
the nitrogen model and assumes that the reactions between the
various forms of phosphorus can be described by a first order
equation. The concentrations of the phosphorus forms are
given by D^ and the rate constants for the reactions are
given by I~.
so
3HC
1
5?HORU5. ;
37 . '
i"^
ScOIMS.'lT
'HOSPHORUS
31
A f
*» L »
! -
\
I
AL2AL ! fc
= HQS?.HORUS ^
aecAYiNS
t 1 or i,« r r
?HOSPHQRUS
31
I
'.Sr^ACTQSr-s
1
j
3QLu3L£
'HCSPHORUS
3?
"
r 3HCSPHQRUS
^ '' Q-
4( 1
3. FiSST OROES PHOSPHORUS -OOSL
SOLUBLE
SHOSPHOSUS
°2
i
1
1 ''Z >
A. Ai.SAL PHOSPHORUS
C. ScCSNQ JSCi?. 'HCSPHORUS
FIGURE 5. PHOSPHORUS MODELS
A-3
-------
-------
Phosphorus was modeled as a set of first ordar reactions pri-
marily because no other models were available in the lireratu:
The first order modal allows a sins is analysis method for
phosphorus behavior in cases where sufficient da-a are avail-
able" co calibra-s the model. In addition, scne past work
has indicated tha-c first order reactions are applicable to
phosphorus behavior in at least some areas .
This model is considerably ' simpler than most proposed phos-
phorus cycles because it would not be possibla to determine
all the rate constants involved in a complicated system,
Phosphorus data are usually incomplete and it is vary un-
likely that enough data would exist to allow a regression
analysis on. many constants. The equations for this modal are:
dD.,
' V -Sp (IS)
^ -I2D2 - 1,0, (13)
dD,
— - - r3D3 + V ^P " (D2 ~ GZ)AD22 (2C)
whara
D^ » soluble phosphorus concentration, mc/1
D-, - sedimentary phosphorus concsn-cr-aricn , -g/1
I, , I^ , Z7 =- rate, constants : 0 . 1-0 . T/day
3 * phytoplankton. concan-tratian , ag-C/1
Z * zooolankton concsnrration , mg-C/1
~ ratio or nitrogsr. to caxbcn in phytoplanJc— on and
rcop lank ton: 0.17, a. IS mg— Vmg-C*
~ growth, rates of phy top lank ton and zcoplankton,/day
* dear:": --tas of phytopianJ-cton and zee? lank-can, /day
D, and D ars formulanad in the algae modal.
A-9
-------
-------
The firs- order modal (Tig-ore So) assumes that only reactions
between soluble and sedimentary phosphorus occur. This model
uses the equations :
2~ = - I1D1 + I2D2 (21)
cD
-nr- - - r2D2 - ITD, (22)
v«fc W £ £• +*. J»
where- the notation is the sane as above .
The second crder model (Figure 5c) uses the equation:
dD ,
rr^- = - IT°L (23)
PHOTOSYNTHESIS AND 3Z3PI3ATION OF aUSrSSDSD ALGAZ
The model chosen to represent the rate of photosynthesis and
respiration of suspended algae is a two term harnonic function
of the form:
(?-R) c = A, cos (cut •-!• i) -r A7 cos(2ut -re:) (24)
where
(?— R) = net rate of thatosvnthesis and respiration:
3 =(3.1-2.0) , mg/l-day
j. = period, days
A. , A-, , p , ct = constants
This formulation was selected for two reasons. First, it is
relatively simple since only three constants need, to be deter-
.TiinecL (A^,A2 and. the difference o-c:} . The: success of previous
authors In using- one, two, and. three terar periodic functions
to model diurnaL variations has- shown that these relatively
simple expressions can. be successfully applied.. Second, the
two term expression is more desirable than the one term expres-
sion since with two terms the maximum rate of photosynthesis
need not equal the maximum rate of respiration.
An option is provided in the program which allows the user to
couple this equation to the phytoplanJctcn density. This is
done simply by:
(?-R)s - 5? (25)
A-10
-------
-------
Where ? is the phytojlaafctcn concentration and 5 is a
proportionality constant.
PHOTOSYNTHESIS AND 3ZS2I3ATION OF ATTACHED ALGAZ AND
3ENTHIC PLANTS
The modal chosen to represent the net rate of photosynthesis and
respiration of attached algae and benthic plants is identical to
that used for the suspended algae model, it is given as
(S-R) , = A a cos(ui t +
-------
-------
wrier a
T = temperature, 3C
Four separate models for estimating the raaeraticn coefficient
have been included in the program. These models cover most
of the work dona on re aeration coef f icients .
The first model for the reaeration coefficient is a general
velocity-depth medal of the fora
k2 = rL — (29)
where
'<-) - reaeration coefficient , /day
v = velocity, ft/ sec
h = depth, ft.
r^r.,,-., = constants-. 2.0-12.0, 0.5-1.0, 0.35-1.35
*
The second model is one proposed by O'Connor and Dobbins^
which is given as
D 1/2-1/4
k = 11Q9- JS — —- - for C <17 (30)
v "
(D
for C >17 (31
where C - ' . /a
(hS)^2-
and
ILj =• liquid film diffusion coefficient: 0.01-0.02 ft /sec
5 =* river" channel slone
The third model, proposed by Dobbins, is given as:
-12
-------
-------
3/8
2772 C,AEJ/8 coth
(32)
where
A - 9.S8 +> 0.054 (T-20)
3 = 0.97S +• 0.0137 (30-TJ3/2
E - 30. v S
5 =» gravita-fcional cons-tan-t, f-t/ssc
T = tamseratura , "G
faiirtii model was oraoosed bv Thacxs-ton and Xrsakel •
0.000299 1 +>
v \ 1/2
^^zb.
, ' A
Scr
(33)
The temperature dependence of the reaeraticn coefficient for
models one, two and. four above is described by the following
relationship:
This relationship was chosen because it is based on large
amounts of data, and has proven, reliable in. many investigations,
The effects of wind on the reaeration coefficient have been
observed; however, only minor theoretical, treatments were
found. Therefore, wind effects were not included in this
model.
'vith the exception of TOC (che contribution of which is handle
through the 3CD models) and toxic compounds/ all models dis-
cussed previously provide inau-r information either directly
or indirectly to the dissolved oxygen model. 'The general
equation describing the dissolved oxygen level is
A-13
-------
-------
»*
where
DC = dissolved oxygen, rag/1
k^,k^,k,, = deoxyger.ation races for carbonaceous,
"~ nitrogeneous and benthic BOD, /day
k7 = reaeration coefficient, /day
(D0_ - DO) = dissolved oxygen deficit
(?-R) _, (?-R) = photosynthesis and respiration rate for
3 a suspended and attached algae, respectively
Values for the rate constants and femulation of (?-R) con-
-ribucions are presen-ed in previous sections of -he report.
?LAN:
-------
-------
Notation
? = phytoplankton concentration, r.g-C/1
G0 = phytoplankton growth ra~a, /day
D., =» chvtoolanktcn daatui rata, /dav
p - - -
Z * zcoplankton concantration, ag-C/1
GZ =» 2ooplank-ton growth rata, /day
. D = zcoplankton dsath rata, /dav
te "*
.M » nitrara-nitrogen ccncantraticn, ag/1
?1 = saturatad growth rata of phytoplankton:0.05-0.20 /day-aC
T = taiaaarattira, 3C
f = photoparicd:0.3-0.7, days
]
-------
-------
A ,A, = ohytooiankton nitroaen-carbon ratio: C.05-0.17
M ^^ jj * ^
"*• ing-N/mg-C and phosphorus-carbon ratio: 0.024-
0.24 mg-D/ir.g-C
A ../-Aj = zooplankton nitrogen-carbon ratio: 0.05-0.17
n"i' mg—M/mg-C and phosphorus-carbon ratio: 0.024-
0.24 mg-D/mg-C*
A = zoopiankton conversion efficiencv: 0.5, mg-C/ma-C
zp
K = phytopiankton Hichaelis constant: 0.06, mg-Chl/1
?. = zooplankton endocencus respiration rata Q.01
/day-°C
?d = zooplankton death rata due to other causes, /day
D = phosphorus concentration, mg/1
C = carbon
Chi = chiorophyll-A
This modal considers two trophic levels (phytoplank-on and zoo-
piankton) and two limiting nutrients (nitrogen and phosphorus).
If carbon is limiting, then additional Michaeiis-Menton terms cai
be added to Equation 33 to adjust the saturated .growth rate.
This model was selected because it is the least complex one
available which accounts for ail of the pertinent ecological
parameters associated with phytopiankton and zooplankton. In
addition, the model has been testad successfully in field
studies. Average values for a number of the constants have
been determined. The remaining" constants have to be evaluated
through calibration with field data.
TOXIC COMPOCKDg
The literature- survey revealed very little information on the
modeling- of toxic compounds.. The behavior of toxic compounds
in an aquatic environment is highly variable and depends upon
the compound being modeled. A model which would predict the
interaction of tcxic compounds with sediments, biota and other
constituents would be a fine scientific tool but impossible to
develop within the scope of this program. "igure 7 illustrates
the. complexity of the problem and traces the distribution and
of toxic materials.
A-16
-------
-------
Therefore, a simple r.~" crdar decay modal was included zo de-
scribe toxic compounds. This model accounts fcr any order of
dacay desired, including zaro order, and can easily acccirmoda~e
nonlinear behavior. The modal is giver, as:
c
,-1
"c
(42)
wnera
T * concan -ration of toxic ccnnound
c
a
rata constant
n =* order of decay
This decay model can be used to represent any of the mechanisms
of exchange such as cheaical transformations, sedimentation,
biological uptake, and sorpticn.
3T
HIGHEH
AOUATTC ?«.A«TS
CHEMICAL ArtO |
(PHYSICAL ?*CCi32c3i
FIGURE 7.
FACTORS AFFECTING THE OISTRI5UTIQN AND FATE OF
TOXIC MATERIALS IN WATER
A-17
-------
-------
SECTION V
HYDRAULIC MODEL SELECTION
Thrae capabilities were required of the hydraulic model to
be practicable for routine use: 1) it must be. applicable
to large scale physical systems (such as entire river basins);
2) it should accommodate as many of the flow regimes as pos-
sible which would be present in a river basin (e.g., small
streams, large rivers, and tidaily influenced regions);
and 3) it must: utilize an efficient and reliable algorithm.
HZDa&GLIC .MODELS CCNSPSS5D
Among the available hydraulic models considered for inclusion
in the EXPLORE-I cede were the following:
Author Description
1. Amein and Fang Implicit Flood Routing
2. Amorocho and Strelkcff Hydraulic Transients
3. Garrison Unsteady Flow Simulation
4. Pisano River 3asin Simulation
5. Watsr Resources Engineers Storm Water Management
A summary of the hydraulic models reviewed is presented in
Appendix A. Few general purpose computer models capable of
handling complex river networks and varying hydraulic regimes
ars available for simulation of open channel flow hydraulics.
In fact, general, practice has relied upon simple regression
relationships for s-cage and velocity as functions of flew.
More scphisticatsd techniques ars required, however, for
estuardne hydrodynamics-
The Water- Rascurcss Engineers (WH2) Dynamic Estuary Model is
to data the only hydrcdynamic medal, capable of handling various
hydraulic regimes and complex system gecmerrias. At least
three versions of chs original program are,,currently available,
including the- FWrCA. Dynamic Estuary Model,' the Columbia
River Mcdel,3 and the Storm Water Management Model.0 Each
version varies in overall generality and in the numerical
aTTcrcximaticns used.
A-19
-------
-------
MODEL SELECTION
The hydraulic simulation module of the EXPLORE-1 computer model
is an adaptation and extension of the receiving water component
of the Storm Water Management Model. This rr.cdel utilizes a
pseudo two-dimensional approach to solve momentum and continuity
equations on an arbitrary computational mesh consisting of a set
of interconnected channels and junctions.
The equations of motion and con-inuity for open channel flow
(Saint-Venant equations) consist of two simultaneous equations:
Momentum
• S - v4? + gl| = g (S -S.-QJ (43)
-------
-------
D7 =-0, for buUc lateral outflow
Dj - TT* — , for seepage outflow
V-uz
D,, s ^ c, ror iararal inflow
i. aq
where
u, =* x-conrponent of the inflow— velocity vector.
The Saint- Venant aquations can also be written with discharge
instead of velocity as the griaiary kinetic variable. For flow
through a rectangular cross section with negligible botton
slope and dynastic affect (of Lateral inflow) these aquations
reduce to
Moaentua
Continuity
44^4|=a • (46)
oC sX
where
*7 =* verticallv averaged channel velocity
h = water" daprih
Q = vclusetric flow rata
A =• channel cross-sec-ional area (A*ch)
h - channel width
h - channel depth
3 = wind stress tarn
w
The wind stress tarn Sw has been added to the traditional fo
Bula-cicn co account for wind af facts. In certain cases wind
induced currants can be significant to -he analysis af -he
flow ractams .
Mcst versions of the W2Z hydraulic, mcdei are f emulated in
tarns of head (potential) instead of wacar depth, which has
A-2I
-------
-------
created some confusion about the correctness and completeness of
the momentum and continuity equations. In previous versions of
the hydraulic model, the following substitutions were generally
made:
3h _ SE (47)
Tt ~" ot
3H =» _ 1J1 j. s ^4S)
Tx 3x ' o
SQ = - || (49)
Convenience and convention, rather than computational reasons,
dictate the use of these approximations. The bottom slope con-
tribution S is generally small in estuarine applications.
However, in°river systems where changes in elevation drive the
motion of water, these contributions can be significant.
The iXPLQRS-I hydraulic simulation program uses the modified
form of the Saint-Venant equations to model the two-dimensional
flow in an estuary or one-dimensional flow in a river system.
A system is geometrically represented by a computational grid
network, using a simple routing concept, the flow in the system.
is prescribed along paths designated by channels con-
nected by junctions. Eacn junction is cnaracterized by its
1) surface area, 2) water surface elevation, 3) bottom elevation,
and 4) the actual (X,£) coordinates. Surface areas are computed
from those polygons formed by the perpendicular bisectors of
channels in the network. Flow between these junctions occurs
through rectangular channels which connect the junctions. The
characteristics of each channel are described by length, width,
Manning coefficient, and average bottom elevation at the mid-
point. Figure S illustrates the channel-junction configuration.
In the numerical, solution scheme, the Saint-Venant equations are
not solved, simultaneously for all points in the water system.
Instead, che equation, of motion, Equation 50, is solved for each.
of the channels in the system on. the basis of present values of
the junction heads.
Momentum
where
*7 =• velocity
H: =• water surface elevation measured from the
datum, plane
A-22
-------
-------
JUNCTION
FIGURE 3 CHANNEL-JUNCTION DEFINITION FOR
COMPUTATIONAL MESH3Q
= wind stress
_
u COS 'i
K » dixensionlass coefficiant with a value cf
0.0026 (3afs. 2 £ 3)
d * depuh. of flow
Pa/p » air and watar densities
War W
CT = wind velocity
;£ =' ancle bettveen the wind direc"tifln and the axis
of the channel
The velocity and flow in each of uhe channels is thus decemined.
The equation cf continuity, Equation. 51, is solved at each of
the junctions on the basis cf predicted flows in. each, cf the
channels.
A-23
-------
-------
Continuitv
:ai _ v- « , oim «ex
where
A • = surface area associated with the junction j
Q?. = flow of a connecting channel
im ^.ex °v
Q^ /Q- rQ^ * water import, export and evaporation
-» - - rates at the junction
lc = number of channels connected to the j ""a junction
The numerical integration technique employed in the hydraulic
model uses a space- and time-staggered scheme based on a sim-
ple Hunge-Kutta approximation. The finite-difference equation
for the momentum equation may be expressed for the i~
channel as:
IV.. AV.
*
where
V. - channel velocity
It = time step
-~=- ~ velocity gradient
X. s» frictional resistance coefficient for channel i
g = acceleration of gravity
^i "" Wi-taj""i -engt.*
h * averace channel denth
The velocity gradien- tern is computed front the continuity
ecuation:
A-24-
-------
-------
AV.
7,-j—s. = V.
-Ax. i
b. Ah,. V. AA. 1
with
b. = channel width
A; = channel cross-sectional area
In addition, the relationships given below
Ah. AHi Aft £S
and
AA. * b.Ah. (55)
are used in intermediate computations of the key variablas:
V, h, Q, H, A. Since heads (potentials) are designated and com-
puted at junctions the terms AH.;/At and AA-j/L^ are computed as
the average of changes at the junctions at bcth ends of the
channel.
The numerical solution technique proceeds as follows:.
Initialization
1. Set initial and boundary flows and stages.
2. Set import and/or export of flows.
3. Set water- surface elevations and/or tidal conditions.
Leap Frog- Z:cslicit Solution
1. Ccapu-ts half-interval velocities, '~. ' "' and quarter-
w ?^-/* **
« 77
r- * gs V
*»
A 7
2. Controlled function calculations for flow at1_cua^-ter
intervai, QT~ ~ , and head at half -interval , HJr ' "/ " -
Head and/cr^flow oay be specified functions of time
or related via a rating- cu.rvs.
A -25
-------
-------
3.
4.
c-1 /2
Coispute head at half- interval, H^ "'
n .
. _
= a . T -
2A
sj
.<
r
i=L
- o
3
-1/2
ev
" QJ
Compute half-interval flows, Q,- ' , and channel.
areas, Ar~^'~ and full-interval velocities, V^-:
= A" + b.
,t+L
.t-f-1/2
- g.
1
Controlled junction calculations for flows and heads
at full tise steo.
Ccnpute full-interval heads, 3^'*/ water depths,
"nP"*"^- and channel areas, Ar^: ^
•
Qt+l/2
7. Update system parasierars and repeat steps 1 through
o until, hydraulic cycle is complete»
3. Average flows and velocities over hydraulic cycle.
This leap-frog schente introduces a certain degree of accuracy
and stability to The explicit numerical solution; however",
there are inheren- conputa-ional lirtizawicns. To maintain
a stable and accurate solution the following criterion must
be aocroxiria^eiv satisfied:
A-26
-------
-------
* \
crR
(37)
where
= raaxiriusi tine star? far stable solution.
The expression In the brackets represents the celerity of a shallow
water water wave.
A-27
-------
------- |