Maryland Coastal Bays Nutrient TMDL Water Quality Modeling Report

Maryland Coastal Bays Nutrient TMDL Water Quality Modeling Report
HYDRODYNAMIC AND WATER QUALITY MODELING AND TMDL
DEVELOPMENT FOR MARYLAND’S COASTAL BAYS SYSTEM
Prepared for:
Maryland Department of the Environment
Science Service Administration
TMDL Technical Development Program
1800 Washington Boulevard, Suite 540
Baltimore, MD 21230
Prepared by:
Harry Wang, Zhengui Wang, Derek Loftis and Yi-cheng Teng
Department of the Physical Sciences, Virginia Institute of Marine Sciences,
College of William and Mary
November 2013
Table of Contents
List of Tables .................................................................................................................................... i
List of Figures ...................................................................................................................................ii
EXECUTIVE SUMMARY .................................................................................................................vi
CHAPTER 1: INTRODUCTION........................................................................................................ 1
1.1 Background ..................................................................................................................................... 1
1.2 Study Approach .............................................................................................................................. 4
CHAPTER 2: GENERAL CHARACTERISTICS OF MARYLAND COASTAL BAYS WATERSHED .. 6
2.1 Physical Setting............................................................................................................................... 6
2.2 The terrestrial and atmospheric loading.......................................................................................... 6
2.2.1 Land use distribution............................................................................................................ 7
2.2.2 Non-point source loading..................................................................................................... 9
2.2.3 Point-source loading .......................................................................................................... 10
2.2.4 Shoreline erosion loading .................................................................................................. 13
2.2.5 Atmospheric loading .......................................................................................................... 16
2.2.6 Discussion of Contributing Loads to the Maryland Coastal Bays ..................................... 18
2.3 Data to Support the Water Quality Modeling ............................................................................... 24
CHAPTER 3: THE THREE-DIMENSIONAL HYDRODYNAMIC MODEL ........................................ 29
3.1 Model description ......................................................................................................................... 29
3.2 Model set-up ................................................................................................................................. 30
3.3 Hydrodynamic model forcing functions ....................................................................................... 30
3.4 Hydrodynamic model calibration ................................................................................................. 33
3.4.1 Water level calibration ....................................................................................................... 33
3.4.2 Velocity calibration ............................................................................................................ 42
3.4.3 Salinity calibration ............................................................................................................. 46
CHAPTER 4: THE COUPLED WATER QUALITY AND SEDIMENT BENTHIC FLUX MODEL ...... 52
4.1 The ICM Water Quality Model..................................................................................................... 52
4.2. The Sediment Benthic Flux Model .............................................................................................. 54
4.3 Water Quality Model Set-up ......................................................................................................... 56
4.3.1 External loading ................................................................................................................. 56
4.3.2 Initial and boundary conditions ......................................................................................... 59
4.3.3 Estimation of parameters ................................................................................................... 59
4.4 Water Quality Model Calibration and Verification ...................................................................... 60
4.4.1 Model calibration ............................................................................................................... 60
4.4.2 Model verification .............................................................................................................. 70
CHAPTER 5: ADDITIONAL MODEL ANALYSES .......................................................................... 94
5.1 Adjustments to incorporate the DO Diel Cycle ............................................................................ 94
5.2 Sensitivity Analyses .................................................................................................................... 112
5.2.1 Ocean City Wastewater Treatment Plant outfall ............................................................. 112
5.2.2 Effect of phytoplankton and organic nutrient settling rate .............................................. 114
5.2.3 Effects of groundwater discharge .................................................................................... 116
6.1. Developing nutrient load reduction scenario ............................................................................. 119
6.1.1 Critical conditions ............................................................................................................ 119
6.1.2 Margin of Safety (MOS) .................................................................................................. 120
6.1.3 Seasonal variation ............................................................................................................ 120
6.2 Developing the TMDL scenario ................................................................................................. 121
6.3 Final TMDL scenario for MD Coastal Bays – geographic isolation method ............................. 134
CHAPTER 7: DISCUSSION AND CONCLUSION ........................................................................ 145
REFERENCES ............................................................................................................................ 148
APPENDIX A: Summary of SELFE hydrodynamic model formulations ........................................ 155
APPENDIX B: The ICM Water Quality Model Formulation ........................................................... 168
APPENDIX C: Summary of Parameters Used for ICM Water Quality Model ................................ 203
List of Tables
Table 1.1: TMDL impaired water listing for the Maryland Coastal Bays (MDE 2010) ........................ 3
Table 1.2: HEM3D model water quality state variables .................................................................... 5
Table 2.1: Drainage and water surface area (in acres) of the Maryland Coastal Bays ...................... 7
Table 2.2: Land use (in acres) for sub-watersheds in MCBs HSPF watershed model ...................... 9
Table 2.3 (a): Major point source facilities in Maryland with permits regulating the discharge of
nutrients. ............................................................................................................................ 11
Table 2.3 (b): Permitted point source facilities in Virginia ............................................................... 12
Table 2.4: TN and TP loads from shoreline erosion for individual bays and entire MCBs ............... 15
Table 2.5: Total baseline nitrogen and phosphorus loads (lbs/year), MD Coastal Bays 2001-2004.23
Table 2.6: Summary of data supporting the hydrodynamic and water quality modeling. ................. 25
Table 3.1: Description and location of field observation stations for current, water level and salinity
in the MCBs. ....................................................................................................................... 34
Table 3.2: Statistical measures for modeled versus observed salinity at five basins in MCBs. ....... 51
Table 4.1: The parameter selected for phytoplankton dynamics ..................................................... 60
Table 4.2: Model and data comparison statistics for DNR stations. ................................................ 92
Table 4.3: Model and data comparison statistics for ASIS stations................................................. 93
Table 4.4: Relative Error in the MCBs model compared with other modeled systems. ................... 93
Table 5.1: Variables, coefficient values and statistical test results for DO diel cycle linear regression
model (Perry et al., 2012). .................................................................................................. 99
Table 5.2: Exceedance rates for Chlorophyll-a under Ocean City WWTP 20%, 40% and 60%
incremental reduction scenarios. ...................................................................................... 113
Table 5.3: Effect of changing settling velocities of phytoplankton and organic nutrients on
chlorophyll-a concentration. .............................................................................................. 116
Table 6.1: Maximum Practicable Anthropogenic Reduction (MPAR) Percentages for each non-point
source sector based on CBP-Phase 5.3.2 scenario results. ............................................. 122
Table 6.2: DO exceedance rate under baseline conditions and reduction scenarios of 20%, 40%,
60%, natural conditions and MPAR for all DNR and ASIS stations. .................................. 124
Table 6.3 (a): Modeled chlorophyll-a exceedance rates under baseline conditions and reduction
scenarios of 20%, 40%, 60%, natural and MPAR conditions for non-SAV growing zone with
Chl-a endpoint greater than 50 μg/l. ................................................................................. 132
Table 6.3 (b): Modeled chlorophyll-a exceedance rates under baseline conditions and reduction
scenarios of 20%, 40%, 60%, natural and MPAR conditions for SAV growing zone with Chla endpoint greater than 15μg/l. ......................................................................................... 133
Table 6.4: Final TMDL reductions needed to meet WQS incorporating Geographic Isolation
Scenarios. ........................................................................................................................ 135
Table 6.5: DO exceedance rates for TMDL IR (Incremental Reduction) and final TMDL (TMDL GI,
incorporating Geographic Isolation Scenario). .................................................................. 136
Table 6.6(a): Chla greater than 50 ug/l exceedance rates for TMDL IR (Incremental Reduction) and
final TMDL (TMDL GI, incorporating Geographic Isolation Scenario). ............................... 140
Table 6.6(b): Chla greater than 15 ug/l exceedance rates for TMDL IR (Incremental Reduction) and
final TMDL (TMDL GI, incorporating Geographic Isolation Scenario). ............................... 141
Table C-1: Parameters related to algae in the water column. ....................................................... 203
Table C-2: Parameters related to organic carbon in the water column. ........................................ 205
Table C-3: Parameters related to nitrogen in the water column. ................................................... 206
Table C-4: Parameters related to phosphorus in the water column. ............................................. 208
Table C-5: Parameters related to silica in the water column. ........................................................ 209
Table C-6. Parameters related to chemical oxygen demand and dissolved oxygen in the water
column.............................................................................................................................. 209
Table C-7: Parameters used in the sediment flux model. ............................................................. 210
i
List of Figures
Figure 1.1: Location map of MD Coastal Bays System ....................................................... 2
Figure 2.1: Land use distribution in the Maryland Coastal Bays watershed. ........................ 8
Figure 2.2: Locations of sampling sites (in red) and land loss polygons (in blue) in
Assawoman and Isle of Wight Bays. ....................................................................... 14
Figure 2.3: Wet deposition of inorganic nitrogen in 2002 and 2003 in the United States ... 17
Figure 2.4: Total nitrogen load (percentage of total) from source sectors
in individual MCBs .................................................................................................. 19
Figure 2.5: Total nitrogen loading from source sectors in individual MCBs (lbs/year). ...... 20
Figure 2.6: The total phosphorus load (percentage of total) from source sectors in
individual MCBs. ...................................................................................................... 21
Figure 2.7: Total phosphorus load from source sectors in individual MCBs (lbs/year) ...... 22
Figure 2.8: Station location for DNR and ASIS monitoring program in MCBs ................... 26
Figure 2.9: DNR monthly DO data from stations in Saint Martin River (2000-2004) .......... 27
Figure 2.10: Combined DNR and ASIS monthly DO data from stations in Newport Bay
(2000-2004) ............................................................................................................. 28
Figure 3.1 Hydrodynamic model grid in the Maryland Coastal Bays (MCBs). .................... 31
Figure 3.2: Bathymetry for the Maryland Coastal Bays ...................................................... 32
Figure 3.3: Map of field observation stations for water elevation, currents and
salinity in the MCBs. ................................................................................................ 35
Figure 3.4 (a) Tidal calibrations near Ocean City Inlet in the northern portion
of the MCBs. ........................................................................................................... 36
Figure 3.4 (b) Tidal calibrations near Chincoteague Inlet in the southern portion of the
MCBs. ...................................................................................................................... 37
Figure 3.4(c): Tidal calibrations across the middle and southern portions of the MCBs. ... 38
Figure 3.5: M2 Tidal profiles along Chincoteague and Sinepuxent Bays transect............. 39
Figure 3.6: Wind record at Ocean City and Wallops Island, September 23 to October 8,
2004. ....................................................................................................................... 40
Figure 3.7: Water level verification at A) Ocean City Inlet;, B) South Point; C) Harbor of
Refuge, and D) Turville Creek. ................................................................................ 41
Figure 3.8: (a) Spatial distribution of surface velocities during maximum flood tide at Ocean
City Inlet, October 2004; and (b) Current speed time series comparison along the
major axis of the local channel; ADCP measurement (red), and model results
(blue). ...................................................................................................................... 43
Figure 3.9: Current calibration results at A) Isle of Wight Channel B) Sinepuxent Bay C C)
Chincoteague Bay Channel; model (red solid line) and data (blue dash line) ......... 44
Figure 3.10: Statistical measures for current comparison at calibration stations A, B. and C,
respectively.............................................................................................................. 45
Figure 3.11 (a) Areally-weighted daily discharge from USGS flow gauges (01484719 and
0148471320) in MCBs, and (b) Salinity calibration in the tributaries of the northern
MCBs. ...................................................................................................................... 47
Figure 3.11 (c): Salinity calibration in the open waters of the northern MCBs . .................. 48
Figure 3.11 (d): Salinity calibration in the southern MCBs.................................................. 49
Figure 3.12: Spatial distribution of averaged salinity contour in MCBs over summers of (a)
2003 and (b) 2004. .................................................................................................. 50
ii
Figure 4.1 Schematic diagram of coupled HSPF watershed, SELFE hydrodynamic, ICM
water quality, and sediment benthic models. ........................................................... 53
Figure 4.2 (a) Schematic diagrams for (a) ICM model water column processes and (b)
sediment digenesis processes................................................................................. 55
Figure 4.3 (a) Example the selected segments of HSPF Coastal Bays model segments in
the Assawoman Bay, Isle of Wight Bay, Sinepuxent Bay and Newport Bays .......... 57
Figure 4.3 (b) Examples of the selected segments of HSPF Coastal Bays model segments
in the Chincoteague Bay ......................................................................................... 58
Figure 4.4 Intensive monitoring stations in (a) Isle of Wight Bay and (b) Maryland Coastal
Bays. ....................................................................................................................... 61
Figure 4.5 (a) and (b): Comparisons of model prediction and field measurement for
chlorophyll-a and DO in the Isle of Wight Bay. ....................................................... 63
Figure 4.5 (c) and (d): Comparisons of model predictions and field measurement for NH4
and NO23 in the Isle of Wight Bay. ........................................................................... 64
Figure 4.5 (e) and (f): Comparisons of model predictions and field measurement for PO4
and DON in the Isle of Wight Bay. ........................................................................... 65
Figure 4.6 (a) and (b): Comparisons of model predictions and field measurement for
Chlorophyll-a and DO in Bay-wide stations. ............................................................ 67
Figure 4.6 (c) and (d): Comparisons of model predictions and field measurements for NH4
and NO23 in Bay-wide stations. ................................................................................ 68
Figure 4.6 (e) and (f): Comparisons of model predictions and field measurement for PO4
and DON in the Bay-wide stations. .......................................................................... 69
Figure 4.7: MD DNR and ASIS monitoring stations in the Maryland Coastal Bays. ........... 72
Figure 4.8 (a) The flow discharges and (b) Chlorophyll-a concentration from Birch Branch
watersheds. ............................................................................................................. 73
Figure 4.9 (a): DO verification with DNR data, Stations 1-12, 2001- 2005. ........................ 76
Figure 4.9 (b): DO verification with DNR data, Stations 13-24 2001-2005. ........................ 77
Figure 4.9 (c): DO verification with DNR data, Stations 25-27, 2001-2005. ....................... 78
Figure 4.10: Statistical comparison of observed versus modeled DO; Assawoman, Isle of
Wight, Newport, and Chincoteague Bays. .............................................................. 79
Figure 4.11 (a): Chlorophyll-a verification with DNR data Stations 1-12, 2001-2005. ........ 80
Figure 4.11(b): Chlorophyll-a verification with DNR data Stations 13-24, 2001-2005. ....... 81
Figure 4.11 (c): Chlorophyll-a verification with DNR data Stations 25-27, 2001-2005. ....... 82
Figure 4.12 (a): Statistical comparison of observed versus modeled Chlorophyll-a data for
Assawoman and Isle of Wight Bays. ....................................................................... 83
Figure 4.12 (b): Statistical comparison of observed versus modeled Chlorophyll-a data for
Newport, and Chincoteague Bays. .......................................................................... 84
Figure 4.13 (a): SOD calculation at DNR Stations 1-12, 2001- August 2005. .................... 85
Figure 4.13 (b): SOD calculation at DNR Stations 13-24, 2001- August 2005. .................. 86
Figure 4.13 (c): SOD calculation at DNR Stations 25-27, 2001- August 2005. .................. 87
Figure 4.14 (a): DO verification at ASIS Stations 1-9, 2001- August 2005. ........................ 88
Figure 4.14 (b): DO verification at ASIS Stations 10-18, 2001- August 2005. .................... 89
Figure 4.15 (a): Chlorophyll verification at ASIS Stations 1-9, 2001- August 2005. ........... 90
Figure 4-15 (b): Chlorophyll verification at ASIS Stations 10-18, 2001-August 2005 ......... 91
Figure 5.1: ConMon data measured at Bishop’s Landing, July – August, 2005; before (left)
and after (right) Fast Fourier transformation. ........................................................... 95
iii
Figure 5.2: Low-pass-filtered daily DO, DO saturation, salinity, temperature, and
Chlorophyll-a data (left); measured (blue) and fitted sinusoidal curve (red) diel cycle
of DO (right). ............................................................................................................ 98
Figure 5.3: DO diel cycle amplitude adjustments for seasonal temperature cycle
proportioned to monthly mean temperature. .......................................................... 100
Figure 5.4 (a): DO daily average (black) and diel (green) time series, Stations XDN7545
and XDN4851, Assawoman Bay, January 1 – December 31 2004. ...................... 101
Figure 5.4 (b): DO daily average (black) and diel (green) time series, Stations XDN4486
and XDN3724, St. Martin River, Isle of Wight Bay,
January 1 – December 31, 2004. .......................................................................... 102
Figure 5.4 (c): DO daily average (black) and diel (green) time series, Stations XDN3445
and TUV0019, Isle of Wight Bay open waters, January 1 – December 31 2004. . 103
Figure 5.4 (e): DO daily average (black) and diel (green) time series, Stations ASSA 4 and
ASSA 3, Newport Bay, January 1 – December 31, 2004. ..................................... 105
Figure 5.4 (f): Daily average (black) and diel (green) time series, Stations XCM1562 and
ASSA 7, Chincoteague Bay January 1 – December 31, 2004. ............................. 106
Figure 5.5 (a): Correlation coefficient (R2) between diel-cycle-adjusted modeled and
observed DO, DNR Stations 1-12, baseline conditions,
January 2001- August 2005................................................................................... 107
Figure 5.5 (b): Correlation coefficient (R2) between diel-cycle-adjusted modeled and
observed DO, DNR Stations 13-24, baseline conditions,
January 2001- August 2005................................................................................... 108
Figure 5.5 (c): Correlation coefficient (R2) between diel-cycle-adjusted modeled and
observed DO, DNR Stations 25-27, baseline conditions,
January 2001- August 2005................................................................................... 109
Figure 5.5 (d): Correlation coefficient (R2) between diel-cycle-adjusted modeled and
observed DO, ASIS Station 1-9, baseline conditions,
January 2001- August 2005................................................................................... 110
Figure 5.5 (e): Correlation coefficient (R2) between diel-cycle-adjusted modeled and
observed DO, ASIS Stations 10-18, baseline conditions,
January 2001- August 2005................................................................................... 111
Figure 5.6: Sensitivity test comparing the effect of settling velocities of diatom species on
Chlorophyll-a concentration at station AYR0017 in 2004. ..................................... 115
Figure 5.7 (a): Location of Johnson Bay in the middle of the Chincoteague Bay and the
nearby DNR stations XBM5932 and XBM3418. .................................................... 117
Figure 5.7(b): Comparison of modeled total nitrogen concentrations under groundwater
release at edge-of-stream vs. bay-floor release, Johnson Bay area,
Chincoteague Bay, 2001-2005. ............................................................................. 118
Figure 6.1: Response of six stations having the highest DO exceedance rates to the
incremental reductions, MPAR and natural conditions scenarios. ......................... 125
Figure 6.2 (a): DO time series, 2001- August 2005, baseline conditions (blue), 60%
reduction scenario (magenta), and observed (symbol) at DNR Stations 1-12. ...... 127
Figure 6.2 (b): DO time series, 2001- August 2005, for baseline conditions (blue), 60%
reduction scenario(magenta), and observed (symbol) at DNR Stations: 13 -24. ... 128
Figure 6.2 (c): DO time series, 2001-August 2005 for baseline conditions (blue), and 60%
reduction scenario (magenta), and observed (symbol) at DNR Stations: 25 - 27. . 129
iv
Figure 6.2 (d): DO time series, 2001-August 2005, baseline conditions (blue), 60%
reduction scenario (magenta) and observed data (symbol) at ASIS Stations:1-9. 130
Figure 6.2 (e): DO time series, 2001-August 2005, baseline conditions (blue), 60%
reduction scenario (magenta), and observed data (symbol)
at ASIS Stations 10 -18. ........................................................................................ 131
Figure 6.3 (a): DO time series, 2001-August 2005, baseline conditions (blue), final TMDL
(red) and observed (symbol) at DNR Stations 1 -12. ............................................. 137
Figure 6.3 (b): DO time series, 2001-August 2005, baseline conditions (blue), final TMDL
(red) and observed (symbol) at DNR Stations 13 -24. ........................................... 138
Figure 6.3 (c): DO time series, 2001-August 2005, baseline conditions (blue), final TMDL
(red) and observed (symbol) at DNR Stations 25-27. ............................................ 139
Figure 6.4 (a): Chlorophyll-a time series, 2001-August 2005, baseline conditions (blue),
final TMDL (red) and observed (symbol) at DNR Stations 1 -12............................ 142
Figure 6.4 (b): Chlorophyll-a time series, 2001-August 2005, baseline conditions (blue),
final TMDL (red) and observed (symbol) at DNR Stations 13 -24.......................... 143
Figure 6.4 (c): Chlorophyll-a time series, 2001-August 2005, baseline conditions (blue),
final TMDL (red) and observed (symbol) at DNR Stations 25 -27.......................... 144
v
EXECUTIVE SUMMARY
Shallow coastal bays and lagoons are important buffer zones between terrestrial and
deeper coastal ecosystems. They are inherently vulnerable to eutrophication,
particularly from anthropogenic influences. The Maryland Coastal Bays (MCBs) is a
collection of shallow coastal basins including Assawoman Bay, Isle of Wight Bay,
Sinepuxent Bay, Newport Bay and Chincoteague Bay adjacent to the Delmarva
Peninsula of the US East Coast. The MCBs have shown signs of water quality
degradation in recent years. Extensive monitoring has been conducted, demonstrating
low dissolved oxygen (DO) and high levels of chlorophyll a. It was determined that the
MCBs water quality conditions exceed the State’s water quality standards, and the
MCBs were placed on the State’s 303(d) List of impaired water bodies in 1996.
In the current effort, a Hydrodynamic Eutrophication Model 3-D (HEM3D) was
developed and used as a tool to simulate the dynamics of physical-biological-chemical
processes in the receiving MCBs waters, using the nutrient loads generated by the
MCBs Hydrologic Simulation Program-FORTRAN (HSPF) watershed model.
The HEM3D modeling system was calibrated and compared exceedingly well with the
intensive field data including water level, current velocity, salinity, chlorophyll a,
dissolved oxygen (DO), ammonia nitrogen, nitrate nitrogen, phosphate and dissolved
organic nitrogen collected during 2001- 2004 by the Maryland Department of Natural
Resources (MD-DNR) and the US National Park Service/Assateague Island National
Seashore (ASIS). The calibrated and verified hydrodynamic model was used to
determine the physical transport time scales for the entire system to be on the order of
2-3 months. Most of the system is nitrogen limited except in the headwaters of the
tributaries where the phosphorus and nitrogen can be co-limiting depending on the flow
and/or season. The predicted daily mean DO from the HEM3D was further adjusted for
the diel cycle using Elgin Perry’s statistical analysis results (2012). The empirical
corrections for diel cycle were made to the daily mean DO by monthly temperature,
daily temperature, daily Photosynthetically Active Radiation (PAR), and daily chlorophyll.
In doing so, the DO variability includes the diel cycle, which provides a better
representation of full spectrum of DO levels in the MCBs. Following calibration and
verification of the HEM3D hydrodynamic and water quality model, sensitivity analyses
were conducted to test the effects of (a) Ocean City wastewater treatment plant outfall,
(2) phytoplankton and organic nutrient settling rate and (c) ground water discharge. The
model was used to evaluate point and nonpoint source loading allocations and
reduction scenarios.
The TMDL endpoint for DO requires that daily mean DO concentrations simulated at the
model cells corresponding to the water quality stations shall not be below 5 mg/l more
than 10 percent of the time, both annually and in the growing season (May 1 – October
31). The TN and TP sources were assessed for the five impaired basins in the MCBs.
It was found that for Assawoman, Isle of Wight, and Newport Bays, the terrestrial
sources are the dominant source of loading, whereas in the Sinepuxent and
Chincoteague Bays the terrestrial source loading are about equal to that of atmospheric
vi
loading. In determining a final TMDL scenario for management action in MCBs,
incremental reductions were conducted. Based on the incremental reduction scenarios
including 20%, 40%, 60%, Maximum Practicable Anthropogenic Reduction (MPAR) and
the natural condition, the northern Bays (north of Ocean City), particularly in the Saint
Martin River, area appear to require the most reductions. The southern portion of the
Bay only requires minor reduction. Given this spatial disparity, it is obvious that the load
reduction scenario for TMDLs requires including the geographic influences. The
“geographic isolation method” was further conducted and the final TMDL reductions
needed to meet water quality standards are: 20% for Assawoman Bay, 40% for Isle of
Wight Bay (Open Waters), 55-58% for Bishopville Prong/Shingle Landing, 0% for
Sinepuxent Bay, 20% for Newport Bay and for the Maryland portion of the
Chincoteague Bay.
vii
CHAPTER 1: INTRODUCTION
1.1 Background
The Maryland Coastal Bays system (MCBs) is located along the US Atlantic coast of the
Delmarva Peninsula on the eastern edge of the Atlantic coastal plain (Figure 1.1). It
consists of a series of coastal bays including the Assateague Island National Seashore.
These bays span across three states: Delaware, Maryland and Virginia, and are
composed of five major separate basins: (1) Assawoman Bay (2) Isle of Wight Bay (3)
Sinepuxent Bay (4) Newport Bay and (5) Chincoteague Bay, from the north to the south
as shown in Figure 1.1. In 1991 the United States Environmental Protection Agency’s
(EPA) Office of Water Assessment and Protection Division published Guidance for
Water Quality Based Decisions: the TMDL Process. In 1992, EPA published the final
Water Quality Planning and Management Regulation (40 CFR Part 130). Together
these documents describe the roles and responsibilities of EPA and the States in
meeting the requirements of Section 303(d) of the Clean Water Act (CWA). Section 303
(d) requires States to: (1) Identify waters that are and will remain polluted after the
application of technology standards; (2) Prioritize these waters, taking into account the
severity of their pollution; and (3) Establish TMDLs for these waters at levels necessary
to meet applicable water quality standards, taking into account seasonal variations and
a protective margin of safety (MOS) to account for uncertainty.
The Maryland Department of the Environment is required to identify waters that are
impaired by pollutants. The MCBs were identified in MD’s 2010 Section 303(d)
Impaired Waters list as impaired and the listings are shown in Table 1.1. The list
highlights nitrogen and phosphorus as the major pollutants in the watershed, stating
water quality standards are not being met because of excess loads of these substances
(Maryland Department of the Environment, 2001). The water quality of this system is
considered degraded, as evidenced by substantial increase of nutrients, seasonal
hypoxia, macroalgae biomass in areas, annual blooms of brown tide (Wazniak and
Gilbert, 2004), and are projected to experience environmental stress due to increased
population and intense development (Boynton et al.,1996; Wazniak et al.,2004;
Maryland Department of the Environment, 1993). Maryland Department of Natural
Resources (2004) documents the most up-to-date (up to 2004) status of the water
quality and living resources in the Coastal Bays. Wazniak et al. (2005) further
investigated the overall ecosystem health including using high frequency DO
measurements to identify diel cycle of DO and the impact of sediment influence from the
shoreline erosion. Maryland Department of the Environment (2004) identified the
priority areas for wetland restoration. Wazniak et al. (2007) linked the water quality
condition to the living resources and Dennison et al. (2009) characterized the physical,
chemical, and biological resources of the Coastal Bays and serves as an additional step
in providing sound management for the future.
1
Figure 1.1: Location map of MD Coastal Bays System
2
Table 1.1: TMDL impaired water listing for the Maryland Coastal Bays (MDE 2010)
Year
listed
1996
Watershed
Assawoman
Bay
Basin
Code
02130102
2010 IR Assessment
Unit ID
MD-02130102-TAssawoman_Bay
MD-02130102-TGreys_Creek
MD-02130103-TTurville_Creek
MD-02130103-TManklin_Creek
1996
1996
1996
1996
Isle Of Wight
Bay
Newport Bay
Sinepuxent
Bay
Chincoteague
Bay
02130103
MD-02130103-THerring_Creek
MD-02130103-TBishopville_Prong
MD-02130103-TStMartin_River
MD-02130103-TShingle_Landing_Prong
02130104
MD-02130103-TIsle_Of_Wight_Bay
MD-02130105-TNewport_Creek
MD-02130105-TMarshall_Creek
MD-02130105-TKitts_Branch
MD-02130105-TAyer_Creek
MD-02130105-TNewport_Bay
MD-02130104-T
02130106
MD-02130106-T
02130105
Specific Area
Open water
Grey’s Creek
Turville Creek
Manklin
Creek
Herring Creek
Bishopville
Prong
St. Martin
River
Shingle
Landing
Prong
Open Water
Newport
Creek
Marshall
Creek
Kitts Branch
Identified
Pollutant
Nitrogen
Phosphorus
Nitrogen
Phosphorus
*Listing
Category
5
5
5
5
Nitrogen
4a
Phosphorus
4a
Nitrogen
5
Phosphorus
Nitrogen
Phosphorus
Nitrogen
Phosphorus
Nitrogen
Phosphorus
Nitrogen
5
4a
4a
4a
4a
4a
4a
4a
Phosphorus
4a
Nitrogen
Phosphorus
5
5
Nitrogen
4a
Nitrogen
Phosphorus
Biochemical
Oxygen Demand
5
5
4a
Ayer Creek
Nitrogen
4a
Newport Bay
Nitrogen
4a
Nitrogen
Phosphorus
Nitrogen
Phosphorus
5
5
5
5
Sinepuxent
Bay
Chincoteague
Bay
*Definition of listing category: 4a – TMDL developed; 5 – TMDL required.
The main goal of TMDLs is to obtain a projected distribution of pollutant loading in each
basin to meet water quality standards. In the MCBs, the determination of loadings and
their allocations rely on the water body’s complex interactions among physical, chemical
and biological processes. Statistical models, expressed in simple mathematical
relationships derived from fitting equation to observed data, are usually easy to use and
require minimal effort. One weakness of statistical models is that they tend to have
large standard errors of prediction, especially when the water quality data have
relatively large range of variations spatially. They are most reliable when applied within
the range of observation and in a relatively homogeneous and well-mixed system such
as lakes. When the interactions of flow, loading, internal chemical and biological
3
processes are too complicated to be solved through the use of statistical and data
analysis techniques, mechanistic computer simulation models are often employed.
Mechanistic models are based on physical, chemical and biological mechanisms that
govern the water systems. It is formulated upon equations that contain directly definable
observable parameters. When properly calibrated and verified, mechanistic models are
usually better at representing the physical chemical and biological processes, and
carries higher prediction skill for the relationships between loading and water quality
condition in the water body systems.
1.2 Study Approach
In order to develop TMDLs in the MCBs that consider complex physical and aquatic
biochemical dynamics, a three-dimensional (3-D) hydrodynamic and eutrophication
model is needed to simulate algal dynamics and dissolved oxygen levels and to
determine acceptable pollutant load allocations for nutrients that can result in attaining
water quality standards. A HEM3D (Hydrodynamic and Eutrophication Modeling 3D)
developed by Virginia Institute of Marine Science (VIMS) was selected as the modeling
framework consisting of the hydrodynamic model SELFE, the water quality model
Integral Compartment Water Quality Model (CE-QUAL-ICM or ICM) and the benthic
sediment flux model all of which was used to simulate the receiving waters of Maryland
Coastal Bays (MCBs).
The hydrodynamic model, an unstructured grid, finite element model SELFE (Semiimplicit, Eulerian, Lagrangian Finite Element model), was selected and used to simulate
tidal inlet and estuarine dynamics in the MCBs. The salient feature of the model is that it
uses an unstructured grid to better simulate the coastline and the tidal inlets. In addition,
the model is capable of simulating the wetting-and-drying process, which is a common
feature occurring in the shallow coastal system. Although the model uses a highresolution grid (on the order of 200-m resolution), it still maintains a relatively large time
step without becoming restricted by the CFL (Courant-Friedrichs-Lewy) condition. It
does this by using a special advection scheme known as the Eulerian-Lagrangian
scheme. In this way, the high-resolution model grids can be used to represent a large
model domain without reducing computational efficiency. The model is a general threedimensional model capable of simulating both 2-dimensional (vertically averaged) and
3-dimensional hydrodynamics and transport processes. In the horizontal, the model
uses an unstructured, triangular grid and in the vertical, a terrain following s-coordinate,
a variation of the sigma coordinate with higher resolution on the surface and at the
bottom. The convective terms are treated by the Eulerian-Lagrangian transport scheme
and a semi-implicit method for implementing 3-D equations.
The eutrophication model is a three-dimensional time-variable eutrophication model
package CE-QUAL-ICM (Integral Compartment Model). The model includes both the
water column eutrophication process and the benthic sediment process, which are
dynamically coupled with hydrodynamic and watershed models. The US Army Corps of
Engineers originally developed the model for the EPA Chesapeake Bay Program for
studying the Chesapeake Bay. In the MCBs, the eutrophication model used has twentyone model state variables, which consist of five interacting systems: i.e., phytoplankton
4
dynamics, nitrogen, phosphorus, and silicate cycles, and oxygen dynamics, as shown in
Table 1.2. The water column eutrophication model solves the mass-balance equation
for each state variable and for each model cell. A detailed description of the water
column eutrophication model framework can be found in Cerco and Cole (1994).
A benthic sediment flux model (DiToro and Fitzpatrick, 1993) was coupled with CEQUAL-ICM. There are two layers; the upper aerobic layer (Layer 1) and the lower
anoxic layer (Layer 2) representing the sediment in this model. The sediment process
is coupled with the water column eutrophication model through depositional and
sediment fluxes. The sediment model is driven by net settling of particulate organic
matter from the overlying water column to the sediments (depositional flux). The
mineralization of particulate organic matter in the lower anoxic sediment layer produces
soluble intermediates, which are quantified as diagenesis fluxes. The intermediates
react in the upper oxic and lower anoxic layers, and portions are returned to the
overlying water column as sediment fluxes. Computation of sediment fluxes requires
mass-balance equations for ammonium, nitrate, phosphate, sulfide/methane, and
available silica. Mass-balance equations are solved for these variables for both the
upper and lower layers.
Table 1.2: HEM3D model water quality state variables
(1) Cyanobacteria
(12) Labile particulate organic nitrogen
(2) Diatom algae
(13) Dissolved organic nitrogen
(3) Green algae
(14) Ammonia nitrogen
(4) Refractory particulate organic carbon
(15) Nitrate nitrogen
(5) Labile particulate organic carbon
(16) Particulate biogenic silica
(6) Dissolved organic carbon
(17) Dissolved available silica
(7) Refractory particulate organic
phosphorus
(18) Chemical oxygen demand
(8) Labile particulate organic phosphorus
(19) Dissolved oxygen
(9) Dissolved organic phosphorus
(20) Salinity
(10)Total phosphate
(21) Temperature
(11)Refractory particulate organic nitrogen
5
CHAPTER 2: GENERAL CHARACTERISTICS OF MARYLAND COASTAL BAYS
WATERSHED
2.1 Physical Setting
The Maryland Coastal Bays (MCBs) system is characterized as a coastal lagoon
system linked to the Atlantic Ocean through two inlets: Ocean City inlet in the north, and
Chincoteague inlet in the south. Tidal height is about 1 – 1.3 m at the Ocean City Inlet,
0.5-0.9 m in the Isle of Wight Bay, 0.3-0.5 m in the Assawoman Bay and only 0.1-0.3 m
in the middle of Chincoteague Bay. Pritchard (1960) postulated that water entering the
two inlets meets in the middle and runs back out the inlets, which explained the low tidal
range in the northern Chincoteague Bay. The bays have depths ranging between 0.5 m
– 3 m, with deepest portions reaching 10 m in the Inlets. The flushing rate has been
estimated to be on the order of 10-30 days in the Northern Coastal Bays and 30-100
days in the Chincoteague Bay (Wang, Taiping, 2009). The MCBs are, in general, poorly
flushed with non-stratified condition; thus nutrients and contaminants entering the bays
tend to stay for a long period of time, especially in the Chincoteague Bay.
Water temperatures in the MCBs generally range from 5 to 29o C, with an annual
average of 14o C. In the Northern Bays there is only a small horizontal gradient in
temperature, while in the Chincoteague Bay the temperatures increase toward the
confluence of Newport and Sinepuxent Bays until reaching Ocean City inlet. In the
individual creeks, however, the temperature can exceed 32o C in the summer. Salinities
in the Northern Bays generally decrease with distance from the Ocean City Inlet. The
lower portion of the St. Martin River has high salinities whereas the upper, headwater
regions and the tributaries can be fresh during the spring season, particularly in wet
years. The Chincoteague Bay exhibits fairly high salinity throughout the year in the
main stem of the Bay. There is a longitudinal gradient with salinities decreasing toward
the confluence of Newport and Sinepuxent Bays. The sediments are mostly sandy on
eastern side, silt within the channel, and sand/silt mix along the western shore. The
region receives approximately 40 inches of precipitation annually and the watershed
has traditionally been dominated by farming and forestry.
2.2 The terrestrial and atmospheric loading
The watershed approach adopted below is the logical basis for managing water
resources environmentally. The important relationship between surface areas of water
body and point source, non-point source, atmospheric, shoreline erosion loads on its
watershed cannot be overemphasized. While point sources can have significant effects,
nonpoint source pollutant inputs have increased in recent decades and have degraded
water quality in many aquatic systems.
One of the uncommon characteristics of the Maryland Coastal Bays’ watershed is that
its watershed area is small relative to the water surface area of the receiving bays.
Based on MDE report (1993), the overall average drainage area is only 1.7 times of the
water surface area. For example, the individual basins’ drainage areas and water
surface areas are: 1.2, 7.7, 1.1, 8.6, and 0.7 for Asswawoman Bay, Isle of Wight Bay,
6
Sinepuxent Bay, Newport Bay, and Chincoteague Bay respectively, as shown in Table
2.1. These ratios are much smaller than those in many estuarine systems. For example,
the drainage area to water surface area ratio for the Chesapeake Bay is 14:1 (EPA
Chesapeake Bay Program, 2009). This is especially true for Sinepuxent and
Chincoteague Bays which have the smaller watershed to surface area ratios, and thus
receive smaller runoff from their basins than do the other Coastal Bays. The rest of the
section provides a summary of the sources of nutrient loading into MCBs, which
includes (1) non-point source (2) point source (3) shoreline erosion, and (4) atmospheric
deposition. For the detail of HSPF watershed modeling, see the companion report
VIMS (2013).
Table 2.1: Drainage and water surface area (in acres) of the Maryland Coastal Bays
Coastal
Drainage
Percent of
Water Surface
Percent of
Ratio of
Bays
Area (DA)
Total DA
Area (WSA)
WSA
DA / WSA
Assawoman
6,094
5
5,174
8
1.2
Isle of Wight
36,184
32
4,697
7
7.7
Sinepuxent
6,602
6
5,958
9
1.1
Newport
27,945
25
3,262
5
8.6
Chincoteague
34,718
31
46,592
71
0.7
Total
111,543
100
65,681
100
1.7
2.2.1 Land use distribution
Pollutant loading from the watersheds to the MCBs is primarily a function of land use
and land cover within the Bay’s watershed. The land use in the Coastal Bays
watershed is diverse. Land use information was derived from the Delaware Office of
Planning Land Use Database (2002), Worcester County (Maryland) 2004 Land Use
database (2009), and for Virginia, 1999 National Land Cover Data [United States
Geological Survey (USGS) 1999]. The geographic distribution of the land uses are
shown in Figure 2.1. The aggregated land use categories for the MCBs is shown in
Table 2.2, which shows that forest and other herbaceous growth occupies 30%, mixed
agriculture 29%, water: 26%, urban 12%, and barren or beaches 3%.
7
Figure 2.1: Land use distribution in the Maryland Coastal Bays watershed.
8
Table 2.2: Land use (in acres) for sub-watersheds in MCBs HSPF watershed model
Land use
Assawoman Isle of Wight Sinepuxent
Newport Bay
Bay
Bay
Bay
Chincoteague
Bay
FOREST
4350.53
12921.6
2340.34
11641.21
31565.53
NHI
2055.03
2368.97
91.78
1398.81
4090.4
NHO
346.23
399.12
15.46
235.67
689.14
NHY
223.37
257.5
9.98
152.04
444.61
NLO
7974.4
9192.64
356.16
5427.99
15872.51
HYO
569.6
656.62
25.44
387.71
1133.75
PAS
306.63
194.31
0
32.26
4850.05
BAR
596.51
823.7
883.99
328.58
3891.35
PERVIOUS
5369.82
6073.28
1335.42
2873.86
3281.46
AFO
594.31
134.57
0
0
28.61
IMPERVIOUS 1468.38
3164.36
502.32
1034.48
757.4
WATER
4874.85
1881.66
4869.3
34963.44
7766.41
2.2.2 Non-point source loading
Non-point source loads come from numerous wide-spread locations or sources that
have no well-defined points of origin. These sources are widespread and more difficult
to identify and quantify than point sources and, cumulatively, threaten water quality and
natural systems. The effects caused by non-point sources differ significantly from those
caused by the point sources in their distribution in time and space and often involve
complex transport through soil, water, and air. Examples of non-point sources include
runoff from urban (lawn care, parking lot, golf course, construction), agriculture
(grassland, agriculture operations), pasture, natural forest, animal feeding operation,
and septic tank through ground water. Most non-point sources are directly or indirectly
driven by precipitation; thus, their loadings are inherently dynamic in nature. The carrier
of the pollutants is water as the water runs through the watershed. Therefore,
watershed processes involve detail hydrological description of the discharge of the
water. From modeling point of view, obtaining the discharge from each of the
watersheds is the first task to be carried out; their results required to be calibrated with
known, measured gauge.
In the Maryland Coastal Bays, there are two USGS stream gauges used for the
calibration: one in Birch Branch watershed and the other in Bassett Creek watershed.
They were used for extensive calibration on flows, nitrogen, phosphorus and sediment
loads. Because the Maryland Coastal Bays watershed has extensive animal feeding
operations, the manure was an important source of non-point load and was estimated
from agricultural census. To estimate the nutrient load from the septic tank, the
database from 2000 US census for Virginia, Maryland and Delaware was used. The
9
final nonpoint source loading to the HEM3D water quality model is received from the
HSPF watershed model developed by VIMS (Virginia Institute of Marine Science, 2013).
For full details about the non-point source loading, please see Chapter 5 of the report
entitled, “Maryland Coastal Bays Watershed Modeling Report” by VIMS dated February
2013.
2.2.3 Point-source loading
Point sources are discharges that can be traced back to a specific location at the end of
a pipe. Examples include sewage treatment plants, industrial plants, livestock facilities
etc. Point sources are regulated by state agency and EPA National Pollutant Discharge
Elimination System (NPDES) for pollutants such as BOD, NH4, TKN, suspended solids
and coliform bacteria. In MCBs, point sources were not part of the calibration for the
HSPF watershed model whose focus was mainly on the non-point source loads. To
account for point sources loads, the point source facilities were incorporated as
additional loads into the HEM3D water quality model. The major point sources in MCBs
have four categories: (1) industrial facilities, (2) municipal facilities, (3) injection wells
and (4) facilities using spray irrigation. The major industrial facilities in Maryland include:
Perdue Farm Inc. - Showell Complex; Kelly Foods Corporation, Berlin Properties North,
LLC, and Hudson/Tyson Foods. The major municipal facilities in Maryland are: Ocean
Pines, Assateague Island National Park, Berlin WWTP, Newark WWTP and Ocean City
WWTP, as shown in Table 2.3 (a). The major municipal facilities in Virginia are: US
NASA Wallops flight facility, Sunset Bay utilities – South, US Coast Guard Group,
Eastern Shore, Comfort Suite Hotel – Chincoteague, Hampton Inn and Suites, Sunset
Bay Utilities –North, Chincoteague Landmark WWTP, Taylor landing, and Rays Shanty,
as shown in Table 2.3 (b). As part of NPDES program, permit is issued by EPA and
MDE that sets specific limits on the type and amount of pollutants these facilities can
discharge into receiving waters. The detail of the permitted TN and TP concentration,
design flow, and the permit TN and TP loads are presented in Tables 2.5 (a) and (b). A
complete list of point source including all minor industrial and municipal facilities can be
found in Chapter 6, Tables 16-21 of VIMS (2013).
10
Table 2.3 (a): Major point source facilities in Maryland with permits regulating the discharge of nutrients.
WS-m odel
segm ent
MDE
Perm it #
NPDES #
Facility
Nam e
Period
Flow
TN
TP
TKN
NH3
mgd
mg/l
mg/l
mg/l
mg/l
TN Load
(lbs/yr)
TP Load
(lb/yr)
12176.4
1217.64
1095.88
36.5292
9741.12
1217.64
72,162
9,132
110
11
3,378
375
3,836
639
767,113
127,852
Industrial Facilities
187
76
76
Perdue
95DP0051 MD00009 Farm, Inc. - May-Oct
A
65
Showell
Nov-Apr
Complex
01DP0266
MD00013 Kelly Foods May-Sep
09
Corporation
Oct-Apr
Berlin
Properties May-Oct
MD00020
North, LLC
96DP0375
71
(Hudson/Ty
Nov-Apr
son Foods)
0.8
5
0.5
2
0.8
5
0.5
5
0.02
18
0.6
0.02
18
0.6
0.8
4
0.5
0.8
4
10
4
0.5
2
5
Municipal Facilities
36
174
Ocean
May-Oct
MD00234
Pines
77
Nov-Apr
WWTP
Assateague
MD00210
Island
05DP2530
Jan-Dec
91
WWTP
05DP0708
2.5
3
1.2
2.5
16
1.2
0.012
3
0.3
76
98DP0669
MD00226 Berlin
32
WWTP
Apr-Oct
0
0
0
Nov-Mar
0.6
4.5
0.5
82
05DP0141
MD00206 Newark
30
WWTP
Apr-Oct
0.07
18
3
8
2.3
Nov-Mar
0.07
18
3
8
7.1
outside
05DP0596
MD00200 Ocean City
44
WWTP
Jan-Dec
14
18
3
0.5
11
Table 2.3 (b): Permitted point source facilities in Virginia
Major/
Facility Nam e 1
Minor
Type
SIC Code SIC Nam e
Outfall
Estim ated
Design
Avg. TN
Flow (mgd)
Conc. (mg/l)
Estim ated
TN Lim it TP Lim it TN Load TP Load
Avg. TP
(mg/l)
(mg/l)
(lbs/yr)
(lbs/yr)
Conc.(mg/l)
0.3
18.7
2.5
18.7
2.5
17154.3
2293.4
0.0395
20
15
20
15
2415.7
1811.8
MINOR
Municipal
3769
Guided
Missile and
1
Space
Vehicle
Sunset Bay
MINOR
Utilities - South3
Municipal
5812
Eating
Places
0.006
18.7
2.5
18.7
2.5
343.1
45.9
US NASA Wallops Flight
Facility 2
1
US Coast
Guard Group - MINOR
Eastern Shore2
Municipal
9621
Regulation&
Administrati
on of
1
Transportati
on
Programs
Comfort Suite
Hotel Chincoteague3
MINOR
Municipal
7011
Hotels and
Motels
1
0.009
20
15
20
15
550.4
412.8
Hampton Inn
and Suites 3
MINOR
Municipal
7011
Hotels and
Motels
1
0.01
20
15
20
15
611.6
458.7
Sunset Bay
MINOR
Utilities - North3
Municipal
8811
Private
1
Households
0.025
20
15
20
15
1528.9
1146.7
Chincoteague
Landmark
WWTP4
MINOR
Municipal
4952
Sew erage
Systems
1
0.035
18.7
2.5
18.7
2.5
2001.3
267.6
Taylor Landing3 MINOR
Municipal
7011
1
0.012
20
15
20
15
733.9
550.4
Rays Shanty 3,5 MINOR
Municipal
5812
1
0.0191
20
15
20
15
1168.1
876.1
Hotels and
Motels
Eating
Places
1
Chincoteague Town WTP was eliminated from the analysis since it is a water supply, surface water discharge permit. Therefore, TN/TP concentrations are expected to be de minimis.
Only TSS concentrations from the discharge would be of any significance.
2
US NASA Wallops Flight Facility and US Coast Guard Group - Eastern are both federal facilities. TN/TP concentrations were estimated based on descriptions of the type of
wastewater treatment at the facilities found in a spreadsheet of southeast Virginia treatment plants on VADEQ's website. Outfall 002 at US NASA Wallops Island did not need to be
included in the analysis, since the discharge has been inactive since 1993, well before the model calibration time period.
3
Estimated TN/TP concentrations associated with the wastewater treatment at these hotels/motels and eateries are based on monitored concentrations at similar facilities in Maryland.
4
Estimated TN/TP concentrations associated with the municipal WWTP is based on Virginia's default Bay Phase I WIP value used for minor municipal facilities in order to characterize
the loadings from these facilities, if they were missing data.
5
The Design Flow for Ray's Shanty was missing from the Accomack County 2008 Comprehensive Plan Update, which was used to gather the design flows for all of the other facilities.
Therefore, to estimate a flow for the facility, the average flow of the other hotel/motel and eatery facilities was applied.
6
Average TKN weekly and monthly limits are identified within the actual permits for the facilities; however, no TN or TP limits are specified.
12
2.2.4 Shoreline erosion loading
An important component of the aquatic ecosystem is the sediments, which have a
significant influence on both the biology and chemistry of the ecosystem. Sediment can
either be suspended in a water column or settle and accumulate at the bottom of a
water body. Total suspended sediments can influence water quality and eutrophication
process through effects on density, light penetration, and nutrient availability. Through
the adsorption-desorption process, it can also regulate the particulate and dissolved
component of the chemical species. The sediments from bank loads serve as sources
for pollutants. Bank loads are the solids, carbon, and nutrient loads contributed to the
water column through shoreline erosion. The bank loads from shoreline erosion can act
as a non-point source of nutrients (for TSS, nitrogen and phosphorus), which affect the
water quality in the MCBs. In particular, many pollutants and nutrients introduced into
the bays can accumulate and remain in the sediment beds which functions as a sink.
On the other hand, the sediments can act as a source of pollutants, either through
remobilization of these pollutants by way of natural processes (i.e., diagenetic reactions),
or by physical disturbance or mixing from the sediment bed. Maryland Geological
Survey (MGS) have conducted a multi-year study to determine the flux of sediments
and nutrients eroding from unprotected shorelines bordering Maryland’s Coastal Bays.
The volume of bank eroded material is quantified from comparison of topographic maps
aided by aerial photos separated by time scales of many years. The erosion estimates
are, consequently, averaged over periods of the years separated. Wells et al. (1998,
2002, and 2003) conducted surveys for northern Coastal Bays, middle Coastal Bays,
and Chincoteague Bays respectively, in which a GIS template of irregular polygons was
constructed section by section to determine total sediment, TN and TP loads. An
example estimate of the shoreline erosion in the Northern Bays is shown in Figure 2.2,
in which the historical shorelines dating from 1942 to 1989 was digitized, classified and
inputs into GIS to compared the bank height and quantify losses due to erosion.
Different stretches of shoreline erode at different rates. To account for this variability,
MGS divided the study area shoreline into 18 segments and land loss polygon was
assigned as a number, P#, in the template. Similar procedures for erosion rates were
applied to the entire unprotected shoreline of the MCBs, which become part of source of
the nutrient loads to the HEM3D water quality model. Table 2.4 shows the TN and TP
erosion rate per unit length of shoreline for individual Bays, the total shoreline length for
each of Bays, and thus the sum of the total TN and TP loads for the entire MCBs.
13
Figure 2.2: Locations of sampling sites (in red) and land loss polygons (in blue) in Assawoman and Isle of Wight Bays.
14
Table 2.4: TN and TP loads from shoreline erosion for individual bays and entire MCBs
Shoreline Length
(Model)
ft
113270.01
214437.16
107994.33
60790.22
796386.3
TN Erosion Load
lb/yr
10924.19
19961.4
9062.32
6221.57
145659.05
Total Coastal Bays
1292878.03
191828.54
Total Phosphorus
from Individual Basin
Assawoman Bay in MD
Isle of Wight Bay
Sinepuxent Bay
Newport Bay
Chincoteague Bay
Shoreline Length
(Model)
ft
113270.01
214437.16
107994.33
60790.22
796386.3
TP Erosion Load
lb/yr
1003.42
2169.23
1473.32
830.55
20068.93
1292878.03
25545.47
Total Nitrogen
from Individual Basin
Assawoman Bay in MD
Isle of Wight Bay
Sinepuxent Bay
Newport Bay
Chincoteague Bay
Total Coastal Bays
Erosion Rate
lb/ft/yr
0.0964
0.2009
0.0839
0.1023
0.1829
Erosion Rate
lb/ft/yr
0.0089
0.022
0.0136
0.0137
0.0252
15
2.2.5 Atmospheric loading
Atmospheric deposition is increasingly recognized as a significant external source of
pollutants to surface waters. A pollutant from the air may be deposited into water bodies
and affect water quality in these systems when pollutants are transferred from the air to
the earth’s surface (either land or water) by dry- and wet-weather periods as referred to
as dry or wet deposition. Nutrients forms in precipitation are generally soluble and
those in dry deposition are generally insoluble. Observations of wet depositions are
frequently available through National Atmospheric Deposition Program
(http://nadp.sws.uiuc.edu/nadpdata/). Atmospheric deposition of excess nitrogen can
be a major contributor to eutrophication: increased primary production, algal blooms,
and changes in algal community composition. Nitrogen deposited from the atmosphere
can be a large percentage of the total nitrogen load. In the Chesapeake Bay, 21% of
the nitrogen delivered to the Bay is from the atmosphere including both direct deposition
to the Bay’s water surface and deposition to the watershed that is later transported to
the Bay as runoff. In the MCBs HEM3D water quality model, the atmospheric loads are
part of the non-point source loads and the deposition on the land becomes part of the
allocated load because the air deposition on the land becomes mixed with the nitrogen
loadings from the land based sources and, therefore, becomes indistinguishable from
land based sources. By contrast, the nitrogen deposition directly onto the MCBs’
surface waters is a direct loading onto the surface water, and therefore needs to be
linked directly to the HEM3D water quality model. This is especially important for MCBs,
for their surface water areas relative to drainage areas are large. The time series of the
dry and wet atmospheric deposition was obtained from U.S. EPA Chesapeake Bay
Program Office (EPA-CBP) as a product of NOAA’s airshed modeling (Grimm and
Lynch, 2005). The atmospheric loading calculated by the airshed model was initially as
an input for the HEM3D water quality model. It includes both wet and dry deposition for
nitrogen and phosphorus. The annual averaged TN and TP loads from the airshed
model are estimated as: 13 lb/acre and 0.57 lb/acre, respectively. However, the TN
measurement at the Assateague Island NADP station was only reported only 4 - 6
lb/acre of annual atmospheric deposition (see Figure 2.3), which is a factor of 2 to 3 less,
for the mid-Atlantic Bight region. It was recognized, however, that the observational
measurements recorded only wet deposition. Given the uncertainty about the
magnitude of the dry deposition and there being is no atmospheric TP measurement, it
was determined that a median value of TN (7.42 lb/acre) and TP (0.37 lb/acre) would be
more appropriate to be used as the atmospheric loading for the TMDL after consultation
with the scientists and stakeholders in the Maryland Coastal Bays region. Although
much less than EPA-CBP airshed modeling results, these loading rates were
considered more consistent with the local measurement at Assateague NADP station.
The time series for the atmospherically-derived TN and TP loading was thus adjusted
down by a factor of 0.57 and 0.65 for TN and TP, respectively.
16
Inorganic Nitrogen Wet Deposition
2002
Inorganic Nitrogen Wet Deposition
2003
Figure 2.3: Wet deposition of inorganic nitrogen in 2002 and 2003 in the United States
(from National Atmospheric Deposition Program: http://nadp.sws.uiuc.edu/)
17
2.2.6 Discussion of Contributing Loads to the Maryland Coastal Bays
In the MCBs, the non-point source and atmospheric deposition loads are the two largest
sources of total nitrogen and total phosphorus to the surface water of the MCBs. Figure
2.4 shows the nitrogen load distribution for the entire MCBs. The non-point source
loads from various land uses (13% from urban, 35% from mixed agriculture, and 2%
from forest and barren) of the watershed constitute 50% of the total; the atmospherically
deposited loads 32%, shoreline erosion 8%, and septic tank 8%, and point source 3%.
For the individual basins, the nitrogen load percentage varies significantly, which is
shown in Figure 2.5. For example, Isle of Wight Bay has the largest non-point source
contribution at 70% versus 12% from atmospheric deposition. In contrast,
Chincoteague Bay, which has the largest water surface area to drainage area ratio,
non-point source loads comprise 35% of the total, versus atmospheric deposition at 44%
Figure 2.6 depicts the phosphorus load distribution for the entire MCBs. The non-point
source loads from various land uses (18% from urban, 34% from mixed agriculture, and
2% from forest and barren) of the watershed constitute 54% of the total; atmospherically
deposited loads 23%, shoreline erosion 16%, septic loads 0%, and point source loads
7%. For the individual basins, again, the phosphorus load percentage varies
significantly, which is shown in Figure 2.7. For example, Isle of Wight Bay has the
largest non-point source contribution at 77% versus 9% from atmospheric deposition.
In contrast, Chincoteague Bay, which has the largest water surface area to drainage
area ratio, non-point source loads comprise 35% of the total, versus atmospheric
deposition at 32%
Overall, the relative contribution of atmospheric and septic loads to the total phosphorus
loads is less than that in the case of nitrogen, but the relative contributions of shoreline
erosion and point source loads are greater. Table 2.5 further provides a synthesized
temporal variation of TN and TP loads from 2001-2004 in which 2002 is the driest year
and 2004 is the wettest year during the 4 year period. It can be seen that 2004 has the
largest loading and 2002 the lowest, which is consistent with the characterization of wet
hydrological year for the former and dry for the latter.
For both nitrogen and phosphorus loads, the point source sector represents a small
portion of loads in contrast to the dominance of non-point source loads. Furthermore,
based on the areal loading rate defined as the total loading divided by the total water
surface area, it is clear that Isle of Wight sub-watershed has the largest TN and TP unit
loads (per water surface water area), followed by Newport Bay and Assawoman Bay.
The Sinepuxent and Chincoteague Bays have the lowest. This theme is in the same
vein as that of drainage to surface water area ratio: the larger drainage area leads to
larger non-point source loads. By contrast, a small drainage area to surface water area
ratio leads to smaller non-point source load (relative to the water surface area), which is
the case for Sinepuxent and Chincoteague Bays.
18
Watershed
Urban Mixed Forest/ Point
Septic Atmos- Shoreline
Agri.
barren
source tank
Phere
Erosion
22%
48%
1%
0%
8%
18%
3%
25%
43%
2%
4%
9%
12%
4%
21%
43%
3%
7%
10%
14%
3%
Sinepuxent Bay 24%
Chincoteague
Bay
4%
7%
2%
1%
8%
48%
10%
29%
2%
2%
7%
44%
12%
Entire MCBs
35%
2%
3%
8%
32%
8%
Assawoman
Bay
Isle of Wight
Bay
Newport Bay
13%
Figure 2.4: Total nitrogen load (percentage of total) from source sectors
in individual MCBs
19
TN-lb/y
Watershed
Urban
Mixed
Forest/
Agriculture
Barren
Point Source*
Septic
Atmospheric
Shoreline
Tanks
Deposition
Erosion
total
Assawoman Bay
79,111
172,120
5,071
183
29,883
63,362
10,923
360,653
Isle of Wight Bay
106,633
184,675
7,123
16,459
39,672
51,901
18,729
425,192
Newport Bay
46,188
92,167
6,203
14,207
21,183
30,214
6,221
216,382
Sinepuxent Bay
21,662
6,054
1,671
1,220
6,971
43,396
9,064
90,037
Chincoteague Bay
50,562
356,197
20,934
26,507
86,358
547,573
145,725
1,233,856
Entire MCBs
304,155
811,212
41,001
58,576
184,066
736,446
190,664
2,326,120
Figure 2.5: Total nitrogen loading from source sectors in individual MCBs (lbs/year).
20
Watershed
Assawoman
Bay
Isle of Wight
Bay
Newport Bay
Sinepuxent Bay
Chincoteague
Bay
Entire MCBs
Urban
Mixed
Agri.
Forest/ Point
Septic Atmos- Shoreline
barren source tank
phere
erosion
31%
50%
2%
0%
0%
13%
4%
33%
31%
33%
42%
41%
6%
2%
4%
2%
6%
8%
0%
0%
0%
0%
9%
11%
35%
7%
6%
24%
6%
18%
27%
34%
2%
2%
9%
7%
0%
0%
32%
23%
24%
16%
Figure 2.6: The total phosphorus load (percentage of total) from source sectors in
individual MCBs.
21
TP-lb/y
Watershed
Urban
Mixed
Forest/
Point
Septic
Atmospheric
Shoreline
Agriculture
barren
Source*
Tanks
Deposition
Erosion
total
Assawoman Bay
7,376
11,939
433
0
0
3,167
1,008
23,924
Isle of Wight Bay
9,861
12,427
607
1,837
0
2,594
2,196
29,523
Newport Bay
4,407
5,927
529
1,081
0
1,510
833
14,287
Sinepuxent Bay
2,060
388
143
0
0
2,169
1,469
6,229
Chincoteague Bay
5,069
22,727
1,704
7,863
0
27,367
20,078
84,809
Entire MCBs
28,773
53,409
3,415
10,781
0
36,807
25,585
158,771
Figure 2.7: Total phosphorus load from source sectors in individual MCBs (lbs/year)
22
Table 2.5: Total baseline nitrogen and phosphorus loads (lbs/year), MD Coastal Bays 2001-2004.
23
2.3 Data to Support the Water Quality Modeling
The calibration and verification of a coupled 3-D hydrodynamic and water quality
modeling system requires sufficient field observation data to quantify an acceptable
level of verification such that confidence is established for use of the modeling system
for evaluating various load and waste load allocation scenarios. The tidal data obtained
from the tidal gauge station, used for water level calibration in the hydrodynamic model
is maintained by USGS at the Ocean City Inlet. The USGS gauge stations in the Birch
Branch and Bassett Creek were used for generating non-point source flow as the
hydrological inputs. The weather data including wind, cloud cover and precipitation,
were obtained from Ocean City Municipal Airport. The topographic data and the
shoreline erosion data were provided by Maryland Geological Survey. Table 2.6
summarizes all the monitoring program and observation data which support the
hydrodynamic and water quality modeling.
The Coastal Bays Eutrophication Monitoring Program, led by Maryland Department of
Natural Resources (DNR), is a cooperative program between State and federal
agencies as well as Universities, which started to measure a variety of ecosystem
variables and indices in 2001. The monitoring data collected in the Coastal Bays are
used to assess the conditions of natural resources and to track the trend changes over
time. The information is vital for evaluating the progress of management actions aimed
at restoring the Coastal Bays and their tributaries, for determining attainment of water
quality criteria and for providing guidance on future actions. Monitoring data are also
used for research and for the calibration and verification of the model in the MCBs
ecosystem. There is a network of twenty-seven (27) fixed stations routinely monitored
by DNR, of which twenty (20) stations are in the Northern Bays and seven (7) in the
Southern Bays of the MCBs as shown as DNR stations in Figure 2.8. The major water
quality parameters collected once a month include: salinity, temperature, dissolved
oxygen, chlorophyll-a, total nitrogen, total phosphorus, dissolved organic carbon, TSS,
Secchi depth, ammonia, nitrate, phosphate, and silica, which are used for water quality
model calibration and verification. In addition, the National Park Service, Assateague
Island (ASIS) has maintained eighteen (18) stations in the Sinepuxent, Newport and
Chincoteague Bays, shown as ASIS stations in Figure 2.8. The frequency that ASIS
collected is similar to DNR stations, but the parameters, analytical methods and the
vertical profile are slightly different. Each of these data sources was useful in the model
development and verification processes. In a bay-wide comparison of surface and
bottom water quality parameters with model results, both DNR and ASIS data were
used. On the other hand, for the southern Bay investigation, beside only ASIS has data
coverage in Sinepuxent Bay, DNR data was strictly located in the middle channel, while
ASIS data has the coverage on both eastern and western side of the Bay. On an
investigation in the Northern Bays, such as Saint Martin River, DNR data provided the
monthly DO data from 2000-2004 continuously, as shown in Figure 2.9, which showed
severe hypoxia in several of the stations in the Saint Martin Rivers particularly for 2003,
2004 and 2005. In the Southern Bays, the combined ASIS and DNR stations showed a
minor hypoxia problem in XCM4878, as shown in Figure 2.10, which was quite different
from those seen in the Isle of Wight Bay. The monitoring data was further integrated in a
broad basis for assessing nutrient impacts in the Maryland Coastal Bays. The DNR and
24
ASIS monitoring program are a great asset to the Maryland Coastal Bays; combined
they provided the scientific basis for safeguarding the health of the Bay.
Table 2.6: Summary of data supporting the hydrodynamic and water quality modeling.
Data type
1. Water quality fixed station
monitoring data
2. Water quality data at
Chincoteague Bay
3. DATAflow high frequency data
4. Topographic data
5. Wind field data
6. PAR (Photosynthetic Active
Radiation) data
7. Stream gauge data
8. Precipitation and cloud cover
9. ADCP current meter data
10. Water quality open boundary
condition
11. Airshed atmospheric
deposition
12. Point source data
13. Non-point source data
14. shoreline erosion data
Source
Modeling Support
MDDNR
Water quality model
ASIS
MDDNR
MGS (Maryland
Geological Survey)
Ocean City municipal
airport
Water quality model
Diel cycle adjustment
University of Maryland
USGS
Ocean City Municipal
airport
University of Maryland
EPA region III
EPA Chesapeake Bay
Program
MDE (MD Dept. of the
Environment)
MDE
MGS
Hydrodynamic model
Hydrodynamic model
Phytoplankton
dynamics
Hydrodynamic/water
shed model
Watershed model
Hydrodynamic model
Water quality model
Water quality model
Watershed model
Watershed model
Water shed model
25
DNR
stations
ASIS
stations
•ASS
Figure 2.8: Station location for DNR and ASIS monitoring program in MCBs
26
Figure 2.9: DNR monthly DO data from stations in Saint Martin River (2000-2004)
27
Figure 2.10: Combined DNR and ASIS monthly DO data from stations in Newport Bay
(2000-2004)
28
CHAPTER 3: THE THREE-DIMENSIONAL HYDRODYNAMIC MODEL
Hydrodynamics is the study of fluid in motion; specifically, the motion and the force
acting on water. In this project, the hydrodynamic model of Maryland Coastal Bay
provides the water velocities, tidal elevation, circulation patterns as well as temperature,
salinity, density stratification, and dispersion to drive the eutrophication model.
Hydrodynamic processes are an integrated component of the complex ecosystem of the
MCBs, which is essential to provide the transport and mixing mechanism for ecological
variables.
3.1 Model description
The MCBs are characterized as a system of coastal lagoons connected to the Atlantic
Ocean by two inlets: Ocean City Inlet in the north and the Chincoteague Inlet in the
south. In order to simulate the MCBs hydrodynamics properly, the model must be able
to resolve the two narrow inlets accurately. Among different models, the finite element
model uses unstructured, triangular grids and has the advantage that its grid resolution
can be flexibly refined in a local region while preserving coarse resolution elsewhere,
thus is most suitable model for inlets in systems such as the MCBs. Due to this reason,
the Semi-implicit, Eulerian, Lagrangian, Finite Element Model (SELFE) was selected to
model hydrodynamics in the MCBs. SELFE is a multifunctional surface water
hydrodynamic model which can be coupled with surface wave, sediment transport, and
eutrophication models, and is currently maintained at Virginia Institute of Marine
Science as a COASTAL OCEAN community model:
http://ccrm.vims.edu/w/index.php/About_SELFE
The SELFE model, depending on its application, can be set up for use in 1-, 2-, and 3dimension applications. The model utilizes an unstructured triangular grid in the
horizontal and hybrid terrain-following S-Z coordinates in the vertical. It uses an
efficient semi-implicit time stepping in conjunction with an Eulerian-Lagrangian method
(ELM) to treat the advection. As a result, numerical stability is greatly enhanced and
errors from the “mode splitting” method are avoided; in fact, the only stability constraints
are related to the explicit treatment of the horizontal viscosity and baroclinic pressure
gradient, which is much milder than the stringent CFL (Courant, Friedrichs, and Lewy)
condition. The default numerical scheme is 2nd-order accurate in space and time, but
optional high-order schemes have been developed as well [e.g., the dual Kriging ELM
proposed by LeRoux et al., (1997)]. The model also incorporates wetting and drying
naturally as part of the semi-implicit scheme, and has been rigorously benchmarked for
inundation problems such as tsunami and storm surge simulation (Zhang et al. 2011;
NTHMP 2011). As an open-source community-supported model, SELFE has been well
demonstrated to be accurate, efficient, robust and flexible, with a wide range of
applications from general circulation (Brovchenko et al. 2011), tsunami inundation
(Zhang et al. 2011), storm surge (Bertin et al., 2012), ecology (Rodrigues et al. 2009),
oil spill (Azevedo et al. 2009), and water quality studies. Details of the SELFE model’s
hydrodynamic capabilities are provided in Appendix A.
29
3.2 Model set-up
The general procedure for the application of the SELFE model in the MCB follows a
sequence of steps beginning with model set-up. Model set-up involves (1) the
construction of a horizontal triangulated grid of the water body; (2) interpolation of
bathymetric data to the grid; (3) generation of the initial and boundary conditions (4)
construction of SELFE input files, (5) selection of model parameters and (5) compilation
of the paralleled source code with the appropriate computer platform to execute the
application. The SELFE input files include the master input file (param.in), files
specifying the grid and bathymetry (hgrid.gr3 and vgrid.in), atmospheric forcing files
(wind.th and sflux), inflow-outflow file (flux.th), salinity and temperature boundary
condition, inflow concentration files (salt.ic and temp.ic), and a hotstart input file
(hotstart.in).
The horizontal grid for the MCBs used unstructured, horizontal triangular grid cells and
was constructed using a utility tool: xmgredit. Figure 3.1 shows the grid of the entire
MCBs model region from Assawoman Bay to Chincoteague Bay including the Ocean
City and Chincoteague Inlets, and the surrounding inner continental shelf. The
horizontal grid has 12,428 active cells, of which 8,348 are inside the MCBs and 4,080
are in the coastal ocean. The horizontal coordinate system used by the model is a
localized UTM system. The water depth data collected by the Maryland Geological
Survey (MGS) was interpolated to the horizontal model grid using an arithmetic average
of all data points falling within a specific cell. The bathymetry of MCBs presented in
Figure 3.2 (using mean sea level as datum) shows most of the bay is shallower than 3
m except near the inlet where it can reach 10 m. The model vertical grid utilizes 5 sigma
layers and has varying thicknesses throughout the horizontal model domain.
3.3 Hydrodynamic model forcing functions
Hydrodynamics in the MCBs model are forced by a tidal open boundary at the inner
continental shelf (about 10 km offshore), the point and non-point source inflows inland,
and the local wind surface wind stress. The hourly tidal elevation open boundary
condition was obtained from the interpolation of NOAA’s observed water level at Ocean
City Inlet and Wachapreague, VA. The monthly salinity and temperature was applied at
the Open boundary condition. Inflows include the upstream river inflow calculated from
the HSPF model at the following Creeks: Greys Creek, Bishopville Prong, Shingle
Landing Prong, St. Martin River, Herring Creek, Turville Creek, Manklin Creek, Ayer
Creek, Newport Creek and Marshall Creek. The Ocean City WWTP has two outfalls
and was designed for a total of 32 MGD capacity with one for 20 MGD and the other 12
MGD. These outfalls are at 10 m depth in a distance about 1.4 km from shore and
within the model’s boundary. The discharge rate provided by the Ocean City Public
Works was used as an inflow discharge.
30
Figure 3.1 Hydrodynamic model grid in the Maryland Coastal Bays (MCBs).
31
Figure 3.2: Bathymetry for the Maryland Coastal Bays
Atmospheric forcing functions for the model were developed from National Climatic
Data Center (NCDC) records from Ocean City Municipal Airport and included wind
speed and direction, atmospheric pressure, air temperature, relative humidity, rainfall,
and cloud cover at approximately hourly intervals. Wind speed and direction are used
internally in the model to provide surface wind stress forcing, while wind speed is used
in the prediction of water surface latent and sensible heat exchange. Wind speed at 10
meter height is also used in determination of surface reaeration rates in the
eutrophication component of the model. Wind speeds were internally adjusted in the
model using input directional sheltering coefficients determined during the calibration.
All model forcing data were assembled for a 4 year period spanning 2001 through 2004.
Initial conditions for the hydrodynamic model included a constant water surface
elevation corresponding to mean water level, the water temperature and salinity
representative of early January 1999. The model was initialized for the entire year of
2000, and the results saved and used as the initial condition of the year 2001.
32
3.4 Hydrodynamic model calibration
The MCBs hydrodynamic model was calibrated for water level, current velocity, and
salinity by comparing model simulation results with field measurements. The calibration
was an iterative process, and parameters for bottom friction and stability function (inside
the turbulence closure model) were adjusted to produce reasonable tidal elevation and
stratification results. Once calibrated, the parameters were not changed and the model
was validated with results from runs in different years. The field observation stations
and the locations for current, wave level and salinity were shown in Table 3.1, and the
map of the field observation stations for water elevation, currents, and salinity in the
MCBs is shown in Figure 3.3.
3.4.1 Water level calibration
There is only one permanent active water level monitoring station maintained by
National Oceanic and Atmospheric Administration (NOAA) at Ocean City Inlet, MD,
(NOAA Station #08570283) inside the MCBs. In order to have enough coverage of
different tidal conditions in the MCBs, the astronomical tidal prediction result produced
by XTIDE (http://tbone.biol.sc.edu/tide/index.html) was used to characterize the spatial
variability of the tidal range (shown at station W1-W11 in Table 3.1). The tide generated
by XTIDE is solely the astronomical tide, which does not include effects of meteorology
condition such as pressure and wind stress on water level. To calibrate surface
elevation changes induced by astronomical tides, the SELFE model was run in a twodimensional vertically integrated mode. The model calculations results of tidal elevation
were compared with XTIDE predictions at nine stations, as shown in Figure 3.4 (a) - (c),
throughout the MCBs. The model results match XTIDE predictions very well except for
the station at Assateague Beach at Tom’s Cove (see Figure 3.4 (b)) where the model
under-predicted the tidal range. Considering from shoreline data collected which
showed around this area is continuously changing, the under-prediction is probably due
to the inaccuracy of the local geometry at this station as represented in the model.
Similar to many inlet systems, the tidal signal dampens quickly as it propagates from the
coastal ocean into the inlets. In the MCBs, the tidal range decreases significantly from
the regions outside the inlets in the Atlantic Ocean (~1.6 m) to less than 0.2 m at the
Public Landing Station, located in the Northern Chincoteague Bay as shown in Figure
3.4 (c). In contrast, Figure 3.4 (a) shows the M2 tidal range from the coastal ocean into
the Ocean City inlet. Tidal range also show similar decrease trend from coastal ocean
into the inlet, but the magnitude of the decrease is less than that in the southern portion
of the MCBs. Along the transect from Chincoteague in the south and Sinepuxent Bays
in the north, the semi-diurnal tidal amplitude show a bi-model features with lowest
amplitude about 0.1 m at the public landing and highest amplitude at Chincoteague and
Sinepuxent at about 0.5 m. The M2 tidal phase, on the other hand, shows a uni-model
distribution with the highest phase at about 150 degree in the public landing.
33
Table 3.1: Description and location of field observation stations for current, water level
and salinity in the MCBs.
Station
ID
Name
Type
Source
Latitude
Longitude
C1
C2
C3
Isle of Wight Bay Channel
Sinepuxent Bay Channel
Chincoteague Bay Inlet
Current
Current
Current
MDNR
MDNR
MDNR
38.3311
38.3237
37.8826
-75.0920
-75.1001
-75.4142
W1
Ocean City (Fishing Pier)
Water Level
XTIDE
38.3267
-75.0833
W2
Ocean City Inlet
Water Level
38.3283
-75.0917
W3
Ocean City (Isle of Wight
Bay)
North Beach Coast Guard
Station
Assateague Beach, Toms
Cove
Wishart Point, Bogues Bay
Wallops Island
Harbor of Refuge
Public Landing
South Point, Sinepuxent
Neck
Turville Creek
XDN2438
XDN6454
XDN7261
XDN5737
XDN4312
TUV0011
XCM4878
XCM0159
XBM1301
Water Level
XTIDE,
NOAA
XTIDE
38.3317
-75.0900
Water Level
XTIDE
38.2000
-75.1500
Water Level
XTIDE
37.8667
-75.3667
Water Level
Water Level
Water Level
Water Level
Water Level
XTIDE
XTIDE
MGS
XTIDE,
MGS
37.8817
37.8417
37.9033
38.1483
38.2150
-75.4917
-75.4783
-75.4067
-75.2850
-75.1917
Water Level
Salinity
Salinity
Salinity
Salinity
Salinity
Salinity
Salinity
Salinity
Salinity
MDNR
MDNR
MDNR
MDNR
MDNR
MDNR
MDNR
MDNR
MDNR
MDNR
38.3554
38.3546
38.4417
38.4528
38.4283
38.4041
38.3585
38.2457
38.1682
38.0215
-75.1499
-75.0891
-75.0775
-75.0639
-75.1050
-75.1473
-75.1314
-75.2033
-75.2369
-75.3332
W4
W5
W6
W7
W8
W9
W10
W11
S1
S2
S3
S4
S5
S6
S7
S8
S9
34
Figure 3.3: Map of field observation stations for water elevation, currents and salinity in
the MCBs.
35
Figure 3.4 (a) Tidal calibrations near Ocean City Inlet in the northern portion
of the MCBs.
36
Figure 3.4 (b) Tidal calibrations near Chincoteague Inlet in the southern portion of the
MCBs.
37
Figure 3.4(c): Tidal calibrations across the middle and southern portions of the MCBs.
38
Figure 3.5: M2 Tidal profiles along Chincoteague and Sinepuxent Bays transect.
The next calibration is aimed at the comparison of model results with the real water
level observation, which is influenced both by astronomical and meteorological-induced
variation. The real time water level data is collected both by the tidal gauge and the
pressure sensor by the National Oceanic and Atmospheric Administration (NOAA), MD
DNR, and Maryland Geological Survey (MGS). NOAA has a permanent tidal gauge
station inside the Ocean City Inlet. MD DNR deployed a YSI telemetry water quality
data logger at its continuous monitoring station at Turville Creek in April 2005. MGS
conducted field water level measurements at South Point in Newport Bay and Harbor of
Refuge in the Chincoteague Bay in August 2004. The data collected from these 4
stations covered the entire MCBs and were used for model verification bay-wide. In
order to model the real water level variation, SELFE require to input the wind speed,
direction, and pressure data measured at the 10 meter standard height. The wind
39
stress is then caluated by appying the wind drag coefficeint to the wind speed. The
hourly wind data at the two stations are available: one in the north at Ocean City
Municipal Airport, MD and the other in the south at Wallops Island, VA were used.
Figure 3.6 shows the examples of the wind record in August 2004. The modeled results
and observed time series of water level variation is shown in Figure 3.7; the top panel
shows the station and lower panel the comparison. The water level variation now
consisted of not only the component induced by the tide but also that by the wind
displayed as low frequency variatons. It can be seen that the amplitude and phase
matched the observation quite well. The wind-induced set down, shown as examples at
day 269-270 for stations A and C that moves the mean sea level below the mean
averaged 0, was captured. A similar phenomenon occurring at day 96 – 98 at Station D
was also correctly simulated by the model. Spatially, the model also captured the larger
ranges near the two inlet stations and smaller range in the interior of Chincoteague Bay.
Proper portrayal of both time series and spatial variation of water level ranges
demonstrates the predict capability of the model for both astronomical and wind-induced
water level variation.
Figure 3.6: Wind record at Ocean City and Wallops Island, September 23 to October 8,
2004.
40
Figure 3.7: Water level verification at A) Ocean City Inlet;, B) South Point; C) Harbor of
Refuge, and D) Turville Creek.
41
3.4.2 Velocity calibration
The hydrodynamic model velocity calibration was also conducted during the year 2004,
when simultaneous measurements of currents by Acoustic Doppler Current Profilers
(ADCP) and water levels were available. The University of Maryland, Center for
Environmental Science (UMCES) utilized ADCP to monitor velocities at two locations in
the Ocean City Inlet and one site in the Chincoteague Bay (see Figure 3.1). The MGS
recorded water levels at two locations. The current measurement locations are shown
as U1, U2, U3 for current and W8 and W10 for tide. Detailed descriptions of the
instrumentation and methods used by MGS can be found in:
http://www.mgs.md.gov/coastal/pub/FR04_06.html
The ADCP deployed at the Ocean City Inlet was bottom mounted while the one
deployed in the Chincoteague was a towed ADCP. The 1200-kHz ADCP manufactured
by RD instruments has self-contained power supply and data recording and storage.
ADCP data processing divides the measurement into uniform segments called depth
cells or bins. Due to interference caused by side-lobe reflection from the surface and
bottom, the cells for the top two meters and those within 1.5 m of the bottom were not
be considered. Given nearly vertically homogeneous conditions at the inlets and the
fact that the model uses sigma coordinate layer, which does not exactly match the
vertical depth of the ADCP layer, it was decided to compare the vertically averaged
velocity to avoid the potential error introduced by the interpolation of the exact vertical
position, and thus achieve a more realistic result. In coastal waters, the data normally
show a rotating current vector in the presence of the oscillating tide. Around the tidal
inlet, because of its narrow width, the current is then dominated by the longitudinal
component along the main axis of the local channel. Thus, the current velocity derived
from the major axis of the scatter plot was convenient to compare with the modeled
results. A spatial distribution of surface velocity distribution during flood tide was shown
Figure 3.8 (a) near Ocean City Inlet. In Figure 3.8 (b), an un-filtered ADCP time series
measured at the south of the Inlet was compared against the modeled current speed at
a point location with reasonable comparison of current on the order of 0.6 m/sec range
from flood to ebb. Figure 3.9 compares a 15-day model simulation of current velocities
and the ADCP measured velocity at Stations A, B and C from 9/23/2004 – 10/08/2014.
Station A is located slightly north of the Ocean City Inlet at the main channel in the Isle
of Wight; Station B is located south of the Inlet at Sinepuxent Bay. Both comparisons
were satisfactory. Station A has higher maximum current of about 1 m/sec as
compared to that of Station B at about 0.5 m/sec near the Ocean City Inlet. At Station C,
the maximum current is larger than 1 m/sec and the model slightly over-estimated the
velocity, presumably due to the complicated geometry and the abundant submerged
aquatic vegetation (SAV) present at the mouth of Chincoteague Bay. The statistical
measures of the comparison areas follow: R2 = 0.94, 0.81, 0.91; and relative error =
8.96%, 12.94%, and 12.01% for Stations A, B. and C, respectively. These measures
were shown in Figure 3.10.
42
(a)
(b)
Figure 3.8: (a) Spatial distribution of surface velocities during maximum flood tide at
Ocean City Inlet, October 2004; and (b) Current speed time series comparison along the
major axis of the local channel; ADCP measurement (red) and model results (blue).
43
Figure 3.9: Current calibration results at A) Isle of Wight Channel; B) Sinepuxent Bay Channel; c) Chincoteague Channel;
model (red solid line) and data (blue dash line).
44
Modeled (m/sec)
RMS=0.14
E=8.96%
2
R = 0.94
(A)
Observed (m/sec)
Modeled
RMS=0.15
E=12.01%
2
R = 0.81
(B)
Observed (m/sec)
Modeled (m/sec)
RMS=0.18
E=12.94%
2
R = 0.91
(C)
Observed (m/sec)
Figure 3.10: Statistical measures for current comparison at calibration stations A, B. and
C, respectively.
45
3.4.3 Salinity calibration
Salinity is a measure of the salt concentration in the water which can affect the density
of water. Because the salinity satisfies the law of mass conservation, it is normally used
as a conservative tracer to provide verification of the transport in the model. The saline
waters from the coastal ocean, passing through Ocean City and Chincoteague Inlets,
move into the MCBs and mixed with the fresher water discharged from the terrestrial
inputs. These terrestrial inputs of the freshwater from 2001-2004 were generated by
HSPF model, which is calibrated with measurements from USGS stream gauges at
Birch Branch and Bassett Creek. Figure 3.11 (a) shows the daily discharge in 2004 for
the Saint Martin River watershed, in which multiple freshwater input events in the form
of pulsation can be clearly seen over the course of the year, for example, at Julian days
35, 105, 215, 230, 320, 330, and 345. Associated with the freshwater events are the
low salinity regimes observed and simulated by the model, as shown in Figure 3.11 (b)
in the northern MCBs’ tributaries. It was seen that each freshwater event created a
salinity drop whose top-to-bottom variation can be as large as 20 ppt, for example, at
Saint Martin River (Station S5) and Turville Creek (Station S6), and with a lesser range
at Greys Creek (Station S4). Similar trends were found at Stations S1 and S2 in the
Assawoman Bay [see Figure 3.11(c)], but with lesser variation in Station S3, which is
close to the Ocean City inlet, and more controlled by the ocean water, and thus shows
less of a drop in salinity compared to the upstream stations. The observed and modeled
salinities exhibit reasonable good agreement. Down to the southern Bay, observed and
modeled salinities in the southern MCBs are compared at stations S7, S8 and S9, as
shown in Figure 3.11 (d). For the station S7, one can see the effect of the freshwater
pulse downstream of the Newport Bay can cause variation of salinity on the order of 10
ppt. The salinity in the middle and southern portions of the Chincoteague Bay, however,
was less affected by the freshwater runoff and the range of the temporal variation of
salinity is smaller. At station S9, model over-predicted the salinity in the spring and in
the fall while under-predicted it in the early summer with mean absolute error around 22.5 ppt. The salinity variation in the southern Chincoteague Bay does not appear to be
influenced directly by the freshwater discharge from the Newport Bay in the north, which
suggests there are other processes at work. One possible mechanism is the ground
water recharge studied by USGS (http://soundwaves.usgs.gov/2002/06/research.html).
Overall, the modeled salinity compared well with the monthly measurement which
reproduced field observations throughout the entire domain from the north to the south.
Considering the total freshwater discharge in the MCBs system is small compared to
other systems such as Chesapeake Bay, the episodic rainfall events still can have
substantial impacts on the salinity field of the creeks in the tributaries. Because the
fresh water influence, there is a clear salinity gradient pattern from the main stem of
MCBs toward the upper tributaries - the saltier salinity near the inlets and main stem of
MCBs and decreased as it moved toward upstream of the tributary, as shown in Figure
3.12. Both 2003 and 2004 are wet hydrological years with 2004 having a larger total
amount of freshwater inputs and, thus, the salinity inside the MCBs was the lowest. The
statistics of salinity comparison for 2001- 2003 at all 5 basins in MCBs are shown in
Table 3.2. The relative errors were in the range of 2–5.4% with Sinepuxent having the
lowest error. The highest relative error occurred in Isle of Wight Bay followed by
Newport Bay due to the greater salinity change resulting from the larger freshwater
46
inputs in that basin. The salinity calibration and validation are considered satisfactory –
the errors are within 5% for the majority of the stations - proved to be suitable for further
use in coupling with the water quality model.
(a)
(S4)
(b)
S4
S5
S6
(S5)
(S6)
Figure 3.11 (a) Areally-weighted daily discharge from USGS flow gauges (01484719 and
0148471320) in MCBs, and (b) Salinity calibration in the tributaries of the northern MCBs.
47
Figure 3.11 (c): Salinity calibration in the open waters of the northern MCBs.
48
MDNR XCM4878
Model
S7
Field
35
Salinity (ppt)
30
S7
25
20
15
10
5
0
0
50
100
150
200
250
300
350
Days of 2004
MDNR XCM0159
Model
S8
S8
Field
35
Salinity (ppt)
30
25
20
15
10
5
0
0
50
100
150
200
250
300
350
Days of 2004
S9
MDNR XBM1301
Model
S9
Field
35
Salinity (ppt)
30
25
20
15
10
5
0
0
50
100
150
200
250
300
350
Days of 2004
Figure 3.11 (d): Salinity calibration in the southern MCBs.
49
Figure 3.12: Spatial distribution of the averaged salinity contour in the MCBs over summers of
(a) 2003 and (b) 2004.
50
Table 3.2: Statistical measures for modeled versus observed salinity at five basins in MCBs.
2001
Mean error
Absolute Mean error
Relative error (%)
Asswoman Bay
-0.72
1.27
3%
Isle of Wight Bay Sinepuxent Bay
-0.91
0.6
2.7
1.1
5.00%
2%
Newport Bay
-0.8
2.7
4%
Chincoteague Bay
0.8
1.8
2.60%
2002
Mean error
Absolute Mean error
Relative error
Asswoman Bay
-0.7
1.17
2.80%
Isle of Wight Bay Sinepuxent Bay
-0.8
0.5
2.5
0.9
4.80%
2%
Newport Bay
-0.76
2.5
3.70%
Chincoteague Bay
0.9
1.9
3.00%
2003
Mean error
Absolute Mean error
Relative error
Asswoman Bay
-0.9
1.4
3.10%
Isle of Wight Bay Sinepuxent Bay
-1.1
0.8
2.9
1.4
5.40%
2.30%
Newport Bay
-0.9
2.9
4.50%
Chincoteague Bay
1.1
2.1
3.60%
51
CHAPTER 4: THE COUPLED WATER QUALITY AND SEDIMENT BENTHIC FLUX
MODEL
The Maryland Coastal Bays (MCBs) are shallow water bodies situated behind barrier
islands, with limited access to ocean exchange. They share many characteristics with
coastal lagoons and are vulnerable to the excess anthropogenic nutrients leading to
eutrophication (Dennison et al, 2012). A HEM3D (Hydrodynamic Eutrophication Model),
which consists of a hydrodynamic model - SELFE, a water quality model ICM (Integrated Compartment Model), and a sediment benthic flux model, was developed
and coupled with the HSPF watershed model to simulate lagoon-type biogeochemical
and water quality transport processes in the MCBs (Figure 4.1). The HSPF model
furnished flows to hydrodynamic model and, at the same time, provided nutrient and
carbon loads to the water column water quality model. The water column water quality
model interacted with sediment benthic flux model by providing particulate organic
matter and fed back with sediment fluxes and sediment oxygen demand.
4.1 The ICM Water Quality Model
The water quality model ICM (Integrated Compartment Model), originally developed by
Cerco and Cole (1994), was used for simulating eutrophication in the water column.
Fall-line load and below-fall line point and non-point source loads were supplied by the
HSPF model. Computation of eutrophication process in ICM was coupled directly with
hydrodynamic model on a time step of every 5 minutes. The ICM model simulates all
the processes occurring in the water body from the sediment interface up to the surface
of the water. It consists of 6 water quality variable groups as follows: (1) algae; (2)
organic carbon; (3) nitrogen; (4) phosphorus, (5) silica and (6) dissolved oxygen. Each
nutrient contains dissolved and particulate phases, and the particulate matter is further
subdivided into refractory and labile particulate forms. Both refractory and labile
particulate organics settle out of the water column and deposit onto the surface layer of
benthic sediments. Refractory organic particulate has a longer degradation time,
whereas labile organic particulate has a shorter degradation time. In terms of chemical
processes, dissolved organic carbon, nitrogen and phosphorus are converted into
inorganic forms by processes including hydrolysis and bacterially-mediated activities.
Utilization of dissolved organic carbon during respiration of heterotrophic bacteria
consumes dissolved oxygen. Similarly, dissolved organic nitrogen and phosphorus are
converted by bacterial activity to ammonium nitrogen (NH4-N) and orthophosphorus
(PO4-P). NH4-N is subsequently oxidized by bacteria to nitrate nitrogen (NO23-N). This
process, which is called nitrification, consumes dissolved oxygen. Under conditions of
extremely low dissolved oxygen, NO3-N may be reduced by bacteria to dissolved
nitrogen gas, which may subsequently be lost to the atmosphere at the air-water
interface. Denitrification is the process whereby dissolved organic carbon is consumed.
The schematic diagram is shown in Figure 4.2 (a).
52
Figure 4.1 Schematic diagram of coupled HSPF watershed, SELFE hydrodynamic, ICM water quality, and sediment benthic
models.
53
In the phytoplankton dynamics of the model, algae species are subdivided into three
forms: diatoms, blue-green algae (cyanobacteria), and green algae (dinoflagellates).
The growth, respiration, and mortality of each of these algal groups are controlled by
optimal water temperature specified in the model. In the MCBs HEM3D model, algae
are growth-limited in a multiplicative manner, based on the Liebig Law via ambient
levels of light, water temperature, and concentrations of inorganic nitrogen (NH4-N and
NO3-N) and phosphorus (dissolved PO4-P). All three algae forms consume nitrogen
(NH4-N and NO23-N) and phosphorus (dissolved PO4-P) during growth. Similarly, due to
respiration and mortality, algae release dissolved and particulate organic carbon,
nitrogen, and phosphorus. Algae consume dissolved oxygen during respiration and
release dissolved oxygen during photosynthetic activity. Blue-green algae exhibit a toxic
response to salinity levels above one (1) part per thousand (ppt). Blue-green algae are
not limited by low inorganic nitrogen concentrations, since they can alternatively utilize
dissolved organic nitrogen in the water column. The particulate forms of the algae also
settle out of the water column, contributing their organic carbon, nitrogen, and
phosphorus contents to the surface layer of benthic sediments. For more details of the
water column eutrophication model formulation, the reader can review Appendix B.
4.2. The Sediment Benthic Flux Model
The sediment benthic flux model was used to predict nutrient fluxes and sediment
oxygen demand at the water-sediment interface. Water column eutrophication is
coupled in real time with a benthic flux sub-model, which has twenty-seven water quality
state variables associated with mass fluxes in a 2-layer sediment compartment (DiToro
and Fitzpatrick, 1993). As shown in Figure 4.2 (b), the upper layer (layer 1) is in contact
with the water column and may be oxic or anoxic depending on dissolved oxygen
concentration in the overlying water. The lower layer (layer 2) is permanently anoxic.
The upper layer depth, which is determined by the penetration of oxygen into the
sediments, is at its maximum only about 1 centimeter (cm) thick. Layer 2 is much thicker,
on the order of 10 cm to 1 meter. The sediment benthic flux sub-model incorporates
three basic processes: (1) depositional flux of particulate organic matter (POM); (2)
diagenesis flux; and (3) sediment flux. The settling of particulate organic matter (POM)
from the overlying water is the main driver. POM fluxes include particulate organic
carbon, nitrogen and phosphorus which are deposition to layer 2 sediments. Because
of the negligible thickness of the upper layer, deposition is considered to be from the
water column directly to the lower layer. Within the lower layer, the model simulates the
diagenesis (mineralization or decay) of deposited POM, which produces inorganic
nutrients flux (diagenesis flux). The third basic process is the flux of substances
exchanged with the overlying water (sediment flux). The sediment fluxes include
ammonium nitrogen, nitrate nitrogen, phosphate phosphorus and sediment oxygen.
54
(a)
(b)
N2
Figure 4.2 (a) Schematic diagrams for (a) ICM model water column processes and (b) sediment digenesis processes.
55
Oxygen demand takes three paths out of the sediments: (1) oxidation at the sedimentwater interface as sediment oxygen demand; (2) export to the water column as
chemical oxygen demand, or (3) burial to deep as inactive sediments. Inorganic
nutrients produced by diagenesis take two paths out of the sediments: (1) release to the
water column, or (2) burial to deep, inactive sediments. The state variables for
describing above processes including: three separate classes (G1, G2, and G3) of
particulate organic carbon, nitrogen and phosphorus, and sulfide/methane, ammonium
nitrogen, nitrate nitrogen, phosphate phosphorus, and temperature in layers 1 and 2.
4.3 Water Quality Model Set-up
4.3.1 External loading
The water quality model receives nutrient loads from (1) non-point sources (2) point
sources (3) shoreline erosion and (4) atmospheric deposition. The non-point source
loads were provided from 225 watersheds delineated in the MCBs, as shown in Figure
4.3 (a) for the northern and Figure 4.3 (b) for the southern Bays. The water quality
model grid was connected with the watersheds, with the green thin lines representing
the major streams inside the watershed. These major streams include: (1) Greys Creek
(in the Assawoman Bay); (2) Bishopville Prong, Shingle Landing Prong, St. Martin
River, Herring Creek, Turville Creek and Manklin Creek (in the Isle of Wight Bay); (3)
Kitts Branch/Ayer Creek, Newport Creek and Marshall Creek (in the Newport Bay).
Examples of watershed segments, their Maryland 8 digit watershed codes, the 8-digit
basin name, and the acreage of the segment are shown in the list table of Figures 4.3 (a)
and (b). A complete list of segment list for the MCBs watershed can be found in VIMS
(2013). In Figure 4.3, there are a total of 231 edge of stream loads points shown as
diamond green symbols along the shoreline representing the point and nonpoint
sources discharge points. The distribution is as follows: 36 in Assawoman Bay, 74 in
Isle of Wight Bay, 25 in Sinepuxent Bay, 18 in Newport Bay and 78 in Chincoteague
Bay. These discharge points are the direct linkage of flow and nutrient loading from the
watershed into the water quality model. The shoreline erosional loads were determined
by the total shoreline length multiplied by the TN and TP loads per unit shoreline length
based on MGS’ sediment erosion studies (Wells, 1998, 2002, 2003). The inputs for
atmospheric loading were determined based on the water surface area of each of the
basins multiplied by the unit loading of TN and TP per unit area. The determination of
the unit loads was described in details in Chapter 2, Section 2.2.5.
56
(a)
SEGMENT
1
2
3
4
5
6
6
7
8
9
9
10
MDE8DIGIT
02130103
02130103
02130103
02130103
02130103
02130102
02130102
02130102
02130102
02130102
02130102
02130102
MDE8NAME
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
State
MD
MD
MD
MD
MD
DE
MD
MD
MD
DE
MD
DE
Area (m^2)
Acres
5877581.487 1452.366
2174529.604 537.3322
396413.9605 97.95497
440934.9881 108.9562
928127.7271 229.3429
4088756.294 1010.343
362549.3402 89.58693
2380970.667 588.3443
2923821.195 722.4842
414929.8449 102.5303
841858.0755 208.0254
1229176.32 303.7328
Figure 4.3 (a) Examples of the selected segments of HSPF Coastal Bays model segments
in the Assawoman Bay, Isle of Wight Bay, Sinepuxent Bay and Newport Bays
57
(b)
SEGMENT
MDE8DIGT
MDE8NAME
STATE
Area(m^2)
Acres
132
02130106
Chincoteague Bay
VA
1664404.256
411.2788
133
02130106
Chincoteague Bay
VA
3698790.703
913.9812
134
02130106
Chincoteague Bay
VA
1112237.58
274.8369
135
02130106
Chincoteague Bay
VA
1783601.063
440.7327
136
02130106
Chincoteague Bay
VA
1454331.688
359.3693
137
02130106
Chincoteague Bay
VA
1124736.705
277.9255
138
02130106
Chincoteague Bay
VA
2081700
514.3937
139
02130106
Chincoteague Bay
VA
6290108.622
1554.303
140
02130106
Chincoteague Bay
VA
403200
99.63182
141
02130106
Chincoteague Bay
VA
42861600
10591.22
142
02130106
Chincoteague Bay
VA
19170900
4737.182
Figure 4.3 (b) Examples of the selected segments of HSPF Coastal Bays model segments
in the Chincoteague Bay
58
4.3.2 Initial and boundary conditions
The initial condition was specified using January and February 2000 monitoring data of
year 2000 conducted by DNR survey. The initial condition for each polygon within the
MCBs was specified through linearly interpolating the survey data between adjacent
stations. The polygons are horizontally variable but homogeneous vertically initially.
The vertical variation was obtained by initializing the model for 5 years during which the
vertical variation of concentration emerges as a result of the interaction with the
sediment benthic model. The initial condition in the coastal ocean was specified based
on the linear interpolation between concentrations at the Ocean City and Chincoteague
Bay Inlets and the associated concentration specified at the open boundary condition.
The model’s open boundary conditions were specified approximately 7 miles offshore
along the east boundary of the coast in addition to the two cross-shore boundaries - one
at the north just south of Indian River, Delaware, and the other in the south, off of NASA
Wallops Island Flight Center. The concentrations specified at these location are longterm nutrient mean conditions obtained from EPA’s report “Criteria Development
Guidance: Estuarine and Coastal Bay Waters” (see http://www2.epa.gov/nutrient-policydata/criteria-development-guidance-estuarine-and-coastal-waters). The key variables
assigned at the boundary are: Chlorophyll = 1 ug/l, DO = 5 mg/l, Salinity = 30 ppt, TN =
0.2 mg/l, TDN = 0.1mg/l, TP = 0.02 mg/l, TDP = 0.01 mg/l and Secchi depth = 4m.
Because no sediment concentrations were measured, the initial condition for the
sediment benthic flux model was not known a priori. Thus, the sediment concentration
will be specified as clean sediments initially and executed along with the watershed
loading into the system long enough until the depositional flux, diagenesis flux, and the
sediment flux are in equilibrium. In a normal practice, the model is run to execute for five
(5) years continuously until the loading and the sediment flux reaches a steady state. At
the end of the five (5)-year simulation, the concentration values from sediment flux
model, which characterize the sediment characteristics of each location, were then
output and used as the initial condition for the different scenarios.
4.3.3 Estimation of parameters
Many parameters need to be specified in the water quality model. Most parameters in
the water quality model CE-QUAL-ICM were adopted from the default parameters for
the Chesapeake Bay (Cerco and Cole, 2004). However, there are a few parameters that
which are re-calibrated because MCBs is a relatively shallow water system as
compared to the Chesapeake Bay. During the testing, some parameters were changed
in order to produce results that more closely simulate the field observations unique to
the MCBs. The parameters that are specified differently from those of the Chesapeake
Bay model are shown in Table 4.1. The parameters 1- 6 are related to algae in the
water column, parameter 7 is related to heterotrophic respiration of DOC and parameter
8 is related to the critical oxygen concentration for releasing of PO4 in the sediment
benthic flux model. The final parameters used for the water quality model of this study
are listed in of Appendix C.
59
Table 4.1: The parameter selected for phytoplankton dynamics
PARAMETER
1. PMd
2. TMc
3. TMg
4. STOX
5. WSd
6. WSg
DESCRIPTION
Maximum growth rate of diatoms
Optimum T for cyanobacteria
Optimum T for green algae
Salinity toxicity
Settling velocity for diatoms
Settling velocity for green algae
VALUE
2.5
25
22.5
5
0.35
0.25
7. Kdcalg
Constant relating respiration of DOC to algal
Biomass
Critical DO concentration for layer 1
incremental
PO4 sorption
0.03
8. O2CRIT
2
UNIT
1/day
Degree
Degree
Ppt
m/day
m/day
per g C
m^3
per day
mg/l
4.4 Water Quality Model Calibration and Verification
4.4.1 Model calibration
The primary means of calibrating the water quality model was through comparison of
modeled and observed water quality variables. Calibration was an iterative process, in
which algal growth and decay rates, chemical kinetic coefficients, partition coefficients,
half saturation constants, and sediment mineralization rates were adjusted to improve
model-observation comparison. For the calibration of the ICM model, the year 2004
was selected as the calibration period, as this is the year with intensive monitoring
observations for both the hydrodynamic [surface water elevation, ADCP (Acoustic
Doppler Current Profiler) measurement of currents], and water quality variables (DO,
Chlorophyll, TN, TP, ammonia, nitrate, DON) as well as high spatial coverage. Figure
4.4 (a) displays six (6) intensive water quality monitoring stations in the Isle of Wight
Bay: Stations 1, 2, 3 are located in Turville Creek and Stations 4, 5, 6 extend throughout
Isle of Wight Bay approaching the Ocean City Inlet. Figure 4.4 (b) shows five (5) baywide stations: Station A at Assawoman Bay, Station B at the middle reach of the St.
Martin River, Station C in northern Chincoteague Bay downstream of Newport Bay, and
Stations D and E in middle and southern Chincoteague Bay. The hydrology of year
2004 in the Saint Martin River is shown in Figure 3-10 (a) in Chapter 3. Compared to
the historical record, year 2004 is a wet hydrologic year with large spring freshwater
inputs at about 200-300 cubic feet per second (cfs) in April, and greater than normal
summer freshwater inputs in July and August. The model predicted water quality
variables in 2004 for the five stations along the transect of Turville Creek in Isle of Wight
Bay were compared to the DNR monthly survey data. Station 1 was not included due to
60
its location in the nontidal headwaters (and thus it is not included in the water quality
model grid).
(a)
(b)
Figure 4.4 Intensive monitoring stations in (a) Isle of Wight Bay and (b) Maryland Coastal
Bays.
61
Figure 4-5 (a) shows the results of the chlorophyll-a simulation, which capture both
temporal and spatial features of chlorophyll-a along the transect of Turville Creek in the
Isle of Wight Bay. Specifically, chlorophyll-a exhibits a strong longitudinal gradient from
Station 2 to Station 6 from about 40μg/l at Station 2 (upper Turville Creek) to less than
10 μg/l at Station 6 (lower Isle of Wight Bay). The model also captures the seasonal
pattern well. For instance, at Station 2, the chlorophyll a levels remain low in winter and
reach maximum in late summer. This general pattern has been well reproduced in the
model. Additionally, the model also catches the spring phytoplankton bloom event
around Day 90-120. An example of this is at Station 2, day 110, when chlorophyll a
suddenly jumps above 40μg/l. The same event was recorded at DNR’s continuous
monitoring site as well (not shown). Apparently, this phytoplankton bloom event was
fueled by a large freshwater pulse and accompanied by a sharp salinity drop as shown
in Chapter 3.
Figure 4.5(b) shows the model is capable of simulating the temporal and spatial
variability of DO correctly as compared well with the observation data. In the upstream
two stations of Turville Creek, the hypoxia and episodic, prolonged (days) low DO
events does occur which is presumably driven by high nutrient load combined with
respiration rate, suggested by the continuous monitoring data. Although the current
model is not fully implemented to simulate diurnal DO diel cycle, the model does have
the capability to capture variation of episodic hypoxic events on time scale of several
days. For example, at Station 2, the monitoring data reveals a DO drop around Day 210
and the same event was also reasonably captured by the model. Lastly, along with field
measurements, the model results reasonably reflect DO seasonality and the longitudinal
gradient (e.g., low DO rarely occur at Stations 4-6 closer to the Ocean City Inlet).
Chapter 5 discusses diurnal DO diel variation, which was reproduced by postprocessing model results using a statistical relationship with the daily mean DO.
In terms of simulating nutrient variables including the nitrogen and phosphorus species,
the model also performs satisfactorily. In Figure 4.5(c) and (d), the model captures the
nutrients’ seasonal pattern. For instance, NO23, the most important land-originated N
source has a distinct seasonal pattern that rose in winter-spring due to watershed inputs,
and depleted in summer due to rapid uptake by autotrophs. This temporal variability is
well produced by the model, as shown in Figure 4-5(d). On the other hand, NH4, which
mainly derives from in situ recycling/regeneration processes in the sediment and water
column, often increases in the summer/fall season, and the model captures this
temporal trend. The PO4 obtained from the watershed source can be used up in the
spring by the phytoplankton bloom and subsequently deposited into sediment and remineralized. In the Chesapeake Bay, when the anoxic condition developed in the late
summer, the stored phosphate will then be released back to the water column in large
quantity and fuels the second peak of phytoplankton bloom. Figure 4.5(e) shows the
seasonal pattern of PO4 in the MCBs, in which, it doesn’t seem to have release in
pulsation like that in Chesapeake Bay presumbly due to milder low DO condition in
MCBs. Lastly, dissolved organic nutrients (e.g., DON), which are controlled by the
balance between in situ production (source) and decomposition (sink) processes, were
maintained at stable but higher concentrations as compared to inorganic nutrients. As
can be seen from Figure 4.5(f), DON does not exhibit substantial variations in either
62
Figure 4.5 (a) and (b): Comparisons of model prediction and field measurement for chlorophyll-a and DO in the Isle of
Wight Bay.
63
Figure 4.5 (c) and (d): Comparisons of model predictions and field measurement for NH4 and NO23 in the Isle of Wight Bay.
64
Figure 4.5 (e) and (f): Comparisons of model predictions and field measurement for PO4 and DON in the Isle of Wight Bay.
65
temporal (seasonal) or spatial (at difference stations) scales unlike inorganic nutrients,
which can change by an order of magnitude. Nevertheless, DON does have a seasonal
trend, increasing in warmer months, presumably fueled by high primary production and
respiration in the system. The model results agree well with field observations.
To expand the water quality calbiration to the bay-wide scale, the model results were
presented in five stations located in the open bay portion of the MCBs as shown in
Figure 4.4 (b). These stations include Station A in Assawoman Bay, Station B in St.
Martin River, Station C is in Newport Bay, Station D in mid -Chincoteague Bay and
Station E in the southern Chincoteague bay. Figure 4.6 (a) shows that chlorophyll-a is
consistently higher in the northern Bays (Stations A and B) than in the southern Bays
(Stations C, D, and E). This is consistent with the fact that the northern Bays have a
larger nutrient loading than do the southern Bays and thus supports higher
concentration of algae. The high frequency variation of chlorophyll-a is also observed in
the northern Bay stations, but not in the southern Bay stations. Figure 4.6 (b) shows
DO concentration has a seasonal pattern, with the highest concentrations in the winter
and the lowest in the summer. When compared with the previously generated DO
results at the tributaries where DO can fall down to hypoxic levels (Figure 4.5 (b)), the
DO at open Bay stations A, B, C, D, E does remains at or above 5 mg/l. It was observed
that DO oscillation was more pronounced in the northern Bays, an indication of the
effect of tidal currents which can induce semi-diurnal oscillation, as well as the
manifestation of more photosythetic acitivity during daytime and more cellular respiraton
at night due to greater algal bomass. Figure 4.6 (c) and 4.6 (d) show the temporal
variation of nitrogen species NH4 and NO23, which have quite differnt seasonal patterns.
NH4 exhibts the highest concentrations in the late summer and early fall, whereas NO23
concentrations are highest in the winter and lowest in the late summer and early fall.
This can be understood from the fact that most important source for NO23 is terrestrial
and thus abundant in the winter due to watershed inputs; NO23 is depleted in the late
spring and summer due to the uptake by phytoplankton. NH4 is mainly maintained by in
situ recycling and regeneration through sediment process. In the spring, NH4 is quickly
used up by the algal spring bloom, but is recycled back in the later summer and early
fall into the water column through the sediment diagensis process. Consistent with the
chlorophyll-a concetrations, NH4 concentrations are higher in the northern Bays than in
the southern Bays. Figure 4.6(e) shows the seasonal pattern of phosphorus; lowest
concentrations are in the spring due to the spring algae bloom, but, like NH4,
concentrations can bounce back in the later summer through the sediment flux. In
additon, phosphorus be delivered from the land in the winter. In the southern Bays,
there is an apparent logitudinal gradient of NH4, NO23 and PO4, higher in the north and
lower in the south along the axis of Stations C, D, and E; this suggests that it is derived
from a source in the Newport Bay. Connecting cause and effect, it can be seen that the
watershed loading, tranport dynamics and bio-chemical processes are all linked
together. The successful coupling of watershed, hydrodynamic and water quality
models in this project, with good calibration, allows the model to produce excellent
results that are consistent with the multi-phase measurements in MCBs.
66
(a)
(b)
Dissolved Oxygen
Chlorophyll a
15
60
Modeled
Observed
A
40
0
0
50
100
150
200
250
300
350
B
40
0
50
0
100
150
200
250
300
350
150
200
250
300
350
50
100
150
200
250
300
350
50
100
150
200
250
300
350
50
100
150
200
250
300
350
50
100
200
150
Julian Days (2004)
250
300
350
B
0
15
DO, mg L-1
Chlorophyll a, µg L-1
100
5
60
C
40
20
50
100
150
200
250
300
C
10
5
0
0
350
0
15
60
D
10
D
40
5
20
0
0
50
10
20
0
0
15
60
0
Modeled
Observed
5
20
0
A
10
0
50
100
150
200
250
300
0
350
15
E
60
10
E
40
5
20
0
0
0
50
100
150
200
Julian Days (2004)
250
300
350
0
Figure 4.6 (a) and (b): Comparisons of model predictions and field measurement for Chlorophyll-a and DO in Bay-wide
stations.
67
(c)
(d)
NH+4
-N
NO-x - N
0.1
0.05
0
1
Modeled
Observed
A
0.5
0
50
100
150
200
250
300
0
350
0.1
0
100
150
200
250
300
350
50
100
150
200
250
300
350
50
100
150
200
250
300
350
50
100
150
200
250
300
350
50
100
150
200
Julian Days (2004)
250
300
350
B
0.05
0.5
50
0
100
150
200
250
300
0
350
NO- - N, mg L-1
0.1
C
0.05
0
0.1
C
0.05
x
4
NH+ - N, mg L-1
50
1
B
0
Modeled
Observed
A
0
0
50
100
150
200
250
300
350
0.1
0
0
0.1
D
D
0.05
0
0.05
0
50
100
150
200
250
300
350
0.1
0
0
0.1
E
E
0.05
0.05
0
0
0
50
100
150
200
Julian Days (2004)
250
300
350
0
Figure 4.6 (c) and (d): Comparisons of model predictions and field measurements for NH4 and NO23 in Bay-wide stations.
68
(e)
(f)
PO3-P
4
DON - N
0.2
2
Modeled
Observed
A
0.1
0
0
50
100
150
200
250
300
1
0
350
0.2
0
50
100
150
200
250
300
DON - N, mg L-1
C
0
0
350
4
PO3- - P, mg L-1
0
0.1
50
100
150
200
250
300
350
0.1
200
250
300
350
0
50
100
150
200
250
300
350
50
100
150
200
250
300
350
50
100
150
200
250
300
350
50
100
150
200
Julian Days (2004)
250
300
350
2
C
1
0
0
1
D
D
0.05
0.5
0
50
100
150
200
250
300
350
0.1
0
0
1
E
E
0.05
0
150
1
0.2
0
100
B
0.1
0
50
2
B
0
Modeled
Observed
A
0.5
0
50
100
150
200
Julian Days (2004)
250
300
350
0
0
Figure 4.6 (e) and (f): Comparisons of model predictions and field measurement for PO4 and DON in the Bay-wide stations.
69
4.4.2 Model verification
Since a water quality model typically requires using many coefficients to parameterize
the ecosystem processes, a comparison of numerical model output with the observation
data is an important and necessary step. Given that prediction capability is an
important component of TMDLs, it is critical to assess the water quality model with an
independent set of data and test the range of validity for the model. The process
involves running the models with the calibrated parameters and comparing the results to
verify that the coefficients used are self- consistent under various dynamic conditions.
The calibration was conducted using intensive survey data collected in 2004; whereas
for verification purposes, the water quality model was further compared with surveys
conducted through 2001-2003 by Maryland DNR and Assateague Island Park Service
(ASIS). With this step, the model is confirmed as valid for different hydrological and
environmental conditions, and the model developed can be applied to investigate
various operational and management scenarios. For the verification process, the
external nutrient loadings, boundary conditions, and all the parameters of the water
quality model were specified using identical values for the calibration. The initial
condition was also specified to be the same at the beginning of both simulations. Figure
4.7 showed station locations where the monthly observed data were collected at the
twenty-seven (27) DNR and eighteen (18) ASIS stations. The DNR stations cover most
of the MCBs except Sinepuxent Bay whereas the ASIS stations only cover stations
south of the Ocean City Inlet including Sinepuxent, Newport and Chincoteague Bays.
An annual runoff cycle exists, with peak flow in the spring and minimum flow in the
summer. However, floods and droughts frequently cause daily and monthly flow
deviations from the long-term pattern. Figures 4.8 (a) show the flow discharge rate and
Figure 4.8 (b) the chlorophyll-a calculated from January 2000 - August 2005 (note: for
2005, only a partial record, through August, was available) at the calibration site of
Birch Branch watershed. The year of 2001 can be characterized as an average year,
2002 a dry year, and 2003 and 2004 were wet years. In 2002, the lower than normal
spring flow was the result of the drought year and in 2003, the higher than normal flow
in September was caused by Hurricane Isabel. As can be seen, the chlorophyll-a level
is proportion to the hydrologic inputs: the larger the nutrient load inputs, the greater the
phytoplankton biomass. The flushing effect of the fast moving stream on the
phytoplankton biomass, which happens in other estuaries, does not seem to apply in
the Maryland Coastal Bays.
The time series comparisons of water quality model results with the observations
measured at DNR and ASIS stations are presented in Figures 4.9 through 4.15. Figure
4-9 (a) (b) and (c) show the dissolved oxygen comparison at all DNR stations from
January 2001 to August 2005. The DO concentrations in Figure 4.9 (a) exhibit various
degrees of hypoxia at the first 6 stations (4 stations in the left panel and the top 2
stations in the middle panel) located in the tributaries of the Saint Martin River. Among
them, the Manklin Creek station demonstrates anoxic conditions in 2004 and 2005,
which the model was able to capture by additional chemical oxygen demand. The
model also simulated temporal variation of the hypoxia well for the other 5 stations. The
DO at station XDM4486, located in the upstream of Saint Martin River, also showed
similar signs of hypoxia both from both observations and model results (Figure 4.9 (b)).
70
For these stations, there are clear signs of increasing summer hypoxia from 2003 2005, coinciding with the high flow of these consecutive years. The rest of the 20 DNR
stations, away from the aforementioned 7 stations in the Upper Saint Martin River
tributaries, in general, do not exhibit persistent hypoxia issues. Figure 4.11 (a) – (c)
shows the comparison of chlorophyll a for all the DNR stations. The pattern of high
chlorophyll-a, in general, is correlated with that of low DO stations. For example, the 7
stations in the upper Saint Martin River tributaries (where hypoxia occurred) showed a
consistent higher chlorophyll pattern. Three additional stations in the middle and lower
Saint Martin River, XDN3724, XDN4312 and XDN4797, also have high chlorophyll a,
exceeding 50 μg in 2004 and 2005. The reason for the correlation between low DO and
high chlorophyll a can be partially explained by the sediment oxygen demand (SOD) as
a result of the deposition of organic matter from phytoplankton blooms, as shown in
Figures 4.13 (a) – (c). It was clear that hypoxia occurred at the stations whose SOD
exceeds 0.5g C per m2. For stations with SOD less than 0.5 g C per m2, hypoxia rarely
occurred. The DO and Chlorophyll a comparison for the ASIS stations is also shown in
Figures 4.14 (a) – (b) and Figure 4.15 (a) – (b). For the stations south of the Ocean
City Inlet, in general, there are no persistent low DO and high chlorophyll a problems.
71
Figure 4.7: MD DNR and ASIS monitoring stations in the Maryland Coastal Bays.
72
(a)
(b)
Figure 4.8 (a) The flow discharges and (b) Chlorophyll-a concentration from Birch Branch
watersheds.
73
Evaluating model performance during the verification process requires statistical
summaries of the comparison of many observations with model results. Summary
statistics of mean error, absolute mean error, relative error and correlation coefficient
are employed to assess the accuracy of the model. The mean error (ME) is defined as:
ME =
1
N
N
∑ (P
n =1
- On )
n
Positive ME indicates the model’s overestimation of the data on the average and
negative ME indicates the model’s underestimation of the data on the average, with
zero ME being ideal. The mean absolute error (MAE), a measure of the absolute
deviation of the model results from the data on the average, is defined as:
1
N
MAE =
N
∑P
n
n =1
- On
where Pn and On = corresponding model results and data; N = number of observations.
The MAE of zero is ideal. Since the MAE cannot be used to discern the overestimation
or underestimation, another measure is desirable. The relative error (RE) is defined as:
RE =
∑P - O
∑O
n
n
n
The RE is the ratio of the MAE to the mean of the data, indicating the magnitude of the
MAE relative to the data on the average. The correlation coefficient defined as:
N
r=
∑ (P
i =1
mod
− P mod )(Oobs − O oobs )
1/2
N
N
2
2
(
P
−
P
mod ) ∑ (Oobs − O obs )
∑
mod


=
 i 1 =i 1

in which the model predictions were treated as independent variables and observations
as dependent variables in a regression analysis.
The mean error describes whether the model over-estimates or under-estimates the
observations, on average. The mean error can achieve its ideal value, zero, while large
discrepancies exist between individual observations and computations. The absolute
mean error is a measure of the characteristic difference between individual observations
74
and computations. An absolute mean error of zero indicates the model perfectly
reproduces each observation. The relative error is the absolute mean error normalized
by the mean concentration. Relative error provides a statistic suitable for comparison
between different variables or systems. Quantitative statistics were determined through
comparison of model results and observations for each of the basins in the MCBs
sampled at approximately monthly intervals. For dissolved oxygen, the concentration
measured at 1.5 meter above the bottom was compared and for chlorophyll a
comparisons for surface samples were examined and presented.
The measure of correlation between modeled and observed DO were above 0.75, as
shown in Figure 4.10, Tables 4.2 and 4.3. The bottom DO is within 1 mg/l of the
observed range and the relative error is within 15%. Examination of relative error of
chlorophyll-a indicating that chlorophyll a has the greatest error around 60-70%. The
chlorophyll a error reflects the difficulty in computing this dynamic biological component
which can attain a very large magnitude. The TN and TP statistics, as shown in Table
4.2, are in the mid-range of 30% to 45%, which total phosphorus, perhaps exhibiting
slightly higher relative error. The higher error in phosphorus reflects the difficulties in
evaluating loads, in simulating re-suspension and in representing particulate
phosphorus transport (Cerco et al., 2004). This is particularly true for the Saint Martin
River, where uncertain loads discharged into the constrained volumes of the tributaries,
are the major reason for higher relative error in the Isle of Wight and Newport Bays.
No standard criteria exist for judging acceptable model performance. One approach is
to compare performance with similar statistics from other model applications. Statistics
comparable to other systems at least indicate the model is in the performance
mainstream. The relative error was compared to the relative errors presented in the
2002 Chesapeake Bay application, Florida Bay (Cerco et al., 2000) and the lower St.
Johns River, Florida (Tillman et al. 2004). Florida Bay is a shallow sub-tropical lagoon.
The St. Johns River is a partially- to well-mixed estuary with a substantial tidal
freshwater extent. When the relative error is inter-compared, as shown in Table 4.4, all
models indicate chlorophyll a has the greatest relative error. Comparison of the
parameters DO, TN and TP indicate that MCBs model results are comparable to the
quality of other TMDL studies.
75
Figure 4.9 (a): DO verification with DNR data, Stations 1-12, 2001- 2005.
76
Figure 4.9 (b): DO verification with DNR data, Stations 13-24 2001-2005.
77
Figure 4.9 (c): DO verification with DNR data, Stations 25-27, 2001-2005.
78
ME=-0.49
AME=0.7
5
ME-=-0.47
AME=1.06
RE=0.15
ME=-0.07
AME=1.1
3
ME=-0.49
AME=0.89
RE=0.12
Figure 4.10: Statistical comparison of observed versus modeled DO; Assawoman, Isle of Wight, Newport, and
Chincoteague Bays.
79
Figure 4.11 (a): Chlorophyll-a verification with DNR data Stations 1-12, 2001-2005.
80
Figure 4.11(b): Chlorophyll-a verification with DNR data Stations 13-24, 2001-2005.
81
Figure 4.11 (c): Chlorophyll-a verification with DNR data Stations 25-27, 2001-2005.
82
Figure 4.12 (a): Statistical comparison of observed versus modeled Chlorophyll-a data for Assawoman and Isle of Wight
Bays.
83
Figure 4.12 (b): Statistical comparison of observed versus modeled Chlorophyll-a data for Newport, and Chincoteague
Bays.
84
Figure 4.13 (a): SOD calculation at DNR Stations 1-12, 2001- August 2005.
85
Figure 4.13 (b): SOD calculation at DNR Stations 13-24, 2001- August 2005.
86
Figure 4.13 (c): SOD calculation at DNR Stations 25-27, 2001- August 2005.
87
Figure 4.14 (a): DO verification at ASIS Stations 1-9, 2001- August 2005.
88
Figure 4.14 (b): DO verification at ASIS Stations 10-18, 2001- August 2005.
89
Figure 4.15 (a): Chlorophyll verification at ASIS Stations 1-9, 2001- August 2005.
90
Figure 4-15 (b): Chlorophyll verification at ASIS Stations 10-18, 2001-August 2005
91
Table 4.2: Model and data comparison statistics for DNR stations.
DO statistics
Assawoman Bay
Isle of Wight Bay
Newport Bay
Chincoteague Bay
R
0.92
0.78
0.79
0.85
RMS
0.93
1.55
1.37
1.04
ME
-0.49
-0.07
-0.47
-0.49
AME
0.75
1.13
1.06
0.89
RE
0.10
0.15
0.15
0.12
TN statistics
Assawoman Bay
Isle of Wight Bay
Newport Bay
Chincoteague Bay
R
0.60
0.67
0.60
0.65
RMS
0.41
0.65
0.75
0.30
ME
0.29
0.35
0.41
0.22
AME
0.32
0.43
0.51
0.25
RE
0.32
0.37
0.37
0.32
TP statistics
Assawoman Bay
Isle of Wight Bay
Newport Bay
Chincoteague Bay
R
0.73
0.45
0.49
0.67
RMS
0.02
0.05
0.07
0.02
ME
0.00
0.01
0.03
0.00
AME
0.01
0.03
0.04
0.02
RE
0.26
0.42
0.43
0.35
CHLA statistics
Assawoman Bay
Isle of Wight Bay
Newport Bay
Chincoteague Bay
R
0.60
0.60
0.56
0.69
RMS
10.84
28.32
30.33
6.62
ME
7.62
8.15
13.26
2.87
AME
8.07
12.75
15.42
4.36
RE
0.61
0.61
0.67
0.49
92
Table 4.3: Model and data comparison statistics for ASIS stations.
DO statistics
R
RMS
ME
AME
RE
Sinepuxent Bay
0.83
1.20
0.35
0.95
0.11
Newport Bay
0.89
1.55
1.11
1.31
0.15
Chincoteague Bay
0.71
1.67
0.20
1.01
0.12
CHLA statistics
R
RMS
ME
AME
RE
Sinepuxent Bay
0.59
5.96
1.02
4.27
0.69
Newport Bay
0.55
5.85
-1.65
4.34
0.70
Chincoteague Bay
0.61
5.27
1.09
3.70
0.60
Table 4.4: Relative Error in the MCBs model compared with other modeled systems.
Dissolved Oxygen (mg/l)
Chlorophyll-a (μg/l)
Total Nitrogen (mg/l)
Total Phosphorus (mg/l)
Chesapeake Bay
0.36
0.58
0.24
0.37
St. Johns River
0.09
0.49
0.29
0.27
Florida Bay
0.07
0.72
0.39
0.31
MD Coastal Bays
0.16
0.58
0.34
0.36
93
CHAPTER 5: ADDITIONAL MODEL ANALYSES
5.1 Adjustments to incorporate the DO Diel Cycle
One of the important considerations for the TMDL assessment with respect to hypoxia
in the MCBs is diel (diurnal) cycle of dissolved oxygen (DO). Oxygen is a by-product of
aquatic plant photosynthesis. Through photosynthesis and respiration, phytoplankton,
periphyton, and rooted aquatic plants can significantly affect the DO levels in a
waterbody with a profound effect on the variability of the DO throughout a day. This is
because photosynthesis only occurs during daylight hours, whereas respiration and
decomposition proceed at all times and are not dependent on solar energy. On a daily
average basis, aquatic plants provide a net addition of DO to a water body through
photosynthesis, yet respiration can cause low DO levels at night. This results in the diel
cycle, whereby daily DO maximum occurs in mid-afternoon, during which time
photosynthesis is the dominant mechanism and the daily DO minimum occurs in the
early morning during which time respiration and decomposition have the greatest effect
on DO. When algae are growing excessively, they can cause large diurnal DO variation,
and lead to violations of DO standards. DO was monitored in the MCBs monthly at the
fixed station network to provide information on the status of water quality condition for
living resources. Traditional monitoring programs collected periodic data at a small
number of fixed sampling locations, often in the deeper channel areas. Since the mid2000s, new monitoring technology allows for continuous monitoring (ConMon) has
become available, with which high frequency (every 15 minutes) temperature, salinity,
DO, fluorescence (converted to chlorophyll), pH and turbidity can be continuously
recorded. ConMon was implemented in the Maryland Coastal Bays in April 2002 and
was reported by Wazniak et al. (2004) in details. Today, MD Department of Natural
Resources (DNR), Delaware Natural Resource and Environmental Control (DNREC)
and Virginia Estuarine and Coastal observation system are using it routinely as a realtime monitoring program; see http://mddnr.chesapeakebay.net/eyesonthebay/index.cfm.
The ConMon data sondes usually are installed in shallow water sites, record data at
approximately 0.5 m above the sediment surface, and thus can provide information for
estimating the total production and respiration in shallow water ecosystems. Figure 5.1
shows an example of the ConMon data measured at Bishop’s Landing in the MCBs in
July and August, 2005. It has a record of 25 days of continuously measured depth, DO,
salinity, temperature, turbidity, chlorophyll-a, and Photosynthetically Active Radiation
(PAR). On the left panel, it can be seen that a daily swing of DO with amplitude 2-3
mg/l can vary between DO saturation and hypoxia within a single daily cycle. Typically,
the diel pattern shows the lowest DO occurring in the early morning and rising through
the day to reach a maximum in the later afternoon as the DO-producing chlorophyll of
phytoplankton is activated by sunlight. It can also be seen that at day 12 and day 13,
the diel cycle amplitude was suppressed by the low light condition by a storm with rain
and cloud cover. On the right panel, the spectral magnitude at the diurnal frequency
was clearly identified by the FFT (Fast Fourier Transform) analysis, indicative of signal
of diel oscillation, followed by a weaker semi-diurnal signal presumably related to the
tide.
94
Figure 5.1: ConMon data measured at Bishop’s Landing, July – August, 2005; before (left)
and after (right) Fast Fourier transformation.
95
Both CH3D/ ICM and HEM3D models used for eutrophication studies in the
Chesapeake Bay and MCBs, respectively, were most suitable for predicting daily
average DO. Although higher frequency DO outputs can be made available in the
current framework, the accuracy of its variation in short time scale is questionable
because of the following reasons. First, the current technology is lacking full
understanding of the short time scale of photosynthesis and respiration processes for
phytoplankton dynamics. The best example is the harmful algal bloom, in which a rapid
increase and accumulation of algae population can occur within very a short time (within
a day). The cause is still an actively researched topic. Second, it lacks robust and
accurate high frequency forcing functions and parameters to specify for growth rate and
mortality. These forcing functions are high frequency PAR, wind speed and direction,
air and water temperature, and wind wave sea state. The parameters include carbon to
chlorophyll-a ratio, heterotrophic respiration, and the organic nitrogen uptake rate for
certain species. Even the best available modeling technology today does not guarantee
the prediction of DO diel cycle as accurate enough for TMDL implementation. In fact, a
national research program, ECOHAB (The Ecology and Oceanography of Harmful Algal
Blooms), is being launched to unlock the factors causing the harmful algal bloom and
the consequence of it. As a result, new information and tools are being developed.
Given the situation, the strategy adopted in this study was to use the process-based
HEM3D model results and to combine with the empirically-derived statistical results.
The empirical relation was developed by Elgin Perry et al. (2012) and used to
incorporate the diel oscillation into the predicted daily DO. The diel adjustment was
implemented in steps. First, a trigonometric time series model using the Sine and
Cosine function of time was fitted to DNR ConMon data to determine the amplitude and
phase of the diel cycle. Second, a separate regression equation was developed to
obtain the amplitude and phase as a function of seasonal water temperature, daily
water temperature, turbidity, salinity, chlorophyll concentration, and PAR. As a third
step, the regression equation is used to estimate the diel cycle of DO for each day using
daily conditions. The incorporation of the diel cycle can be used for adjusting the fixed
station monthly DO measurements to a fixed time of day. Or it can be used for TMDL
worst-case scenario evaluation. The idea behind the diel DO adjustment essentially is
to take the total DO and split it into the daily mean and diel DO component (expressed
as DO’).
DO
= DO + DO '
Where DO is the daily mean, and DO ' is the diel component, which will be fitted by
the Fourier series for its amplitude and the phase.
96
By the splitting and then fitting with Fourier series, the resultant formula is:
DO =
µi + β1 sin(
t * 2π
t * 2π
) + β 2 cos(
)
24
24
Where µi is mean DO for the 24 hour period, β1 and β 2 are regression coefficients, t is
time of the day and i is the day.
The phase and amplitude of the diel cycle are computed as:
β 
24
pˆ =
arctan  1  ×
2π
 β2 
based on the Fourier series formulation.
aˆ =
β12 + β 2 2
and
In practice, the diel DO measured each day was fitted with sinusoidal curves using a 24
hour period and with different amplitude and phase, as shown on the right panel of
Figure 5.2. The amplitude measures how far the diel curve deviates from the mean line,
while the phase measures shift in time between the beginning of the diel cycle on each
day and the nominal start time of 6:00 am. Perry et al. (2012) used a 5-year dataset to
fit all the amplitude and phase and relate them to the contributing variables. In many
cases the fit was good (as in the right top panel in Figure 5.2) and in some cases the fit
was poor (as in the right bottom panel in Figure 5.2). The amplitude tends to be skewed
to the right and it was found that a 1/5th power transformation provides symmetric and
approximately normal residuals distribution. Therefore, the amplitude was transformed
by the 5th root first and then used by a linear regression model as dependent variable.
97
Figure 5.2: Low-pass-filtered daily DO, DO saturation, salinity, temperature, and
chlorophyll-a data (left); measured (blue) and fitted sinusoidal curve (red) diel cycle of
DO (right).
Different variables - seasonal water temperature, daily water temperature, turbidity,
salinity, chlorophyll concentration, and PAR-were added as independent variables by
the stepwise procedure to ensure that the linear regression model covers enough
variables to adequately represent the cause-and-effect relationship. The final variables
used for the linear regression model and the corresponding coefficients are listed in
Table 5.1. Also shown are the associated parameters and their variance. As can be
seen, the seasonal temperature has the largest variance, followed by daily temperature,
turbidity, chlorophyll-a and PAR. It was observed that the variance of the seasonal
temperature likely overlaps with the variance of PAR for both sharing a strong seasonal
component.
98
Table 5.1: Variables, coefficient values and statistical test results for DO diel cycle linear
regression model (Perry et al., 2012).
Source variables
Seasonal temperature
Daily temperature
Log turbidity
Log Chlorophyll
PAR (photosynthetic
active radiation)
Residual
Sum of
squares
Mean
square
F-stat
P-value
6.18
1.17
0.65
0.39
0.8833
1.1676
0.6546
0.3857
287.321
379.8237
212.9489
125.4532
<0.0001
<0.0001
<0.0001
<0.0001
0.23
2.78
0.2332
0.0031
75.8538
<0.0001
The amplitude adjustments for the seasonal temperature cycle are apportioned to the
monthly mean temperature with the mean amplitude in the range of 0.5 – 2.5 mg/l
depending on the season. It has the largest amplitude in July and August with 2.5 m/l in
amplitude, and smallest in November and December with less than 1 m/l, as shown in
Figure 5.3. An empirical relationship between light extinction coefficient:
Ke (ft-1) and turbidity (NTU)
=
K e 0.543 + 0.0177 Turb
was used to calculate the turbidity correction. The log transformation was made to the
turbidity and chlorophyll-a for the amplitude adjustment in the regression equation. It
should be noted that since the phase does not exhibit a systematic variation, the
adjustment was not considered in this study. In the end, the diel adjustment was
performed for the entire model calibration years from 2001-2004. Examples of diel
adjustments were shown in Figure 5.4 (a)-(f) for stations located in Assawoman Bay,
Isle of Wight Bay (inside St. Martin River), Isle of Wight Bay (outside St. Martin River),
Sinepuxent Bay, Newport Bay, and Chincoteague Bay. It can be seen that the DO diel
oscillation are embedded in the mean daily DO and with larger amplitude in the summer
months and smaller amplitude in the winter, as expected. With the diel adjustment
procedure, the skill score for the final calibration of the DO results for all DNR and
Assateague Island National Seashore (ASIS) stations are presented in Figure 5.5 (a) (e). The correlation coefficient (R2) for model and data comparison are mostly in the
range of 0.8 - 0.9. This skill score for model and data comparison was improved using
the modeled, adjusted DO (over that without the DO adjustment) because the adjusted
DO now have a range of values within a day in comparing with the observed data. In
other words, it was improved for the observed data can compare with a range of the
simulated adjusted diel DO rather than just single comparison.
99
Figure 5.3: DO diel cycle amplitude adjustments for seasonal temperature cycle
proportioned to monthly mean temperature.
.
100
Assawoman Bay
XDN7545
DO (mg/l)
Julian Day 2004
XDN4851
DO (mg/l)
Figure 5.4 (a): DO daily average (black) and diel (green) time series, Stations XDN7545 and XDN4851, Assawoman Bay,
January 1 – December 31 2004.
101
Isle of Wight Bay – inside St. Martin
XDN4486
DO mg/l
Julian Day
XDN3724
DO (mg/l)
Figure 5.4 (b): DO daily average (black) and diel (green) time series, Stations XDN4486 and XDN3724, St. Martin River, Isle of
Wight Bay, January 1 – December 31, 2004.
102
Figure 5.4 (c): DO daily average (black) and diel (green) time series, Stations XDN3445 and TUV0019, Isle of Wight Bay open
waters, January 1 – December 31 2004.
103
Sinepuxent Bay
ASSA
DO (mg/l)
Julian Day 2004
ASSA 2
DO (mg/l)
Julian Day 2004
Figure 5.4 (d): DO daily average (black) and diel (green) time series, Stations ASSA 16 and ASSA 2, Sinepuxent Bay,
January 1 – December 31, 2004.
104
Newport Bay
ASSA 4
DO (mg/l)
Julian Day 2004
ASSA 3
DO (mg/l)
Julian Day 2004
Figure 5.4 (e): DO daily average (black) and diel (green) time series, Stations ASSA 4 and ASSA 3, Newport Bay, January 1 –
December 31, 2004.
105
XCM1562
Chincoteague Bay
DO (mg/l) (mg/l)
Julian Day 2004
ASSA 7
DO (mg/l) (mg/l)
Julian Day 2004
Figure 5.4 (f): Daily average (black) and diel (green) time series, Stations XCM1562 and ASSA 7, Chincoteague Bay, January
1 – December 31, 2004.
106
Predicted (Mg/L)
Predicted (Mg/L)
Figure 5.5 (a): Correlation coefficient (R2) between diel-cycle-adjusted modeled and observed DO, DNR
Stations 1-12, baseline conditions, January 2001- August 2005.
107
Figure 5.5 (b): Correlation coefficient (R2) between diel-cycle-adjusted modeled and observed DO, DNR
Stations 13-24, baseline conditions, January 2001- August 2005.
108
Figure 5.5 (c): Correlation coefficient (R2) between diel-cycle-adjusted modeled and observed DO, DNR Stations 25-27,
baseline conditions, January 2001- August 2005.
109
Figure 5.5 (d): Correlation coefficient (R2) between diel-cycle-adjusted modeled and observed DO, ASIS Stations 1-9,
baseline conditions, January 2001- August 2005.
110
Figure 5.5 (e): Correlation coefficient (R2) between diel-cycle-adjusted modeled and observed DO, ASIS Stations 10-18,
baseline conditions, January 2001- August 2005.
111
5.2 Sensitivity Analyses
Model sensitivity to variations in model parameters is an important characteristic of a
model. Often, one needs to find out how model results vary as model parameters are
changed and to identify the most influential parameters in determining the accuracy of
model results. The accuracy of model output is influenced by a number of uncertainties
from measured data, model formulations, and model parameters. Sensitivity analysis is
a useful tool to clarify the relationship between uncertainty in parameter values and
model results. With the efforts of HEM3D model calibration and verification, sensitivity
analyses were conducted to test the following effects: (1) the Ocean City Wastewater
Treatment Plant outfall (2) phytoplankton and organic nutrient settling rate and (3) the
inputs of ground water discharge.
5.2.1 Ocean City Wastewater Treatment Plant outfall
Within the MCBs modeling domain, there is a major point source outfall offshore of the
Ocean City Inlet -The Ocean City Wastewater Treatment Plant (OC WWTP). The
outfall is approximately 4600 feet offshore of Ocean City Inlet in the Atlantic Ocean at
the depth of 30 feet. This is a secondary treatment plant which treats the wastewater by
biological method after the primary treatment. The NPDES (National Pollutant
Discharge Elimination System) permitted flow is 14 MGD with a TN concentration of 19
mg/l, and a TP concentration of 3 mg/l. To be consistent with the most conservative
approach, the flow rate and TN and TP concentrations were set at the permit level in all
the HEM3D model runs conducted. This sensitivity analysis was conducted in order to
test whether this facility has the potential to significantly affect the water quality
condition inside the MCBs. The sensitivity tests were performed by reducing TN and TP
loads by 20%, 40% and 60%, from the base condition over 2001-2004 with all other
conditions remaining the same.
The chlorophyll-a modeling result for base and incremental reduction scenarios at DNR
and ASIS stations were examined. The exceedance rates for criteria (15 μg/l or 50 μg/l,
depending on location) are shown in Table 5.2. When comparing exceedance rates of
these criteria in both the baseline scenario and scenarios with reductions of 20%, 40%
and 60% from the Ocean City WWTP, there is practically no change. When the actual
chlorophyll concentrations were examined, the change is on the order of 0.0001μg/l,
which is within the numerical error and considered statistically no difference from
baseline condition. The DO exceedance results displayed similar results (not shown).
Thus, it is concluded that given the present setting of the model domain and the open
boundary condition, the Ocean City WWTP outfall does not affect chlorophyll-a and DO
inside the MCBs. It should be noted that the present modeling was conducted in a
model domain limited by the cross-shore ocean extent, and the boundary conditions
used are based on climatologically-averaged values, and thus do not explicitly account
for event-driven coastal ocean phenomena.
112
Table 5.2: Exceedance rates for Chlorophyll-a under Ocean City WWTP 20%, 40% and
60% incremental reduction scenarios.
Percent Chla > 15
Station TMDL-basin
(SAV growing stations)
XBM1301 Chincoteague Bay
XBM8149 Chincoteague Bay
XCM0159 Chincoteague Bay
XDN0146 Isle of Wight Bay
XDN2340 Isle of Wight Bay
XDN2438 Isle of Wight Bay
XDN3445 Assawoman Bay
XDN4851 Assawoman Bay
XDN6454 Assawoman Bay
XDN7261 Assawoman Bay
XDN7545 Assawoman Bay
ASSA 1. Sinepuxent Bay
ASSA 2. Sinepuxent Bay
ASSA 3. Newport Bay
ASSA 5. Chincoteague Bay
ASSA 6. Chincoteague Bay
ASSA 7. Chincoteague Bay
ASSA 8. Chincoteague Bay
ASSA 9. Chincoteague Bay
ASSA 11. Chincoteague Bay
ASSA 12. Chincoteague Bay
ASSA 13. Chincoteague Bay
ASSA 14. Chincoteague Bay
ASSA 15. Chincoteague Bay
ASSA 16. Sinepuxent Bay
ASSA 17. Sinepuxent Bay
ASSA 18. Sinepuxent Bay
Grow Season 2001-2004
Base
OCwwtp- OCwwtp20%
40%
0.00%
0.00%
0.00%
2.31%
2.31%
2.31%
2.85%
2.72%
2.72%
0.00%
0.00%
0.00%
4.21%
4.21%
4.21%
0.00%
0.00%
0.00%
3.94%
3.94%
3.94%
4.48%
4.48%
4.48%
2.99%
2.99%
2.99%
3.26%
3.26%
3.26%
1.22%
1.22%
1.22%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
8.83%
8.83%
8.83%
4.48%
4.48%
4.48%
0.00%
0.00%
0.00%
8.15%
8.15%
8.15%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
3.53%
3.53%
3.53%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
Percent Chla>50
Grow Season 2001-2004
(Non-SAV growing stations)
BSH0008 Bishopville Prong 19.43% 19.43% 19.43%
MKL0010 Manklin Creek
2.45%
2.45%
2.45%
SPR0002 Shingle Landing Pron5.57%
5.57%
5.57%
SPR0009 Shingle Landing Pron6.66%
6.66%
6.66%
TUV0011 Turville Creek
0.54%
0.54%
0.54%
TUV0019 Turville Creek
1.36%
1.36%
1.36%
XBM3418 Chincoteague Bay 0.00%
0.00%
0.00%
XBM5932 Chincoteague Bay 0.00%
0.00%
0.00%
XCM1562 Chincoteague Bay 0.00%
0.00%
0.00%
XCM4878 Newport Bay
0.00%
0.00%
0.00%
XDM4486 Bishopville Prong 38.59% 38.59% 38.59%
XDN3724 St. Martin River
1.36%
1.36%
1.36%
XDN4312 St. Martin River
2.72%
2.72%
2.72%
XDN4797 St. Martin River
5.98%
5.98%
5.98%
XDN5737 Assawoman Bay
0.00%
0.00%
0.00%
AYR0017 Ayer Creek
0.14%
0.14%
0.14%
ASSA 4. Newport Bay
0.00%
0.00%
0.00%
ASSA 10. Chincoteague Bay 0.00%
0.00%
0.00%
OCwwtp60%
0.00%
2.31%
2.72%
0.00%
4.21%
0.00%
3.94%
4.48%
2.99%
3.26%
1.22%
0.00%
0.00%
8.83%
4.48%
0.00%
8.15%
0.00%
0.00%
0.00%
0.00%
0.00%
3.53%
0.00%
0.00%
0.00%
0.00%
Annual average 2001-2004
Base
OCwwtp- OCwwtp20%
40%
0.00%
0.00%
0.00%
1.16%
1.16%
1.16%
1.44%
1.37%
1.37%
0.00%
0.00%
0.00%
2.12%
2.12%
2.12%
0.00%
0.00%
0.00%
1.99%
1.99%
1.99%
2.26%
2.26%
2.26%
1.51%
1.51%
1.51%
1.64%
1.64%
1.64%
0.62%
0.62%
0.62%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
4.73%
4.73%
4.73%
2.26%
2.26%
2.26%
0.00%
0.00%
0.00%
4.79%
4.79%
4.79%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
1.78%
1.78%
1.78%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
OCwwtp60%
0.00%
1.16%
1.37%
0.00%
2.12%
0.00%
1.99%
2.26%
1.51%
1.64%
0.62%
0.00%
0.00%
4.73%
2.26%
0.00%
4.79%
0.00%
0.00%
0.00%
0.00%
0.00%
1.78%
0.00%
0.00%
0.00%
0.00%
Annual average 2001-2004
19.43%
2.31%
5.57%
6.66%
0.54%
1.36%
0.00%
0.00%
0.00%
0.00%
38.59%
1.36%
2.72%
5.98%
0.00%
0.14%
0.00%
0.00%
10.00%
1.23%
2.81%
3.36%
0.27%
0.68%
0.00%
0.00%
0.00%
0.00%
19.66%
0.68%
1.37%
3.01%
0.00%
0.07%
0.00%
0.00%
10.00%
1.23%
2.81%
3.36%
0.27%
0.68%
0.00%
0.00%
0.00%
0.00%
19.66%
0.68%
1.37%
3.01%
0.00%
0.07%
0.00%
0.00%
10.00%
1.23%
2.81%
3.36%
0.27%
0.68%
0.00%
0.00%
0.00%
0.00%
19.66%
0.68%
1.37%
3.01%
0.00%
0.07%
0.00%
0.00%
10.00%
1.16%
2.81%
3.36%
0.27%
0.68%
0.00%
0.00%
0.00%
0.00%
19.66%
0.68%
1.37%
3.01%
0.00%
0.07%
0.00%
0.00%
113
5.2.2 Effect of phytoplankton and organic nutrient settling rate
The settling velocity is the fundamental property governing the motion of the particles in
water. The settling rate (the product of settling velocity and the concentration) of
phytoplankton and organic matter links the particulate matter in the water column with
the sediment processes. For TN, TP and total carbon (TC), approximately 40% of the
nonliving organic components are in the particulate forms. There are also the living
organic TN, TP and TC, which are part of the phytoplankton biomass. The living and
nonliving particulate TN, TP and TC are the primary source of deposition flux into the
sediment. Once settled in the sediment, the particulate organic nitrogen (PON) and
phosphorus (POP), and POC are transformed into dissolved forms through the
sediment diagenesis process. The ammonia, nitrate, phosphorus, SOD and methane
fluxes are generated as a result of interactions between sediment and the overlying
water column. When nutrient loads are reduced as the TMDL is implemented, in the
MCBs, the reduced particulate portion of the load will have effects on the settling rate
and in turn affect the SOD and nutrient fluxes. Since settling velocity is extremely
difficult to measure and the exact forms are not known, it is normal practice to specify
them as constants. In the HEM3D model, the phytoplankton’s settling velocities are set
at 0.25, 0.15 and 0.01 m/day for diatoms, green algae and cyanobacteria respectively.
For organic carbon, nitrogen and phosphorus, the settling velocities are set at 1 m/day.
It is well known that the diatom component of phytoplankton can change the settling
velocity appreciably by orders of magnitude (Collins and Wlosinski, 1983; Jorgensen,
1979). A sensitivity test was conducted by varying the diatom settling velocity from 0.25
m/day to 0.50 m/day, 100% larger than the prescribed value. The results at station
AYR0017 for 2004 are shown in Figure 5.6. Due to the high settling velocity (0.5 m/day),
the phytoplankton are not retained in the water column long enough to undergo net
growth. As a consequence, the chlorophyll-a concentration became lower, which is not
reflective of the observed value. In this case, by increasing the settling velocity of
diatoms by 100%, the spring chlorophyll-a concentration was reduced by about 50%
(from 36 μg/l to 18 μg/l). In general, concentrations of chlorophyll-a, total N, P and C
reduced when the settling velocity increased as an inversely relationship. For different
phytoplankton species, the responses to changes in settling velocities vary through the
different seasons. For example, green algae and cyanobacteria tend to be affected
more in the summer and fall while diatoms are affected in the spring.
114
Figure 5.6: Sensitivity test comparing the effect of settling velocities of diatom species
on Chlorophyll-a concentration at station AYR0017 in 2004.
Nine sensitivity runs were conducted by changing the settling velocities for
phytoplankton, organic carbon, organic nitrogen, and organic phosphorus one each at a
time. Table 5.3 summarizes the results of changing the settling velocities on
chlorophyll-a concentration. Overall, the chlorophyll-a concentrations are sensitive to
the phytoplankton settling velocities. The change of organic N, P and C particulate
settling velocity has less effect on the chlorophyll-a concentration. The effect of labile
organic component on chlorophyll concentration is slightly higher, while the refractory
component has almost no effect. Thomann et al. (1975) reported that phytoplankton
settling velocity can have an effect on the nutrient limitation function and thus on the
nutrient uptake, which has more of a nonlinear effect. An important question for the
sensitivity analysis is: when the TMDL scenarios are implemented, does the settling
velocities need to be changed? The answer is no. When the TMDL reduction scenarios
were conducted, the settling velocity does not change, but the settling “rate” will be
changed due to the change of the concentration in the water column. During the TMDL
reduction scenario, if the load reduction leads to concentration reduction, the settling
rate (the product of settling velocity and the concentration) will be reduced accordingly
and approximately in proportion to the change of the particulate concentration in the
water column. This is one of the ways that sediment concentration will be gradually
improved.
115
Table 5.3: Effect of changing settling velocities of phytoplankton and organic nutrients
on chlorophyll-a concentration.
Parameter Description
in unit m/day
Settling
Net settling Actual Settling
Input
Response of ChlA at MCBs stations (%)
velocity
velcoity
Change
BUSH0008
SPR0009
XDM4486 AYR0017
(Base)
velocity
(Sensitivty test)
Algal settling rate
Diatom
0.25
0.25
0.5
100%
-41%
-49%
-51%
-46%
Green algae
0.15
0.15
0.075
-50%
25%
23%
21%
18%
Cyanobacteria
0.01
0.01
0.005
-50%
15%
13%
12%
10%
labile
1
1
0.5
-50%
1%
1%
1%
0%
refractory
1
1
0.5
-50%
0%
0%
0%
0%
labile
1
1
2
100%
-6%
-7%
-7%
-9%
refractory
1
1
2
100%
0%
1%
1%
0%
labile
1
1
2
100%
-4%
-3%
-3%
-2%
refractory
1
1
0.5
-50%
0%
0%
0%
0%
Organic carbon settling rate
Organic nitrogen settling rate
Organic phosphorus settling rate
5.2.3 Effects of groundwater discharge
In the Maryland Coastal Bays, it was reported that groundwater discharge in the Atlantic
Coastal Bays basin can enhance nitrogen load (Dillow and Greene, 1999). The
pathways that groundwater can deliver nitrogen from the land to the Bay are either
through direct discharge of groundwater or through base flow to streams that discharge
to the Bays. Recently, Fertig et al. (2013) suggested that there is terrestrial nutrient
source discharge into the Chincoteague Bay near Johnson Bay, as shown in Figure 5.7
(a); the source is unknown. Cornwell and Owens (2013) further suggested that
sediments are the key source and have strong seasonality. They hypothesize that
nitrate can be de-nitrified or converted to ammonium at the groundwater/ wetland
interface, and subsequently released from the underlying aquifer into Chincoteague Bay
proper, with a seasonal switch mechanism in the later summer. For the present water
quality model setup, the watershed model input has already generated the interflows,
which discharge into the edge of the stream. Thus, the component of groundwater that
was not considered is the groundwater discharge directly from the aquifer underneath
the Bay into the surface water. In order to test how sensitive the present water quality
calibration is to the nitrogen directly discharged from the groundwater to surface water,
the interflow from the edge of the land-water margin was re-distributed to the open
water of Johnson Bay. The model was re-run and the total nitrogen concentrations
were compared at XBM5932 and XBM3418 for before-and-after conditions. The left
panel of the Figure 5.7 (b) shows the results before re-distribution, and the right panel of
Figure 5.7 (b) shows the result after the interflow was re-distributed to the open water.
116
It was recognized that the right panel results, which mimic the groundwater release from
the underneath of the open bay, has slightly better comparison with observed data in
terms of magnitude and phase, especially at the maximum in later summer. This
suggests that direct groundwater discharge through under-the-Bay aquifers to the Bay
is a plausible hypothesis, which is suggested by the field experiment. From the TMDL
point of view, while this is of scientific interest, the full understanding of the process is
yet to be uncovered to answer the questions as to actual ammonia release mechanism,
the amount of release, and the extent to which it affects the MCBs. The percentage
difference in Figure 5.7 (b) is on the order of 5-7%, which is not substantial and also
may be localized in the vicinity of Johnson Bay only.
Johnson Bay
Figure 5.7 (a): Location of Johnson Bay in the middle of the Chincoteague Bay and the
nearby DNR stations XBM5932 and XBM3418.
117
Total nitrogen (mg/l)
Figure 5.7(b): Comparison of modeled total nitrogen concentrations under groundwater
release at edge-of-stream vs. bay-floor release, Johnson Bay area, Chincoteague Bay,
2001-2005.
118
CHAPTER 6: TMDL SCENARIO DEVELOPMENT
The purpose for developing coupled hydrodynamic and water quality models, which link
with the watershed model, was to determine the total load of nutrients that the MCBs
could assimilate while maintaining the State’s water quality standards (WQS). To
support the TMDL analysis, the calibrated model was used to simulate the effect of
reduced nutrient loading on DO and chlorophyll concentrations, and the exceedance
frequency of relevant water quality criteria. The base condition was conducted from
January 2000 through December 2004, while the load reduction analysis time period
was calendar years 2001 – 2004.
6.1. Developing nutrient load reduction scenario
6.1.1 Critical conditions
One of the primary concerns in the MCBs is its recurring hypoxia in the bottom waters.
Upon completion of the calibrated model from 2000 to 2004 as described in Chapters 4
and 5, the model was used to develop the TMDL that would attain water quality
standards under the critical conditions. Estuaries and coastal waters are complex
systems and present a challenge for defining the critical conditions. The goal is to
estimate the loading capacity during periods when estuaries and coastal waters are
most vulnerable to pollutant sources. Based on data analysis and model runs, the
factors contributing to low DO and high chlorophyll-a in MCBs are (1) nutrient loads
entering into the Bays; (2) hydrodynamic residence time and vertical stratification as a
result of freshwater and tides; (3) water temperature and plant respiration; (4) sediment
oxygen demand and (5) carbonaceous oxygen demand. All these factors can
contribution to low DO and high chlorophyll in MCBs, but not all contribute equally,
depending on the time, locations and the associated mechanisms. In term of temporal
variation of DO and chlorophyll-a, it has been found that as flow and nitrogen and
phosphorus loads increase, the DO levels decrease and chlorophyll-a level increase.
This is clearly demonstrated in Chapter 4; DO is lower and chlorophyll-a is higher in the
wet years (2003-2004) versus the average and low flow years (2001-2002). Therefore,
high-flow years are the critical periods. A strong spatial gradient exist for both physical
characteristics and water quality condition. For example, Allen et al. (2007) showed the
gradient of tidal range, salinity, and wave energy from the south to the north of
Chincoteague Bay to reach Newport Bay where it has relatively low salinity, low tide,
and low wave energy. The water quality condition in the MCBs also has a spatial
gradient with the worst condition is generally occurred in the head water of the streams
as part of the tributaries of the open Bay, and the best condition generally near the
inlets (DNR, 2004). There are two basins in the MCBs where hypoxia was consistently
observed: Saint Martin River of the Isle of Wight Bay and Newport Bay. The high nonpoint source loads from the upper reach of the St. Martin River and headwater of the
tributaries are believed to be the dominant causes of low DO. On the other hand, in
Newport Bay, the combination of the moderately high nutrient load coupled with the long
residence time is believed to be the root cause. Hypoxia and anoxia occur most
commonly in systems that are seasonally stratified, such as Chesapeake Bay. For a
shallow, well mixed water column, the reaeration from the air-water exchange would
tend to be sufficient to prevent the development of hypoxic conditions. In metabolically
active aquatic systems such as the MCBs, however, a dissolved oxygen diel cycle can
119
develop due to the presence of large concentrations of algae (Boynton, 1996; Dennison
et al, 2012). Under diel oscillation, hypoxia can develop in the early morning and swing
to the DO saturation condition in the afternoon within a diel time scale. The diel oxygen
change was found to be correlated closely with the chlorophyll concentration, an
indication that hypoxia is largely controlled by the algal concentration. In terms of
nutrient limitation, N:P ratios tend to be low in the MCBs with nitrogen tending to be
limiting in most of the open waters. In the isolated headwaters of some tributaries,
phosphorus can be limiting. The nitrogen and phosphorus limitations can vary with the
season and the flow as well.
6.1.2 Margin of Safety (MOS)
In the development of a TMDL, a Margin of Safety (MOS) is required to account for the
uncertainties involved. In general this uncertainty can be caused by insufficient or poor
quality data, a lack of knowledge about the pollution effects, and errors in estimating the
loading of the water body. In this approach in the MCBs, the uncertainties were
accounted for implicitly by using conservative assumptions. Specifically, the
conservative modeling assumptions used include: (1) incorporating a wide range of
meteorological conditions - extremely dry to extremely wet—for continuous simulation;
(2) the organic material originated from the land was directly transported into the stream,
without considering the loss of organic material from the land surface on the way to the
stream; (3) all land areas are considered directly connected to streams; (4) nutrient
sequestration or transformation occurring in retention ponds or wetlands is not
considered; (5) point sources discharges and concentrations were set at the permit
value; and (6) the DO diel cycle was incorporated. The effect of DO diel cycle was
discussed in detail in Chapter 5.1.
6.1.3 Seasonal variation
The seasonal variations of the phytoplankton optimal growth rate and the bio-chemical
reactions are important factor affecting the water quality. These rates in general are a
function of water temperature and light condition, both of which change daily and the
seasonally. The hydrology in the region also exhibits a seasonal cycle, with the spring
having the largest runoff, followed by the winter, with less runoff during the summer and
fall seasons. In developing the TMDLs, the seasonal variation was considered for four
periods: December through February for winter, March through May for spring, June
through August for summer, and September through November for fall. Model runs
were performed with direct inputs of temperature, light as well as various hydrologic,
meteorological, and loading conditions to simulate the water quality conditions, thus fully
evaluating the response to the seasonal variations. A key aspect of nitrogen and
phosphorus dynamics in the MCBs modeling system is that daily loads were generated
and used as a forcing function to the ecosystem simulated in the HEM3D model. The
model can thus be viewed as an integrator of the day-to-day nutrient variation, which
incorporated the temperature dependent physical bio-chemical processes, described
above, and presented the results in a seasonal-dependent time scale. Problems
associated with eutrophication are most likely to occur during the growing season,
during which there is typically less stream flow available to flush the system, more
sunlight to grow aquatic plants, and warmer temperature, which are favorable conditions
120
for the biological processes of plant growth and decay of dead plant matter. Therefore,
the load reduction scenarios based on the temporal analysis were broken into annual
average as well as growing seasons (May 1 – October 31).
6.2 Developing the TMDL scenario
In the watershed load analysis (Chapter 2.2), it was shown that the nutrient loads of
MCBs are from five major sources: (1) atmospheric deposition; (2) non-point sources
loading (including those from different land uses and the interflow from the shallow
groundwater aquifer); (3) point sources; (4) shoreline erosion and (5) septic tanks.
Among theses sources, the non-point source was the most significant portion of loads
that caused the high algal concentration and hypoxia in the MCBs. Therefore, the
management strategies for the MCBs is to incrementally reduce the non-point source
nutrient loading and to find out to what extent the modeled DO concentration will satisfy
the TMDL target. Thus, the TMDL development requires the identification and
evaluation by model of various management alternatives for achieving water quality
goals. The HEM3D model is used to determine the relationships between changing
pollutant loads and the water quality response, and to project the future water quality
conditions under “what-if” scenario condition. For the scenario model set up, the water
quality boundary condition remained unchanged for various scenarios, which was
reasonable since open boundary condition is located 7-9 km away from the open coast
(and the inlet), and the concentration is much less than the riverine boundaries in the
upstream. The effects of changes in nutrient fluxes from the sediment were
incorporated by using the results of the benthic sediment flux model. For each scenario,
the load reduction was simulated by running a 5-year simulation three times to achieve
equilibrium condition for the concentration in the sediment and overlaying water. The
reductions were then applied to all of the simulated years (calendar years 2001-2004)
during the TMDL analysis. To represent the simulated DO and chlorophyll-a
concentration for each of the Water Quality Limited Segments (WQLS), a volumeweighted calculation was made with chlorophyll-a and DO for each grid cell. In order to
encompass a wide range of hydrologic conditions during the TMDL analysis period, the
full data set was divided into three hydrology years: 2001 - an average year; 2002 - a
dry year; and 2003-2004, wet years.
In developing the TMDL scenario, multiple reduction scenarios were run to determine
the assimilative capacity of the waterbody. These scenarios include baseline conditions,
natural conditions, incremental reduction and geographic isolation scenarios to
determine the best possible combination of load- and source-reduction to attain water
quality standards. A set of model scenarios were developed based on the percentage
reduction to simulate the changes of chlorophyll-a and DO as the nitrogen and
phosphorus loads coming into the MCBs were reduced. The natural conditions
scenario was also conducted to check whether there are areas where the water quality
endpoints would not be met even if the watershed was returned to a natural state of allforested and beach areas. In the natural conditions scenario, the atmospheric
deposition was reduced to 10% of the baseline atmospheric load, all land uses except
beach were changed to forest, and septic loads as well as all point sources were
removed. Incremental load reduction scenarios were conducted by reducing the
121
loading of dissolved inorganic nitrogen: ammonia, nitrate and nitrite, particulate organic
nitrogen, dissolved organic nitrogen, phosphate, and organic phosphorus by 20%, 40%,
60% and Maximum Practicable Anthropogenic Reduction (MPAR) to determine which
loading conditions in these watersheds would result in the attainment of water quality
standards (WQS). The 20%, 40% and 60% scenario are reductions straight from the
total loading of the base condition. For the MPAR scenario, percent reductions are
calculated from CBP-P5 scenario results for the Eastern Shore for total nitrogen and
total phosphorus. CBP-P5 scenario results are available for the following scenarios: noaction (no reductions applied to the baseline); E-3 (Everyone, Everything, Everywhere –
maximum reductions from all sources); 2009 Progress (incorporates reductions from
implementation through 2009); and 2010 progress (incorporates reductions from
implementation through 2010). For each land use sector, the mean percent reduction
from the baseline and the three available reduction scenarios was used to calculate the
reduction rate for the Coastal Bays watershed model: no-action to E3; 2009 progress to
E3; and 2010 progress to E3. The MPAR reductions from specific non-point source
sectors are shown in Table 6.1.
Table 6.1: Maximum Practicable Anthropogenic Reduction (MPAR) Percentages for each
non-point source sector based on CBP-Phase 5.3.2 scenario results.
Non-Point Source Sector
Animal Feeding Operations (AFO)
Crop
Pasture
Urban
Septic
Forest
TN Reduction
67%
64%
45%
51%
57%
0%
TP Reduction
69%
34%
46%
68%
0%
0%
TSS Reduction
29%
50%
54%
73%
0%
0%
The model results, showing WQS exceedance rates under baseline and incremental
scenarios: 20%, 40%, 60%, natural condition, and MPAR are presented in Table 6.2 for
all 27 DNR and 18 ASIS stations. The exceedance rate was determined for growing
season as well as annual average conditions, and spatially aggregated into five
assessment basins—Assawoman Bay, Isle of Wight Bay, Sinepuxent Bay, Newport Bay
and Chincoteague Bay (in columns 2 and 3), consistent with Maryland’s WQLS listing.
Table 6.2 demonstrates that there are 6 stations: MKL0010, XDM4486, BSH0008,
SPR0009, AYR0017 and ASSA4, located in the Isle of Wight Bay and in Newport Bay,
having the highest exceedance rate during both growing season and average annual
conditions. Among these six stations, the first four stations are in the Saint Martin River
of the Isle of Wight; and the other two stations are in the Newport Bay. The response of
these 6 stations to the incremental reductions of 20%, 40%, 60%, MPAR and natural
conditions are shown in Figure 6.1, which shows a decreasing trend of the exceedance
rate as the reduction increases, with an asymptotic approach to the natural condition. It
can be seen that DO exceedance rates are below 10% criteria for all 6 stations when 60%
incremental reduction is executed. The 60% is, thus, considered to be the theoretical
maximum reduction that is required for any place within the MCBs. To demonstrate that
the ideal 60% reduction would improve the water quality condition in real time, the time
122
series results for DO base condition and 60% reduction from 2001-2004 are shown in
Figure 6.2 (a)–(e). In these figures, the time series with 60% reduction (represented by
magenta) and without 60% reduction (represented by blue) were directly compared. It
can be seen that under the 60% reduction scenario all of the stations were improved
and meet the DO criteria. The improvements were particularly obvious for the 6 most
severely impaired stations identified in Figure 6.1.
123
Table 6.2: DO exceedance rate under baseline conditions and reduction scenarios of 20%, 40%, 60%, natural conditions and
MPAR for all DNR and ASIS stations.
Stations
TMDL-basin
MD-8digit-basin
BSH0008
MKL0010
SPR0002
SPR0009
TUV0011
TUV0019
XBM1301
XBM3418
XBM5932
XBM8149
XCM0159
XCM1562
XCM4878
XDM4486
XDN0146
XDN2340
XDN2438
XDN3445
XDN3724
XDN4312
XDN4797
XDN4851
XDN5737
XDN6454
XDN7261
XDN7545
AYR0017
ASSA 1.
ASSA 2.
ASSA 3.
ASSA 4.
ASSA 5.
ASSA 6.
ASSA 7.
ASSA 8.
ASSA 9.
ASSA 10.
ASSA 11.
ASSA 12.
ASSA 13.
ASSA 14.
ASSA 15.
ASSA 16.
ASSA 17.
ASSA 18.
Bishopville Prong
Manklin Creek
Shingle Landing Prong
Shingle Landing Prong
Turville Creek
Turville Creek
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Newport Bay
Bishopville Prong
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Assawoman Bay
St. Martin River
St. Martin River
St. Martin River
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Ayer Creek
Sinepuxent Bay
Sinepuxent Bay
Newport Bay
Newport Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Sinepuxent Bay
Sinepuxent Bay
Sinepuxent Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Newport Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Asswoman Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Newport Bay
Sinepuxent Bay
Sinepuxent Bay
Newport Bay
Newport Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Sinepuxent Bay
Sinepuxent Bay
Sinepuxent Bay
DO exceedance rate (Growing Season 2001-2004)
Base
20%
40%
60%
natural C MPAR
20.79%
11.55%
3.53%
1.77%
1.77%
2.72%
54.18%
32.20%
17.80%
5.16%
0.27%
1.36%
1.90%
0.95%
0.41%
0.27%
0.68%
0.00%
19.84%
13.32%
8.42%
2.99%
1.63%
6.11%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.54%
0.00%
0.00%
0.41%
0.68%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
5.03%
0.68%
0.00%
0.14%
0.82%
0.00%
42.66%
34.38%
20.79%
5.03%
2.58%
18.48%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
2.04%
1.09%
0.54%
0.41%
0.68%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
8.42%
3.67%
0.54%
0.82%
1.36%
0.68%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.27%
0.00%
0.00%
0.00%
0.00%
0.00%
5.57%
1.09%
0.27%
0.54%
0.95%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
DO exceedance rate (Annual 2001-2004)
Base
20%
40%
60%
natural C MPAR
10.48%
5.82% 1.78%
0.89%
1.03%
1.37%
40.89%
36.99% 17.81%
5.32%
0.14%
9.32%
0.96%
0.48% 0.21%
0.14%
0.34%
0.00%
10.00%
6.71% 4.25%
1.51%
0.82%
3.08%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.27%
0.00% 0.00%
0.21%
0.34%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
2.53%
0.34% 0.00%
0.07%
0.41%
0.00%
21.51%
17.33% 10.48%
2.53%
2.26%
9.32%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
1.03%
0.55% 0.27%
0.21%
0.34%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
4.25%
1.85% 0.27%
0.41%
0.68%
0.34%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.14%
0.00% 0.00%
0.00%
0.00%
0.00%
2.81%
0.55% 0.14%
0.27%
0.48%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
0.00%
0.00% 0.00%
0.00%
0.00%
0.00%
124
DO Growing Season 2001-2004
Figure 6.1: Response of six stations having the highest DO exceedance rates to the incremental reductions, MPAR and
natural conditions scenarios.
125
In addition to the DO criteria, Table 6.3 (a) and (b) present the chlorophyll-a
exceedance rates under baseline, reductions of 20%, 40% and 60%, MPAR, natural
conditions, growing season TMDL, and average annual TMDL scenarios. There are
two endpoints for the chlorophyll-a concentration: 50 μg/l for the non-SAV growing area,
and 15μg/l for the SAV growing area. Tables 6.3 (a) and (b) show the exceedance
rates for all of the stations in the MCBs under non-SAV growing and SAV growing areas
chlorophyll-a endpoint respectively. From Table 6.3 (a), it was identified that there were
two stations: BSH008 and XDM4488, both located near Bishopville Prong of the Isle of
Wight Bay, exceeded the 50ug/l chlorophyll-a concentration endpoint in the non-SAV
growing areas. For the 4 year time period simulated, the model predicted that a 40%
reduction in nutrient loading would have eliminated the water quality impairment for
chlorophyll-a in all of the basins in MCBs, as the frequency of chlorophyll-a
exceedances fell below 10% in each area during the growing season. It is noted that
the two stations that exceeded the chlorophyll-a endpoint were also the stations
corresponding to the lower DO condition and thus reducing nutrient loading will produce
lower chlorophyll-a concentrations as well as improving the daily oxygen concentration.
It was estimated that the 40% nutrient reduction reduced average chlorophyll a
concentrations by approximately 60 ug/l. For the SAV growing areas (and the
surrounding 2500-ft buffer zone), the more stringent endpoint, a maximum chlorophyll-a
concentration not to exceed 15ug/l is applied. Table 6.3 (b) shows that the highest
exceedance rates under base conditions were 8.83% and 8.15% in the ASIS Stations 3
and 7, located near Newport Bay and Johnson Bay, respectively. MDE’s assessment
methodology requires that the chlorophyll-a concentration not exceed 15μg/l more than
10 percent of the time under both average annual and growing season conditions.
Therefore, no TMDL action was required to meet the chlorophyll a TMDL endpoint with
respect to SAV grow zones.
126
DO 60% Reduction
Figure 6.2 (a): DO time series, 2001- August 2005, baseline conditions (blue), 60% reduction scenario (magenta), and
observed (symbol) at DNR Stations 1-12.
127
DO 60% Reduction
Figure 6.2 (b): DO time series, 2001- August 2005, for baseline conditions (blue), 60% reduction scenario(magenta), and
observed (symbol) at DNR Stations: 13 -24.
128
DO 60% Reduction
Figure 6.2 (c): DO time series, 2001-August 2005 for baseline conditions (blue), and 60% reduction scenario (magenta), and
observed (symbol) at DNR Stations 25 - 27.
129
DO 60% Reduction
Figure 6.2 (d): DO time series, 2001-August 2005, baseline conditions (blue), 60% reduction scenario (magenta) and
observed data (symbol) at ASIS Stations 1-9.
130
DO 60% Reduction
Figure 6.2 (e): DO time series, 2001-August 2005, baseline conditions (blue), 60% reduction scenario (magenta), and
observed data (symbol) at ASIS Stations 10 -18.
131
Table 6.3 (a): Modeled chlorophyll-a exceedance rates under baseline conditions and reduction scenarios of 20%, 40%,
60%, natural and MPAR conditions for non-SAV growing zone with Chl-a endpoint greater than 50 μg/l.
Percent Chla > 50
Station
TMDL-basin
MD-8digit-basin
BSH0008
MKL0010
SPR0002
SPR0009
TUV0011
TUV0019
XBM3418
XBM5932
XCM1562
XCM4878
XDM4486
XDN3724
XDN4312
XDN4797
XDN5737
AYR0017
ASSA 4.
ASSA 10.
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Newport Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Asswoman Bay
Newport Bay
Newport Bay
Chincoteague Bay
Bishopville Prong
Manklin Creek
Shingle Landing Prong
Shingle Landing Prong
Turville Creek
Turville Creek
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Newport Bay
Bishopville Prong
St. Martin River
St. Martin River
St. Martin River
Assawoman Bay
Ayer Creek
Newport Bay
Chincoteague Bay
Chla exceedance rate (Growing Season 2001-2004)
Base
20%
40%
60%
natural C MPAR
19.43% 8.70% 3.40%
0.82% 0.00%
6.66%
2.45% 0.00% 0.00%
0.00% 0.00%
0.00%
5.57% 2.72% 0.68%
0.00% 0.00%
2.58%
6.66% 4.08% 0.00%
0.00% 0.00%
2.04%
0.54% 0.00% 0.00%
0.00% 0.00%
0.00%
1.36% 0.00% 0.00%
0.00% 0.00%
0.00%
0.00% 0.00% 0.00%
0.00% 0.00%
0.00%
0.00% 0.00% 0.00%
0.00% 0.00%
0.00%
0.00% 0.00% 0.00%
0.00% 0.00%
0.00%
0.00% 0.00% 0.00%
0.00% 0.00%
0.00%
38.59% 18.07% 5.16%
1.49% 0.00%
10.05%
1.36% 0.54% 0.00%
0.00% 0.00%
0.41%
2.72% 0.95% 0.27%
0.00% 0.00%
0.82%
5.98% 3.53% 0.68%
0.00% 0.00%
2.85%
0.00% 0.00% 0.00%
0.00% 0.00%
0.00%
0.14% 0.00% 0.00%
0.00% 0.00%
0.00%
0.00% 0.00% 0.00%
0.00% 0.00%
0.00%
0.00% 0.00% 0.00%
0.00% 0.00%
0.00%
Chla exceedance rate (Annual 2001-2004)
natural MPAR
base
20%
40%
60%
10.00% 4.38% 1.71% 0.41% 0.00% 3.36%
1.23% 0.00% 0.00% 0.00% 0.00% 0.00%
2.81% 1.37% 0.34% 0.00% 0.00% 1.30%
3.36% 2.05% 0.00% 0.00% 0.00% 1.03%
0.27% 0.00% 0.00% 0.00% 0.00% 0.00%
0.68% 0.00% 0.00% 0.00% 0.00% 0.00%
0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
19.66% 9.18% 2.60% 0.75% 0.00% 5.21%
0.68% 0.27% 0.00% 0.00% 0.00% 0.21%
1.37% 0.48% 0.14% 0.00% 0.00% 0.41%
3.01% 1.78% 0.34% 0.00% 0.00% 1.44%
0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
0.07% 0.00% 0.00% 0.00% 0.00% 0.00%
0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
132
Table 6.3 (b): Modeled chlorophyll-a exceedance rates under baseline conditions and reduction scenarios of 20%, 40%,
60%, natural and MPAR conditions for SAV growing zone with Chl-a endpoint greater than 15μg/l.
Percent Chla >15
Station
TMDL-basin
MD-8digit-basin
XBM1301
XBM8149
XCM0159
XDN0146
XDN2340
XDN2438
XDN3445
XDN4851
XDN6454
XDN7261
XDN7545
ASSA 1.
ASSA 2.
ASSA 3.
ASSA 5.
ASSA 6.
ASSA 7.
ASSA 8.
ASSA 9.
ASSA 11.
ASSA 12.
ASSA 13.
ASSA 14.
ASSA 15.
ASSA 16.
ASSA 17.
ASSA 18.
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Sinepuxent Bay
Sinepuxent Bay
Newport Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Sinepuxent Bay
Sinepuxent Bay
Sinepuxent Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Sinepuxent Bay
Sinepuxent Bay
Newport Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Sinepuxent Bay
Sinepuxent Bay
Sinepuxent Bay
Chla exceedance rate (Growing Season 2001-2004)
Base
20%
40%
60%
natural C MPAR
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
2.31% 0.00% 0.00% 0.00% 0.00%
0.54%
2.85% 1.36% 0.00% 0.00% 0.00%
2.31%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
4.21% 3.53% 2.04% 0.14% 0.00%
3.67%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
3.94% 2.45% 0.00% 0.00% 0.00%
3.53%
4.48% 2.72% 0.00% 0.00% 0.00%
0.82%
2.99% 0.41% 0.00% 0.00% 0.00%
0.00%
3.26% 1.09% 0.00% 0.00% 0.00%
0.00%
1.22% 0.54% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
8.83% 4.89% 3.26% 0.27% 0.00%
5.43%
4.48% 3.40% 1.49% 0.00% 0.00%
3.67%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
8.15% 6.66% 3.40% 0.54% 0.00%
4.62%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
3.53% 1.49% 0.00% 0.00% 0.00%
0.41%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
0.00% 0.00% 0.00% 0.00% 0.00%
0.00%
Chla exceedance rate
Base
20%
40%
0.00% 0.00% 0.00%
1.16% 0.00% 0.00%
1.44% 0.68% 0.00%
0.00% 0.00% 0.00%
2.12% 1.78% 1.03%
0.00% 0.00% 0.00%
1.99% 1.23% 0.00%
2.26% 1.37% 0.00%
1.51% 0.21% 0.00%
1.64% 0.55% 0.00%
0.62% 0.27% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
4.73% 2.47% 1.64%
2.26% 1.71% 0.75%
0.00% 0.00% 0.00%
4.79% 3.49% 1.71%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
1.78% 0.75% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
(Annual 2001-2004)
natural C MPAR
60%
0.00% 0.00% 0.00%
0.00% 0.00% 0.27%
0.00% 0.00% 1.16%
0.00% 0.00% 0.00%
0.07% 0.00% 1.85%
0.00% 0.00% 0.00%
0.00% 0.00% 1.78%
0.00% 0.00% 0.41%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.14% 0.00% 3.22%
0.00% 0.00% 1.85%
0.00% 0.00% 0.00%
0.27% 0.00% 2.95%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.21%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
0.00% 0.00% 0.00%
133
6.3 Final TMDL scenario for MD Coastal Bays – geographic isolation method
In the MCBs, different basins have different assimilation capacities and respond
differently to the inputs of nutrient loading. Based on the incremental reduction studies
in Section 6.1, it is apparent that an across-the-board 60% reduction is sufficient to
bring all stations into attainment. On the other hand, many of the stations in different
basins do not (or only slightly) exceed DO or chlorophyll-a endpoints; so they do not
need (or only need minor) reductions. This geographic difference may be due to
various factors such as in-stream transport, geographic location of the nutrient source,
the estuarine transport, the residence time, different phytoplankton response, and
variations in the nitrogen to phosphorus ratio. For example, upper reaches of the Saint
Martin River in Isle of Wight Bay have the highest chlorophyll-a and lowest DO, because
large nutrient loads are discharged into a narrow headstream of the river where there is
very limited transport and dilution capacity leading, to high chlorophyll-a and low DO. In
contrast, the areas close to the inlets have much larger transport and mixing leading to
the lower chlorophyll-a and higher DO conditions. In determining final TMDL scenario
for management action in the MCBs, it is necessary to consider incremental reductions
in the context of these differing geographic characteristics. To determine the magnitude
of the load reductions needed in six basins due to different variability in each of the
basins, the “geographic isolation method” was used. The essence of geographic
isolation reductions is to hold the target basin in MPAR and all other basin at calibration
levels. In doing so, the spatial impact of loads on water quality was revealed. Based on
the analysis, the stations in the headwater areas of the Saint Martin River require a
significantly higher reduction (55%-58% or MPAR) than the other watersheds, followed
by 40% reduction in the open water of the Isle of Wight Bay in order to achieve the DO
criteria. This is mainly because Isle of Wight Bay has a relatively low assimilative
capacity, especially in the headwater stream, the Saint Martin River.
Combining the incremental reduction, namely the base, 20%, 40%, and 60% with the
geographic isolation reduction, it was determined that Assawoman Bay, Newport Bay
and Chincoteague Bay require 20% reduction while Sinepuxent Bay requires no
reduction. In the case of Chincoteague Bay, the final TMDL scenario entailed a 20%
reduction applied to Maryland’s portion of the watershed, with assessment of attainment
conducted at stations within Maryland’s waters. Final TMDL scenarios are shown in
Table 6.4. For the final TMDL, the reductions in Bishopville Prong and Shingle Landing
Prong used MPAR, which is approximately 55 – 58%. The resulting exceedances
associated with the final TMDL are shown in Table 6.5. Under the final TMDL
incorporating geographic isolation scenarios (TMDL-GI in the tables), the projected DO
exceedance for all stations under growing season condition are under 1% except for
Bishopville Prong and Shingle Landing Prong, which are 8.56% and 6.66 %,
respectively. Under the average annual conditions, the exceedance is reduced to 4.32%
and 3.36%, respectively. The model-simulated DO and chlorophyll-a concentration
under the final TMDL are shown in Figures 6.3 (a) – (c) and Figures 6.4 (a)-(c). Both
figures show the time series from January 2001-August 2005 for baseline versus final
TMDL conditions over spatially distributed individual stations in the MCBs. The TMDL
reduction clearly results in improving conditions and in meeting the water quality
standards. The geographic isolation method combined with incremental reductions thus
134
provides a way to address temporal and spatial variation of loading capacity for
achieving both DO and chlorophyll-a endpoints in the MCBs.
Table 6.4: Final TMDL reductions needed to meet WQS incorporating Geographic
Isolation Scenarios.
Water Body/Water Quality
Limited Segment
Sinepuxent Bay (All)
Newport Bay (All)
Bishopville Prong/Shingle
Landing Prong
Assawoman Bay (Open
Waters)
Isle of Wight Bay (All areas
except those identified above)
Chincoteague Bay (Maryland
portion only)
TMDL - IR
(Incremental Reduction)
The incremental
reductions
needed to meet WQS
using Incremental
Reduction Scenarios
0%
20%
TMDL – GI
(Geographic Isolation)
Final TMDL -Reduction
needed to meet WQS
incorporating Geographic
Isolation Scenarios
0%
20%
60%
MPAR (55-58%)
20%
20%
40%
40%
20%
20%
135
Table 6.5: DO exceedance rates for TMDL IR (Incremental Reduction) and final TMDL
(TMDL GI, incorporating Geographic Isolation Scenario).
Station
TMDL-basin
MD-8digit-basin
BSH0008
MKL0010
SPR0002
SPR0009
TUV0011
TUV0019
XBM1301
XBM3418
XBM5932
XBM8149
XCM0159
XCM1562
XCM4878
XDM4486
XDN0146
XDN2340
XDN2438
XDN3445
XDN3724
XDN4312
XDN4797
XDN4851
XDN5737
XDN6454
XDN7261
XDN7545
AYR0017
ASSA 1.
ASSA 2.
ASSA 3.
ASSA 4.
ASSA 5.
ASSA 6.
ASSA 7.
ASSA 8.
ASSA 9.
ASSA 10.
ASSA 11.
ASSA 12.
ASSA 13.
ASSA 14.
ASSA 15.
ASSA 16.
ASSA 17.
ASSA 18.
Bishopville Prong
Manklin Creek
Shingle Landing Prong
Shingle Landing Prong
Turville Creek
Turville Creek
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Newport Bay
Bishopville Prong
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Assawoman Bay
St. Martin River
St. Martin River
St. Martin River
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Assawoman Bay
Ayer Creek
Sinepuxent Bay
Sinepuxent Bay
Newport Bay
Newport Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Sinepuxent Bay
Sinepuxent Bay
Sinepuxent Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Newport Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Asswoman Bay
Isle of Wight Bay
Isle of Wight Bay
Isle of Wight Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Asswoman Bay
Newport Bay
Sinepuxent Bay
Sinepuxent Bay
Newport Bay
Newport Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Chincoteague Bay
Sinepuxent Bay
Sinepuxent Bay
Sinepuxent Bay
Percent DO less than 5mg/l
Growing Season (2001-2004)
Base
TMDL IR TMDL GI
(Increment (Geographic
Reduction) Isolation)
15.49%
0.41%
0.00%
78.26%
0.00%
0.27%
1.36%
0.00%
0.00%
15.76%
1.77%
6.66%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
1.77%
0.00%
0.00%
37.09%
2.72%
8.56%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
1.63%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
5.98%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
2.58%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
Percent DO less than 5mg/l
Average Annual (2001-2004)
Base
TMDL IR
TMDL GI
(Increment (Geographic
Reduction) Isolation)
7.81%
0.21%
0.00%
39.45%
0.00%
0.14%
0.68%
0.00%
0.00%
7.95%
0.89%
3.36%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.89%
0.00%
0.00%
18.70%
1.37%
4.32%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.82%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
3.01%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
1.30%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
136
Figure 6.3 (a): DO time series, 2001-August 2005, baseline conditions (blue), final TMDL (red) and observed (symbol) at DNR
Stations 1 -12.
137
Figure 6.3 (b): DO time series, 2001-August 2005, baseline conditions (blue), final TMDL (red) and observed
(symbol) at DNR Stations 13 -24.
138
Figure 6.3 (c): DO time series, 2001-August 2005, baseline conditions (blue), final TMDL (red) and observed (symbol) at DNR
Stations 25-27.
139
Table 6.6(a): Chla greater than 50 ug/l exceedance rates for TMDL IR (Incremental Reduction) and final TMDL (TMDL GI,
incorporating Geographic Isolation Scenario).
Percent Chla greater than 50 ug/l
Station
TMDL-basin
MD-8digit-basin
Percent Chla greater than 50 ug/l
Growing Season (2001-2004)
Average Annual (2001-2004)
Base
Base
TMDL IR
TMDL GI
TMDL IR
TMDL GI
(Increment
(Geographic
(Increment
(Geographic
Reduction)
Isolation)
Reduction)
Isolation)
BSH0008
Bishopville Prong
Isle of Wight Bay
19.43%
3.40%
3.40%
10.00%
1.71%
1.71%
MKL0010
Isle of Wight Bay
2.45%
0.00%
0.00%
1.23%
0.00%
0.00%
Isle of Wight Bay
5.57%
0.82%
0.82%
2.81%
0.41%
0.41%
SPR0009
Manklin Creek
Shingle Landing
Prong
Shingle Landing
Prong
Isle of Wight Bay
6.66%
3.40%
3.40%
3.36%
1.71%
1.71%
TUV0011
Turville Creek
Isle of Wight Bay
0.54%
0.00%
0.00%
0.27%
0.00%
0.00%
TUV0019
Turville Creek
Isle of Wight Bay
1.36%
0.00%
0.00%
0.68%
0.00%
0.00%
SPR0002
XBM3418
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
XBM5932
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
XCM1562
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
XCM4878
Newport Bay
Newport Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
XDM4486
Bishopville Prong
Isle of Wight Bay
38.59%
5.57%
5.57%
19.66%
2.81%
2.81%
XDN3724
St. Martin River
Isle of Wight Bay
1.36%
0.00%
0.00%
0.68%
0.00%
0.00%
XDN4312
St. Martin River
Isle of Wight Bay
2.72%
0.00%
0.00%
1.37%
0.00%
0.00%
XDN4797
St. Martin River
Isle of Wight Bay
5.98%
0.95%
0.95%
3.01%
0.48%
0.48%
XDN5737
Assawoman Bay
Asswoman Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
AYR0017
Ayer Creek
Newport Bay
0.14%
0.14%
0.14%
0.07%
0.07%
0.07%
ASSA 4.
Newport Bay
Newport Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 10.
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
140
Table 6.6(b): Chla greater than 15 ug/l exceedance rates for TMDL IR (Incremental Reduction) and final TMDL (TMDL GI,
incorporating Geographic Isolation Scenario).
TMDL-basin
MD-8digit-basin
Percent Chla greater than 15 ug/l
Percent Chla greater than 15 ug/l
Growing Season (2001-2004)
Average Annual (2001-2004)
Base
Base
Station
TMDL IR
TMDL GI
(Increment
(Geographic
Reduction)
Isolation)
TMDL IR
TMDL GI
(Increment
(Geographic
Reduction)
Ioslation)
XBM1301
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
XBM8149
Chincoteague Bay
Chincoteague Bay
2.31%
0.00%
0.00%
1.16%
0.00%
0.00%
XCM0159
Chincoteague Bay
Chincoteague Bay
2.85%
0.00%
0.00%
1.44%
0.00%
0.00%
XDN0146
Isle of Wight Bay
Isle of Wight Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
XDN2340
Isle of Wight Bay
Isle of Wight Bay
4.21%
1.22%
1.22%
2.12%
0.62%
0.62%
XDN2438
Isle of Wight Bay
Isle of Wight Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
XDN3445
Assawoman Bay
Asswoman Bay
3.94%
0.00%
0.00%
1.99%
0.00%
XDN4851
Assawoman Bay
Asswoman Bay
4.48%
1.36%
1.36%
2.26%
0.68%
0.68%
XDN6454
Assawoman Bay
Asswoman Bay
2.99%
1.22%
1.22%
1.51%
0.62%
0.62%
XDN7261
Assawoman Bay
Asswoman Bay
3.26%
3.13%
3.13%
1.64%
1.58%
1.58%
XDN7545
Assawoman Bay
Asswoman Bay
1.22%
5.03%
5.03%
0.62%
2.74%
2.74%
ASSA 1.
Sinepuxent Bay
Sinepuxent Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 2.
Sinepuxent Bay
Sinepuxent Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 3.
Newport Bay
Newport Bay
8.83%
4.62%
4.62%
4.73%
2.74%
2.74%
ASSA 5.
Chincoteague Bay
Chincoteague Bay
4.48%
0.00%
0.00%
2.26%
0.00%
0.00%
ASSA 6.
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.07%
ASSA 7.
Chincoteague Bay
Chincoteague Bay
8.15%
0.14%
0.14%
4.79%
0.07%
ASSA 8.
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 9.
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 11.
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 12.
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 13.
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 14.
Chincoteague Bay
Chincoteague Bay
3.53%
0.00%
0.00%
1.78%
0.00%
0.00%
ASSA 15.
Chincoteague Bay
Chincoteague Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 16.
Sinepuxent Bay
Sinepuxent Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 17.
Sinepuxent Bay
Sinepuxent Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ASSA 18.
Sinepuxent Bay
Sinepuxent Bay
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
141
Figure 6.4 (a): Chlorophyll-a time series, 2001-August 2005, baseline conditions (blue), final TMDL (red) and observed
(symbol) at DNR Stations 1-12.
142
Figure 6.4 (b): Chlorophyll-a time series, 2001-August 2005, baseline conditions (blue), final TMDL (red) and observed
(symbol) at DNR Stations 13-24.
143
Figure 6.4 (c): Chlorophyll-a time series, 2001-August 2005, baseline conditions (blue), final TMDL (red) and observed
(symbol) at DNR Stations 25-27.
144
CHAPTER 7: DISCUSSION AND CONCLUSION
The MCBs are subject to anthropogenic impacts from the agricultural practices and
expanding population in the coastal watersheds. As a result, the individual bays are
showing signs of increasing eutrophication. In particular, the hypoxic conditions in the
tributaries pose the most serious threat to the long-term health and function of the bays.
As a component of developing the MCB TMDLs, a HEM3D (Hydrodynamic
Eutrophication Model) was developed, consisting of a hydrodynamic model SELFE
(Semi-implicit, Eulerian-Lagrangian Finite Element model), a water quality model ICM
(Integral Compartment Model), and a sediment benthic flux model. The HEM3D was
coupled with an HSPF watershed model to simulate water quality transport and
biochemical ecosystem processes in the MCBs.
In terms of nutrient loading generated by the HSPF watershed model, the non-point
source and atmospheric loads dominate the total loads with point-source loads
comprising only a small portion. Based on the areal loading rate defined as the total
loading divided by the total water surface area, it is clear that the Isle of Wight subwatershed has the largest TN and TP unit loads (per water surface water area),
followed by Newport Bay, Assawoman Bay, Sinepuxent and Chincoteague Bays. The
relatively larger drainage areas lead to the greater non-point source loads, causing the
receiving waterbody to be vulnerable to eutrophication problems. By contrast,
waterbodies with relatively smaller drainage areas receive correspondingly less nonpoint source loading, and are less susceptible to eutrophication problems.
Each component of the HEM3D is calibrated and verified with field measured data
collected in the MCBs. The hydrodynamic model was calibrated with astronomical tide,
the measured water level, the intensive ADCP current velocities (measurement
conducted in 2004), and salinity observations across MCB stations. Although the total
freshwater discharge in the MCBs system is small as compared to other systems (for
example, the Chesapeake Bay), episodic rainfall events can still have substantial
impacts on the runoff and salinity distribution in the creeks and tributaries. The model
captured the large rainfall-induced salinity variability as large as 10-20 ppt. For the
water quality model calibration, the chlorophyll-a, dissolved oxygen, ammonia, nitrite
and nitrate, phosphate and DON data compared well with the model result over a
transect in the Isle of Wight Bay. For validation, this analysis was extended throughout
the MCBs, and comparison with data at all 27 DNR and 18 ASIS stations yielding
reasonable results with satisfactory skill scores.
It is should be noted that a stand-alone macroalgae sub-model in the MCBs has been
developed (Wang, Taiping, 2009). The model includes two macroalgae species, Ulva
lactuca and Gracilaria vermiculophylla, and the Droop formulation was used to account
for the luxury uptake. Using a vertical one-dimensional box model, the well-known
boom-and-bust life cycle of macroalgae was qualitatively simulated. The macroalgae
sub-model, however, was used primarily in a research capacity; it was not included in
the current TMDL, for with the current technology, a thorough understanding of the
macroalgae life cycle is still lacking. For example, little is known about a suite of factors
contributing to the boom-and bust life cycle, the fate of the macroalgae biomass after
145
the bust, and the species interaction of macroalgae with other phytoplankton species.
These gaps in knowledge about macroalgae hampered the ability of the model to
formulate a proper relationship between phosphorus and nitrogen loading to the
macroalgae sub-model, and hence the model was not implemented in this study.
However, the micro-algae, a single-celled ‘plant-like’ organism was included by
imbedding it in the benthic sediment flux model which acted as proxies for
macrobenthos. As a result, some effects of the macroalgae, such as its interaction with
the overlying water column, are partially simulated.
After the calibration and verification of the HEM3D model, additional modeling analyses
were conducted to substantiate the effect of diel DO cycle and perform the sensitivity
analysis in the MCBs. First, the daily mean DO (from HEM3D) was adjusted to
incorporate diel oscillation based on the work of Elgin Perry (2012). The empirical
corrections were made based on monthly temperature, daily temperature, daily
Photosynthetically Active Radiation (PAR), and daily chlorophyll. In doing so, the DO
variation includes the diel oscillation, providing a better representation of the full
spectrum of DO in the MCBs. Secondly, sensitivity analyses were conducted to test the
effects of (a) the Ocean City wastewater treatment plant outfall, (b) phytoplankton and
organic nutrient settling rate and (c) ground water discharge so their potential effects
are checked and measured.
For the TMDL scenario development, the target is such that the daily mean DO
concentration everywhere shall not be below 5 mg/l more than 10 percent of the time,
both annually and during the growing season (May 1 – October 31). The TN and TP
sources were assessed for the five impaired basins in the MCBs. It was found that for
Assawoman, Isle of Wight, and Newport Bays, the terrestrial sources are the dominant
source of loading, whereas in the Sinepuxent and Chincoteague Bays the terrestrial
source loading are about equal to that of atmospheric loading. To further determine the
assimilative capacity of the MCBs for nutrients, incremental reductions of 20%, 40%,
60%, natural conditions and MPAR (Maximum Practicable Anthropogenic Reduction) of
the total loading were conducted. It was revealed that there is a large difference in the
response from the different basins given the similar amount of reduction. In the end, the
final TMDL scenario was determined by the geographic isolation method to reflect the
relative impact of source sectors from different basin in the Maryland Coastal Bays to
meet the TMDL endpoints. A spatial pattern emerges that Bishopville Prong of the
Upper Saint Martin River watershed requires the most reduction (55-60%) followed by
the other areas of the Isle of Wight Bay watershed (40%). Assawoman Bay, Newport
Bay and Chincoteague Bay only require only a 20% reduction to meet the water quality
targets. It was determined that Sinepuxent Bay does not require a reduction from the
baseline loading in order to meet the water quality endpoints.
146
Acknowledgements:
The work was supported by Maryland Department of the Environment TMDL Technical
Development Program. The water quality data was provided by Maryland Coastal Bay
Program, Maryland Department of Natural Resources, and Assateague Island National
Seashore, US National Park Services. Additionally, Maryland Geological Survey
provided the assistance on processing the shoreline erosion data and University of
Maryland Horn Point Laboratory the ADCP data. During the course of the project, we
appreciated the valuable inputs and comments by the members of the Maryland Coastal
Bays Program Scientific and Technical Advisory Committee (STAC). Lastly, but not
least, the authors are grateful to Ms. Rou Shi, Ms. Melissa Chatham and Mr. Tim Rule
for their technical guidance and project development.
147
REFERENCES
Allan, Thomas R., H. T. Tolvanen, G. F. Oertel, and G. M. McLeod (2007): Spatial
characterization of environmental gradients in a coastal lagoon. Chincoteague Bay.
Estuaries and Coasts, Vol. 30. No.6, pp. 959-977.
Ambrose, R. B.,T. A. Wool, and J. L. Martin (1992): The water quality Analysis and
Simulation Program, WASP5; Part A, Model Documentation. US EPA Athens,
Environmental Research Laboratory, 2010 pp.
Azevedo, A., Oliveira, A., Fortunato, A.B. and Bertin, X. (2009): Application of an
Eulerian-Lagrangian oil spill modeling system to the Prestige accident: trajectory
analysis. J. Coastal Res., SI 56, 777-781.
Banks, R.B. & Herrera, F.F. (1977): Effect of wind and rain on surface reaeration. J. of
the Environmental Engineering Division, ASCE, 103(EE3): 489-504.
Bertin, X., N. Bruneau, J.-F. Breilh, A.B. Fortunato, and M. Karpytchev (2012):
Importance of wave age and resonance in storm surges: The case Xynthia, Bay of
Biscay, Ocean Modelling, 42(0), 16-30.
Blumberg, A.F. and G.L. Mellor (1987): A description of a three-dimensional coastal
ocean circulation model. In: Three-Dimensional Coastal Ocean Models, vol. 4,
Coastal and Estuarine Studies, N. Heaps, editor, Washington, D.C.: AGU, pp. 1-16.
Boynton, W. R. (1996): A comparative analysis of eutrophication patterns in a temperate
coastal lagoon. Estuaries, Vol. 10, No. 2B, p408-421.
Brovchenko I., V. Maderich, and K . Terletska (2011): Numerical simulations of 3D
structure of currents in the region of deep canyons on the east coast of the Black
Sea. International Journal for Computational Civil and Structural Engineering , 7 (2):
47-53.
Canuto, V.M., A. Howard, Y. Cheng and M.S. Dubovikov (2001): Ocean turbulence I:
one-point closure model. Momentum and heat vertical diffusivities. J. Phys. Oceano.,
31, pp. 1413-1426.
Casulli, V. and P. Zanolli (2005): High resolution methods for multidimensional
advection–diffusion problems in free-surface hydrodynamics. Ocean Modelling, 10,
pp.137-151.
Casulli, V. and E. Cattani (1994): Stability, accuracy and efficiency of a semi-implicit
method for 3D shallow water flow. Computers & Mathematics with Applications, 27,
pp. 99-112.
148
Cerco, C. F. and Mark Noel (2004): The 2002 Chesapeake Bay eutrophication model .
Contract report EPA-903-R-04-004 to Chesapeake Bay Program Office, US
Environmental Protection Agency, 410 Severn Avenue, Annapolis, MD 21401,
administrated by the Baltimore District, US Army Corps of Engineers.
Cerco, C. F. and T. Cole (1994): Three-dimensional eutrophication model of
Chesapeake Bay/ Volume 1: Main report. Technical report EL-94-4, U. S. Army
Corps of Engineers, Waterways Experiment Stations, Vicksburg, MS.
Cerco, C. F. B Bunch, A. Teeter, and M. Dortch (2000): Water quality model of Florida
Bay: Technical report ERDC/EL-00-10, U. S. Army Corps of Engineers, Waterways
Experiment Stations, Vicksburg, MS.
Collins, C. D. and J. H. Wlosinski (1983): Coefficients for Use in the U.S. Army Corps of
Engineers Reservoir Model, CE-QUAL-R1. Tech. Rept. E-83-15, Environmental
Laboratory, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Cornwell, J. and M. Owens (2013): Benthic nutrient cycling at the coastal bays land-sea
interface: investigation of ammonium sources. Maryland Coastal Bay STAC meeting
presentation, January, 2013.
Dennison, W. C. J. E. Thomas, C. J. Cain, T, U. B. Carruthers, M. R. Hall, R. V. Jesien,
C. E. Wazniak, D. E. Wilson (2009): Shifting Sands, Environmental and cultural
change in Maryland’s coastal Bays. pp396, IAN press, University of Maryland,
Center for Environmental Science.
Di Toro, D. M. (2001): Sediment flux modeling. John Wiley and Sons, Inc. New York.
pp624.
Di Toro, D. M. and J. Fitzpatrick (1993): Chesapeake Bay sediment flux model. Contract
report EL-93-2, U. S. Army Corps of Engineers, Waterways Experiment Stations,
Vicksburg, MS.
Dillow, J.J.A. and E.A. Greene (1999): Groundwater Discharge and Nitrate Loadings to
the Coastal Bays of Maryland. U.S. Geological Survey Water-Resources
Investigations Report 99-4167.
Department of Natural Resources (2004): State of the Maryland Coastal Bays. Maryland
Department of Natural Resources, Maryland Coastal Bays Program, C. E. Wazniak,
C., M. Hall, C. Cain, D. Wilson, R., Jesien, J. Thomas, T. Carruthers, W. Dennison.
2004. and University of Maryland Center for Environmental Science.
http://dnrweb.dnr.state.md.us/pressroom/MCB.pdf
EPA Chesapeake Bay Program (2009): Bay barometer: A health and restoration
assessment of the Chesapeake Bay and watershed in 2008. Chesapeake Bay
Program report TRS 293-09: US EPA-903-R-09-001, March 2009.
149
Fertig, B. M, J. O’Neil , K. A. Beckert , C. J. Cain, D. M. Needham, T. J. B. Carruthers,
W. C. Dennison WC (2013): Elucidating terrestrial nutrient sources to a coastal
lagoon, Chincoteague Bay, Maryland, USA. Estuarine,.Coast. Shelf Sci. 116:1–10.
Flather, R.A. (1987): A tidal model of Northeast Pacific. Atmosphere–Ocean. Vol. 25,
22–45.
Fortunato, A.B. and A.M. Baptista (1996): Evaluation of Horizontal Gradients in SigmaCoordinate Shallow Water Models. Atmosphere-Ocean, 34, pp. 489-514.
Galperin, B., L. H. Kantha, S. Hassid and A. Rosati (1988): A quasi-equilibrium turbulent
energy model for geophysical flows. J. Atmos. Sci., 45, pp. 55-62.
Ham, D.A., J. Pietrzak, J. and G.S. Stelling (2005): A scalable unstructured grid 3dimensional finite volume model for the shallow water equations. Ocean Modelling,
10, pp. 153-169.
HydroQual (1995): A water quality model for Massachusetts and Cape Cod Bays:
Calibration of the Bays eutrophication model (BEM). Technical report, HydroQual,
Inc. Mahwah, NJ.
Ji, Zhen-Gang (2008): Hydrodynamics and water quality modeling rivers, lakes, and
estuaries. Wiley-Interscience, John Wiley and Sons, Inc. pp676.
Jørgensen, S. E., editor. (1979): Handbook of Environmental Data and Ecological
Parameters. Pergamon Press, Oxford.
July O’Neil, C. Wazniak, C. McCollough, D. McKay, K. Meyer, B. Walsh (2013):
Offshore nutrient sampling R/V Rachel Carson: August 2012. Maryland Coastal Bay
STAC meeting presentation, January, 2013.
Kantha, L.H. and C.A. Clayson (1994): An improved mixed layer model for geophysical
applications. J. Geophy. Res, 99(25), pp. 235-266.
Le Roux, D. Y., C. A. Lin, and A. Staniforth (1997): An accurate interpolating scheme for
semi-Lagrangian advection on an unstructured mesh for ocean modelling, Tellus A,
49(1), 119-138.
Luettich, R.A., J.C. Muccino and M.G.G. Foreman (2002): Considerations in the
calculation of vertical velocity in three-dimensional circulation models. Journal of
Atmospheric and Oceanic Technology, 19, pp. 2063-2076.
Mellor, G.L. and T. Yamada (1982): Development of a turbulence closure model for
geophysical fluid problems. Rev. Geophys., 20, pp. 851-875.
150
Maryland Department of the Environment (1993): An assessment of aquatic
ecosystems, pollutant loadings and management options. University of Maryland
System Centers for Environmental and Estuarine Studies and Coastal
Environmental Services Inc.
Maryland Department of the Environment (2001): Total maximum daily loads of nitrogen
and phosphorus for 5 tidal tributaries in the Northern Coastal Bays, Worcester
County, MD. Final report submitted to EPA Region III, Watershed Protection Division,
Philadelphia, PA. http://www.mde.state.md.us/assets/document/NCB_main_final.pdf.
Maryland Department of the Environment (2004): Priority areas for wetland restoration,
preservation, and mitigation in Maryland’s Coastal Bays. Final report submitted to
US Environmental Protection Agency, State Wetland Program development grant.
Maryland Department of the Environment) (2010): Final integrated report of surface
water quality in Maryland. Baltimore, MD: Maryland Department of the Environment.
Morel, F. (1983): Principles of Aquatic Chemistry. John Wiley & Sons, N.Y, NY, 446 pp.
NTHMP (2011): NTHMP MMS Tsunami Inundation Model Validation Workshop,
Galveston, April 2011. NOAA internal report (in press).
O'Connor, D.J. & Dobbins, W.E. (1958): Mechanism of reaeration in natural streams.
Transactions of the Americal Society of Civil Engineers, 123(2934): 641-684
Oliveira, A. and A.M. Baptista (1998): On the role of tracking on Eulerian-Lagrangian
solutions of the transport equation. Advances in Water Resources, 21, pp. 539-554.
Park, K., A. Y. Kuo, J. Shen, and J. M. Hamrick (1995): A three-dimensional
hydrodynamiceutrophication model (HEM3D): description of water quality and
sediment processes submodels. The College of William and Mary, Virginia Institute
of Marine Science. Special Report 327, 113pp.
Perry, Elgin (2012): Adjustment of dissolved oxygen observations for diel cycles in
Maryland Coastal Bays; contract report submitted to Maryland Department of natural
Resources, Tidewater Ecosystem Assessment, administrated by Catherine E.
Wazniak and Carol B. McCollough.
Pietrzak, J., J.B. Jakobson, H. Burchard, H.J. Vested and O. Petersen (2002): A threedimensional hydrostatic model for coastal and ocean modelling using a generalised
topography following coordinate system. Ocean Modelling, 4, pp. 173-205.
Pond, S. and G.L. Pickard (1998): Introductory Dynamical Oceanography, ButterworthHeinmann.
Pritchard, D. W. (1960). Salt balance and exchange rate for Chincoteague Bay.
Chesapeake Science 1:48-57.
151
Rodi, W. (1984): Turbulence models and their applications in hydraulics: a state of the
art review. Delft, The Netherlands, International Association for Hydraulics Research.
Rodrigues M., A. Oliveira, H. Queiroga, A.B. Fortunato, Y. Zhang (2009): Threedimensional modeling of the lower trophic levels in the Ria de Aveiro (Portugal),
Ecological Modelling, 220(9-10),1274-1290.
Shapiro, R. (1970): Smoothing, filtering and boundary effects. Rev. Geophy. and Space
Physics, 8(2), pp. 359-387.
Shchepetkin, A.F. and J.C. McWilliams (2005): The regional oceanic modeling system
(ROMS): a split-explicit, free-surface, topography-following-coordinate, oceanic
model. Ocean Modelling, 9, pp. 347-404.
Song, Y., Haidvogel, D. (1994). A semi-implicit ocean circulation model using a
generalized topography-following coordinate system. J. Comput. Phys. 115, 228–
244.
Sweby, P.K. (1984): High resolution schemes using flux limiters for hypobolic
conservation laws. SIAM J. Num. Analysis, 21(5), pp. 995-1011.
Thomann, Robert V., D. M. Di Toro, R. P. Winfield, D. J. O’Connor (1975): Mathematical
modeling of phytoplankton in Lake Ontario. Part 1: Model development and
verification.
Tillman, Dottie H., Carl F. Cerco, Mark R. Noel, J. L. Martin, and John Hamrick (2004):
Three-dimensional eutrophication model of the lower St. Johns River, Florida.
Environmental Lab. ERDC/EL TR-04-13, US Army Corps of Engineering Research
and Development Center.
Umlauf, L. and H. Burchard (2003): A generic length-scale equation for geophysical
turbulence models. J. Mar. Res., 6, pp. 235-265.
VIMS (Virginia Institute of Marine Science), (2013): Maryland Coastal Bays Watershed
Modeling Report conducted by Hongzhou Xu and Jian Shen, submitted to MDE on
February 2013.
Walters, Roy, A. (2005): Coastal ocean models: two useful finite element methods.
Continental Shelf Research, Vol. 25, p775-793.
Wang, Taiping (2009): Numerical modeling of eutrophication dynamics in the shallow
coastal ecosystem: a case study in the Maryland and Virginia Coastal Bays. Virginia
Institute of Marine Sciences, Ph.D. thesis, pp236.
152
Wazniak, C.E., and P.M. Glibert. (2004): Potential impacts of brown tide, Aureococcus
anophagefferens, on juvenile hard clams, Mercenaria mercenaria, in the Coastal
Bays of Maryland, USA. Harmful Algae 3:321-329.
Wazniak, C., M. Trice, B. Sturgis, W. Romano, and M. Hall. (2004): Status and trends of
phytoplankton abundance in the Maryland Coastal Bays. In: Wazniak, C. and M. Hall
[eds.] Maryland’s Coastal Bays: Ecosystem health assessment 2004. Maryland
Department of Natural Resources, document number: DNR-12-1202-0009.
Wazniak, C.E. and M.R. Hall [Ed]. (2005): Maryland's Coastal Bays: Ecosystem Health
Assessment 2004. DNR-12-1202-0009. Maryland Department of Natural Resources,
Tidewater Ecosystem Assessment, Annapolis, MD.
Wazniak, C.E., M.R. Hall, T. Carruthers, and R. Sturgis (2007): Linking water quality to
living resources in a mid-Atlantic lagoon system, USA. Ecological Applications.
17(5):S64-S78.
Wezernak, C. T. and Gannon, J. J. (1968): Evaluation of nitrification in streams. Journal
of the Sanitary Engineering Division, ASCE, 94(SA5): p883-895.
Wells, Darlene V., J. M. Hill, M. J. Park, and C. P. Williams (1998): The shallow
sediments of the middle Chincoteague Bay area in Maryland: physical and chemical
characteristics. Coastal and estuarine geology, file report no. 98-1. Maryland
Geological Survey.
Wells, Darlene. V. 2002, E. L. Hennessee, and J. M. Hill (2002): Shoreline erosion as a
source of sediments and nutrients, Northern Coastal Bay, Maryland. Coastal and
estuarine geology, file report no. 02-05. Maryland Geological Survey.
Wells, Darlene. V. 2003, E. L. Hennessee, and J. M. Hill (2003): Shoreline erosion as a
source of sediments and nutrients, Middle Coastal Bays, Maryland. Coastal and
estuarine geology, file report no. 03-07. Maryland Geological Survey.
Wicker, L.J. and Skamarock, W.C. (1998). A time-splitting scheme for the elastic
equations incorporating second-order Runge–Kutta time differencing. Monthly
Weather Rev. 126, 1992–1999.
Wilcox, D.C. (1998): Reassessment of scale determining equation for advance
turbulence models. AIAA J., 26, pp. 1299-1310.
Zeng, X., M. Zhao and R.E. Dickinson (1998): Intercomparison of bulk aerodynamic
algorithms for the computation of sea surface fluxes using TOGA COARE and TAO
data. J. Clim., 11, pp. 2628-2644.
153
Zhang, Y.L., A.M. Baptista and E.P. Myers (2004): A cross-scale model for 3D
baroclinic circulation in estuary-plume-shelf systems: I. Formulation and skill
assessment. Continental Shelf Research, 24, pp. 2187-2214.
Zhang, Y., and A. M. Baptista (2008a): SELFE: A semi-implicit Eulerian–Lagrangian
finite-element model for cross-scale ocean circulation, Ocean Modelling, 21(3–4),
71-96.
Zhang, Y., R. C. Witter, and G. R. Priest (2011): Tsunami–tide interaction in 1964
Prince William Sound tsunami, Ocean Modelling, 40(3–4), 246-259.
Zhang, Y., and A. Baptista (2008b). An efficient and robust tsunami model on
unstructured grids. Part I: inundation benchmarks, Pure and Applied Geophysics,
165(11), 2229-2248.
154
APPENDIX A: Summary of SELFE hydrodynamic model formulations
A1. Physical formulation of SELFE
SELFE solves the 3D shallow-water equations, with hydrostatic and Boussinesq
approximations, and transport equations for salt and heat. The primary variables that
SELFE solves are free-surface elevation, 3D velocity, 3D salinity, and 3D temperature
of the water. In a Cartesian frame, the equations read:
∂w
(1) ∇u +
=
0
∂z
η
∂η
+ ∇∫ udz = 0
(2)
∂t
−h
(3)
Du
∂  ∂u 
1
g
= f − g ∇η + ν
 ; f = − fk × u + α g ∇ψˆ − ∇p A −
Dt
∂z  ∂z 
ρ0
ρ0
∫
η
z
∇ρ dζ + ∇( µ∇u)
DS ∂  ∂S 
(4)
=
κ
 + Fs
Dt ∂z  ∂z 
DT ∂  ∂T 
Q
(5) =
+ Fh
κ
+
Dt ∂z  ∂z  ρ0C p
where
(x,y)
horizontal Cartesian coordinates, in [m]
z
vertical coordinate, positive upward, in [m]
∇
 ∂ ∂ 
 , 
 ∂x ∂y 
t
time [s]
η ( x, y , t )
free-surface elevation, in [m]
h ( x, y )
bathymetric depth, in [m]
u( x, y, z , t ) horizontal velocity, with Cartesian components (u,v), in [ms-1]
w
vertical velocity, in [ms-1]
f
Coriolis factor, in [s-1] (Section 2.5)
g
acceleration of gravity, in [ms-2]
ψˆ (φ , λ )
earth-tidal potential, in [m] (Section 2.5)
α
effective earth-elasticity factor
155
ρ (x, t )
water density; by default, reference value ρ0 is set as 1025 kgm-3
p A ( x, y , t )
atmospheric pressure at the free surface, in [Nm-2]
S,T
salinity and temperature of the water [practical salinity units (psu), oC]
ν
vertical eddy viscosity, in [m2s-1]
µ
horizontal eddy viscosity, in [m2s-1]
κ
vertical eddy diffusivity, for salt and heat, in [m2s-1]
Fs , Fh
horizontal diffusion for transport equations (neglected in SELFE)
Q
rate of absorption of solar radiation [Wm-2]
Cp
specific heat of water [JKg-1K-1]
The differential system Eqs. (1-5) are closed with: (a) the equation of state describing
the water density as a function of salinity and temperature, (b) the definition of the tidal
potential and Coriolis factor; (c) parameterizations for horizontal and vertical mixing, via
turbulence closure equations, and (d) appropriate initial and boundary conditions.
Futher details can be found in Zhang et al. (2004).
A1.1 Turbulence closure model
SELFE uses the Generic Length Scale (GLS) turbulence closure of Umlauf and
Burchard (2003), which has the advantage of encompassing most of the 2.5-equation
closure models: k-ε (Rodi 1984); k-ω (Wilcox 1998); k- l (Me llor a nd Ya m a da 1982).
In this framework, the transport, production, and dissipation of the turbulent kinetic
energy (K) and of a generic length-scale variable (ψ ) are governed by:
DK ∂  ψ ∂K 
2
2
(6) =
ν k
 +ν M + µ N − ε ,
Dt ∂z  ∂z 
(7)
Dψ
∂  ∂ψ
=
νψ
Dt ∂z 
∂z
 ψ
2
2
 + ( cψ 1ν M + cψ 3 µ N − cψ 2 Fwε ) ,
 K
whereν ψk and ν ψ are vertical turbulent diffusivities, cψ 1 , cψ 2 and cψ 3 are model-specific
constants (Umlauf and Burchard 2003; Zhang et al. 2004), Fw is a wall proximity
function, M and N are shear and buoyancy frequencies, and ε is a dissipation rate. The
generic length-scale is defined as
(8) ψ = ( cµ0 ) K m  n ,
p
156
where cµ0 =0.31/2and  is the turbulence mixing length. The specific choices of the
constants p, m and n lead to the different closure models mentioned above. Finally,
vertical viscosities and diffusivities as appeared in Eqs. (3-5) are related to K,   and
stability functions:
ν = 2 sm K 1/ 2
µ = 2sh K 1/ 2
(9)
ν
σ ψk
ν
νψ =
σψ
ν ψk =
,
where the Schmidt numbers σ ψk and σ ψ are model-specific constants. The stability
functions (sm and sh) are given by an Algebraic Stress Model (e.g.: Kantha and Clayson
1994, Canuto et al. 2001, or Galperin et al. 1988).
At the free surface and at the bottom of rivers and oceans, the turbulent kinetic energy
and the mixing length are specified as Direchlet boundary conditions:
1 2/3
B1 | τ b |2 ,
2
(10)
K=
(11)
 = κ 0 db or κ 0 d s ,
where τ b is a bottom frictional stress (Eq. (14)),κ0=0.4 is the von Karman’s constant, B1 is
a constant, and db and ds are the distances to the bottom and the free surface,
respectively.
A1.2 Vertical boundary conditions for the momentum equation
The vertical boundary conditions for the momentum equation – especially the bottom
boundary condition - play an important role in the SELFE numerical formulation, as it
involves the unknown velocity (see Section A2). In fact, as a crucial step in solving the
differential system, SELFE uses the bottom boundary condition to decouple the freesurface Eq. (2) from the momentum Eq. (3).
At the sea surface, SELFE enforces the balance between the internal Reynolds stress
and the applied shear stress:
(12)
ν
∂u
= τw ,
∂z
at z = η
where the stress τw can be parameterized using the approach of Zeng et al. (1998) or
the simpler approach of Pond and Pickard (1998).
157
Because the bottom boundary layer is usually not well resolved in ocean models, the
no-slip condition at the sea or river bottom (u = w= 0) is replaced by a balance between
the internal Reynolds stress and the bottom frictional stress,
(13)
ν
∂u
= τ b , at z = − h .
∂z
The specific form of the bottom stress τ b depends on the type of boundary layer used.
While the numerical method for SELFE as outlined in Section A2 can be applied to
other types of bottom boundary layer (e.g., laminar boundary layer), only the turbulent
boundary layer below (Blumberg and Mellor 1987) is discussed, given its prevalent
usage in ocean modeling. The bottom stress in Eq. (13) is then:
(14)
τ b = CD | u b | u b .
The velocity profile in the interior of the bottom boundary layer obeys the logarithmic
law:
(15) u
=
ln [ ( z + h) / z0 ]
ln(δ b / z0 )
u b , ( z0 − h ≤ z ≤ δ b − h ) ,
which is smoothly matched to the exterior flow at the top of the boundary layer. In Eq.
(15),δb is the thickness of the bottom computational cell (assuming that the bottom is
sufficiently resolved in SELFE that the bottom cell is inside the boundary layer), z0 is the
bottom roughness, and ub is the velocity measured at the top of the bottom
computational cell. Therefore, the Reynolds stress inside the boundary layer is derived
from Eq. (15) as:
(16)
ν
∂u
ν
=
ub .
∂z ( z + h) ln(δ b / z0 )
Utilizing the turbulence closure theory discussed in Section 2.1, the eddy viscosity can
be found from Eq. (9), with the stability function, the turbulent kinetic energy, and the
meso-scale mixing length given by:
sm = g 2 ,
(17)
1 2/3
B1 CD | ub |2 ,
2
=
 κ 0 ( z + h)
K=
where g2 and B1 are constants with g 2 B11/ 3 = 1 . Therefore, the Reynolds stress is constant
inside the boundary layer:
158
κ0
∂u
(18) ν
=
CD1/ 2 | ub | ub , ( z0 − h ≤ z ≤ δ b − h) ,
∂z ln(δ b / z0 )
and the drag coefficient is calculated from Eqs. (13), (14), and (18) as:
−2
(19)
 1 δ 
CD =  ln b  ,
 κ 0 z0 
a drag coefficient formula as discussed in Blumberg and Mellor (1987). Eq. (18) also
shows that the vertical viscosity term in the momentum equation Eq. (3) vanishes inside
the boundary layer. This fact will be utilized in the numerical model of SELFE in Section
A2.
A2. Numerical formulation of SELFE
Numerical efficiency and accuracy consideration dictates the numerical formulation of
SELFE. SELFE solves the differential equation system described in Section 2 with
finite-element and finite-volume schemes. No mode splitting is used in SELFE, thus
eliminating the errors associated with the splitting between internal and external modes
(Shchepetkin and McWilliams 2005). Semi-implicit schemes are applied to all equations;
the continuity and momentum equations (Eqs. (2-3)) are solved simultaneously, thus
bypassing the most severe stability restrictions (e.g. CFL). A key step in SELFE is to
decouple the continuity and momentum equations (Eqs. (2-3)) via the bottom boundary
layer, as shown in Section A2.2. SELFE uses an Eulerian-Lagrangian method (ELM) to
treat the advection in the momentum equation, thus further relaxing the numerical
stability constraints. The advection terms in the transport equations (Eqs. (4-5)) are
treated with either ELM or a finite-volume upwind method (FVUM), the latter being mass
conservative.
A2.1. Domain discretization
In SELFE, unstructured triangular grids are used in the horizontal direction, while hybrid
vertical coordinates – partly terrain-following S coordinates and partly Z coordinates –
are used in the vertical direction. The origin of the z-axis is at the undisturbed Mean Sea
Level (MSL). The terrain-following S layers (Song and Haidvogel 1994) are placed on
top of a series of Z layers with the demarcation line between S and Z layers located at
level kz (z=-hs). That is to say, the vertical grid is allowed to follow the terrain up to a
maximum depth of hs. The free surface is at level Nz throughout the domain (for all wet
points), but the bottom level indices, kb, may vary in space due to the staircase
representation of the bottom in Z layers. Note that kb < or equal kz and the equality
occurs when the local depth h< or equal to hs . A "pure S" representation is a special
case with kb=kz=1 and hs greater than the maximum depth in the domain, but a "pure
Z" model is not a special case in SELFE. The details of the terrain-following coordinates
used in SELFE can be found in Appendix A. The rationale for using such a hybrid
coordinate system is discussed below.
159
The "pure S" representation of SELFE was initially chosen by the authors to avoid the
staircase representation of the bottom and surface, and thus loss of accuracy commonly
associated with the Z coordinates. While sufficient and preferable for some applications,
"pure S" SELFE suffers from the so-called hydrostatic inconsistency commonly
associated with the terrain-following coordinate models, and fails in applications
involving steep bathymetry and strong stratification, as found in freshwater plumes of
large rivers like the Columbia River. The inclusion of Z layers effectively alleviates the
hydrostatic inconsistency and results in a physically more realistic plume. Therefore, the
hybrid vertical coordinate system has the benefits of both S and Z coordinates: the S
layers used in the shallow region resolves the bottom efficiently and the Z layers, which
are only used in the deep region with h>hs, fend off the hydrostatic inconsistency. The
effects of the staircase representation of the bottom are arguably small in the deep
region because the velocities there are small; the effects can also be minimized by
choosing the largest possible value for hs for a given application.
The use of a hybrid vertical coordinate system raises the issue of in which coordinate
system the equations should be solved. All equations are solved in their original forms
in the untransformed Z coordinates and use the transformation only to generate a
vertical grid and to evaluate the horizontal derivatives (such as the horizontal viscosity).
The main reason for not transforming the equations into S coordinates is that the
transformation degenerates under the special circumstances described. Therefore, the
role of vertical coordinates is mostly hidden in SELFE; all equations but one (the
integrated continuity equation) are solved along the vertical direction only, which can be
done in any vertical grid (including, in theory, an unstructured grid). The liberal
treatment of the vertical coordinates makes the implementation of the hybrid vertical
coordinates (SZ) system easier. A similar approach was also used by Shchepetkin and
McWilliams (2005), who solved the equations in the Z space, despite the S coordinates
being used in the vertical direction.
Strictly speaking, since the free surface is moving and so are the upper S levels (in the
original Z space), all variables need to be re-interpolated onto the new vertical grid after
the levels are updated at the end of each time step. However, the effects of movement
of the S levels from one time step to the next are negligible, as long as the vertical
movement of the free surface within a step is much smaller than the minimum layer
thickness. This condition is easily satisfied in most practical applications; for example, in
typical tidal-driven circulations, the maximum displacement of the free surface in a time
step as large as five (5) minutes is only a few centimeters or less, which is much smaller
than a typical top layer thickness of a few meters or more. Therefore, this interpolation
step was skipped in SELFE, as a linear interpolation would introduce additional
numerical diffusion, and a higher-order interpolation would introduce numerical
dispersion into the solution. Note that a similar omission also occurs in many Z
coordinate models, where the top layers also change with time.
In many parts of SELFE, interpolation at an arbitrary location in 3D space is necessary;
examples include the interpolation at the foot of the characteristic line and the
conversion of velocity from element sides to nodes. The horizontal interpolation is
160
usually done on a fixed Z plane (instead of along an S plane). One problem with this
approach is the loss of accuracy near the bottom and the free surface (Fortunato and
Baptista 1996). Therefore, in SELFE, the interpolation can be optionally done in the
transformed S space in regions where no Z layers are used ("pure S region" with h≤hs).
The latter approach is more accurate in shallow regions where rapid changes in
bathymetry are common.
In the horizontal dimension, unstructured triangular grids are used, and the connectivity
of the grid is defined as follows: the three sides of an element i are enumerated as js (i,l)
(l = 1,2,3). The surrounding elements of a particular node i are enumerated as ine(i,l) (l
= 1, …, nne(i)), where nne(i) is the total number of elements in the “ball” of the node.
After the domain is discretized horizontally and vertically, the basic 3D computational
units of SELFE are triangular prisms. In the original Z space, the prisms may not have
level bottom and top surfaces. A staggering scheme is used to define variables. The
surface elevations are defined at the nodes. The horizontal velocities are defined at the
side centers and whole levels. The vertical velocities are defined at the element centers
and whole levels as the equations are solved with a finite-volume method. The linear
shape functions are used for elevations and velocities; note, however, that for velocities,
shape functions are only used for interpolation at the feet of characteristic lines. Note
that the shape functions used here are different from those in a lowest-order RaviartThomas element (Walters 2005), in that the elevations are not constant within an
element but continuous across elements. The locations where salinities and
temperatures are defined depend on the method used to solve the transport equations;
they are defined at the prism centers if the FVUM is used, and at both nodes and side
centers, at whole levels, if the ELM is used.
A2.2 Barotropic module
SELFE solves the barotropic Eqs. (1-3) first, as the transport and turbulent closure
equations lag one time step behind (in other words, the baroclinic pressure gradient
term in the momentum equation is treated explicitly in SELFE). Due to the hydrostatic
approximation, the vertical velocity w is solved from Eq. (1) after the horizontal velocity
is found. To solve the coupled Eqs. (2-3), it was first discretized and combined with the
vertical boundary conditions Eqs. (12-13) to be solved semi-implicitly in time as:
(20)
η
η
η n +1 − η n
+ θ∇∫ u n +1dz + (1 − θ )∇∫ u n dz =0
−h
−h
∆t
(21)
u n +1 − u*
∂  ∂u n +1 
= f n − gθ∇η n +1 − g (1 − θ )∇η n + ν n

∆t
∂z 
∂z 
(22)
 n ∂u n +1
n +1
ν
ηn;
τ=
w , at z
 =
∂z

n +1
ν n ∂u = χ nu n +1 , at z = −h,
b

∂z
,
161
where superscripts denote the time step, 0 ≤ θ ≤ 1 is the implicitness factor, u*(x,y,z,tn) is
the back-tracked value calculated with ELM , and χ n = CD | ubn | . The elevations in the 2nd
and 3rd terms of Eq. (20) are treated explicitly, which effectively amounts to a
linearization procedure.
A Galerkin weighted residual statement in the weak form for Eq. (20) reads:
(23)


d Ω + θ  − ∫ ∇φi U n +1d Ω + ∫ φiUˆ nn +1d Γ v + ∫ φiU nn +1d Γ v  +
∆t
 Ω

Ω
Γv
Γv


(1 − θ )  − ∫ ∇φi U n d Ω + ∫ φiU nn d=
(i 1,..., N p )
Γ  0,=
Γ
 Ω

∫ φi
η n +1 − η n
where Np is the total number of nodes,Γ ≡ Γ v + Γ v is the boundary of the entire
domain, with Γ v corresponding to the boundary segments where natural boundary
η
conditions are specified, U = ∫ udz is the depth-integrated velocity, Un is its normal
−h
component along the boundary, and Uˆ n is the boundary condition. In SELFE, linear
shape functions are used; thus, φi are conventional “hat” functions.
Integrating the momentum Eq. (21) along the vertical direction leads to:
(24)
1
U n +=
G n − gθ H n ∆t∇η n +1 − χ n ∆tubn +1
with
(25)
n
G=
τU* + ∆t F n +
n +1
w
− g (1 − θ ) H n∇η n  ,
ηn
ηn
−h
−h
Hn =
h +η n , Fn =
∫ fdz, U* =
∫ u*dz
Note that Eq. (24) involves no vertical discretization as it is merely an analytical
integration of Eq. (21).
To eliminate the unknown ubn +1 in Eq. (24), the discretized momentum equation form was
invoked, as applied to the top of the bottom cell:
162
(26)
ubn +1 − u*b
∂  ∂u n +1 
= fbn − gθ∇η n +1 − g (1 − θ )∇η n + ν n
 , at z = δ b − h
∆t
∂z 
∂z 
.
However, since the viscosity term vanishes inside the bottom boundary layer (Eq. (18)),
the bottom velocity can be formally solved as:
(27)
ubn +1= fˆbn − gθ∆t∇η n +1 ,
where:
(28)
fˆbn= u*b + fbn ∆t − g ∆t (1 − θ )∇η n .
Note that although the vertical viscosity is not explicitly present in Eq. (27), it is indirectly
involved through termsu*b and the Coriolis term in fbn .
Substituting Eq. (27) into (24) results in:
(29)
1
ˆ n − gθ Hˆ n ∆t∇η n +1 ,
U n +=
G
where:
(30)
ˆ n = G n − χ n ∆tfˆ n , Hˆ n = H n − χ n ∆t .
G
b
It is interesting to note from Eq. (30) that the bottom friction reduces the total depth by
an amount that is proportional to the drag coefficient and the bottom velocity. For
simplicity the Coriolis terms are treated explicitly in SELFE. It is well known that the
explicit treatment of the Coriolis terms is stable but introduces damping (Wicker and
Skamarock 1998). SELFE could have instead been formulated to treat the Coriolis
terms implicitly, in which case, the two components of U would become coupled in Eq.
(29), but could still be solved simultaneously from this equation.
163
Since SELFE uses linear shape functions for the elevations, the two components of the
horizontal velocity, u and v, are solved from the momentum equation independently
from each other after the elevations are found. This approach has important implications
as far as the Coriolis is concerned, and is different from that used in ELCIRC (Eulerian
Lagrangian Circulation). As a matter of fact, special treatment must be made to find the
tangential velocity components in UnTRIM-like models after the normal velocities are
found, as discussed in Zhang et al. (2004) and Ham et al. (2005).
Finally, substitution of Eq. (29) into (23) leads to an equation for elevations alone:
n +1
+ gθ 2 ∆t 2 Hˆ n∇φi ∇η n +1 d Ω − gθ 2 ∆t 2 ∫ φi Hˆ n
(31) ∫ φη
i
Ω
Γv
∂η n +1
d Γ v + θ∆t ∫ φiUˆ nn +1d Γ v = I n
∂n
Γv
where In consists of some explicit terms:
In
=
(32)
∫ φη
i
Ω
n
n
ˆ n d Ω − (1 − θ )∆t φ U n d Γ − θ∆t φ nG
+ (1 − θ )∆t∇φi U n + θ∆t∇φi G
∫i n
∫ i ˆ d Γv

Γ
Γv
Following standard finite-element procedures, and using appropriate essential and
natural boundary conditions, SELFE solves Eq. (31) to determine the elevations at all
nodes. For example, the integrals on Γ v need not be evaluated if the essential
boundary conditions are imposed by eliminating corresponding rows and columns of the
matrix. Natural boundary conditions are used to evaluate the integral on Γ v on the lefthand side of Eq. (31). If a Flather-type radiation condition (Flather 1987) needs to be
applied, it can be done in the following fashion:
+1
−Un
(33) Uˆ nn=
g / H (η n +1 − η ) ,
where U n and η are specified incoming current.
The matrix resulting from Eq. (31) is sparse and symmetric. It is also positive-definite
if a mild restriction is placed on the friction-reduced depth in the form of Hˆ n ≥ 0 .
Numerical experiments (not shown) indicated that even this restriction can be relaxed
for many practical applications that include shallow areas. The matrix can be efficiently
solved using a pre-conditioned Conjugate Gradient method (Casulli and Cattani 1994).
164
After the elevations are found, SELFE solves the momentum Eq. (3) along each vertical
column at side centers. A semi-implicit Galerkin finite-element method is used, with the
pressure gradient and the vertical viscosity terms being treated implicitly, and other
terms treated explicitly:
(34)
n +1

∂  ∂u  
  dz
∫− h γ k u − ∆t ∂z ν ∂z =
  j ,k
η
∫
η
−h
{
}
γ k u* + ∆t f jn,k − gθ∇η nj +1 − g (1 − θ )∇η nj  dz ,
where γ k ( z ) is the hat function in the vertical dimension. The two terms that are treated
implicitly would have imposed the most severe stability constraints. The explicit
treatment of the baroclinic pressure gradient and the horizontal viscosity terms,
however, does impose mild stability constraints.
After the velocities at all sides are found, the velocity at a node, which is needed in
ELM, is evaluated by a weighted average of all surrounding sides in its ball, aided by
proper interpolation in the vertical. The procedure to average the velocities (or
alternatively calculating the velocity at a node based on a least-square fit from all
surrounding sides) introduces numerical diffusion of the same order as the ELM. This is
because the velocities at nodes are not used anywhere else in the model except in ELM
tracking and interpolation. As an alternative to the averaging procedure, the velocity at a
node is computed within each element from the three sides using the linear shape
function and is kept discontinuous between elements. This approach leads to parasitic
oscillations, but a Shapiro filter (Shapiro 1970) can be used to suppress the noise, with
minimum distortion of physical features. The preliminary results indicate that the filter
approach induces less numerical diffusion.
The vertical velocity serves as a diagnostic variable for local volume conservation, but is
a physically important quantity, especially when a steep slope is present (Zhang et al.
2004). To solve the vertical velocity, we apply a finite-volume method to a typical prism,
assuming that w is constant within an element i, and obtain:
Sˆk +1 (ukn++11nkx+1 + vkn++11nky+1 + win,k+1+1nkz+1 ) − Sˆk (ukn +1nkx + vkn +1nky + win,k+1nkz ) +
(35)
3
∑ Pˆ
m =1
js ( i , m )
k b ,..., N z − 1)
(qˆ njs+(1i ,m ),k + qˆ njs+(1i ,m ),k +1 ) / 2 =
0, (k =
where Ŝ and P̂ are the areas of the five prism surfaces, (nx, ny, nz), are the
normal vector (pointing upward), u and v the averaged horizontal velocities at the
top and bottom surfaces, and q̂ is the outward normal velocity at each side center.
The vertical velocity is then solved from the bottom to the surface, in conjunction with
the bottom boundary condition. The closure error between the calculated w at the free
surface and the surface kinematic boundary condition is an indication of the local
165
volume conservation error (Luettich et al. 2002). Because the primitive form of the
continuity equation is solved in the model, this closure error is in general negligible.
A2.3 Baroclinic module
The barotropic module is one of the core parts of SELFE. To complete the model,
SELFE solves two more sets of equations: transport and turbulence closure equations.
The advection in the transport equations is usually a dominant process. SELFE treats
the advection in the transport equations with either an ELM or FVUM. If the ELM is
used, the transport equations are solved at nodes and side centers along each vertical
column using a finite-element method, with the lumping of the mass matrix to minimize
numerical dispersion (in the form of under- or over-shoots). In order to interpolation
used in ELM is important since linear interpolation leads to excessive numerical
diffusion. To reduce the numerical diffusion, element-splitting or quadratic interpolation
is used in ELM (Zhang et al. 2004).
Despite its efficiency, one of the main drawbacks of the ELM approach is its disregard
for mass conservation (Oliveira and Baptista 1998). On the other hand, FVUM
guarantees mass conservation. In FVUM, the scalar variables (salinity or temperature)
are defined at the center of a prism, (i,k), which has 5 exterior faces (top and bottom
Sˆ and Sˆi ,k −1
Pˆ
with areas i ,k
, and 3 vertical faces with areas jsj ,k ; The discretized
temperature equation reads:
Ti ,nk+1Vi ,nk + ∆t (un )in,+k1 Sˆi ,kTupn+(1i ,k ) + ∆t (un )in,+k1−1 Sˆi ,k −1Tupn+(1i ,k −1) =
(36)
 n Ti ,nk++11 − Ti ,nk+1
Ti ,nk+1 − Ti ,nk+−11 
n
,
∆tAi κ i ,k
− κ i ,k −1
+
n
n
∆
∆
z
z
i ,k +1/ 2
i ,k −1/ 2


3
 n

Q
n
k k b + 1,, N z )
(=
∆t  − ∆t ∑ q ln+1Tupn ( jsj ,k ) ,
Vi ,k  Ti ,k +


ρ 0C p 
l =1

where "up(i,j,k)" indicates upwinding, Vi ,k is the volume of the prism, un is the outward
normal velocity, jsj=js(i,l) are 3 sides, and q ln +1 = Pˆjsj ,k (un ) njsj+1,k are 3 horizontal advective
fluxes. The salinity equation is similarly discretized. Note that Eq. (36) reduces to Eq.
(35) when T=const. and Q = 0 .
The stability condition for the upwind scheme, the Courant number restriction, is given
by:
166
(37)
∆t ≤
Vi ,k
,
q
|
|
∑ j
j∈S +
where S+ indicates all outflow horizontal faces. Were the vertical advective fluxes on the
left-hand side of Eq. (36) treated explicitly, the denominator in Eq. (37) would include
the outflow faces for the top and bottom faces as well (Sweby 1984; Casulli and Zanolli
2005). But since the advective fluxes at the top and bottom faces are treated implicitly,
S+ excludes the top and bottom faces, and thus the more stringent stability constraints
associated with the vertical advective fluxes are by-passed. The Courant number
restriction (Eq. (37)) may still be too severe, and in this case the sub-division of a time
step is necessary. Despite the fact that Eq. (36) does not conform to the depth
integrated continuity Eq. (31), the FVUM guarantees mass conservation and the
maximum principle (i.e., the solution is bounded by the maximum and minimum of the
initial and boundary conditions; Casulli and Zanolli 2005), and thus is usually preferred
over the ELM approach. To further reduce the numerical diffusion, we have recently
implemented a higher-order finite-volume TVD scheme in SELFE (Sweby 1984).
SELFE solves the turbulence closure equations (Eqs. (6-7)) along each vertical column
at each node with a finite-element method. The vertical mixing terms and the dissipation
term in these equations are treated implicitly, but the production and buoyancy terms
are treated either implicitly or explicitly, depending on the sign of their total contribution
(Zhang et al. 2004). The advection terms in the turbulence closure equations are small
compared to other terms, and are therefore neglected in SELFE.
167
APPENDIX B: The ICM Water Quality Model Formulation
This section summarizes water quality and eutrophication processes and their
mathematical formulation in the ICM water quality-eutrophication model. The central
issues in the water quality model are primary production of carbon by algae and
concentration of dissolved oxygen. Primary production provides the energy required by
the ecosystem to function. Dissolved oxygen is necessary to support the life functions
of higher organisms and is considered an indicator of the health of estuarine systems.
Different from earlier water quality model such as WASP (Ambrose et al. 1992), which
use biochemical oxygen demand to represent oxygen demanding organic material, the
ICM water quality model is carbon based. The four algae species are represented in
carbon units and the three organic carbon variables play an equivalent role to BOD.
Organic carbon, nitrogen and phosphorous were represented by up to three reactive
sub-classes, refractory particulate, labile particulate and labile dissolve. Table B-1 lists
the model's complete set of state variables and their interactions are illustrated in Figure
4.2 in the main text. The use of the sub-classes allows a more realistic distribution of
organic material by reactive classes when data is to estimate distribution factors. The
following sub-sections discuss the role of each variable and summarize there kinetic
interaction processes. The kinetic sources and sinks, as well as the external loads for
each state variable, are described in details in Chapter 2 and 4 in the main text. The
kinetic processes include the horizontal transport fluxes described in Chapter 3 by
hydrodynamic process as well as at the sediment-water interface, including sediment
oxygen demand. The kinetic processes included in the ICM water quality model are
mostly from the Chesapeake Bay three-dimensional water quality model, CE-QUALICM (Cerco and Cole 1994). The description of the ICM water column water quality
model in this section is from Park et al. (1995).
Table B-1. ICM model water quality state variables
(1) cyanobacteria
(2) diatom algae
(3) green algae
(4) refractory particulate organic carbon
(5) labile particulate organic carbon
(6) dissolved organic carbon
(7) refractory particulate organic
phosphorus
(8) labile particulate organic phosphorus
(9) dissolved organic phosphorus
(10) total phosphate
(11) refractory particulate organic nitrogen
(12) labile particulate organic nitrogen
(13) dissolved organic nitrogen
(14) ammonia nitrogen
(15) nitrate nitrogen
(16) particulate biogenic silica
(17) dissolved available silica
(18) chemical oxygen demand
(19) dissolved oxygen
(20) salinity
(21) temperature
168
B1 Model State Variables
B1.1 Algae
Algae are grouped into three model classes: cyanobacteria, diatoms, and green algae.
The grouping is based upon the distinctive characteristics of each class and upon the
significant role the characteristics play in the ecosystem. Cyanobacteria are
characterized by their bloom-forming characteristics in fresh water. Cyanobacteria are
unique in that some species fix atmospheric nitrogen, although nitrogen fixers are not
believed to be predominant in many river systems. Diatoms are distinguished by their
requirement of silica as a nutrient to form cell walls. Diatoms are large algae
characterized by high settling velocities. Settling of spring diatom blooms to the
sediments may be a significant source of carbon for sediment oxygen demand. Algae
that do not fall into the preceding two groups are lumped into the heading of green
algae. Green algae settle at a rate intermediate between cyanobacteria and diatoms
and are subject to greater grazing pressure than cyanobacteria.
B1.2 Organic Carbon
Three organic carbon state variables are considered: dissolved, labile particulate, and
refractory particulate. Labile and refractory distinctions are based upon the time scale
of decomposition. Labile organic carbon decomposes on a time scale of days to weeks
whereas refractory organic carbon requires more time. Labile organic carbon
decomposes rapidly in the water column or the sediments. Refractory organic carbon
decomposes slowly, primarily in the sediments, and may contribute to sediment oxygen
demand years after deposition.
B1.3 Nitrogen
Nitrogen is first divided into organic and mineral fractions. Organic nitrogen state
variables are dissolved organic nitrogen, labile particulate organic nitrogen, and
refractory particulate organic nitrogen. Two mineral nitrogen forms are considered:
ammonium and nitrite and nitrate combined. Both ammonium and nitrate are utilized to
satisfy algal nutrient requirements, although ammonium is preferred. The primary
reason for distinguishing the two is that ammonium is oxidized by nitrifying bacteria into
nitrate. This oxidation can be a significant sink of oxygen in the water column and
sediments. An intermediate in the complete oxidation of ammonium, nitrite, also exists.
Nitrite concentrations are usually much less than nitrate, and for modeling purposes,
nitrite is combined with nitrate. Hence the nitrate state variable actually represents the
sum of nitrate plus nitrite.
B1.4 Phosphorus
As with carbon and nitrogen, organic phosphorus is considered in three states:
dissolved, labile particulate, and refractory particulate. Only a single mineral form, total
phosphate, is considered. Total phosphate exists as several states within the model
ecosystem: dissolved phosphate, phosphate sorbed to inorganic solids, and phosphate
incorporated in algal cells. Equilibrium partition coefficients are used to distribute the
total among the three states.
169
B1.5 Silica
Silica is divided into two state variables: available silica and particulate biogenic silica.
Available silica is primarily dissolved and can be utilized by diatoms. Particulate
biogenic silica cannot be utilized. In the model, particulate biogenic silica is produced
through diatom mortality. Particulate biogenic silica undergoes dissolution to available
silica or else settles to the bottom sediments.
B1.6 Chemical Oxygen Demand
In the context of this study, chemical oxygen demand is the concentration of reduced
substances that are oxidizable by inorganic means. The primary component of chemical
oxygen demand is sulfide released from sediments. Oxidation of sulfide to sulfate may
remove substantial quantities of dissolved oxygen from the water column.
B1.7 Dissolved Oxygen
Dissolved oxygen is one of the most important parameters of water quality.It is a basic
requirement for a healthy aquatic ecosystem and is used to measure the amount of
oxygen available for biochemical activity in water.
B1.8 Salinity
Salinity is a conservative tracer that provides verification of the transport component of
the model and facilitates examination of conservation of mass. Salinity also influences
the dissolved oxygen saturation concentration and is used in the determination of
kinetics constants that differ in saline and fresh water. Salinity is simulated in the
hydrodynamic component of the model.
B1.9 Temperature
Temperature is a primary determinant of the rate of biochemical reactions. Reaction
rates increase as a function of temperature, although extreme temperatures result in the
mortality of organisms. Temperature is simulated in the hydrodynamic component of the
model.
B2 Conservation of Mass Equation
The governing mass-balance equation for each of the water quality state variables may
be expressed as
∂C ∂ (uC ) ∂ (vC ) ∂ ( wC ) ∂  ∂C  ∂ 
∂C  ∂  ∂C 
+
+
+ =
 +  Kz
 Kx
 +  Ky
 + SC
∂t
∂x
∂y
∂z
∂x 
∂x  ∂y 
∂y  ∂z 
∂z 
(B.1)
where
C = concentration of a water quality state variable.
u, v, w = velocity components in the x-, y- and z-directions, respectively.
Kx, Ky, Kz = turbulent diffusivities in the x-, y- and z-directions, respectively.
SC = internal and external sources and sinks per unit volume.
170
The equation (B.1) incorporates transport due to flow advection and dispersion, external
pollutant inputs and the kinetic interaction between the water quality variables. The last
three terms on the left-hand side (LHS) of equation (B.1) account for the advective
transport, and the first three terms on the right-hand side (RHS) of equation (B.1)
account for the diffusive transport. These six terms for physical transport are analogous
to, and thus the numerical method of solution is the same as, those in the mass-balance
equation for salinity in the hydrodynamic model. The last term in equation (B.1)
represents the kinetic processes and external loads for each of the state variables. The
present model solves equation (B.1) after decoupling the kinetic terms from the physical
transport terms.
The governing mass-balance equation for water quality state variables (equation B.1)
consists of physical transport, advective and diffusive, and kinetic processes. When
solving equation B.1, the kinetic terms are decoupled from the physical transport terms.
The mass-balance equation for physical transport only, which takes the same form as
the salt-balance equation, is:
∂C ∂ (uC ) ∂ (vC ) ∂ ( wC ) ∂  ∂C  ∂ 
∂C  ∂  ∂C 
+
+
+ =
 +  Kz
 Kx
 +  Ky

∂t
∂x
∂y
∂z
∂x 
∂x  ∂y 
∂y  ∂z 
∂z 
(B.2)
The equation for kinetic processes only, which will be referred to as the kinetic equation,
is:
∂C
= SC
∂t
(B.3)
which may be expressed as:
∂C
= k ⋅C + R
∂t
(B.4)
-1
-1
-1
where K is kinetic rate (time ) and R is source/sink term (mass volume time ).
Equation (B.4) is obtained by linearizing some terms in the kinetic equations, mostly
Monod type expressions. Hence, K and R are known values in equation (B.4).
The remainder of this section details the kinetics portion of the mass-conservation
equation for each state variable. For consistency with reported rate coefficients, kinetics
are detailed using a temporal dimension of days. Within the CE-QUAL-ICM computer
code, kinetics sources and sinks are converted to a dimension of seconds before
employment in the mass-conservation equation.
171
B2.1 Algae
Algae are primary producers which are able to utilize light, carbon dioxide, and nutrients
to synthesize new organic material. Algae play a key role in the eutrophication process
and are essential for water quality modeling. Algae affect the nitrogen cycle, the
phosphorus cycle, the DO balance, and the food chain, primarily through nutrient uptake
and algae death. As algae grow and die, they form part of the nutrient cycles.
Algae are grouped into three model state variables: cyanobacteria, diatoms and green
algae. The algae kinetics is governed by the following processes:
1.
2.
3.
4.
5.
Algal growth
Metabolism, including respiration and excretion
Predation
Settling
External sources
and a general equation that includes all of these processes can be expressed as:
Net algal production = algal growth – metabolism – predation – settling + external
sources
The kinetic equation can now be written for algae as:
∂Bx
WBx
∂
=( Px − BM x − PRx ) Bx + (WS x ·Bx ) +
V
∂t
∂z
(B.5)
where
Bx = algal biomass of algal group x (g C m-3)
t = time (day)
Px = production rate of algal group x (day-1)
BMx = basal metabolism rate of algal group x (day-1)
PRx = predation rate of algal group x (day-1)
WSx = settling velocity of algal group x (m day-1)
WBx = external loads of algal group x (g C day-1)
V = volume (m3)
subscript x = c, d, g are cyanobacteria, diatoms and green algae, respectively.
172
Algal growth (production)
Algal growth is the most important process for algae modeling. The algal growth rate is
a complicated function of temperature, light, and nutrients and is often the determining
factor for the net algal production. The effects of these processes are considered to be
multiplicative:
Px = PM x · f1 ( N )· f 2 ( I )· f3 (T )
(B.6)
where
PMx = maximum growth rate for algal group x (day-1)
f1(N) = growth limiting function for nutrients (0 ≤ f1 ≤ 1)
f2(I) = growth limiting function for light intensity (0 ≤ f2 ≤ 1)
f3(T) = growth limiting function for temperature (0 ≤ f3 ≤ 1)
The freshwater cyanobacteria coming from upstream rivers can undergo rapid mortality
in salt water. Hence, the growth of freshwater blue-green algae in saline water can be
limited by (Cerco and Cole, 1994):
Px = PM x · f1 ( N )· f 2 ( I )· f 3 (T )· f 4 ( S )
(B.7)
where f4(S) = growth limiting function for salinity (0 ≤ f4 ≤ 1) and
Effects of nutrients for algal growth
Using Liebig's Law of the Minimum, which states that growth is determined by the
nutrient in least supply, the nutrient limitation for growth of blue-green algae and green
algae is expressed as:


NH 4 + NO3
PO 4d
,
f1 ( N ) = minimum 

 KHN x + NH 4 + NO3 KHPx + PO 4d 
(B.8)
where
KHNx = nitrogen half saturation constant for algal group x, mass/volume
KHPx = phosphorus half saturation constant for algal group x, mass/volume
subscript x = c for cyanobacteria and g for green algae, respectively, mass/volume
173
When diatoms are considered, silicon limitation should be included, and Eq. (B.8) is
modified to:


NH 4 + NO3
PO 4d
SAd
f1 ( N ) = minimum 
,
,

 KHN d + NH 4 + NO3 KHPd + PO 4d KHS + SAd 
(B.9)
where KHNd = nitrogen half saturation constant for diatoms, mass/volume
KHPd = phosphorus half saturation constant for diatoms, mass/volume
Effect of light on algal growth
The daily and vertically integrated form of Steel’s equation is:
f2 (I )
=
2.718·FD −α B
e − e −αT )
(
Kess·∆z
(B.10)
with
=
αB
=
αT
Io
·exp ( − Kess[ H T + ∆z ])
FD·( I s ) x
(B.11)
Io
·exp ( − Kess·H T )
FD·( I s ) x
(B.12)
where
FD = fractional daylength (0 ≤ F D ≤1)
Kess = Ke = total light extinction coefficient (m-1)
Δz = layer thickness (m)
Io = daily total light intensity at water surface (langleys day-1)
(Is)x = optimal light intensity for algal group x (langleys day-1)
HT = depth from the free surface to the top of the layer (m).
The total light extinction coefficient, Kess, is given by
Kess =
Keb + KeTSS ·TSS + KeChl ·
∑
x =c ,d , g
(
Bx
)
CChlx
(B.13)
where
Keb = background light extinction, m-1
KeTSS = light extinction coefficient for total suspended solid, m-1 per g m-3
TSS = total suspended solid concentration provided from the hydrodynamic model in
unit of g m-3
KeChl = light extinction coefficient for chlorophyll a, m-1 per mg Chl m-3
CChlx = carbon-to-chlorophyll ratio in algal group x, g C per mg Chl.
174
The optimal light intensity (Is) for photosynthesis is expressed as:
− Kess ·( Dopt ) x
( I s ) x = minimum ( I o ) avg ·e
, ( I s ) min 
(B.14)
where
(Dopt)x = depth of maximum algal growth for algal group x (m)
(Io)avg = adjusted surface light intensity (langleys day-1).
A minimum, (Is)min, in Eq. (B.14) is specified so that algae do not thrive at extremely lowlight levels. The adjusted surface light intensity, (Io)avg, is estimated as:
( I o ) avg = CI a ·I o + CI b ·I1 + CI c ·I 2
(B.15)
where
I1 = daily light intensity one day preceding model day (langleys day-1)
I2 = daily light intensity two days preceding model day (langleys day-1)
CIa, CIb, and CIc = weighting factors for I0, I1 and I2, respectively: CIa + CIb + CIc = 1.
Effects of temperature on algal growth
Most water quality processes are temperature dependent. Temperature significantly
influences the kinetic rates of nutrient transformations; the rate of chemical reactions
increases with temperature.
Algal growth rate is controlled by temperature, water movement, nutrients, and light. It
increases with temperature until an optimum is reached, after which further temperature
increase will inhibit growth. The value of this optimum temperature varies with the
species concerned and with light and nutrients. This temperature effect can be
expressed as:
e − KTG1x (T −TM 1x )

1.0
 − KTG 2x (T −TM 2x )2
e
2
f3 (T )
if T < TM 1x
if TM1x ≤ T ≤ TM 2 x
(B.16)
if T > TM 2 x
where
f3 (T ) = algal growth function
TM1x = lower end of optimal temperature range for algal growth for algal group x
TM2x = upper end of optimal temperature range for algal growth for algal group x
KTG1x = effect of temperature below TM1x on growth for algal group x
KTG2x = effect of temperature above TM2x on growth for algal group x
subscript x = c for cyanobacteria, d for diatom, and g for green algae
175
Effects of salinity on growth of freshwater Cyanobacteria
The growth of freshwater cyanobacteria in salt water is limited by:
STOX 2
f4 (S ) =
STOX 2 + S 2
(B.17)
where STOX = salinity at which algal growth is halved, ppt
S = salinity in water column, ppt
Algal basal metabolism
Basal metabolism is a general term for biochemical processes that occur in living
organisms by which energy is provided for vital processes and activities. Basal
metabolism in the present model is the sum of all internal processes that decrease algal
biomass and consists of two parts: respiration and excretion. In basal metabolism,
algal matter (carbon, nitrogen, phosphorus, and silica) is returned to organic and
inorganic pools in the environment, mainly to dissolved organic and inorganic matter.
Respiration, which may be viewed as a reversal of production, consumes dissolved
oxygen. Basal metabolism is considered to be an exponentially increasing function of
temperature
=
BM x BMRx  exp( KTBx [T − TRx ])
(B.18)
where
BMRx = basal metabolism rate at TRx for algal group x, 1/time
KTBx = temperature function for basal metabolism, dimensionless
TRx = reference temperature for basal metabolism for algal group x, oC
176
Algal predation
Zooplankton is the plankton consisting of animal life that is moved by flows. It includes
the larval forms of large adult organisms (e.g., crabs, fish) and small animals that never
get larger than several millimeters. Zooplankton consumes algae, bacteria, detritus, and
sometimes other zooplankton, and is in turn eaten by small fish. Algal predation is the
consumption of algae by zooplankton or other aquatic organisms. Present model does
not include zooplankton; instead a temperature dependent rate is specified for algal
predation:
=
PRx PRRx  exp( KTBx [T − TRx ])
(B.19)
where
PRx = predation rate of algal group x, day-1
KTBx = temperature function for predation, dimensionless
Algal settling
Algal settling in natural waters is a complex phenomenon and depends on many factors,
such as:
1. The density, size, and shape of the algae
2. The density, velocity, turbulence strength, and viscosity of the water.
It is impractical to calculate the algal setting velocity in a water quality model. Settling
velocities for three algal groups, WSc, WSd , WSg, are specified as an input. Seasonal
variations in settling velocity of diatoms can be accounted for by specifying time-varying
WSd.
B2.2 Organic Carbon
The production of organic carbon is a key process in eutrophication study. The organic
carbon cycle consists of photosynthesis, respiration, and decomposition. Because some
organic carbons decay at faster rates than others, organic carbon can be divided into
those that decay at a fast rate (labile) and those that decay at a slower rate (refractory).
In water quality models, organic carbon can be categorized as:
1. Refractory particulate organic carbon (RPOC)
2. Labile particulate organic carbon (LPOC)
3. Dissolved organic carbon (DOC)
Total organic carbon (TOC) is the sum of all organic carbon compounds and can be
expressed as: TOC = RPOC + LPOC + DOC
177
Particulate organic carbon
Labile and refractory distinctions are based on the time scale of decomposition. LPOC
has a decomposition time scale of days to weeks and decays rapidly either in the water
column or sediment bed. RPOC has a decomposition time scale of months to seasons,
after being settled to the sediment bed. Through the sediment diagenesis processes,
the settled RPOC in the bed may affect the water quality in the water column for a long
time (seasons and even years). Sources of organic carbon include excretion and death
of living organisms (such as algae) and external loadings. The discharges of organic
matter from point sources (such as wastewater treatment plants) can be a major source
of organic carbon, leading to large DO deficits and violation of water quality standards.
The governing equation for RPOC and LPOC are:
∂RPOC
=
∂t
∑
FCRP·PRx ·Bx − K RPOC ·RPOC +
∑
FCLP·PRx ·Bx − K LPOC ·LPOC +
x =c ,d , g
∂
WRPOC
(WS RP ·RPOC ) +
∂z
V
(B.20)
and
∂LPOC
=
∂t
x =c ,d , g
∂
WLPOC
(WS LP ·LPOC ) +
∂z
V
(B.21)
where
RPOC = concentration of refractory particulate organic carbon (g C m-3)
LPOC = concentration of labile particulate organic carbon (g C m-3)
FCRP = fraction of predated carbon produced as refractory particulate organic carbon
FCLP = fraction of predated carbon produced as labile particulate organic carbon
KRPOC = hydrolysis rate of refractory particulate organic carbon (day-1)
KLPOC = hydrolysis rate of labile particulate organic carbon (day-1)
WSRP = settling velocity of refractory particulate organic matter (m day-1)
WSLP = settling velocity of labile particulate organic matter (m day-1)
WRPOC = external loads of refractory particulate organic carbon (g C day-1)
WLPOC = external loads of labile particulate organic carbon (g C day-1).
178
Dissolved organic carbon
Sources and sinks for dissolved organic carbon include:
1.
2.
3.
4.
5.
6.
7.
Algal excretion
Algal predation
Hydrolysis from RPOC
Hydrolysis from LPOC
Heterotrophic respiration of DOC
Denitrification
External loads
This yields: Net change of DOC = algal excretion + algal predation
+ RPOC hydrolysis + LPOC hydrolysis
– DOC heterotrophic respiration - denitrification
+ external loads
The governing equation for DOC is:
∂DOC
=
∂t

KHRx 
 FCDx + (1 − FCDx )
BM x ⋅ Bx + ∑ FCDP·PRx ·Bx
KHRx + DO 
,d , g 
x c=
x c ,d , g
WDOC
+ K RPOC ·RPOC + K LPOC ·LPOC − K HR ·DOC − Denit ·DOC +
V
∑
(B.22)
where
DOC = concentration of dissolved organic carbon (g C m-3)
FCDx = a constant for algal group x (0 < FCDx <1)
KHRx =half-saturation constant of dissolved oxygen for algal dissolved organic carbon
excretion for group x (g O2 m-3)
DO = dissolved oxygen concentration (g O2 m-3)
FCDP = fraction of predated carbon produced as dissolved organic carbon
KHR = heterotrophic respiration rate of dissolved organic carbon (day-1)
Denit = denitrification rate (day-1)
WDOC = external loads of dissolved organic carbon (g C day-1).
The remainder of this section explains the processes expressed in equations (B.20) (B.22). Two algal processes affect organic carbon concentrations: algal excretion and
algal predation by zooplankton are represented by the terms with summation (
Eqs. (B.20), (B.21), and (B.22).
∑
x =c ,d , g
) in
179
Basal metabolism
Basal metabolism, consisting of respiration and excretion, returns algal matter (carbon,
nitrogen, phosphorus, and silica) back to the environment. Loss of algal biomass
through basal metabolism is expressed as first order reaction equation:
∂Bx
= − BM x  Bx
∂t
(B.23)
which indicates that the total loss of algal biomass due to basal metabolism is
independent of ambient dissolved oxygen concentration.
In the governing equation for algae, Eq. (B.5), the basal metabolism term ( − BM x ⋅ Bx )
actually includes two separated processes: the algal excretion and respiration. In this
model, it is assumed that the distribution of total loss between respiration and excretion
is constant as long as there is sufficient dissolved oxygen for algae to respire. Under
that condition, the losses by respiration and excretion may be written as:
(1 − FCDx )  BM x  Bx due to respiration
(B.24)
FCDx  BM x  Bx due to excretion
(B.25)
where FCDx is a constant of value between 0 and 1. However, algae cannot respire in
the absence of oxygen. Although the total loss of algal biomass due to basal
metabolism is oxygen independent (equation B.23), the distribution of total loss between
respiration and excretion is oxygen-dependent. When oxygen level is high, respiration
is a large fraction of the total. As dissolved oxygen becomes scarce, excretion
becomes dominant. Thus, equation (B.24) represents the loss by respiration only at
high oxygen levels. In general, equation (B.24) can be decomposed into two fractions
as a function of dissolved oxygen availability:
DO
due to respiration
(B.26)
BM x  Bx
(1 − FCDx ) 
KHRx + DO
(1 − FCDx ) 
KHRx
BM x  Bx
DO + KHR
due to excretion
(B.27)
Equation (B.26) represents the loss of algal biomass by respiration, and equation (B.27)
represents additional excretion due to insufficient dissolved oxygen concentration. The
parameter KHRx, which is defined as the half-saturation constant of dissolved oxygen
for algal dissolved organic carbon excretion in equation (B.22), can also be defined as
the half-saturation constant of dissolved oxygen for algal respiration in equation (B.26).
180
Combining equations (B.25) and (B.27), the total loss due to excretion is:

KHRx 
(1 − FCDx ) 
 BM x  Bx
+
DO
KHR


(B.28)
Equations (B.26) and (B.28) combine to give the total loss of algal biomass due to basal
metabolism. The definition of FCDx in equation (B.22) becomes apparent in equation
(B.28) (i.e., fraction of basal metabolism exuded as dissolved organic carbon at infinite
dissolved oxygen concentration). At zero oxygen level, 100 percent of total loss due to
basal metabolism is by excretion regardless of FCDx. The end carbon product of
respiration is primarily carbon dioxide, an inorganic form not considered in the present
model, while the end carbon product of excretion is primarily dissolved organic carbon.
Therefore, equation (B.28), that appears in equation (B.22), represents the contribution
of excretion to dissolved organic carbon, and there is no source term for particulate
organic carbon from algal basal metabolism in equations (B.20) and (B.21).
Predation
Algae produce organic carbon through the effects of predation. Zooplankton take up
and redistribute algal carbon through grazing, assimilation, respiration, and excretion.
Since zooplankton are not included in the model, routing of algal carbon through
zooplankton predation is simulated by empirical distribution coefficients in equations
(B.20) to (B.22): FCRP, FCLP and FCDP. The sum of these three predation fractions
should be unity.
Heterotrophic respiration and dissolution
The fifth term on the RHS of Eq. (B.22), − K HR ⋅ DOC , represents the heterotrophic
respiration that converts DOC into CO2. Heterotrophic respiration needs oxygen. A
Michaelis-Menten function can be used to represent the dependency of heterotrophic
respiration rate, KHR, on DO concentration. It has the form:
K HR =
DO
K DOC
KHORDO + DO
(B.29)
where
KHORDO = oxic respiration half-saturation constant for dissolved oxygen (g O2 m-3)
KDOC = heterotrophic respiration rate of dissolved organic carbon at infinite dissolved
oxygen concentration (day-1).
The dissolution (hydrolysis) rates of RPOC and LPOC and the heterotrophic respiration
181
rate of DOC, KRPOC, KLPOC, and KDOC, can be specified by the following:
K RPOC
= ( K RC + K RCalg
K LPOC
= ( K LC + K LCalg
K=
( K DC + K DCalg
DOC
∑
Bx )·e KTHDR (T −TRHDR )
∑
Bx )·e KTHDR (T −TRHDR )
∑
Bx )·e KTMNL (T −TRMNL )
x =c ,d , g
x =c ,d , g
x =c ,d , g
(B.30)
(B.31)
(B.32)
where
KRC = minimum dissolution rate of refractory particulate organic carbon (day-1)
KLC = minimum dissolution rate of labile particulate organic carbon (day-1)
KDC = minimum respiration rate of dissolved organic carbon (day-1)
KRCalg & KLCalg = constants that relate dissolution of refractory and labile particulate
organic carbon, respectively, to algal biomass (day-1 per g C m-3)
KDCalg = constant that relates respiration to algal biomass (day-1 per g C m-3)
KTHDR = effect of temperature on hydrolysis of particulate organic matter (oC-1)
TRHDR = reference temperature for hydrolysis of particulate organic matter (oC)
KTMNL = effect of temperature on mineralization of dissolved organic matter (oC-1)
TRMNL = reference temperature for mineralization of dissolved organic matter (oC).
Eqs. (B.30) - (B.32) indicate that RPOC and LPOC are converted to DOC via an
hydrolysis process, while DOC is converted to CO2 via a mineralization process.
Hydrolysis and mineralization will also be used to describe the conversions of organic
phosphorus and organic nitrogen later in this chapter.
Effects of denitrification on dissolved organic carbon
As oxygen is depleted from natural systems, organic matter is oxidized by the reduction
of alternate electron acceptors. Thermodynamically, the first alternate acceptor reduced
in the absence of oxygen is nitrate. The reduction of nitrate by a large number of
heterotrophic anaerobes is referred to as denitrification, and the stoichiometry of this
reaction is:
5 CH 2O + 4 NO 3 − + 4 H + → 5 CO 2 + 2 N 2 + 7 H 2O
(B.33)
182
The 4th term in equation (B.22) accounts for the effect of denitrification on dissolved
organic carbon. The kinetics of denitrification in the model are first-order:
Denit =
KHORDO
NO3
AANOX ·K DOC
KHORDO + DO KHDN N + NO3
(B.34)
where
KHDNN = denitrification half saturation constant for nitrate (g N m-3)
AANOX = ratio of denitrification rate to oxic dissolved organic carbon respiration rate.
In equation (B.34), the dissolved organic carbon respiration rate, KDOC, is modified so
that significant decomposition via denitrification occurs only when nitrate is freely
available and dissolved oxygen is depleted. The ratio, AANOX, makes the anoxic
respiration slower than oxic respiration. Note that KDOC, defined in equation (B.32),
includes the temperature effect on denitrification.
B2.3 Phosphorus
Phosphorus exists in organic and inorganic forms. Both forms include particulate and
dissolved phases. Total phosphorus (TP) is a measure of all forms of phosphorus and is
widely used for setting trophic state criteria. In the present model, TP was split into the
following state variables in a water quality model:
1.
2.
3.
4.
Refractory particulate organic phosphorus (RPOP)
Labile particulate organic phosphorus (LPOP)
Dissolved organic phosphorus (DOP)
Total phosphate (PO4t)
Phosphorus processes are closely linked to sediment processes, especially shallow
waters. It is critical to have a good representation of sediment processes, before
phosphorus processes can be described realistically.
183
Particulate organic phosphorus
For particulate organic phosphorus (POP), the source and sinks are:
1.
2.
3.
4.
5.
Algal metabolism
Algal predation
Hydrolysis of POP to dissolved organic phosphorus
Settling
External loads
and can be described as:
The change of POP = Algal basal metabolism + algal predation – POP hydrolysis
– settling + external source
The kinetic equations for RPOP and LPOP are thus (Cerco and Cole, 1994; Park et al.,
1995):
∂RPOP
=∑ ( FPRx ·BM x + FPRP·PRx ) APC ·Bx − K RPOP ·RPOP
∂t
x =c ,d , g
+
∂
WRPOP
(WS RP ·RPOP ) +
∂z
V
(B.35)
and
∂LPOP
=∑ ( FPLx ·BM x + FPLP·PRx ) APC ·Bx − K LPOP ·LPOP
∂t
x =c ,d , g
+
∂
WLPOP
(WS LP ·LPOP ) +
∂z
V
(B.36)
where
RPOP = concentration of refractory particulate organic phosphorus (g P m-3)
LPOP = concentration of labile particulate organic phosphorus (g P m-3)
FPRx = fraction of metabolized phosphorus by algal group x produced as refractory
particulate organic phosphorus
FPLx = fraction of metabolized phosphorus by algal group x produced as labile
particulate organic phosphorus
FPRP = fraction of predated phosphorus produced as refractory particulate organic
phosphorus
FPLP = fraction of predated phosphorus produced as labile particulate organic
phosphorus
184
APC = mean phosphorus-to-carbon ratio in all algal groups (g P per g C)
KRPOP = hydrolysis rate of refractory particulate organic phosphorus (day-1)
KLPOP = hydrolysis rate of labile particulate organic phosphorus (day-1)
WRPOP = external loads of refractory particulate organic phosphorus (g P day-1)
WLPOP = external loads of labile particulate organic phosphorus (g P day-1).
Dissolved organic phosphorus
For dissolved organic phosphorus, the major source and sinks are:
1.
2.
3.
4.
5.
Algal metabolism
Algal predation
Hydrolysis from RPOP and LPOP
Mineralization to phosphate phosphorus
External loads
These processes can be expressed as:
The change of DOP = Algal basal metabolism + algal predation + POP hydrolysis
– mineralization + external source
The corresponding kinetic equation is:
∂DOP
=
∂t
∑
x =c ,d , g
( FPDx ·BM x + FPDP·PRx ) APC ·Bx
+ K RPOP ·RPOP + K LPOP ·LPOP − K DOP ·DOP +
WDOP
V
(B.37)
where
DOP = concentration of dissolved organic phosphorus (g P m-3)
FPDx =fraction of metabolized phosphorus by algal group x produced as dissolved
organic phosphorus
FPDP = fraction of predated phosphorus produced as dissolved organic phosphorus
KDOP = mineralization rate of dissolved organic phosphorus (day-1)
WDOP = external loads of dissolved organic phosphorus (g P day-1).
185
Total phosphate
Total phosphate (PO4t) includes dissolved phosphate (PO4d) and sorbed phosphate
(PO4p) or PO4t = PO4d + PO4p. The amount of total phosphate in a water body
depends on:
1.
2.
3.
4.
5.
Algal metabolism, predation, and uptake
Mineralization from dissolved organic phosphorus
Settling of sorbed phosphate
Exchange of dissolved phosphate at the sediment bed - water column interface
External loads
The corresponding kinetic equation is:
∂PO 4t
=
∂t
∑
x =c ,d , g
+
( FPI x ·BM x + FPIP·PRx − Px ) APC ·Bx + K DOP ·DOP
∂
BFPO 4d WPO 4t
+
(WSTSS ·PO 4 p ) +
∂z
∆z
V
(B.38)
where
PO4t = total phosphate (g P m-3)
PO4p = particulate (sorbed) phosphate (g P m-3)
FPIx = fraction of metabolized phosphorus by algal group x produced as inorganic
phosphorus
FPIP = fraction of predated phosphorus produced as inorganic phosphorus
WSTSS = settling velocity of suspended sediment (m day-1), provided by the sediment
model
BFPO4d = sediment-water exchange flux of phosphate (g P m-2 day-1), applied to the
bottom layer only
WPO4t = external loads of total phosphate (g P day-1).
Sorption and desorption of phosphate
In the presence of oxygen, dissolved phosphates combine with suspended particles.
These particles eventually settle to the sediment bed and are temporarily removed from
the cycling process. The settling of suspended solids and sorbed phosphorus can
provide a significant loss mechanism of phosphorus from the water column to the bed.
The sorption-desorption processes of phosphate are much faster than those for
biological kinetics. The former are on the order of minutes; the latter are on the order of
days. This difference permits an instantaneous equilibrium assumption for the
calculation of phosphate. The dissolved phosphate and the particulate (sorbed)
phosphate is treated as a single state variable. The dissolved and particulate
186
phosphates may be expressed as:
PO 4 p =
PO 4d =
K PO 4 p ·S
1 + K PO 4 p ·S
1
1 + K PO 4 p ·S
PO 4t
(B.39)
PO 4t
(B.40)
where
KPO4p = partition coefficient of phosphate (m3/g)
S = sediment concentration (g/m3)
Dividing Eq. (B.39) by Eq. (B.40) gives:
K PO 4 p =
PO 4 p 1
PO 4d S
(B.41)
The meaning of KPO4p becomes apparent in Eq. (B.41): the partition coefficient is the
ratio of the particulate concentration to the dissolved concentration per unit
concentration of suspended solid.
Algal phosphorus-to-carbon ratio (APC)
Algal biomass is often expressed in units of carbon per volume of water. In order to
estimate the nutrients contained in algal biomass, the ratio of phosphorus-to-carbon,
APC, should be known.
Algal composition varies as a function of nutrient availability and adapts to ambient
phosphorus concentration. When the concentrations of available phosphorus and
nitrogen are low, algae adjust their composition so that smaller quantities of these
nutrients are needed to produce carbonaceous biomass. Algal phosphorus content is
high when ambient phosphorus is high, and is low when ambient phosphorus is low.
Based on measured data, Cerco and Cole (1994) reported large variations of the algal
phosphorus-to-carbon ratio and used the following empirical formulation to estimate the
algal phosphorus-to-carbon ratio:
1
APC =
− CP
·PO 4 d
CPprm1 + CPprm 2 ·e prm 3
(B.42)
where
CPprm1 = minimum carbon-to-phosphorus ratio (g C per g P)
CPprm2 = difference between minimum and maximum carbon-to-phosphorus ratio (g C
per g P)
CPprm3 = effect of dissolved phosphate concentration on carbon-to-phosphorus ratio
(per g P m-3).
187
Effects of algae on phosphorus
As algae grow, dissolved inorganic phosphorus (PO4d) is taken up, stored, and
incorporated into algal biomass. Living algal cells are a major component of the total
phosphorus pool in the water. Settling of algae to the bottom sediments is a major loss
pathway of phosphorus from the water column. As algae respire and die, algal biomass
(and the phosphorus) is recycled to nonliving organic and inorganic matters. The effects
∑
of algae are represented by the summation terms ( x =c ,d , g ) in Eqs. (B.35), (B.36), (B.37),
and (B.38). The total algal loss by basal metabolism in Eq. (B.5), is split using
distribution coefficients FPRx, FPLx, FPDx and FPIx. The algal predation is accounted for
by the terms associated with PRx, the predation rate of algal group x. The total loss by
predation, the term of PRx·Bx in Eq. (B.5), is split using distribution coefficients, FPRP,
FPLP, FPDP, and FPIP, in which FPRP + FPLP + FPDP + FPIP = 1.
Mineralization and hydrolysis
Organic nutrients undergo hydrolysis and mineralization to become inorganic nutrients
before being consumed by algae. The hydrolysis of particulate organic phosphorus is
represented by the term of KRPOP in Eq. (B.35) and the term of KLPOP in Eq. (B.36). The
mineralization of dissolved organic phosphorus is represented by the term of KDOP in Eq.
(B.37). The formulations for hydrolysis and mineralization rates are (Park et al., 1995):
K RPOP
= ( K RP +
K LPOP
= ( K LP +
K=
( K DP +
DOP
KHP
K RPalg ∑ Bx )·e KTHDR (T −TRHDR )
KHP + PO 4d
x =c ,d , g
(B.43)
KHP
K LPalg ∑ Bx )·e KTHDR (T −TRHDR )
KHP + PO 4d
x =c ,d , g
(B.44)
KHP
K DPalg ∑ Bx )·e KTMNL (T −TRMNL )
KHP + PO 4d
x =c ,d , g
(B.45)
where
KRP = minimum hydrolysis rate of refractory particulate organic phosphorus (day-1)
KLP = minimum hydrolysis rate of labile particulate organic phosphorus (day-1)
KDP = minimum mineralization rate of dissolved organic phosphorus (day-1)
KRPalg and KLPalg = constants that relate the hydrolysis of refractory and labile particulate
organic phosphorus, respectively, to algal biomass (day-1 per g C m-3)
KDPalg = constant that relates mineralization to algal biomass (day-1 per g C m-3)
KHP = mean half saturation constant for algal phosphorus uptake (g P m-3)
188
The mean half saturation constant for algal phosphorus uptake, KHP, is calculated
using:
KHP =
1
∑ KHPx
3 x =c ,d , g
(B.46)
Eqs. (B.43) – (B.45) reveal that these rates are functions of water temperature and
dissolved phosphate, and their values increase exponentially with water temperature.
B2.4 Nitrogen
The forms of nitrogen modeled are grouped into 5 categories:
1.
2.
3.
4.
5.
Refractory particulate organic nitrogen (RPON)
Labile particulate organic nitrogen (LPON)
Dissolved organic nitrogen (DON)
Ammonium (NH4)
Nitrate and nitrite (NO3)
Two of the nitrogen state variables are in inorganic forms: NH4 and NO3. The other
three are in organic forms: refractory, labile, and dissolved. The nitrate state variable in
the model represents the sum of nitrate and nitrite.
Particulate organic nitrogen
Particulate organic nitrogen, including RPON and LPON, has the following sources and
sinks:
1.
2.
3.
4.
5.
Algal basal metabolism
Algal predation
Hydrolysis to DON
Settling
External loads
The kinetic equations for RPON and LPON are (Park et al., 1995):
∂RPON
=
∑ ( FNRx ·BM x + FNRP·PRx ) ANCx ·Bx − K RPON ·RPON
∂t
x =c ,d , g
+
WRPON
∂
(WS RP ·RPON ) +
∂z
V
(B.47)
189
and
∂LPON
=
∑ ( FNLx ·BM x + FNLP·PRx ) ANCx ·Bx − K LPON ·LPON
∂t
x =c ,d , g
+
WLPON
∂
(WS LP ·LPON ) +
∂z
V
(B.48)
where
RPON = concentration of refractory particulate organic nitrogen (g N m-3)
LPON = concentration of labile particulate organic nitrogen (g N m-3)
FNRx = fraction of metabolized nitrogen by algal group x as refractory particulate
organic nitrogen
FNLx = fraction of metabolized nitrogen by algal group x produced as labile particulate
organic nitrogen
FNRP = fraction of predated nitrogen produced as refractory particulate organic nitrogen
FNLP = fraction of predated nitrogen produced as labile particulate organic nitrogen
ANCx = nitrogen-to-carbon ratio in algal group x (g N per g C)
KRPON = hydrolysis rate of refractory particulate organic nitrogen (day-1)
KLPON = hydrolysis rate of labile particulate organic nitrogen (day-1)
WRPON = external loads of refractory particulate organic nitrogen (g N day-1)
WLPON = external loads of labile particulate organic nitrogen (g N day-1).
By examining the field data in the Chesapeake Bay, Cerco and Cole (1994) showed that
the variation of nitrogen-to-carbon stoichiometry was small and thus used a constant
algal nitrogen-to-carbon ratio, ANCx.
Dissolved organic nitrogen
Sources and sinks for dissolved organic nitrogen include:
1.
2.
3.
4.
5.
Algal basal metabolism
Algal predation
Hydrolysis from RPON and LPON
Mineralization to ammonium
External loads
190
The kinetic equation describing these processes is:
∂DON
=
∂t
∑
x =c ,d , g
( FNDx ·BM x + FNDP·PRx ) ANCx ·Bx
+ K RPON ·RPON + K LPON ·LPON − K DON ·DON +
WDON
V
(B.49)
where
DON = concentration of dissolved organic nitrogen (g N m-3)
FNDx = fraction of metabolized nitrogen by algal group x produced as dissolved organic
nitrogen
FNDP = fraction of predated nitrogen produced as dissolved organic nitrogen
KDON = mineralization rate of dissolved organic nitrogen (day-1)
WDON = external loads of dissolved organic nitrogen (g N day-1).
Ammonium nitrogen
Major sources and sinks for ammonia nitrogen include:
1.
2.
3.
4.
5.
Algal basal metabolism, predation, and uptake
Mineralization from dissolved organic nitrogen
Nitrification to nitrate
Exchange at the sediment bed -water column interface
External loads
The kinetic equation for NH4 described the process is:
∂NH 4
=
∂t
∑
x =c ,d , g
( FNI x ·BM x + FNIP·PRx − PN x ·Px ) ANC x ·Bx + K DON ·DON
− Nit ·NH 4 +
BFNH 4 WNH 4
+
V
∆z
(B.50)
where
FNIx = fraction of metabolized nitrogen by algal group x produced as inorganic nitrogen
FNIP = fraction of predated nitrogen produced as inorganic nitrogen
PNx = preference for ammonium uptake by algal group x (0 < P Nx < 1), give n by Eq.
(B.52)
Nit = nitrification rate (day-1) given in Eq. (B.58).
BFNH4 = sediment-water exchange flux of ammonium (g N m-2 day-1), applied to the
bottom layer only
WNH4 = external loads of ammonium (g N day-1).
191
Algae can uptake both ammonia and nitrate; however, ammonia is the preferred form of
nitrogen for algal growth and is characterized by the parameter PNx. The NH4 flux from
the sediment bed, BFNH4 is calculated by simulating the sediment diagenesis process.
Nitrate nitrogen
Major sources and sinks for nitrate nitrogen include:
1.
2.
3.
4.
5.
Algal uptake
Nitrification from ammonium
Denitrification to nitrogen gas
NO3 flux at the sediment bed -water column interface
External source
∂NO3
=
− ∑ (1 − PN x ) Px ·ANCx ·Bx + Nit ·NH 4 − ANDC ·Denit ·DOC
∂t
x =c ,d , g
+
BFNO3 WNO3
+
V
∆z
(B.51)
where
ANDC = mass of nitrate nitrogen reduced per mass of dissolved organic carbon
oxidized (0.933 g N per g C)
BFNO3 = sediment-water exchange flux of nitrate (g N m-2 day-1), applied to the bottom
layer only
WNO3 = external loads of nitrate (g N day-1).
The NO3 flux from the sediment bed, BFNO3 is calculated by simulating the sediment
diagenesis process.
Effects of algae on nitrogen
∑
The terms within summation ( x =c ,d , g ) in Eqs. (B.47) – (B.51) represent the effects of
algae on nitrogen. The nitrogen of algal biomass can be recycled to organic nitrogen
and inorganic nitrogen, and is represented by the distribution coefficients. For algal
basal metabolism, the distribution coefficients are: FNRx, FNLx, FNDx, FNIx, in which
FNRx + FNLx + FNDx + FNIx =1
and for algal predation, the distribution coefficients are: FNRP, FNLP, FNDP, FNIP,
in which
FNRP + FNLP + FNDP +FNIP = 1
Two forms of nitrogen, ammonia (NH4) and nitrate (NO3), are used during algal uptake
192
and growth, and NH4 is the preferred form of nitrogen over NO3 for algal growth. The
value of the ammonia preference factor, PNx is a function of the ammonia and nitrate
concentrations, and is expressed as:
PN x = NH4
NO3
KHN x
+ NH4
(KHN x +NH4)(KHN x +NO3)
(NH4+NO3)(KHN x +NO3)
(B.52)
Eq. (B.52) is somewhat similar to the Michaelis-Menten formulation that has been used
to describe limiting functions. The preference for ammonium is 1 when nitrate is absent
and is 0 when ammonium is absent. At PNx =1, NO3 is zero and algae uptake nitrogen
only in the form of NH4. At PNx =0, NH4 is zero and algae uptake nitrogen only in the
form of NO3.
Mineralization and hydrolysis
The third term on the RHS of equation (B.47) and (B.48) represents hydroloysis of
particulate organic nitrogen, and the 3rd term in the 2nd line of equation (B.49)
represents mineralization of dissolved organic nitrogen. The three parameters, KRPON,
KLPON, and KDON, have the following formulations:
K RPON = (K RN +
K LPON = (K LN +
K DON = (K DN +
KHN
K RNalg ∑ Bx )·e KTHDR (T - TR HDR )
KHN+NH4+NO3
x=c,d,g
(B.53)
KHN
K LNalg ∑ Bx )·e KTHDR (T - TR HDR )
KHN+NH4+NO3
x=c,d,g
(B.54)
KHN
K DNalg ∑ Bx )·e KTMNL (T - TR MNL )
KHN+NH4+NO3
x=c,d,g
(B.55)
where
KRN = minimum hydrolysis rate of refractory particulate organic nitrogen (day-1)
KLN = minimum hydrolysis rate of labile particulate organic nitrogen (day-1)
KDN = minimum mineralization rate of dissolved organic nitrogen (day-1)
KRNalg and KLNalg = constants that relate hydrolysis of refractory and labile particulate
organic nitrogen, respectively, to algal biomass (day-1 per g C m-3)
KDNalg = constant that relates mineralization to algal biomass (day-1 per g C m-3)
KHN = mean half-saturation constant for algal nitrogen uptake (g N m-3), which has the
form:
KHN =
1
∑ KHN x
3 x =c ,d , g
(B.56)
Equations (B.53) – (B. 55) have exponential functions relates rates to temperature.
193
Nitrification
Nitrification is the process in which an ammonium ion (NH4+) is oxidized to nitrite (NO2-)
and then to nitrate (NO3-). The stoichiometry equation for nitrification can be expressed
as:
NH 4+ + 2O2 → NO3− + H 2O + 2 H +
(B.57)
This is related to the first term in the second line of equation (B.50) and its
corresponding term in equation (B.51) representing the effect of nitrification on
ammonium and nitrate, respectively. The kinetics of complete nitrification processes
are formulated as a function of available ammonium, dissolved oxygen and temperature:
Nit =
DO
NH 4
Nitm · f Nit (T )
KHNit DO + DO KHNit N + NH 4
(B.58)
and
e − KNit1⋅(T −TNit )
f Nit (T ) = 
− KNit 2⋅(TNit −T )2
e
2
if T ≤ TNit
if T > TNit
(B.59)
where
KHNitDO = nitrification half saturation constant for dissolved oxygen (g O2 m-3)
KHNitN = nitrification half saturation constant for ammonium (g N m-3)
Nitm = maximum nitrification rate at TNit (day-1)
TNit = optimum temperature for nitrification (oC)
KNit1 = effect of temperature below TNit on nitrification rate (oC-2)
KNit2 = effect of temperature above TNit on nitrification rate (oC-2).
Eq. (B.58) shows that the nitrification process can be limited by low concentrations of
DO and NH4.
Denitrification
Denitrification is the process in which nitrate is reduced to nitrite and then to nitrogen
gas by bacteria. The stoichiometry relation for net denitrification reaction is described
by the following equation:
5 CH 2O + 4 NO 3 − + 4 H + → 5 CO 2 + 2 N 2 + 7 H 2O
194
In water columns, denitrification is usually not responsible for a significant nitrogen loss.
However, under the anaerobic conditions found in the sediment bed or during extremely
low oxygen conditions in the water column, denitrification can be important and may
remove a substantial fraction of the nitrogen from a waterbody by converting nitrate and
nitrite into nitrogen gas. Denitrification oxidizes dissolved organic carbon (DOC) and
converts nitrate (NO3) to nitrite (NO2) and then to nitrogen gas (N2).
B2.5. Silica
Silica is included in water quality modeling only when diatoms are considered. Silica is
represented by two state variables: particulate biogenic silica (SU) and available silica
(SA). SU represents the silica unavailable to diatom growth. The sources and sinks for
particulate biogenic silica included in the model are:
1.
2.
3.
4.
Diatom basal metabolism (BMd) and predation (PRd)
Dissolution to available silica
Settling
External loads
The corresponding kinetic equation is:
∂SU
∂
WSU
= ( FSPd ·BM d + FSPP·PRd ) ASCd ·Bd − K SUA ·SU + ( ws ·SU ) +
V
∂t
∂z
(B. 60)
where
SU = concentration of particulate biogenic silica (g Si m-3).
FSPd = fraction of metabolized silica by diatoms produced as particulate biogenic silica
FSPP = fraction of predated diatom silica produced as particulate biogenic silica
ASCd = silica-to-carbon ratio of diatoms (g Si per g C)
KSUA = dissolution rate of particulate biogenic silica (day-1)
ws = settling velocity of cohesive sediment, m/s
WSU = external loads of particulate biogenic silica (g Si day-1).
The available silica includes both the dissolved (SAd) and the particulate (SAp), where
SA = SAd + Sap. SA includes the following sources and sinks:
1.
2.
3.
4.
5.
Diatom basal metabolism (BMd), predation (PRd), and uptake (Pd)
Settling of sorbed (particulate) available silica
Dissolution from particulate biogenic silica
Sediment-water exchange of dissolved silica in the bottom layer
External loads
The kinetic equation describing these processes is:
195
∂SA
∂
= ( FSI d ·BM d + FSIP·PRd − Pd ) ASCd ·Bd + K SUA ·SU + ( ws ·SAp )
∂t
∂z
BFSAd WSA
+
+
∆z
V
(B.61)
where
SA = concentration of available silica (g Si m-3)
SAd = dissolved available silica (g Si m-3)
SAp = particulate (sorbed) available silica (g Si m-3)
FSId = fraction of metabolized silica by diatoms produced as available silica
FSIP = fraction of predated diatom silica produced as available silica
BFSAd = sediment-water exchange flux of available silica (g Si m-2 day-1), applied to
bottom layer only.
∆z = the thickness of the bottom layer in the numerical model
WSA = external loads of available silica (g Si day-1)
Effects of diatoms on silica
In equations (B.60) and (B.61), those terms expressed as a function of diatom biomass
(Bd) account for the effects of diatoms on silica. As in phosphorus and nitrogen, both
basal metabolism (respiration and excretion) and predation are considered, and thus
silica is formulated, to contribute to particulate biogenic and available silica. That is,
diatom silica released by both basal metabolism and predation are represented by
distribution coefficients (FSPd, FSId) and (FSPP, FSIP). The sum of two distribution
coefficients for basal metabolism should be unity and so is that for predation. Diatoms
require silica as well as phosphorus and nitrogen, and diatom uptake of available silica
is represented by (- Pd • ASCd • Bd) in equation (B.61).
Dissolution
The term (- KSUA • SU) in equation (B.60) and its corresponding term in equation (B.61)
represent dissolution of particulate biogenic silica to available silica. The dissolution rate
is expressed as an exponential function of temperature:
K SUA = K SU ·e KTSUA (T −TRSUA )
(B.64)
where
KSU = dissolution rate of particulate biogenic silica at TRSUA (day-1)
KTSUA = effect of temperature on dissolution of particulate biogenic silica (oC-1)
TRSUA = reference temperature for dissolution of particulate biogenic silica (oC).
B2.6.Chemical Oxygen Demand
196
In the present model, chemical oxygen demand is the concentration of reduced
substances that are oxidizable through inorganic means. The COD source is from the
sediment diagenesis process in the sediment bed. The kinetic equation is:
∂COD
DO
BFCOD WCOD
=
−
KCOD·COD +
+
∂t
KH COD + DO
∆z
V
(B.65)
where
COD = COD concentration (g O2-equivalents m-3)
KHCOD = half-saturation constant of DO required for oxidation of COD (g O2 m-3)
KCOD = oxidation rate of COD (day-1)
BFCOD = COD sediment flux (g O2-equivalents m-2 day-1), applied to the bottom layer
only
WCOD = external loads of COD (g O2-equivalents day-1).
An exponential function is used to describe the temperature effect on the oxidation rate
of COD:
KCOD = K CD ·e KTCOD (T −TRCOD )
(B.66)
where
KCD = oxidation rate of COD at TRCOD (day-1)
KTCOD = effect of temperature on oxidation of COD (°C-1)
TRCOD = reference temperature for oxidation of COD (°C)
B.2.7. Dissolved Oxygen
Water obtains oxygen directly from the atmosphere via reaeration and from plants via
photosynthesis. Vertical mixing between surface and deep waters transfers DO to lower
levels. With adequate sunlight, algae and aquatic plants consume nutrients and produce
oxygen as a result of photosynthesis. In water layers where photosynthetic rates are
very high, such as during an algal bloom, the water may become supersaturated, i.e.,
the oxygen content may exceed the DO saturation concentration. During periods of
strong stratification, photosynthesis is the only potential source of DO in the deeper
waters, and this occurs only if light penetrates to the deeper layers. External loads can
be either a DO source increasing the DO concentration in the receiving water or a DO
sink decreasing the DO concentration, depending on the inflow DO concentration.
The oxidation and decomposition of organic matter consume oxygen. The nitrification
process uptakes oxygen and oxidize ammonium (NH4+) to nitrite (NO2-) and then to
nitrate (NO3-). Algal respiration needs oxygen to convert organic carbon to carbon
dioxide and water. Chemical and biological processes in the sediment bed often uptake
oxygen from the water column. Oxygen is consumed by the sediment organism
respiration and the benthic decomposition of organic material, which can be a significant
fraction of the total oxygen demand in a waterbody. Sediment oxygen demand (SOD) is
197
used to represent the oxygen depletion due to benthic reactions. It is the rate of oxygen
consumption exerted by the bottom sediment on the overlying water. Sulfide and
methane provide additional oxygen demands. Microbial activities tend to increase with
increased temperature. The stratification may prevent DO in the surface layer from
reaching the bottom. Therefore, the benthic effects can be particularly acute in summer
under low-flow conditions or highly stratified conditions.
Processes and equations of dissolved oxygen
Based on the description above, the major sources of DO consisted of: (1) Reaeration
(2) Photosynthesis (3) External loads and major sinks of DO consist of:
1.
2.
3.
4.
5.
Oxidation of organic matter
Nitrification
Algal respiration
Sediment oxygen demand due to sediment diagenesis in the bed
Chemical oxygen demand due to reduced substances released from the
sediment bed
If the contribution of DO sources is less than the summation of DO sinks, there is an
oxygen deficit in the waterbody. The DO deficit is the difference between the saturated
DO concentration and the existing DO concentration.
198
The corresponding DO kinetic equation is:
∂DO
=
∂t


DO
BM x  AOCR·Bx
 (1.3 − 0.3·PN x ) Px − (1 − FCDx )
KHRx + DO
x =c ,d , g 

DO
KCOD·COD
− AONT ·Nit ·NH 4 − AOCR·K HR ·DOC −
KH COD + DO
∑
+ K r ( DOs − DO) +
SOD WDO
+
∆z
V
(B.67)
where
PNx = preference for ammonium uptake by algal group x (0 ≤ PNx ≤ 1), given by Eq.
(B.52)
AONT = mass of DO consumed per unit mass of ammonium nitrogen nitrified (4.33 g O2
per g N)
AOCR = dissolved oxygen-to-carbon ratio in respiration (2.67 g O2 per g C)
Kr = reaeration coefficient (day-1), applied to the surface layer only
DOs = saturation concentration of dissolved oxygen (g O2 m-3)
SOD =sediment oxygen demand (g O2 m-2 day-1), applied to the bottom layer only;
a direction of positive is towards the water column
WDO = external loads of dissolved oxygen (g O2 day-1)
The two sink terms in equation (B.67), heterotrophic respiration and chemical oxygen
demand, are explained equation (B.29) and equation (B.66), respectively. The
remainder of this section explains the effects of algae, nitrification, and surface
reaeration on dissolved oxygen.
Effects of algae on dissolved oxygen
The first line on the RHS of (B.67) accounts for the effects for the effects of algae on
dissolved oxygen. In water quality modeling, respiration and photosynthesis are
considered as the same reaction but occur in opposite directions. However,
photosynthesis only occurs during daylight hours, whereas respiration and
decomposition proceed at all times and are not dependent on solar energy. These
reactions can be represented by the following simplified stoichiometry reaction
relationship:
Photosynthesis

→ C6 H12 O6 +6O 2
6CO 2 +6H 2 0 ←

Respiration
where glucose, C6H12O6, represents organic compounds in plants. In this reaction,
photosynthesis converts carbon dioxide and water into glucose and oxygen and leads to
a net gain of DO in the waterbody. Conversely, respiration converts glucose and
oxygen into carbon dioxide and water resulting in a net loss of DO in the waterbody.
Plants generally produce more organic matter and oxygen than they use.
199
The quantity of DO produced also depends on the form of nitrogen utilized for algal
growth. Morel (1983) gave the following equations for DO production:
−
+
106CO2 + 16 NH 4 + H 2 PO4 + 106 H 2O → protoplasm + 106O2 + 15 H +
−
(B.68)
−
106CO2 + 16 NO3 + H 2 PO4 + 122 H 2O + 17 H + → protoplasm + 138O2
(B.69)
where protoplasm is the living substance of algae cells. It is a chemically active mixture
of protein, fats, and many other complex substances suspended in water.
Eq. (B.68) indicates that, when ammonium is the nitrogen source, one mole of oxygen is
produced per mole of carbon dioxide fixed. Eq. (B.69) shows that, when nitrate is the
nitrogen source, 1.3 (= 138/106) moles of oxygen are produced per mole of carbon
dioxide fixed. These two equations are reflected in the first term on the RHS of Eq.
(B.67) by the quantity of (1.3 - 0.3·PNx), which is the photosynthesis ratio and
represents the molar quantity of oxygen produced per mole of carbon dioxide fixed.
When the entire nitrogen source is from ammonium (ammonium preference factor, PNx,
is equal to 1.0), the quantity is 1.0. When the entire nitrogen source is from nitrate (PNx
= 0.0), the quantity is 1.3.
The last term in the first line of equation (B.67) accounts for oxygen consumption due to
algal respiration (equation B.26). Again, representation of respiration process is:
energy released
C6 H12O6 + O2 
→ CO2 + H 2O
(B.70)
The rate of oxygen production (and nutrient uptake) is proportional to the algal growth
rate for each gram of algae carbon produced by photosynthesis, 32/12 (or 2.67
approximately) grams of O2 are produced. Conversely, for every gram of algae carbon
consumed by respiration, 32/12 grams of oxygen are also consumed. Hence, the
dissolved oxygen-to-carbon ratio, AOCR, in Eq. (B.67) should have: AOCR = 2.67 g O2
per g C.
Effect of Nitrification on dissolved oxygen
The nitrification of ammonia has the potential for removing large amounts of oxygen
from a waterbody. The stoichiometry of reactions indicates that two moles of oxygen are
required to nitrify one mole of ammonium into nitrate: 3.43 (= 1.5×32/14) g O2 per g N
for transforming ammonia to nitrite and 1.14 (= 0.5×32/14) g O2 per g N for transforming
nitrite to nitrate. Thus, for every gram of ammonium nitrogen oxidized, 4.57 (= 2 ×
32/14) grams of oxygen are consumed. However, Wezernak and Gannon (1968)
reported that due to the effect of nitrifying bacteria, less than two moles of oxygen are
actually consumed per mole of ammonium nitrified, and a total of 4.33 grams of oxygen
200
is required to oxidize 1.0 gram of ammonia nitrogen. This explains why AONT has the
value of 4.33 (instead of 4.57) g O2 per g N in the DO equation Eq.(B.67).
Effect of Reaeration on dissolved oxygen
The rate of reaeration is proportional to the DO deficit, which is the difference between
the DO concentration and the oxygen saturation value. The DO deficit is a useful water
quality parameter and is influenced by temperature, salinity, and atmospheric pressure.
The saturated concentration of dissolved oxygen, which decrease as temperature and
salinity increase, is specified using an empirical formula by Hyer et al. (1971):
=
DOs 14.6244 − 0.367134 ⋅ T + 0.0044972 ⋅ T 2
+S ⋅ (-0.0966+0.00205 ⋅ T + 0.0002739 ⋅ S )
(B.71)
Typically, oxygen is transferred from the atmosphere into the water, since DO levels in
natural waters are generally below saturation. However, when photosynthesis produces
supersaturated DO levels (e.g., in the afternoon of a eutrophic reservoir) the net transfer
of oxygen can be from the water into the atmosphere.
Reaeration occurs by diffusion of oxygen from the atmosphere into the water (when DO
is not saturated) and by the turbulent mixing of water and air. In general, the reaeration
rate in natural waters depends on:
1. Water flow speed and wind speed
2. Water temperature and salinity
3. Water depth
201
The reaeration coefficient includes the effect of turbulence generated by bottom friction
(O’Connor and Dobbins, 1958) and that by surface wind stress (Banks and Herrera,
1977):

 1
u
T −20
K r  K ro eq + Wrea  ·KTr
=

 ∆z
heq


(B.72)
where
Kro = proportionality constant = 3.933 in MKS units
ueq = weighted velocity over cross-section (m sec-1)
heq = weighted depth over cross-section (m)
Bη= width at the free surface (m)
Wrea = wind-induced reaeration (m day-1)
KTr = constant for temperature adjustment of DO reaeration rate
∆z = the thickness of the surface layer in numerical model
The wind-induced reaeration from (B.72) can be expressed as:
Wrea = 0.728U w
1
2
− 0.317U w + 0.0372U w
2
(B.73)
with Uw = wind speed (m sec-1) at the height of 10 m above surface
Other relationships also exist for estimating the reaeration rate and DO saturated
concentration (refer to Ji, 2008).
Sediment flux
The sediment flux obtained in HEM3D model is primarily based on the Chesapeake Bay
Sediment Flux Model developed by Di Toro and Fitzpatrick (1993), which is now
commonly accepted and used in water quality modeling (e.g., Cerco and Cole, 1994;
Park et al., 1995; HydroQual, 1995). Many discussions and equations in this chapter
originate from the report by Park et al. (1995). Complete model documentation can be
found in Di Toro and Fitzpatrick (1993) and Di Toro (2001).
202
APPENDIX C: Summary of Parameters Used for ICM Water Quality Model
Table C-1: Parameters related to algae in the water column.
Parameter
PMc
PMd
PMg
KHNx
KHPx
KHS
KHRx
IHc
IHd
IHg
KEB
KECHL
KETSS
CCHLx
TMc
TMd
TMg
KTG1c
KTG2c
KTG1d
KTG2d
Description
maximum growth rate
of algae group 1
maximum growth rate
of algae group 2
maximum growth rate
of algae group 3
half-saturation constant
of N uptake by algae
half-saturation constant
of P uptake by algae
half-saturation constant
of Si uptake by diatoms
half-saturation constant of DO
for algal excretion of DOC
half-saturation light intensity
for algal group 1 growth
half-saturation light intensity
for algal group 2 growth
half-saturation light intensity
for algal group 3 growth
background light attenuation
coefficient
light attenuation coefficient
due to self-shading of algae
light attenuation coefficient
due to TSS
C-to-CHL ratio in algae
optimum T for algal group 1
growth
optimum T for algal group 2
growth
optimum T for algal group 3
growth
effect of T below optimum T
on algal group 1 growth
effect of T above optimum T
on algal group 1 growth
effect of T below optimum T
on algal group 2 growth
effect of T above optimum T
on algal group 2 growth
Value
Unit
2.5
day-1
2.5
day-1
2.5
day-1
0.01
g N m-3
0.001
g P m-3
0.05
g Si m-3
0.5
g O2 m-3
50
langley day-1
30
langley day-1
40
langley day-1
0.12 - 0.15
m-1
0.017
m2 per mg CHL
0.07
m2 per g TSS
60.0
g C per g CHL
25.0
°C
20.0
°C
22.5
°C
0.006
°C-2
0.006
°C-2
0.004
°C-2
0.006
°C-2
203
Table C-1: Parameters related to algae in the water column.
KTG1g
KTG2g
BMRc
BMRd
BMRg
PRRc
PRRd
PRRg
KTBx
TRx
WSc
WSd
WSg
effect of T below optimum T
on algal group 3 growth
effect of T above optimum T
on algal group3 growth
basal metabolism rate
of algae group 1 at reference T
basal metabolism rate
of algae group 2 at reference T
basal metabolism rate
of algae group 3 at reference T
predation rate of algae group 1
at reference T
predation rate of algae group 2
at reference T
predation rate of algae group 3
at reference T
effect of T on basal metabolism
of algae
reference T for basal metabolism
of algae
settling velocity for algal group 1
settling velocity for algal group 2
settling velocity for algal group 3
0.012
°C-2
0.007
°C-2
0.05
day-1
0.05
day-1
0.05
day-1
0.05
day-1
0.05
day-1
0.20
day-1
0.069
v
20.0
°C
0.01
0.25
0.1
m day-1
m day-1
m day-1
204
Table C-2: Parameters related to organic carbon in the water column.
_________________________________________________________________________
Parameters description
value
unit
______________________________________________________________________
___
FCRP
fraction of predated algal C
produced as RPOC
0.35
none
FCLP
fraction of predated algal C
produced as LPOC
0.55
none
FCDP
fraction of predated algal C
produced as DOC
0.10
none
FCDx
fraction of metabolized C by algae
produced as DOC
0.0
none
half-saturation constant of DO for
KHRx
algal excretion of DOC
0.5
g O2 m-3
half-saturation constant of DO for
KHODOC
oxic respiration of DOC
0.5
g O2 m-3
KRC
minimum respiration rate of RPOC
0.005
day-1
KLC
minimum respiration rate of LPOC
0.075
day-1
KDC
minimum respiration rate of DOC
0.020
day-1
constant relating respiration
KRcalg
of RPOC to algal biomass
0.0
day-1 per g C m-3
KLcalg
constant relating respiration
of LPOC to algal biomass
0.0
day-1 per g C m-3
KDcalg
constant relating respiration
of DOC to algal biomass
0.0
day-1 per g C m-3
KTHDR
effect of T on hydrolysis/
mineralization of POM/DOM
0.069
°C-1
KTMNL
effect of T on hydrolysis/
mineralization of POM/DOM
0.069
°C-1
reference T for hydrolysis of POM
20.0
°C
TRHDR
TRMNL
reference T for mineralization of DOM 20.0
°C
KHNDNN
half-saturation constant of NO23 for
Denitrification
0.1
g N m-3
AANOX
ratio of denitrification to oxic DOC
respiration rate
0.5
none
_________________________________________________________________________
205
Table C-3: Parameters related to nitrogen in the water column.
_______________________________________________________________
Parameters
description
Value
unit
_______________________________________________________________
FNRP
fraction of predated algal N produced as
RPON
0.35
none
FNLP
fraction of predated algal N produced as
LPON
0.55
none
FNDP
fraction of predated algal N produced as
DON
0.10
none
FNIP
FNR
FNL
FND
FNI
fraction of predated algal N produced as
NH4
fraction of metabolized algal N produced
as RPON
fraction of metabolized algal N produced
as LPON
fraction of metabolized algal N produced
as DON
fraction of metabolized algal N produced
as NH4
ANCx
ANDC
KRN
KLN
KDN
KRnalg
KLnalg
KDnalg
KHDONIT
KHNNIT
NTM
N-to-C ratio in algae
mass of NO23-N consumed per mass
DOC oxidized
minimum hydrolysis/mineralization rate
of RPON
minimum hydrolysis/mineralization rate
of LPON
minimum hydrolysis/mineralization rate
of DON
constant relating hydrolysis/mineralization
of RPON to algal biomass
constant relating hydrolysis/mineralization
of LPON to algal biomass
constant relating hydrolysis/mineralization
of DON to algal biomass
half-saturation constant of DO for
nitrification
half-saturation constant of NH4 for
nitrification
maximum nitrification at optimum T
0.00
none
0.0
none
0.0
none
1.0
none
0.0
none
0.167
g N per g C
0.933
g N per g C
0.005
day-1
0.075
day-1
0.015
day-1
0.0
day-1 per g N m-3
0.0
day-1 per g N m-3
0.0
day-1 per g N m-3
1.0
g O2 m-3
1.0
0.007
g N m-3
day-1
206
Table C-3 (con’t)
_______________________________________________________________
Parameters
description
Value
unit
_______________________________________________________________
KTNT1
effect of T below optimum T on
nitrification rate
0.0045 °C-2
KTNT1
effect of T above optimum T on
nitrification rate
0.0045 °C-2
TMNT
optimum T for nitrification rate
27.0
°C
______________________________________________________________
207
Table C-4: Parameters related to phosphorus in the water column.
_________________________________________________________________
Parameter
description
Value
unit
_________________________________________________________________
FPRP
fraction of predated algal P produced
as RPOP
0.1
none
FPLP
fraction of predated algal P produced
as LPOP
0.2
none
FPDP
fraction of predated algal P produced
as DOP
0.5
none
fraction of metabolized P by algae
FPRx
produced as RPOP
0.0
none
FPLx
fraction of metabolized P by algae
produced as LPOP
0.0
none
fraction of metabolized P by algae
FPDx
produced DOP
0.5
none
APCMIN
minimum P-to-C ratio in algae
0.01
g P per g C
APCMAX
maximum P-to-C ratio in algae
0.024
g P per g C
PO4DMAX maximum PO4d beyond which
APC = APCMAX
0.01
g P m-3
minimum hydrolysis/mineralization
KRP
rate of RPOP
0.005 day-1
KLP
minimum hydrolysis/mineralization
rate of LPOP
0.075
day-1
KDP
minimum hydrolysis/mineralization
rate of DOP
0.1
day-1
KRpalg
constant relating hydrolysis/
mineralization of RPOP to algal biomass 0.0
day-1 per g P m-3
KLpalg
constant relating hydrolysis/
mineralization of LPOP to algal biomass 0.0
day-1 per g P m-3
constant relating hydrolysis/
KDpalg
mineralization of DOP to algal biomass 0.0
day-1 per g P m-3
____________________________________________________________________
208
Table C-5: Parameters related to silica in the water column.
_________________________________________________________________
Parameter
description
Value
unit
_________________________________________________________________
FSA
fraction of predated diatom Si
produced as SA
0.0
none
ASCd
Si-to-C ratio in diatoms
0.5
g Si per g C
KSU
dissolution rate of SU at reference T
0.025
day-1
KTSUA
effect of T on dissolution of SU
0.092
°C-1
TRSUA
reference T for dissolution of SU
20.0
°C
_________________________________________________________________
Table C-6. Parameters related to chemical oxygen demand and dissolved oxygen in the
water column.
_________________________________________________________________________
Parameters
description
Value
unit
_________________________________________________________________________
KHOCOD
half-saturation constant of DO for
oxidation of COD
1.5
g O2 m-3
KCD
oxidation rate of COD at reference
temperature
20.0
day-1
KTCOD
effect of T on oxidation of COD
0.041
°C-1
TRCOD
reference T for oxidation of COD
20.0
°C
reaeration coefficient
2.4
m day-1
KRDO
AOCR
mass DO consumed per mass C
respired by algae
2.67
g O2 per g C
ANOT
mass DO consumed per mass
NH4- N nitrified
4.33
g O2 per g N
_________________________________________________________________________
209
Table C-7: Parameters used in the sediment flux model.
_______________________________________________________________
Parameter
description
value
unit
_______________________________________________________________
HSEDALL
DIFFT
SALTSW
SALTND
FRPPH1(1)
FRPPH1(2)
FRPPH1(3)
FRPPH2(1)
FRPPH2(2)
FRPPH2(3)
FRPPH3(1)
FRPPH3(2)
FRPPH3(3)
FRNPH1(1)
FRNPH1(2)
FRNPH1(3)
FRNPH2(1)
FRNPH2(2)
depth of sediment
heat diffusion coefficient between water
column and sediment
salinity for dividing fresh and saltwater
for SOD kinetics (sulfide in saltwater or
methane in freshwater) and for PO4
sorption coefficients
salinity for dividing fresh or saltwater
for nitrification/denitrification rates
(larger values for freshwater)
fraction of POP in algal group No 1
routed into G1 class
fraction of POP in algal group No 1
routed into G2 class
fraction of POP in algal group No 1
routed into G3 class
fraction of POP in algal group No 2
routed into G1 class
fraction of POP in algal group No 2
routed into G2 class
fraction of POP in algal group No 2
routed into G3 class
fraction of POP in algal group No 3
routed into G1 class
fraction of POP in algal group No 3
routed into G2 class
fraction of POP in algal group No 3
routed into G3 class
fraction of PON in algal group No 1
routed into G1 class
fraction of PON in algal group No 1
routed into G2 class
fraction of PON in algal group No 1
routed into G3 class
fraction of PON in algal group No 2
routed into G1 class
fraction of PON in algal group No 2
routed into G2 class
10
cm
0.0018
cm2 sec-1
1.0
ppt
1.0
ppt
0.65
none
0.255
none
0.095
none
0.65
none
0.255
none
0.095
none
0.65
none
0.255
none
0.095
none
0.65
none
0.28
none
0.07
none
0.65
none
0.28
none
210
Table C-7 (con’t)
_______________________________________________________________
Parameter
description
value
unit
_______________________________________________________________
FRNPH2(3)
fraction of PON in algal group No 2
0.07
none
routed into G3 class
FRNPH3(1)
fraction of PON in algal group No 3
routed into G1 class
0.65
none
FRNPH3(2)
fraction of PON in algal group No 3
routed into G2 class
0.28
none
FRNPH3(3)
fraction of PON in algal group No 3
routed into G3 class
0.07
none
FRCPH1(1)
fraction of POC in algal group No 1
0.65
none
routed into G1 class
FRCPH1(2)
fraction of POC in algal group No 1
0.255
none
routed into G2 class
FRCPH1(3)
fraction of POC in algal group No 1
routed into G3 class
0.095
none
FRCPH2(1)
fraction of POC in algal group No 2
routed into G1 class
0.65
none
FRCPH2(2)
fraction of POC in algal group No 2
0.255
none
routed into G2 class
FRCPH2(3)
fraction of POC in algal group No 2
routed into G3 class
0.095
none
FRCPH3(1)
fraction of POC in algal group No 3
0.65
none
routed into G1 class
FRCPH3(2)
fraction of POC in algal group No 3
routed into G2 class
0.255
none
FRCPH3(3)
fraction of POC in algal group No 3
routed into G3 class
0.095
none
KPDIAG(1)
reaction (decay) rates for G1 class
POP at 20°C
0.035
day-1
KPDIAG(2)
reaction (decay) rates for G2 class
POP at 20°C
0.0018 day-1
KPDIAG(3)
reaction (decay) rates for G3 class
POP at 20°C
0.0
day-1
DPTHTA(1)
constant for T adjustment for G1
class POP decay
1.10
none
DPTHTA(2)
constant for T adjustment for G2
class POP decay
1.15
none
KNDIAG(1)
reaction (decay) rates for G1 class
PON at 20°C
0.035
day-1
KNDIAG(2)
reaction (decay) rates for G2 class
PON at 20°C
0.0018 day-1
211
Table C-7 (con’t)
_______________________________________________________________
Parameter
description
value
unit
_______________________________________________________________
KNDIAG(3)
reaction (decay) rates for G3 class
PON at 20°C
0.0
day-1
DNTHTA(1)
constant for T adjustment for G1
class PON decay
1.10
none
DNTHTA(2)
constant for T adjustment for G2
class PON decay
1.15
none
KCDIAG(1)
reaction (decay) rates for G1 class
POC at 20°C
0.035
(day-1)
KCDIAG(2)
reaction (decay) rates for G2 class
POC at 20°C
0.0018 (day-1)
KCDIAG(3)
reaction (decay) rates for G3 class
POC at 20°C
0.0
(day-1)
DCTHTA(1)
constant for T adjustment for G1
class POC decay
1.10
none
DCTHTA(2)
constant for T adjustment for G2
class POC decay
1.15
none
st
KSI
1 -order reaction (dissolution) rate
of PSi at 20°C
0.5
day-1
THTASI
constant for T adjustment for PSi
dissolution
1.1
none
M1
solid concentrations in Layer 1
0.5
kg l-1
M2
solid concentrations in Layer 2
0.5
kg l-1
THTADP
constant for T adjustment for
diffusion coefficient for particle
mixing
1.117
none
THTADD
constant for T adjustment for
diffusion coefficient for dissolved
phase
1.08
none
KAPPNH4F
optimum reaction velocity for
nitrification in Layer 1 for
freshwater
0.20
m day-1
KAPPNH4S
optimum reaction velocity for
nitrification in Layer 1 for saltwater
0.14
m day-1
THTANH4
constant for T adjustment for
nitrification
1.08
none
KMNH4
half-saturation constant of NH4
for nitrification
1500.0 mg N m-3
KMNH4O2
half-saturation constant of DO
for nitrification
1.0
g O2 m-3
212
Table C-7 (con’t)
_______________________________________________________________
Parameter
description
value
unit
_______________________________________________________________
PIENH4
partition coefficient for NH4 in
both layers
1.0
per kg l-1
KAPPNO3F
reaction velocity for denitrification
in Layer 1 at 20°C for freshwater
0.3
m day-1
KAPPNO3S
reaction velocity for denitrification
in Layer 1 at 20°C for saltwater
0.125
m day-1
K2NO3
reaction velocity for denitrification
in Layer 2 at 20°C
0.25
m day-1
THTANO3
constant for T adjustment for
denitrification
1.08
none
KAPPD1
reaction velocity for dissolved
0.2
m day-1
H2S oxidation in Layer 1 at 20°C
KAPPP1
reaction velocity for particulate
H2S oxidation in Layer 1 at 20°C
0.4
m day-1
PIE1S
partition coefficient for H2S in Layer 1
100.0
per kg l-1
PIE2S
partition coefficient for H2S in Layer 2
100.0
per kg l-1
THTAPD1
constant for T adjustment for both
dissolved & particulate H2S oxidation 1.08
none
KMHSO2
constant to normalize H2S oxidation
rate for oxygen
4.0
g O2 m-3
CSISAT
saturation concentration of Si in the
pore water
40000.0 mg Si m-3
DPIE1SI
incremental partition coefficient for
Si in Layer 1
10.0
per kg l-1
PIE2SI 2
partition coefficient for Si in Layer 2
100.0
per kg l-1
O2CRITSI
critical DO concentration for Layer 1
incremental Si sorption
1.0
g O2 m-3
KMPSI
half-saturation constant of PSi for Si
dissolution
5 × 107 mg Si m-3
JSIDETR
detrital flux of PSi to account for PSi
settling to the sediment that is not
associated with algal flux of PSi
100.0
mg Si m-2 day-1
DPIE1PO4F
incremental partition coefficient
for PO4 in Layer 1 for freshwater
3000.0 per kg l-1
DPIE1PO4S
incremental partition coefficient for
PO4 in Layer 1 for saltwater
300.0
per kg l-1
PIE2PO4
partition coefficient for PO4 in Layer 2
100 per kg l-1
O2CRIT
critical DO concentration for Layer 1
incremental PO4 sorption
2.0
g O2 m-3
213
Table C-7 (con’t)
_______________________________________________________________
Parameter
description
value
unit
_______________________________________________________________
KMO2DP
half-saturation constant of DO for
particle mixing
4.0
g O2 m-3
TEMPBEN
temperature at which benthic stress
accumulation is reset to zero
10.0
°C
KBENSTR
1st-order decay rate for benthic stress 0.03
day-1
KLBNTH
ratio of bio-irrigation to bioturbation
DPMIN
minimum diffusion coefficient for
particle mixing
reaction velocity for dissolved CH4
oxidation in Layer 1 at 20°C
KAPPCH4
THTACH4
constant for T adjustment for dissolved
0.0
none
3×10-6
m2 day-1
0.2
m day-1
CH4 oxidation
1.08
none
VSED
net burial (sedimentation) rate
0.25
cm yr-1
VPMIX
diffusion coefficient for particle mixing 1.2×10-4 m2 day-1
VDMIX
diffusion coefficient in pore water
0.001
m2 day-1
WSCNET
net settling velocity for algal group 1
0.1
m day-1
WSDNET
net settling velocity for algal group 2
0.3
m day-1
WSGNET
net settling velocity for algal group 3
0.1
m day-1
_____________________________________________________________
214
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement