Lower Cape Fear River Estuary Model Progress Report Jim Bowen, UNC Charlotte October 11, 2007...

Lower Cape Fear River EstuaryModel Progress Report

Jim Bowen, UNC Charlotte

October 11, 2007

Charlotte, NC (via Centra)

Description of Model Application

Open BoundaryElevation Cond.

Lower Cape Fear RiverEstuary Schematic

Black River, FlowBoundary Cond.

Cape Fear R. Flow Boundary Cond.

NE Cape FearFlow Boundary Cond.

DO Conceptual ModelBOD Sources

Sediment Sediment O2 Demand

Cape Fear BOD Load

NECF & Black R. BOD Load

Muni & Ind. BOD Load

decaying phytopl.

DO Conceptual ModelBOD Sources, DO Sources

Sediment Sediment O2 Demand

Ocean Inflows

SurfaceReaeration

Phytoplank. Productivity

MCFR Inflows

BOD Consumption

DO Conceptual ModelBOD Sources, DO Sources & Sinks

Sediment Sediment O2 Demand

Ocean Inflows

SurfaceReaeration

Input of NECF & Black R. Low DO Water

BOD Consumption

DO Conceptual ModelBOD Sources, DO Sources & Sinks

Sediment Sediment O2 Demand

Cape Fear BOD Load

NECF & Black R. BOD Load

Ocean Inflows

SurfaceReaeration

Input of NECF & Black R. Low DO Water

Phytoplank. ProductivityMuni & Ind.

BOD Load

decaying phytopl.

MCFR Inflows

Modeling Developments

1. Finished Defining Model Grid• Bottom roughness investigation

• Finished sizing marsh cells

2. Further Developed Salinity Boundary Condition at Estuary Mouth

3. Finished Hydrodynamic Model Calibration

4. WQ Predictions Using New Grid

Hydrodynamic Calibration - Summary, 8/07

• Excellent agreement w/ temperature and salinity

• Elevation agreement (not shown) still needs some work to get predicted tidal amplitude attenuation to match observed attenuation

Model Grid Definition, Procedure

• Objective was to match tidal amplitudes at USGS Stations– Upper Estuary ( Lock and Dam 1, Black at

Currie, NECF at Burgaw)– Middle Estuary (Navassa, NECF at

Wilmington)– Lower Estuary (Marker 12)

Sizing Marsh Cells, Procedure

• Systematically varied grid parameters to match observed elevation data– Rougher bottoms damp tidal amplitudes– More off-channel storage in wetland cells

damps tidal amplitudes

Step 1. Can changes in channel roughness produce desired amplitude attenuation?• Used existing model grid

Existing Model Grid• w/ changes described

in August update• 1004 water cells• Has “marsh cells” in

Black and NECF• Marsh cells 2.0 m

deep• All cells have the

same roughness• See kmz file for

more detail

Step 1. Can changes in channel roughness produce desired amplitude attenuation?• Used existing model grid

• Varied roughness across grid – Typical value = 0.02– Minimum = 0.015 m (very smooth)– Maximum = 0.045 m (very rough)

• Looked at changes in amplitude as bottom roughness increased

Results, Variable Bottom Roughness

M2 Tidal Amplitudes, LCFR

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1 2 3 4 5 6 7

Amplitude (m)

n=0.015

n=0.025

n=0.035

n=0.045

observed

Upper Estuary Middle Estuary Lower EstuaryL&D#1, Currie, Burgaw Navassa, Wilmington M12, M18

Results, Variable Bottom Roughness

M2 Tidal Amplitudes, LCFR

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1 2 3 4 5 6 7

Amplitude (m)

n=0.015

n=0.025

n=0.035

n=0.045

observed

Upper Estuary Middle Estuary Lower EstuaryL&D#1, Currie, Burgaw Navassa, Wilmington M12, M18

Upper Estuary

Stations Underdamped

Results, Variable Bottom Roughness

M2 Tidal Amplitudes, LCFR

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1 2 3 4 5 6 7

Amplitude (m)

n=0.015

n=0.025

n=0.035

n=0.045

observed

Upper Estuary Middle Estuary Lower EstuaryL&D#1, Currie, Burgaw Navassa, Wilmington M12, M18

Middle Estuary

Stations Underdamped

Results, Variable Bottom Roughness

M2 Tidal Amplitudes, LCFR

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1 2 3 4 5 6 7

Amplitude (m)

n=0.015

n=0.025

n=0.035

n=0.045

observed

Upper Estuary Middle Estuary Lower EstuaryL&D#1, Currie, Burgaw Navassa, Wilmington M12, M18

Very little sensitivity to bottom roughness

Sizing Off-Channel Storage, Procedure

1. Went through model grid and resized “marsh cells” to roughly fit wetland delineations

2. Developed a method to quickly vary width and roughness of marsh cells, create EFDC grid files, and see results w/ Google Earth

3. Ran model many times w/ various marsh configurations and observed results

First step, try various marsh cell widths• Varied marsh cell widths

– Base case– Base case * 2– Base case * 5– Base case * 10

• Determined how width changes affected tidal amplitudes

Base Case

Base Case, Width x 2

Base Case, Width x 5

Base Case, Width x 10

Results, Marsh Width VariationM2 Amplitudes at Seven Locations

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1 2 3 4 5 6 7

M2 Amplitude (m)

base

width *2

width * 5

width * 10

observed

Lock andDam 1

Black at Currie

NECF atBurgaw

CF at Navassa

NortheastCF @ Wilm.

Marker12

CF at mouth

Results, Marsh Width VariationM2 Amplitudes at Seven Locations

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1 2 3 4 5 6 7

M2 Amplitude (m)

base

width *2

width * 5

width * 10

observed

Lock andDam 1

Black at Currie

NECF atBurgaw

CF at Navassa

NortheastCF @ Wilm.

Marker12

CF at mouth

• Width ratio = 2.0 gives best results overall

• Need additional damping at Navassa

• Added additional marsh cells in middle estuary (V1, V2)

• Also tried smaller changes in marsh width (1.5, 2.0)

Version 1, Width x 2

Version 2, Width x 1.0/2.0

Version 3, Width x 1.5

Results, Tidal AmplitudesM2 Amplitudes at Seven Locations

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1 2 3 4 5 6 7

M2 Amplitude (m)

width *2

width *2, v1

v2, width = 1.0/2.0

v3, width * 1.5

observed

Lock andDam 1

Black at Currie

NECF atBurgaw

CF at Navassa

NortheastCF @ Wilm.

Marker12

CF at mouth

Results, Tidal AmplitudesM2 Amplitudes at Seven Locations

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1 2 3 4 5 6 7

M2 Amplitude (m)

width *2

width *2, v1

v2, width = 1.0/2.0

v3, width * 1.5

observed

Lock andDam 1

Black at Currie

NECF atBurgaw

CF at Navassa

NortheastCF @ Wilm.

Marker12

CF at mouth

• Unable to match exactly the observed pattern in amplitude reduction

• V2, Width ratio = 2.0 (in green) determined to give the best results overall

Version 1, Width x 2

Previous Model Grid

New Grid Characteristics

• Off-channel storage locations based on wetland delineations

• 46 additional marsh cells added (1050 total cells)

• Additional off-channel storage added to each basin (Cape Fear, Black, NECF)

• Significant amount of marsh area added to middle and lower estuary

Results for New Grid

• Also investigated alternate boundary condition specification– Now use AM and PM max salinity at station

M12 rather than daily max

• Now use hourly rather than 12-hour averaged monitoring data

• Looked at observed vs. predicted temperatures, salinities, elevations

• Compared results to those obtained previously w/ previous model grid

Elevations, Currie, June 04

Elevations, Burgaw, June 04

Elevations, Navassa, June 04

Elevations, NECF Wilm., June 04

Elevations, Mrkr 12, June 04

April - November 2004 Temp., 8/07

April - November 2004 Temp., New

April - November 2004 Temp., 8/07

Statistical Measures of Fit (units of deg C)

mean(pred-obs) =0.10046 ME_norm =0.0043473 RMSE =0.96224 MAE =0.71269 MAE_norm =0.030841 RMSE_norm =0.041639 r_squared =0.97272 num data comparisons = 4150 r2 adjusted for bias = 0.96465

April - November 2004 Temp., New

Statistical Measures of Fit (units of deg C)

mean(pred-obs) =0.034678 ME_norm =0.0015146 RMSE =0.95803 MAE =0.71946 MAE_norm =0.031423 RMSE_norm =0.041843 r_squared =0.97439 num data comparisons = 4579 1-mse/var(obs) = 0.96604

April - November 2004 Salinity, 8/07

April - November 2004 Salinity, New

April - November 2004 Salinity, 8/07

April - November 2004 Salinity, New

April - November 2004 Salinity, 8/07

April - November 2004 Salinity, New

April - November 2004 Salinity, 8/07

April - November 2004 Salinity, New

April - November 2004 Salinity, 8/07

Statistical Measures of Fit (units of PSU)

mean(pred-obs) =-0.25797 ME_norm =-0.043321 RMSE =2.6493 MAE =1.6424 MAE_norm =0.27581 RMSE_norm =0.4449 r_squared =0.87049 num data comparisons = 3517 1-mse/var(obs) = 0.84804

April - November 2004 Salinity, 8/07

Statistical Measures of Fit (units of PSU)

mean(pred-obs) =-0.0044639 ME_norm =-0.00083995 RMSE =2.64 MAE =1.5505 MAE_norm =0.29176 RMSE_norm =0.49675 r_squared =0.87044 num data comparisons = 3953 1-mse/var(obs) = 0.85345

Summary of Progress

• Model grid now includes a significant amount of off-channel storage

• Salinity mean errors now very low (important for predicting dilution)

• Tidal elevation attenuation now well simulated

• Hydrodynamic & conservative transport submodels now calibrated

Summary of Progress, cont’d

• Also have developed a program for animating horizontal contour in Google Earth (good for showing DO results)

• Benoit Duclaud finished Masters thesis on new method for predicting reaeration (thesis, paper available next month)

Information Available Online

• See LCFR website for more infowww.coe.uncc.edu/~jdbowen/LCFR

• This presentation is available

• Google Earth files available for download– Grid and wetland data from presentation– Monitoring stations, point sources– Final EFDC grid information– NOAA bathymetry

Present Work

• Running water quality model now w/ new grid• Still waiting to get BOD data from LCFR Program• Finish assigning decay rates and redefining loads

once additional BOD data are available• Work on incorporating SOD data in a more

detailed way• Do additional model/data comparisons w/ DWQ

special study data