Hydrodynamic Processes
Modeling the Delaware Hydrodynamic Processes Influencing Transport in the Upper Delaware Estuary January 2017
Ramona McCullough, PhD: Sci‐Tek Consultants Paula Kulis, PhD, PE: CDM Smith Philip Duzinski, PE: Philadelphia Water Department
Background
Introduction Develop a 3D hydrodynamic and water quality model of the upper Delaware Estuary for PWD’s Green City, Clean Waters program Main purpose: modeling dissolved oxygen and bacteria Correct transport dependent on accurate representation of • Area of interest (grid resolution, bathymetry, etc.) • Realistic boundary conditions (tributaries, run off, industrial and municipal inflows) • Hydrodynamics water level and velocity tidal asymmetry and overtides non‐tidal processes (local and remote wind effects, river flows) PHILADELPHIA WATER DEPARTMENT | BACKGROUND
Study Area RM 61.8 – RM 134.4 City of Philadelphia at RM 91‐111 Key processes • Reflected progressive tidal wave • Increasing tidal head with maximum at Trenton, NJ • Vertically well mixed • Salinity intrusion (has little hydrodynamic influence in our model area)
PHILADELPHIA WATER DEPARTMENT | BACKGROUND
Model Selection Processes that need to be well Environmental Fluid Dynamics Code represented in numerical model: (EFDC) • Includes hydrodynamic and • Energy dissipation water quality component • Propagation of tidal wave • Part of USEPA Total along estuary Maximum Daily Load • Conversion to turbulent (TMDL) toolbox kinetic energy available for • Modified version of the EFDC USEPA version 1.01 mixing • Curvilinear, orthogonal grid • Net non‐tidal transport with sigma vertical layers
PHILADELPHIA WATER DEPARTMENT | BACKGROUND
Model Setup
Model Domain fine
coarse
Delaware River from RM 61 – 134 (117 km) including tidal portion of: • Schuylkill River • Darby/Cobbs Creek • Frankford Creek • Pennypack Creek all receive CSO discharges 2 grid configurations (5 vertical layers) • Fine: for detailed representation of model domain and validation 9746 elements 17 m – 650 m
• Coarse: higher model efficiency for Water Quality study 2860 elements 33 m‐1140 m 4:1 grid cell ratio with fine grid PHILADELPHIA WATER DEPARTMENT | MODEL SETUP
Boundary Conditions – Hydrological, Oceanographic, Meteorological • Tributaries: USGS discharge • Southern open boundary: water level from NOAA’s Delaware City tidal gage • Wind: (NCDC/NOAA Ports/PWD) stations at • • • • •
PHILADELPHIA WATER DEPARTMENT | MODEL SETUP
Trenton‐Mercer Philadelphia Airport Wilmington Burlington Buoy C
Boundary Conditions ‐ Anthropogenic • Combined sewer outfalls from Philadelphia, Camden, Chester and Wilmington • Industrial and municipal Permitted Dischargers • Direct storm water runoff
PHILADELPHIA WATER DEPARTMENT | MODEL SETUP
Calibration ‐ Bottom Roughness • Spatially variable • Roughness increases going upstream • Based on 2003 sediment inventory study (Sommerfield &Madsen)
PHILADELPHIA WATER DEPARTMENT | MODEL SETUP
Fine Grid Results
Validation Data Tide Gages and Current Meters • NOAA • PWD (Woods Hole Group)
PHILADELPHIA WATER DEPARTMENT | FINE GRID RESULTS
Statistical and Harmonic Analysis Water Level
Station Marcus Hook Philadelphia Burlington Newbold
RMSE [m] 0.038 0.050 0.081 0.102
Skill Factor [‐] 0.999 0.999 0.997 0.996
Amplitude Error
Phase Error Phase Error
PHILADELPHIA WATER DEPARTMENT | FINE GRID RESULTS
Velocity
Station Buoy C Buoy B db0301 Buoy A Amplitude Error
Phase Error
RMSE (m/s) 0.073 0.059 0.093 0.094
Skill Factor 0.997 0.993 0.993 0.988
Tidal Analysis Water Level
Progressive Wave
Velocity
Tidal Asymmetry – Overtide ratios M4/M2 and M6/M2
PHILADELPHIA WATER DEPARTMENT | FINE GRID RESULTS
Subtidal Analysis no Local Wind
with Local Wind
PHILADELPHIA WATER DEPARTMENT | FINE GRID RESULTS
Coarse Grid Results
Statistical and Harmonic Analysis Water Level Station Marcus Hook Philadelphia Burlington Newbold
RMSE (m) 0.060 0.043 0.090 0.105
Velocity Skill Factor 0.998 0.999 0.997 0.996
Station Buoy C Buoy B db0301 Buoy A
Amplitude Error
Phase Error
PHILADELPHIA WATER DEPARTMENT | COARSE GRID RESULTS
RMSE (m/s) 0.038 0.050 0.081 0.102 Amplitude Error
Phase Error
Skill Factor 0.999 0.999 0.997 0.996
Tidal Analysis Water Level
Progressive Wave
Velocity
Tidal Asymmetry – Overtide ratios M4/M2 and M6/M2
PHILADELPHIA WATER DEPARTMENT | COARSE GRID RESULTS
Discussion • • •
•
Model can represent complex processes and resulting transport dynamics Harmonics, Statistical Analysis • RMSE, amplitude error within acceptable range • Harmonic phase errors are small compared to periods Validated shallow water overtides – good representation of: • Progressive waves • Tidal asymmetry • Subtidal signal • Tidal energy transfer captured well important for transport modeling Key findings: • Local winds have non‐negligible impact on transport • Coarse grid resolution does not negatively impact validation criteria
PHILADELPHIA WATER DEPARTMENT | DISCUSSION
Thank You! Questions?
Ramona McCullough, PhD: Sci‐Tek Consultants Paula Kulis, PhD, PE: CDM Smith Philip Duzinski, PE: Philadelphia Water Department
Background
Introduction Develop a 3D hydrodynamic and water quality model of the upper Delaware Estuary for PWD’s Green City, Clean Waters program Main purpose: modeling dissolved oxygen and bacteria Correct transport dependent on accurate representation of • Area of interest (grid resolution, bathymetry, etc.) • Realistic boundary conditions (tributaries, run off, industrial and municipal inflows) • Hydrodynamics water level and velocity tidal asymmetry and overtides non‐tidal processes (local and remote wind effects, river flows) PHILADELPHIA WATER DEPARTMENT | BACKGROUND
Study Area RM 61.8 – RM 134.4 City of Philadelphia at RM 91‐111 Key processes • Reflected progressive tidal wave • Increasing tidal head with maximum at Trenton, NJ • Vertically well mixed • Salinity intrusion (has little hydrodynamic influence in our model area)
PHILADELPHIA WATER DEPARTMENT | BACKGROUND
Model Selection Processes that need to be well Environmental Fluid Dynamics Code represented in numerical model: (EFDC) • Includes hydrodynamic and • Energy dissipation water quality component • Propagation of tidal wave • Part of USEPA Total along estuary Maximum Daily Load • Conversion to turbulent (TMDL) toolbox kinetic energy available for • Modified version of the EFDC USEPA version 1.01 mixing • Curvilinear, orthogonal grid • Net non‐tidal transport with sigma vertical layers
PHILADELPHIA WATER DEPARTMENT | BACKGROUND
Model Setup
Model Domain fine
coarse
Delaware River from RM 61 – 134 (117 km) including tidal portion of: • Schuylkill River • Darby/Cobbs Creek • Frankford Creek • Pennypack Creek all receive CSO discharges 2 grid configurations (5 vertical layers) • Fine: for detailed representation of model domain and validation 9746 elements 17 m – 650 m
• Coarse: higher model efficiency for Water Quality study 2860 elements 33 m‐1140 m 4:1 grid cell ratio with fine grid PHILADELPHIA WATER DEPARTMENT | MODEL SETUP
Boundary Conditions – Hydrological, Oceanographic, Meteorological • Tributaries: USGS discharge • Southern open boundary: water level from NOAA’s Delaware City tidal gage • Wind: (NCDC/NOAA Ports/PWD) stations at • • • • •
PHILADELPHIA WATER DEPARTMENT | MODEL SETUP
Trenton‐Mercer Philadelphia Airport Wilmington Burlington Buoy C
Boundary Conditions ‐ Anthropogenic • Combined sewer outfalls from Philadelphia, Camden, Chester and Wilmington • Industrial and municipal Permitted Dischargers • Direct storm water runoff
PHILADELPHIA WATER DEPARTMENT | MODEL SETUP
Calibration ‐ Bottom Roughness • Spatially variable • Roughness increases going upstream • Based on 2003 sediment inventory study (Sommerfield &Madsen)
PHILADELPHIA WATER DEPARTMENT | MODEL SETUP
Fine Grid Results
Validation Data Tide Gages and Current Meters • NOAA • PWD (Woods Hole Group)
PHILADELPHIA WATER DEPARTMENT | FINE GRID RESULTS
Statistical and Harmonic Analysis Water Level
Station Marcus Hook Philadelphia Burlington Newbold
RMSE [m] 0.038 0.050 0.081 0.102
Skill Factor [‐] 0.999 0.999 0.997 0.996
Amplitude Error
Phase Error Phase Error
PHILADELPHIA WATER DEPARTMENT | FINE GRID RESULTS
Velocity
Station Buoy C Buoy B db0301 Buoy A Amplitude Error
Phase Error
RMSE (m/s) 0.073 0.059 0.093 0.094
Skill Factor 0.997 0.993 0.993 0.988
Tidal Analysis Water Level
Progressive Wave
Velocity
Tidal Asymmetry – Overtide ratios M4/M2 and M6/M2
PHILADELPHIA WATER DEPARTMENT | FINE GRID RESULTS
Subtidal Analysis no Local Wind
with Local Wind
PHILADELPHIA WATER DEPARTMENT | FINE GRID RESULTS
Coarse Grid Results
Statistical and Harmonic Analysis Water Level Station Marcus Hook Philadelphia Burlington Newbold
RMSE (m) 0.060 0.043 0.090 0.105
Velocity Skill Factor 0.998 0.999 0.997 0.996
Station Buoy C Buoy B db0301 Buoy A
Amplitude Error
Phase Error
PHILADELPHIA WATER DEPARTMENT | COARSE GRID RESULTS
RMSE (m/s) 0.038 0.050 0.081 0.102 Amplitude Error
Phase Error
Skill Factor 0.999 0.999 0.997 0.996
Tidal Analysis Water Level
Progressive Wave
Velocity
Tidal Asymmetry – Overtide ratios M4/M2 and M6/M2
PHILADELPHIA WATER DEPARTMENT | COARSE GRID RESULTS
Discussion • • •
•
Model can represent complex processes and resulting transport dynamics Harmonics, Statistical Analysis • RMSE, amplitude error within acceptable range • Harmonic phase errors are small compared to periods Validated shallow water overtides – good representation of: • Progressive waves • Tidal asymmetry • Subtidal signal • Tidal energy transfer captured well important for transport modeling Key findings: • Local winds have non‐negligible impact on transport • Coarse grid resolution does not negatively impact validation criteria
PHILADELPHIA WATER DEPARTMENT | DISCUSSION
Thank You! Questions?
