SUNTANS simulation of the Snohomish River estuary Bing Wang Oliver Fringer, Bob Street Environmental

1
SUNTANS simulation of the Snohomish River
estuaryBing WangOliver Fringer, Bob
StreetEnvironmental Fluid Mechanics
LaboratoryStanford UniversityMURI Site visit
1-2 October 2007
2
Modeling Challenges

• MULTISCALE
• coherent structures 1m
• sill of interest 50 m by 5 m by 10 m
• river width 400 m
• tidal excursion 10,000 m
• Puget sound 50,000 m
• Strong currents 2 m/s
• Large tidal range (5 m) extensive intertidal
  zones/wetting and drying
• Complex bathymetry
• Nonhydrostatic
• Complex boundary conditions (tidal scarce river
  discharge data)

3
SUNTANS model domain and boundary Conditions
Open boundaries forcing 1) at the inlet of
Snohomish River fresh river discharge 2) at
the open boundary in the sea channel that
connects to Skagit Bay saline water, tidally
forced fluxes river discharge 3) at the open
boundary in the sea channel that connects to
Puget Sound Main Basin saline water, tidally
forced fluxes river discharge.
?force cross-sectional averaged velocity.
Open boundary fluxes based on Puget Sound
circulation
4
Unstructured grid
Coarse grid Number of cells in plan 70,000
(85 near study site and channels) Total 3D
cells 3 million Finest resolution 8m Coarsest
resolution 300 m Vertical Resolution 0.25 m
(z gt -5 m) 0.25-0.6 m (-17 lt z lt -5 m) 0.6-30m
(-200 lt z lt 17 m) Run times X days for 1 day of
simulation on X processors.
The coarser grid
5
Bathymetry

• 30m resolution bathymetry and topology data by
  Finlayson (2000), which has large coverage but is
  not reliable for shallow estuarine regions.
• 10m resolution bathymetry data surveyed by our
  field team in 2005, which covers the Snohomish
  River channel downstream of the bridge of highway
  529.
• 10m resolution bathymetry data surveyed by
  Rusello et al in 2007, which covers the
  intertidal mudflats around the bypass.
• For the upstream parts of the river channels,
  where detailed bathymetry is not available, a
  constant north-south bottom slope (1e-4) is
  specified based on the work by Fram (2002). The
  cross-section of the channel is assumed to be
  rectangular.

Interpolated depth on the grid (left) and an
aerial image (right) (data is from Kate Edwards).
SILL CREST
zoomed-in view of sill.
6
Free-surface comparison
Blue Prediction Red Measurements
In the sound
USGS Gage station
Mooring M3b
7
Along-channel depth-averaged velocity comparison
Free surface in the sound (for reference)
Green Predictions Blue Measurements
M3b
M4
8
Lateral variability in the bend

Inner (shallow) edge is faster than outer
(deeper) edge bathymetric effects stronger
than centripetal effects.
9
Comparison of salinity time series
Free surface
Free surface
Model
Data
Mooring M4 (bypass)
Mooring M3b (channel)
10
Salinity over two tidal cycles
11
Surface salinity field
EBB
LLW
HHW
FLOOD
12
Surface salinity field at the study site
early ebb
late ebb
early flood
Salinity in bypass during late flood is lower
than in channel due to advection of more saline
water from the sound resulting from fresh water
"storage" on the flats.
later flood
late flood
13
Surface temperature field at the study site
Comparison Remote Sensing
Model surface temperature (oC)
Aerial IR image
14
Summary and future directions

• SUNTANS simulations accurately reproduce currents
  and salinity fields.
• High-resolution simulations of estuaries are
  hard!
• Bathymetry has the largest effect on accuracy of
  results, followed by (in order of importance)
  grid, boundary conditions, and momentum advection
  scheme.
• Plan
• 3D velocity, salinity, and density fields will be
  used to aid in planning future study sites.
• Higher-resolution grid (1 m near study site) to
  simulate coherent structures near bypass and over
  sill and in the main channel.

15
The importance of good bathymetry
Measurements indicate a flow-reversal during ebb
tide in the bypass channel.
Peak ebb
Peak flood
reversal
Depth-averaged current at Mooring M4 Positive
current indicated by arrow.
"expected"
16
The Importance of Good Bathymetry
Bypass
FLOOD
HHW
17
The Importance of Good Bathymetry
Bypass
EARLY EBB
HHW
18
The Importance of Good Bathymetry
Bypass shoal becomes dry, inducing a reverse
free surface gradient during ebb.
Bypass
LATE EBB
HHW
19
The Importance of Good Bathymetry
Bypass
LATER EBB
HHW
20
(No Transcript)
21
(No Transcript)
22
Comparison to field data
Depth-averaged currents Positive upstream
direction indicated by arrows.