A brief introduction to UMCES Chesapeake Bay Model

1
A brief introduction toUMCES Chesapeake Bay
Model

• Yun Li and Ming Li
• University of Maryland Center for Environmental
  Science
• VIMS, SURA Meeting
• Oct-1-2010

2
Outline

• Grid
• Forcing
• wind
• river
• SST (sea surface temperature)
• Boundary
• Model Sensitivity
• wind
• background diffusivity
• resolution

3
Model domain includes mainstem, 8 major
tributaries a piece of coastal ocean
Curvilinear orthogonal grid 20 stretched
layers Resolution Low (80x120) lt1km
in cross-channel 23km in along-channel
deepest point is 26m. High (160x240)
lt500m in cross-channel 0.52km in
along-channel deepest point is 40.5m. Deep
channel is well resolved in the high-resolution
model
4
Wind
Wind Data Source http//www.wunderground.com
6 wind stations, hourly data
Interpolation wind speed is linearly interpolated
along latitude and longitude Wind speed to
stress where Amplification Factor
(Xu et al. 2002 Wang and Johnson 2000)
Station N-S comp E-W comp
NIA(KORF) 1.37 1.25
PRS(KNHK) 2.05 1.43
BWI(KBWI) 1.50 1.00
5
River
River Data Source USGS monitoring stations,
discharge (m3/s) and temperature (ltmonthly)
salinity is set to zero Major
Tributaries Susquehanna (1) 01578310 Patapsco
(3) 01583500 01586210 01586610 Patuxent
(1) 01594440 Potomac (1)
01646500 Rappahannock (1) 01668000 York (2)
01673000 01674500 James (3)
02037500 02041650 02042500 Choptank (1)
01491000 Sea Level at Riverine Boundary
New CPP PSOURCE_FSCHAPMAN allow incoming
wave but avoid reflection
6
SST
SST Data Source Chesapeake Bay Program
along-channel observation (monthly or biweekly)
Interpolation Surface temperature (lt1m) is
linearly interpolated along latitude
Configuration New CPP SST_RELAXATION (must undef
QCORRECTION) Nudging is performed every 6 hours.
7
Model only has open boundary at eastern edge.
Data Sources Tides Five major components M2, S2,
N2, K1, O1 from Oregon State U. global inverse
tidal model TPXO Subtidal Sea Level detided
component from NOAA historical data at Duck, NC T
and S linear interpolation from WOA2005, monthly
Levitus climatology ubar and vbar zeros at
boundary Configuration sea level
FSCHAPMAN 2D momentum EAST_M2FLATHER
3D momentum EAST_M3RADIATION T and S
EAST_TRADIATION
EAST_TNUDGING (1 day)
8
Wind
Along channel distribution in a low-runoff
period. Using the amplification factors, the
model produces a well-mixed surface layer in
better agreement with the observation
9
Vertical Diffusivity
A comparison of along-channel salinity
distribution between four model runs with
different vertical diffusivity (Ks) shows that
the stratification increases as Ks decreases.
the along-channel salinity gradient increases as
Ks decreases. model prediction becomes less
sensitive when Ks is reduced below 10-5m2/s.
10
Resolution
Salinity, April 23, 1997
Increasing model resolution is important to
resolve the narrow deep channel, which is the
main conduit for the landward salt transport.
k-kl,80x120
k-kl,160x240
observation
11
Summary

• Special features in UMCES Chesapeake Bay Model
• Both low- and high-resolution configurations are
  available
• Apply Chapmans condition for sea level elevation
  at Riverine boundary to avoid wave reflection.
• SST nudging to observation with a time scale of
  6hr.
• Using wind amplification factors, the model
  produces a well-mixed surface layer
• Areas of Improvement
• Model resolution in the deep channel
• Turbulent mixing near the pycnocline (Li et al.
  2005)
• Adjustment of observational wind (Xu et al. 2002
  Li et al. 2005)

12
Salinity, Apr 25, 1997
observation
amplified wind
original wind