Regional coupled ocean modelling

1
Regional coupled ocean modelling Karen
Assmann Bjerknes Centre for Climate Research,
Bergen, Norway
2
(No Transcript)
3
Atmospheric forcing Surface fresh water fluxes
4
Sea ice drift in the Amundsen Sea
Assmann et al., JGR, 2005
5
NCEP Forcing in the Amundsen Sea
NCEP 2m Air Temperatures vs Air Temperatures from
a sea ice buoy in the Amundsen sea (March 2000
Oct 2001)
NCEP 10m winds vs NB Palmer Shipboard
data (Amundsen Sea, February/March 2000)
Assmann et al., JGR, 2005
6
Previous evaluations of Reanalysis data
Wind speed bias (NCEP-WOCE ship)
Smith et al., J.Clim, 2001
7
Previous evaluations of Reanalysis data
Correlation of Reanalyses 2m air temperatures
with observations at Antarctic stations
Bromwich Fogt, J.Clim, 2004
8
Limitations of ice-ocean models forced by
Reanalyses The Weddell Polynya
Sea ice thickness anomaly in BRIOS2.2 August 1975
R. Timmermann, pers.comm.
9
Limitations of ice-ocean models forced by
Reanalyses The Weddell Polynya
NCEP 2-m Temperature for August 1975
Sea ice thickness anomaly in BRIOS2.2 August 1975
R. Timmermann, pers.comm.
10
Limitations of ice-ocean models forced by
Reanalyses The Weddell Polynya
NCEP 2-m Temperature for August 1975
Sea ice thickness anomaly in BRIOS2.2 August 1975
Simulated MLD September 1975
R. Timmermann, pers.comm.
11
Sensitivity of ice extent and thickness to air
temperature
Sea ice thickness Amundsen Sea
Mean annual cycle of sea ice extent Amundsen Sea
(100 150 W) BRIOS2.2 vs SSM/I
Assmann et al., JGR, 2005
12
Precipitation its influence on sea ice
e.g. Amundsen Sea Ice Shelves Model 3
mSv Estimate from Observations 4.9 mSv (Jacobs
et al., 1996) P-E (NCEP mean 1978-2001) 13.4
mSv Snow ice Model 25 of ice volume Obs
14-60 (Jeffries et. al, 19971998)
Assmann et al., JGR, 2005
13
P-E Sea Ice
ECMWF
CSIRO9
GASP
Mean September sea ice concentration
Marsland Wolff, JGR, 2001
14
Surface fresh water fluxes sea ice
Mean September sea ice concentration
CSIRO9
GASP
ECMWF
P-E
10 cm/a Glacial FW flux (68 mSv S of 60 S)
Marsland Wolff, JGR, 2001
15
Ice shelf fresh water flux, sea ice thickness
water mass characteristics
Difference in TS-Characteristics Without cavities
- with Weddell Sea Ross Sea
Sea Ice thickness Without ice shelf cavities
with cavities

Hellmer, GRL, 2004
16
Regional Modelling
17
Cross-shelf exchange in the Ross Sea
Resolution 5 km Run for 1 year, start
mid-September Open boundaries, WOA ST,
OCCAM velocities
Total cross-shelf transport over 1 year Red
on-shelf, Blue off-shelf
T at 300m at model day 350 (early September)
Dinniman et. Al, DSRII, 2003
18
Regional modelling of circulation and water
masses in ice shelf cavities
e.g. FRIS Jenkins et al., JGR, 2004
19
Regional coupled biological-physical modelling
Ross Sea Simulated surface Cl a concentration
ACC Simulated summer surface concentrations Phytop
lankton (?g Cl a/l) Zooplankton
(?C/l) Dissolved iron (nmol/l) Silicate
(?mol/l)
Hense et al., JMS, 2003
Arrigo et al., JGR, 2003
20
Regional Models

• Mostly focussed on continental shelf
• Coupled to
• gt Biology
• gt Ice shelf cavities
• Effect of boundaries
• gt strong control on model circulation
• Useful to directly supplement observations
  identify processes affecting observations
• Separate from larger models - link through
  nesting?

21
Circumpolar Modelling
22
Parametrization of vertical mixing in the
Weddell Sea
Observed Temperature section HASO
SSM/I sea ice concentration Winter 1993
Model Sensitivity studies using BRIOS
Timmermann Beckmann, OM, 2004
23
Dense shelf water variability in the Ross Sea
Simulated HSSW salinity and temperature in the SW
Ross Sea
Correlation of HSSW signal with 150m S and 400m T
No lag
5-year lag
Assmann Timmermann, OD, 2005
24
Transport origin of dense water masses crossing
the South Scotia Ridge
Integrated transports
Philip Passage
Model grid in the Weddell Sea
Orkney Passage
Bransfield Strait
Schodlok et al., DSR, 2002
25
FESOM - applications
I. Circumpolar Southern Ocean configuration
seasonal cycle of ice thickness

• follows BRIOS concept
• 1.5 x 1.5cos ? horizontal resolution
  additional nodes at coast
• open boundary with sponge zone at 48S
• 9000 surface nodes, 26 layers
• has been run for many 100yrs by now

R. Timmermann, pers. comm.
0 1.5
3.0 m
26
Circumpolar Models

• Control on inflow water masses to SO
• ( ACC transport)
• No boundaries for sea ice marginal seas
• Allow investigation of interaction between shelf
  deep basin

27
Global Modelling
28
Regionally focussed global modelsHOPE the
Mertz Polynya
Simulated sea ice concentration drift
Grid resolution km
Simulated salinity at 400 m
Marsland et al., JGR, 2004
29
Antarctic sea ice in a global model
Model (ORCALIM)
Satellite data (SSM/I)
Timmermann et al., Ocean Modelling, 2005
30
Validation of sea ice thickness- The ASPeCt data
set -
Mean sea ice thickness from ASPeCt data
set (Gridded onto ORCALIM2 grid)
Number of observations in each grid cell
Timmermann et al., JGR, 2004
31
Model vs Observations
Difference from ASPeCt data
Real-time sampled model sea ice thickness
Timmermann et al., JGR, 2004
32
Observations Model
ASPECT data set corrected by the bias between RTS
and long-term mean model results
Timmermann et al., JGR, 2004
33
Evaluating global ocean carbon modelsThe
importance of realistic physics
Doney et al., GBC, 2004
34
Ocean acidification in the SO
Aragonite Saturation State in the Year
2100 (Model Composite) Atlantic Pacific
Orr et al., Nature, 2005
35
Global Models

• No need for boundary conditions
• Not necessarily SO focussed, but widely available
• Resolution generally sufficient to investigate
  both basin-scale processes in SO as well as SO
  role in global ocean
• Larger community - good basis for comparison
• OCMIP - strong interest of carbon community for
  SO