Arctic circulation, freshwater, and a regional arctic model

1
Arctic circulation, freshwater, and a regional
arctic model
Alan Condron Peter Winsor WHOI/MIT
2
The arctic in global climate
The arctic is a source of deep water for the
North Atlantic shelf convection (Aagaard et al.
1985). 1/3 of the Denmark overflow is from the
Arctic It is a pathway for freshwater into the
Nordic Sea, ? capping of deep convection ? MOC
link (Dickson et al. 1988 Hakkinen, 1993). MOC
strength also sensitive to flow through Bering
and Canadian Archipelago (Wadley and Bigg, 2002).

3
The model as a tool to explain Arctic dynamics

FRESHWATER An accurate representation of the
arctic freshwater budget will be vital for
determining climate/MOC sensitivity. Changes in
the atmospheric circulation and storage of
freshwater in Beaufort gyre are important. SEA
ICE Shrinking (3 per decade) and thinning
(Rothrock et al. 1999). Many feedbacks and
uncertainties (changes in heat loss etc.)
Critical ice thickness for ice free summer? Cold
Halocline Layer (CHL) Large fluctuations in
thickness important for insulating sea ice from
the warm Atlantic water (Steele et al. 1998).
Must be accurately represented in models.
CIRCULATION What is the role of wind and
thermohaline circulation in the arctic? (arctic
regional model) The ECCO2 (regional arctic)
model will be used to provide further insight and
understanding into why/how these changes occur.
4
In this talk..

• Climatological Arctic atmospheric circulation
• Arctic ocean circulation observed and modeled
• The Arctic freshwater budget
• Moving towards a regional arctic set-up to
  understand arctic sensitivity to atmospheric
  forcing etc..

5
Climatological Arctic atmospheric circulation
High (1014 hPa)
High (1019.8 hPa)
Low (1006 hPa)
Winter (DJF) Anticyclonic circulation
Summer (JJA) Cyclonic wind circulation
(sea-level pressure av. 1979-1998 for NCEP
reanalysis)
6
The North Atlantic Oscillation
NAO POSITIVE
Weaker High
Intensified High
Above NAO winter index (Hurrell et al.) blue/red
is 5 yr mean Positive NAO (persisted for 1990s)
associated with weakening of the Beaufort high
pressure cell.
NAO NEGATIVE
7
Arctic Ocean Bathymetry
Nansen Basin
Amundsen Basin
Makarov Basin
Area of arctic 9.6x106 km2
Canadian Basin
8
Schematic of surface arctic circulation

Observations through Straits
Volume Trans. (Sv)
Fram -4.2 (-1.9 to -6.5)
Bering 1.0 (0.4 to1.2)
Barents 1.5 (0.6 to 2.6)
Canadian Arch. -1.2 (-0.7 to -1.7)
Surface circulation similar to ice drift
patterns, and influenced by the atmospheric
circulation
Surface flow (0-200 m).
(Figure from Rudels, 2006)
9
Model ocean currents(upper 200m)Cube 43 (1992)
BG
Velocity (m/s)
TPD
10
Schematic of deep arctic ocean circulation
(av. 500-1500m)
Figure from Rudels et al. 2006
11
Vertical structure of the water column
Polar mixed layer (PML)
Salinity stratified cold layer. Cold halocline
layer (CHL)
Warm Atlantic water (the most prominent feature
in all arctic profiles) enters from Fram strait
and Barents Sea
(From Jones,2001)
Cold, saltier cold halocline insulates the
surface ice in the PML from the deeper, warmer
Atlantic layer A lack of upper-ocean
stratification will melt sea ice rapidly due to
upward mixing of heat from the Atlantic Layer
(Bjork et al. 2002).
12
Part II Model changes and observations
13
Cube 43 (av. 1992)
Av. upper 200 m circulation
Cube 43 (av. 2002)
Reduced Beaufort gyre. What dynamics in the model
have led to this change?
14
The circulation change is associated with
increased influx of Warm Atlantic water into the
arctic
Av. 1992
Barents Sea
Average temperature (degrees C)
Av. 2002
FRAM
Average temperature (degrees C)
15
Observed changes in vertically integrated
freshwater
Time
Polar hydrographic climatology (Steele et
al.)
1992
2002
Total freshwater 67552 km3
Total freshwater 74714 km3
Total freshwater 74345 km3
16
Some comments on the changes in circulation and
freshwater

• In disagreement with the model, observations
  by Proshutinsky et al. suggest freshwater
  thickness in Beaufort gyre of 21 m in 2003, and
  strong anticyclonic ocean circulation.
• Why the difference?
• WGHC and PHC biased to measurements from the
  1970s-1980s a period of weak NAO and stronger
  anticyclonic atmospheric circulation. WGHC
  conditions look most like 1975.
• The weaker atmospheric circulation of the 1990s
  may not be able to maintain the thermo-haline
  structure of the Beaufort gyre initially
  prescribed in the model.
• Problems in the air-ice-ocean momentum transfer?
• A regional arctic model will enable us to test
  the response of the ocean to different
  atmospheric forcing/initial conditions

17
An arctic budget
Getting a better idea of what is being
imported/exported in the arctic.
Bering Strait
Amundsen Sea
MClure Strait
Canadian Archipelago
Queen Elizabeth Islands
Nares Strait
Barents Sea
Fram Strait
18
Examples
Fram Strait
Fram Strait cross-section
Bering Strait
19
Volume transports (monthly data)
Volume Trans. (Sv) Volume Trans. (Sv)
Strait Cube43 (mean) Obs .
Fram -2.89 -4.2 (-1.9 to -6.5)
Bering 0.96 0.8 (0.4 to1.2)
Barents 2.98 1.5 (0.6 to 2.6)
Canadian Arch. -1.12 -1.2 (-0.7 to -1.7)
Fahrbach et al., Woodgateet al, Schauer et
al., Rudels et al.,
20
Bering strait transport Observations Vs cube43
Woodgate et al. 2005 monthly climatological (av.
1990-2002) from moorings.
Model is in good agreement (r 0.92)!
21
Sea Ice freshwater transport (monthly cube 43)
Ice Freshwater Transport (km3/yr) Ice Freshwater Transport (km3/yr)
Strait Cube 43 Obs.
Fram -2491 -2265 (-1300 to -4000)
Bering 79 0
Barents -110 0
Canadian Arch. -6 -160 (-160 to -410)
Total -2528 -2423
Rothrock et al., Prinsenberg et al.
Main ice export via Fram Strait. Export through
Canadian Archipelago is primarily in liquid form.
22
Liquid freshwater transport(monthly output cube
43)
Liquid Freshwater Transport (km3/yr) Liquid Freshwater Transport (km3/yr)
Strait Cube43 (mean) Obs.
Fram -2555 -2405340
Bering 2263 2489 300
Barents -398 -340 80
Canadian Arch. -2174 -3153 1300
Total -2864 -3409
Importing Freshwater
FT U (1-(sal/34.8) where U is volume
transport. Note that positive and negative FT are
considered
Exporting Freshwater
23
Total (ice water) Freshwater Transport
(anomaly from 1992-2001 av.)
24
Arctic freshwater leaves via the Canadian
Archipelago
Total freshwater transport anomalies (from
1992-2001 av.)
25
NAO and Freshwater Transport
NAO
Fram -0.09
Bering -0.35
Barents -0.37
Can. -0.25
Total -0.39
Monthly freshwater transport
Monthly NAO
26
Towards an Arctic basin freshwater budget
ATMOSPHERE
ATMOSPHERE
Precip 3253
Precip 3253
Evap1299
Evap1299
OCEAN OUT 9033 IN
8595 Imbalance -438
OCEAN OUT 10372 IN
9494 Imbalance -878
Bering 2263
LAND Run-off
LAND Run-off
Bering 2489
R 3000
R 3000
Bering (ice) 79
Barents Sea 398
Barents Sea 340
Canadian Arch. (Water) 2174
Canadian Arch. (Water) 3153
Canadian Arch. (Ice) 6
Canadian Arch. (Ice) 158
Barents (Ice) 110
Fram Strait (water) 2405
Fram Strait (water) 2555
Fram Strait (ice) 2265
Fram Strait (ice) 2491
Cube 43
From Serreze et al. (in press)
27
The arctic regional model

• A solution for exploring the sensitivity of
  the arctic, and understanding observed changes
• Two resolutions
• standard 18-km resolution with 420x384 grid
  points
• Coarse resolution 220x192
• Run-time
• 420x384 approximately 1/12 the resources
  of the global domain e.g. on 36-CPU an 11 yr
  integration takes 8 days,
• Coarse resolution run-time is about 1/10 of
  this.
• The arctic straits (Fram, Bering etc) are
  away from the model open boundaries, so flow in
  these regions will respond to the model dynamics.

28
What to investigate

1. role of wind driven versus thermohaline
   circulation
2. Atmospheric-ocean forcing on arctic freshwater
3. sensitivity to sea ice thickness critical sea
   ice threshold at which point arctic becomes ice
   free in summer.
4. Can an ice age model be added? What is the age of
   ice advected, and how does this change over time?
5. Stability of the cold halocline layer heat
   exchange will it reform if destroyed? What
   determines its horizontal extent?