ECC02

1

ECC02 Towards an Arctic Freshwater Budget
Alan Condron Peter Winsor Woods Hole
Oceanographic Institute
2
Why are we interested in modeling the arctic?

• Over last 2-3 decades, the arctic has warmed more
  than any other region of the world. Models are
  indicating that anthropogenic warming will be
  enhanced in the arctic.
• There are observed increases in precipitation,
  permafrost melt, river discharge, sea-ice melt,
  glacial melt ? alter freshwater budget of the
  arctic. The freshwater anomaly in 1990s 1650
  km3/yr (Peterson et al., 2006).
• A global change
• The release of arctic freshwater into the Nordic
  Seas is believed to be critically important for
  the strength of the MOC.
• And, therefore, a realistic freshwater budget
  will allow a better estimate of the sensitive of
  the global circulation to freshwater
  perturbations originating from high latitudes.

3
Global impacts of arctic freshwater export
Freshwater perturbation Capping of convective
areas Reduced/shut-down MOC Climate impacts
From Curry and Mauritzen (2005)
4
Schematic of arctic circulation

Surface flow (0-200 m).
(Figure from Rudels, 2006)
5
Working towards a freshwater budget in ECCO2

Is the Arctic ECCO2 consistent with
observations? Can ECCO2 move towards a decent
representation of the Arctic freshwater budget?
Observational efforts in the Arctic will
allow us to look at how well the model behaves in
this region. River discharge, sea ice,
ocean freshwater storage and export/import and
P-E need to be evaluated. Examples of existing
data (1) Recent estimates of the Arctic
freshwater budget (see Serreze et al., 2006)
allow us to validate the model in an integrated
way. (2) long-term observations of flux
estimates through Fram and Bering Straits for the
ocean and sea ice (Woodgate et al., Kwok et
al.) (3) long-term basin-wide salinity and
temperature profiles and observational data from
moorings .
6
Observations Arctic basin freshwater budget
(km3/yr)
ATMOSPHERE E40 205
E40 4973
(P-ET)L E40(A)2932
Po Obs 3253
ETo Resid.1299
ETL 4813
Bering Inflow 2489
OCEAN Ocean 74345 As sea ice 10449 OUT
8282 IN 9072 Imbalance
-791
LAND Total IN (P-ET)L 2932 Total OUT (R)
3000 Imbalance 68
NCC 254
R 3000
FRAM Strait 498
Barents Sea 340
W. Spitz Curr. 340
Canadian Arch. Ice 158
Canadian Arch. Water 3153
Fram Strait (upper) water 2400
Fram Strait (upper) ice 2265
From Serreze et al. (in press)
7
Towards determining whether the Arctic in ECCO2
is consistent with observations
So far 3 different cube runs examined Cube 22
Started from end of cube 19 (run for 30 years
before cube 22 run) Cube 43 Starting from WOCE
Climatology Cube 47 Same as Cube 43, but with
arctic rivers (freshwater perturbation experiment)
8
Observed freshwater
Salinity
Vertically integrated freshwater
Beaufort gyre
(meters)
Most freshwater stored in the Beaufort gyre.
Periodic release linked to Great salinity
anomalies (Proshutinsky et al., 2002) Data from
Polar hydrographic climatology (Steele et al.).
1x1 degree dataset.
Total freshwater Ocean 74345 km3 Sea ice
10450 km3
9
Freshwater cube 22 (annual mean 1992-2003)

Salinity
Vertically integrated freshwater
Cube 22 is too salty in the arctic by 3
PSU. Fresh pool in Beaufort gyre is absent. Total
freshwater volume approx 32,000 km3
10
Vertical stratification

• Can any numerical models retain the Arctic
  stratification?
• Recent AOMIP (Artic Ocean Model Inter-comparison
  Project) show that models lack a Cold Halocline
  Layer and struggle with the overall
  stratification.
• A lack of upper-ocean stratification will melt
  sea ice rapidly due to upward mixing of heat from
  the Atlantic Layer. Is this true in the
  cube-sphere runs too?

Cold, saltier cold halocline insulates the
surface ice from the deeper, warmer Atlantic layer
Cold, saltier layer
Warm Atlantic water
Stratification maintained by freshwater (runoff,
P-E, Bering Strait inflow)
11
Vertical stratification in cube 22
Canada Basin
Nansen Basin

• Weak vertical stratification.
• Atlantic water missing.

12
Cube 43
13
Vertically integrated freshwater and upper 100m
velocity field
Start to loose Beaufort gyre circulation
Time
2002
1992
Total freshwater 67552 km3
Total freshwater 74714 km3
There is a 10 loss of freshwater from the model.
14
Is the change in circulation a response to the
positive NAO of the 1990s?
Low NAO
High NAO
Ice drift (Figures from Rigor 2006)
Strong BG circulation
Weaker BG circulation
15

Cube 47 Freshwater forcing experiment
16
3400 km3/yr (0.1 Sv) freshwater added from 54
arctic rivers for 11 years into cube 47. Approx.
60 of discharge is from May-July
Lena
Mackenzie
Ob and Yenisey
2002
1992
Surface salinity anomalies (cube 47-cube 43)
(negative fresher)
River discharge penetrating into interior.
17
Changes in freshwater volume
Observed 74345 km3
4500 km3
Cube 22
Cube 43 (blue) Cube 47 (red) gt Cube 43 has
an initial (1993-1992) freshwater loss of 3146 km3
18
Comments on the change in freshwater volume

The cube runs all show an initial loss of
freshwater from the arctic. The difference in
freshwater volume between cube 47 and cube 43 is
not equivalent to the annual freshwater input of
the rivers. In fact, the difference for 1992 is
only 290 km3 (approx. 1/12 of the total of 3400
km3/yr from rivers) Can sea ice growth account
for the initial loss of freshwater in the cube
runs, and the lack of increase in freshwater
volume in cube 47?
19
Observed 10500 km3
Rapid Sea ice growth ? Brine rejection ? increase
in salinity
Initial sea ice growth (1993-1992) extracts 1382
km3 of freshwater from the ocean (av. 13 cm of
ice growth in one year), and 1/3 of the initial
loss of freshwater from the ocean in cube 43.
There is no large difference in sea ice
concentration between cube 43 and cube 47 to
account for the small increase in arctic
freshwater in cube 47, despite the addition of
3400 km3/yr from rivers. Where is the
freshwater going?
20
Sea ice (cube 43)
Growth
Loss
2000-1992
1992
21

Future plans/ideas
Run cube simulations for longer as the freshwater
residence time of the arctic is 10 years, and
close to the length of the simulation.
Calculate a full freshwater budget for the cube
arctic that can be compared with observations.
Determine fluxes of freshwater through straits
Canadian Archipelago (e.g Nares Strait etc),
Fram, Bering, Barents Sea etc. Calculate sea ice
freshwater budget Determine why and how the cube
arctic rapidly losses freshwater.