1. Circulation and overturning in the northern North Atlantic
Circulation and overturning in the northern North Atlantic
Jan Even Ø. Nilsen
Nansen Environmental and Remote Sensing Center,
and Bjerknes Centre for Climate Research, Bergen, Norway
Jan Even Nilsen
At Nansen and Berknes centres in Bergen, Norway.
In addition to presenting this current work,
I will take the opportunity to give you an overview of the Nordic Seas
As well as some of my previous work.

2. Outline Background and traditional concepts
Introduction to the Nordic Seas
Transformation and ventilation processes in the Nordic Seas
Variabilty of the Northern Thermohaline Circulation

3. Background and traditional concepts 1
The dense water formation in the Nordic Seas and Arctic is an important part of the global conveyor belt
Fresh water may limit dense-water formation

4. Background and traditional concepts 2
The Nordic Seas limb of the conveyor belt is the Greenland (and Iceland) Sea Gyre
The Nordic Seas limb of the conveyor belt is the Greenland (and Iceland) Sea Gyre

5. Background and traditional concepts 3
Freshwater from the Arctic to the Nordic Seas can (and will) slow down the conveyor there
Freshwater from the Arctic to the Nordic Seas can (and will) slow down the conveyor there

6. Introduction to the Nordic Seas

7. Johannessen et al. (2014)

8. The Nordic Seas The warm and saline Atlantic Water
Temperature (d=200 m)
The warm and saline Atlantic Water
transports heat to the Arctic
maintains the ice-free oceans and regulates sea-ice extent
influences the region’s relatively mild climate
is the northern branch of the global thermohaline overturning circulation
Heat loss in the Norwegian Sea is key for both heat transport and deep water formation
Eldevik et al. (2009)
Eldevik & Nilsen (2013)

9. Bathymetry Svalbard Greenland Barents Sea Greenland Basin
Norway
Greenland
Greenland Basin
Lofoten Basin
Norwegian Basin
Iceland Sea
Barents Sea
Svalbard
Bathymetry
Ths is the bathymetry of the Nordic Seas
NAMES #
Note the deep basins, mid atlantic ridge,
and steep continental slopes

10. Furevik & Nilsen (2005)

11. Topographically trapped waves/fronts
Low values of H/Hx show where topographically trapped waves most likely will exist and what their horizontal scales will be.
These perturbations also serve to trap the frontal and current system above the slopes.
When the variability of the wind stress curl field contains the length- and time scales supported by the slope, the response will primarily be a barotropic perturbation of the water column. Hence, information of a sloping ocean floor is communicated to the dynamics of the whole water column. Although these perturbations can initiate current meandering and eddy formations they also serve to trap the frontal- and current system above the slope.
Nilsen & Nilsen (DSR 2007)

12. Wind stress and wind stress variability
Winter-mean surface wind stress for 1949–2004
… regressed on the NAO-index
Furevik & Nilsen (2005)

13. Wind governs variability of currents
Summer to winter difference
in surface geostrophic velocities from altimetry.
Mean geostrophic velocity and speed from altimetry.
See, e.g., Raj et al. (2015)

14. Transformation and ventilation processes in the Nordic Seas

15. Overview of transformation and ventilation processes
Mixing between Atlantic and Arctic Water
Freshwater from Coastal Currents
Atmospheric Cooling
Shelf and Ice Processes
Greenland Sea Gyre
Iceland Sea
Objectives:
Review modification and ventilation processes
Quantify using hydrography and models
Relative contributions
Eastern view.
Now also including the Coastal Currents.
Eddy mixing along the Arctic Front
Freshwater makes lighter, but still has to be mixed into the AW and central areas. However will not talk about FW here (reer to ASOF reports).
AW is cooled from above, as well as by lateral mixing.
Cooling and brine release
Deep open ocean convection.
Review current knowledge about water modification and dense water formation (ventilation) in the Nordic Seas
Use some simple approaches to quantify further using observations and models
Synthesize with emphasis on the processes’ relative contributions to OW
Brief preview of that in this talk.
Pinpoint key geographical areas
Give short overview of forcing
Quantify fluxes/contribution from process

16. Cooling of Atlantic Water
4 Sv / 40 TW
2 Sv / 40 TW
? TW
40 TW
Average temp.
and energy content of AW volume by latitude.
Temperature decrease
Yes, total energy (used Kelvin)
The northward temperature decrease is strong
and in the Fram Strait the AW is at OW density.
The volumes and heat are shwn (relative to

17. Exchange across the Arctic Front
Eddy flux (by thermal wind from hydrography):
u*~ m/s.
Preliminary estimates:
Eddy circulation volume flux of F~1.5 Sv.
Then T~4K gives Hf~30TW.
Cross frontal eddy exchanges
Heat (and salt) to central regions
We saw on previous slide the depth of the fronts/ from hydrography
From the observations:
Parametrized eddy flux, calculated by thermal wind from hydrography and empirical constant c=0.03 is u*~ m/s.
Preliminary estimates:
A frontal area of ~109 m2, leads to an eddy circulation volume flux of F~1.5 Sv.
Using a temperature difference over the front of ~4K gives Hf~30TW.

18. Deep Open Ocean Convection
Greenland Sea Gyre:
~0.5 Sv
~2 TW
R=100 km; H1=H2=500 m; F=0.5 Sv; dT=0.8 K; QT=1.7 TW
4 TW = 1 Sv 1 K : Areal + kjent W/m2 => 2TW. Har minst 1K => max 0.5 Sv.
Also Between buoant currents. A=4A ... Finn ut TW
Strong wintertime convection
Ventilated water inside
Cross frontal eddy circulation
Buoyant water in and ventilated water out
Balanced by heat loss to atmosphere

19. Exchanges with Coastal Currents
ESC/Spitsbergen shelf water:
Freshwater to AW
Heat sink
Capping of AW
MECHANISMS:
Cross frontal eddy exchanges (velocity shear)
Ekman transports followed by vertical mixing (wind forcing)
EGC:
Freshwater to central regions
+ Sea ice?
Heat sink
Advective export (JMC+EIC)
NCC:
Freshwater to AW
Heat to Barents Sea!

20. Shelf and Ice Processes
Shelf Convection: Polynyas, brine release, pooling on shelf
Slope Convection:
Overflow, cascading, entrainment
Key locations:
Storfjorden
Franz Josef Land
Novaya Zemlya
1.1
0.1
0.8
Numbers: Mostly from literature, but also from colleagues ... Such as this.

21. Volume transports 6 Sv 1 4 2 Referanser!!
1+4+2~6 Independent estimates

22. Variability of the Northern Thermohaline Circulation

23. First, the way of the overturning variability
Along which pathway can we trace anomalies in the overturning circulation?

24. The hydrography of the AW-in-OW-out loop
@ 200 m
Nilsen et al. (2008)
T @ 200 m
RAW m ; GSW m ; OW < 27.8 < AW
The collection of time series are unique! The rough composition is in agreement with more detailed assessments (e.g., Mauritzen 1996; Isachsen et al. 2007) - GSW is here understood to represent AIW in general! Long time series of composition is also unique.
Eldevik et al. (Nature geo. 2009)

25. The observed hydrography 1950-2005
No clear correlation with Greenland Sea
(wo/w lag)
– in contrast to what models generally show
The collection of time series are unique!
Salinity
Temperature
The NISE dataset (Nilsen et al.,2008) Eldevik et al. (Nature geo. 2009)

26. Anomalies travel the loop AW-in-OW-out
0.49; 0.46
0.47; 0.42
NAW  1 yr  RAW  2 yr  DS
Eldevik et al. (Nature geo. 2009)

27. Anomalies travel the loop AW-in-OW-out
0.60; 0.50
A “new” overturning loop
NAW  1 yr  NNAW  1 yr  FSC
Eldevik et al. (Nature geo. 2009)

28. Anomalies travel the loop AW-in-OW-outX
0.49; 0.46
0.60; 0.50
0.47; 0.42
Eldevik et al. (Nature geo. 2009)

29. Findings – based on the observational record
Anomalies are carried with the Nordic Seas’ loop of Atlantic-derived waters
Predictability: Atlantic inflow #1 candidate
For the Faroe Shetland Channel a local Norwegian Basin overturning loop is suggested
Overflow variability is not associated with deep convective mixing/Greenland Sea
as opposed to the traditional understanding and climate model diagnoses S
First observation-based synthesis of the thermohaline variability of overflows and sources on interannual to decadal timescales
Salinity (normalised)
Eldevik et al. (Nature geo. 2009)

30. Next, the fresh water in the overturning
What about the freshwater changes in the bulk of the Nordic Seas
(where they may affect overturning circulation)?

31. Between 1965 and 1990, the waters of the Nordic Seas and the North Atlantic Ocean freshened substantially.
The Nordic Seas are key sites in the Atlantic Meridional Overturning Circulation, with deep water formation inhibited if the surface salinity is too low.
The Arctic Ocean also became less saline over this time, as a consequence of increasing runoff.
Not clear whether flow from the Arctic was the main source of the Nordic Seas salinity anomaly.
See also Curry & Mauritzen (SCI, 2005)
Glessmer et al. (Nature geo. 2014)

32. Sources of freshwater anomalies to the Nordic Seas
Atlantic Ocean
Arctic Ocean ice
Arctic Ocean liquid
Glessmer et al. (Nature geo. 2014)

33. Accumulation of freshwater in the Nordic Seas
From the south:
Integrating inflow salinities and assuming constant volume inflow (8 Sv).
Observations
Model
Here we use the instrumental record of the Nordic Seas, along with a global ocean–sea-ice model hindcast simulation, to identify the sources and magnitude of freshwater accumulating in the Nordic Seas.
Glessmer et al. (Nature geo. 2014)

34. Accumulation of freshwater in the Nordic Seas
From the north:
Using variable volume transports and inflow salinities to calculate freshwater transports.
Model
Glessmer et al. (Nature geo. 2014)

35. Conclusions
Freshwater anomalies within the Nordic Seas can mostly be explained by less salt entering with the relatively saline Atlantic inflow
Seemingly little contribution from the Arctic.
Hydrographic changes in the Nordic Seas are primarily related to changes in the Atlantic Ocean.
If the Atlantic inflow and Nordic Seas both freshen similarly, Atlantic Meridional Overturning Circulation is relatively insensitive to Nordic Seas freshwater content.
Glessmer et al. (Nature geo. 2014)

36. Next, the fresh water and cooling in concert
The overturning circulation is only one branch of the thermohaline circulation.
Freshwater’s main role is likely through the estuarine circulation.

37. The estuarine circulation & the overturning circulation
Arctic–Atlantic THC
= two branch system
A circulation of heat and salt
=> two degrees of freedom
Courtesy of Bogi Hansen
Estuarine circulation
bottom
surface
river
1. Fresh water outflow
2. Mixing
Estuarine circulation results from FW flow out of the system – entrainment of deeper water – compensating inflow.
This is also at play in the Arctic Mediterranean, where AW is entrained into the FW to make the outflowing PW.
The overturning branch is well studied, but the Arctic–Atlantic THC is a two branch system.
A combined circulation of heat and salt generically has two degrees of freedom.
3. Compensating inflow
Eldevik & Nilsen (J. Clim. 2014)

38. Our system: The Arctic–Atlantic THC constrained at GSR
Salinity
GSR
heat loss
freshwater
Input
Pot. temperature
GSR OW PW AW
As indicated in previous slide, the main exchanges are at the GSR – Volume fluxes.
The system also involves the TH-forcing.
We need the TH properties of the flow. These are found from climatology of salinity and temperature at the ridge [explain sections]. It is a three layered system, which can be separated by ‘standard definitions’.
Averaging the hydrography in the three layers, we find the TH properties of the exchanges: Warm, saline inflow, colder overflow, and colder and fresher surface polar outflow.
THIS IS THE SYSTEM! (a priori)
Note 1: Bering Strait and Canadian Archipelago = small. (Formally volumes in the PW (balance) and any buoyancy contribution in the buoyancy forcing.)
Thermohaline
Circulation
Eldevik & Nilsen (J. Clim. 2014)

39. Our model: Conservation equations at GSR
U1 + U2 + U3 = 0
S1U1+S2U2+S3U3 = qS
T1U1+T2U2+T3U3 = qT
GSR OW PW AW
Now we can set up conservation equations of volume, salt and heat:
- Note the colors throughout: red=AW=1, blue=PW=2, black=OW=3
- T and S from climatology (TS-diagram) – the THC in today’s climate.
- Left: fluxes carried by the volume transports [into the basin]
- Right: vertical fluxes of salt and heat up and out of the ocean
“Knutsen equations” used for studying estuarine circ for decades. Adding conservation of heat allows us to diagnose both estuarine and overturning circ. at the same time.
THIS IS THE MODEL! A simple balance between buoyancy forcing and advection of heat and salt!
Note 2: No STORAGE (formally included in forcing terms)! However, observed trends in heat (3 TW; yr; Norwegian Sea; Skag&mork12) and FW (0.003 Sv; yr; Nordic Seas; Curry&maur05) + Beaufort Gyre (Prosh09), are all 2 orders of magnitude less than the mean balance!
BALANCE IS DOMINATING on DECADAL timescales and more, and (unless we want to study perturbations less than 1% of the mean balance, 0.1 Sv / 3 TW / 1 mSv) storage is negligible.
+ GCMs tend to balance FROM REST within 30 yr.
Eldevik & Nilsen (J. Clim. 2014)

40. Our model: Solved for the circulation (i.e., THC)
U1 = (qS(T2-T3)-qT(S2-S3))/D
U2 = (qS(T3-T1)-qT(S3-S1))/D
U3 = (qS(T1-T2)-qT(S1-S2))/D PW
Inverting => expressions for the volume transports, i.e., the circulation.
- Common denominator (Delta) is linearly (2x) related to the area of the triangle in the TS diagram => The smaller the contrasts, the higher the sensitivity!
- Furthermore, the response of a volume transport to forcing is dependent on the contrasts between the two other water masses.
Note 3: NOT CAUSALITY! Just DIAGNOSING volume transports consistent with observed or estimated hydrography and buoyancy forcing.
∆ = ( S1 - S2 )( T1 - T3 ) – ( S1 - S3 )( T1 - T2 ) AW OW
GSR
Eldevik & Nilsen (J. Clim. 2014)

41. Sensitivity to northern heat and FW forcing anomalies
U1 = (qS(T2-T3)-qT(S2-S3))/D
U1 = 0  qT/qS= (T2-T3)/(S2-S3)
qS=0 , qT>0 => U1 > 0 qS qT PW
We can span out the forcing space, showing the volume transports’ relation to the forcing.
Which combinations of forcing anomalies will give no change in inflow? On the line that has the same relation as T/S contrast between the two other water masses.
Which combinations of forcing anomalies correspond to increased inflow? Solving for increased heat loss shows that on the cooling side we get increased Atlantic inflow.
= COOLING AW OW
GSR
Eldevik & Nilsen (J. Clim. 2014)

42. Sensitivity to northern heat and FW forcing anomalies
U1 = (qS(T2-T3)-qT(S2-S3))/D
U2 = (qS(T3-T1)-qT(S3-S1))/D qS qT PW
Analogous for the PW … AW OW
GSR
Eldevik & Nilsen (J. Clim. 2014)

43. Sensitivity to northern heat and FW forcing anomalies
U1 = (qS(T2-T3)-qT(S2-S3))/D
U2 = (qS(T3-T1)-qT(S3-S1))/D
U3 = (qS(T1-T2)-qT(S1-S2))/D qS qT PW
And for the overflows …
THE WHEEL IS THE EQUATONS, solved for U = f(qS,qT).
Shown here for the current climate (TS-properties).
Different climates, or climate models, may have different hydrography and different orientation of the sectors. AW OW
GSR
Eldevik & Nilsen (J. Clim. 2014)

44. Sensitivity to northern heat and FW forcing anomalies
U1 = (qS(T2-T3)-qT(S2-S3))/D
U2 = (qS(T3-T1)-qT(S3-S1))/D
U3 = (qS(T1-T2)-qT(S1-S2))/D PW
Some cases:
COOL FW
REF (linear, so wheel works for mean circulation as well) – All flows as known. Forcing from literature (but also from our model with transports from literature, Fref = 0.14 Sv and Qref = 282 TW).
The wheel is a qualitative representation; the strength of the circulation changes are of course spanned out in forcing space, not just the sign.
Also, responses to changes in the hydrography of the flows can be diagnosed. [point to equations]
Next slide will show both … AW OW
GSR
Eldevik & Nilsen (J. Clim. 2014)

45. Strengthening of the Atlantic Inflow with a blue Arctic?
FW: Sv [ ref. 1965; Peterson et al 2006 ] weakening of THC
COOL: +50 TW, 40 yr of sea ice retreat [ -0.4 mill km2 ; 1967–present; Kvingedal 2005 ; W/m2 ; Årthun et al 2012 ] strengthening of THC
FW + COOL strengthening of THC
Peterson et al. 2006
0.03 Sv
Jan 2009
Jan 1979 + +
Polar outflow
The background:
a) present focus on Arctic climate change in general, and sea ice retreat in particular, and
b) any Arctic influence on THC is commonly associated with anomalous fresh water input and
c) a weakened THC is inferred accordingly.
An observed FW increase in the Arctic over the latest decades, alone implies a weakening THC.
The contemporary sea ice retreat – more open water in the warm ocean region of the Arctic and corresponding increased heat loss, alone implies a strengthening of THC.
[animate SI]
The combined effect is – not trivial – but found to correspond to a strengthening of THC (No obs., 10 yrs here and there, and for PW nothing.)
[animate SI+flows]
Eldevik and Nilsen is an analytical model that quantify the changes in the Arctic/Atlantic THC (i.e., in Atlantic inflow, overflow and polar outflow at GSR) consistent with any specified changes in the salt and heat budgets of the Arctic Mediterranean.
In reality, there is synoptic variability, but that is not known from observations.
Overflow
GSR
+1Sv
+1.5 Sv
+0.5Sv
Atlantic inflow
GSR
Eldevik & Nilsen (J. Clim. 2014)

46. Summary and implications
U1 + U2 + U3 = 0
S1U1+S2U2+S3U3 = qS
T1U1+T2U2+T3U3 = qT
Consistent framework to diagnose Arctic– Atlantic THC and -change
THC is not overflow
FW increase does not imply weaker THC
The relative strength of estuarine vs overturning circulation reflects FW input
Present changes in the Arctic contributes to increased THC
Observed increasing TH-contrasts implies a more robust THC
Last point (arrows):
-Instrumental record since the ‘70s show warmer and more saline inflow (Holliday, 2008), comparatively constant OW (Eldevik et al, 2009), and likely fresher PW (Curry & Mauritzen, 2005).
Taken out:
Present variability appears proportional to the mean circulations
THC sensitivity mainly reflects heat loss
Climates – or climate models – that differ distinctly in water masses at GSR, would differ distinctly in their THC
Eldevik & Nilsen (J. Clim. 2014)

47. Overall summary
Hydrographic anomalies are carried to the overflows with the Nordic Seas’ loop of Atlantic-derived waters
Overflow variability is not associated with the Greenland Sea
Freshwater anomalies within the Nordic Seas mostly explained by less salt entering with the Atlantic inflow
Thermohaline circulation (THC) is not overflow
FW increase does not imply weaker THC
Present changes in the Arctic contributes to increased THC
Observed increasing TH-contrasts implies more robust THC

48. References (extended list)
Eldevik, T. and J.E.Ø. Nilsen (2013). The Arctic-Atlantic Thermohaline Circulation. Journal of Climate, 26(21), /JCLI-D
Eldevik, T., J.E.Ø. Nilsen, D. Iovino, K.A. Olsson, A.B. Sandø and H. Drange (2009). Observed sources and variability of Nordic Seas overflow. Nature Geoscience, 2 (6), p , doi: /ngeo518.
Furevik, T. and J.E.Ø. Nilsen (2005). Large-Scale Atmospheric Circulation Variability and its Impacts on the Nordic Seas Ocean Climate - a Review. The Nordic Seas: An Integrated Perspective, AGU Geophysical Monograph Series, 158. p
Glessmer, M.S., T. Eldevik, K. Våge, J.E.Ø. Nilsen and E. Behrens (2014). Atlantic origin of observed and modelled freshwater anomalies in the Nordic Seas. Nature Geoscience, 7, doi: /ngeo2259.
Johannessen, J.A., R.P. Raj, J.E.Ø. Nilsen, T. Pripp, P. Knudsen, F. Counillon, D. Stammer, L. Bertino and N. Serra (2014). Toward Improved Estimation of the Dynamic Topography and Ocean Circulation in the High Latitude and Arctic Ocean: the Importance og GOCE. Surveys in Geophysics, 35(3), doi: /s y.
Mork, K. A., and Ø. Skagseth (2010), A quantitative description of the Norwegian Atlantic Current by combining altimetry and hydrography, Ocean Sci., 6, 901–911.
Nilsen, J. E. Ø., Y. Gao, H. Drange, T. Furevik, and M. Bentsen (2003). Simulated North Atlantic-Nordic Seas water mass exchanges in an isopycnic coordinate OGCM, Geophys. Res. Lett., 30(10), 1536, doi: /2002GL
Nilsen, J.E.Ø. and F. Nilsen (2007). The Atlantic Water Flow along the Vøring Plateau: Detecting Frontal Structures in Oceanic Station Time Series. Deep Sea Research Part I, 54(3), p , doi: /j.dsr
Nilsen, J.E.Ø., H. Hátún, K. A. Mork and H. Valdimarsson (2008). The NISE Dataset. Tech. Rep Faroese Fisheries Laboratory, Box 3051, Tórshavn, Faroe Islands. 17 pp
Nilsen, J.E.Ø. (2016). The North Atlantic and Nordic Seas hydrography collection. NERSC Technical Report no. 372.
Raj, R. P., J. A. Johannessen, J. E. Ø. Nilsen, O.B. Andersen (2015). Quantifying variability of the surface currents in the Norwegian Sea: Estimation based on different gravity models and mean sea surface datasets. The 36th International Symposium on Remote Sensing of Environment, 11 – 15 May 2015, Berlin, Germany, ISRSE
Raj, R. P., J. A. Johannessen, T. Eldevik, J. E. Ø. Nilsen, and I. Halo (2016). Quantifying mesoscale eddies in the Lofoten Basin, J. Geophys. Res. Oceans, 121, doi: /2016JC
Raj, R. P., L. Chafik, J. E. Ø. Nilsen, T. Eldevik, and I. Halo (2015a). The Lofoten vortex of the Nordic Seas, Deep Sea Res., Part I, 96, doi: /j.dsr
Sandø, A. B., J. E. Ø. Nilsen, T. Eldevik, and M. Bentsen (2012), Mechanisms for variable North Atlantic–Nordic seas exchanges, J. Geophys. Res., 117, C12006, doi: /2012JC
Hansen MW, Collard F, Dagestad K-F, Johannessen JA, Fabry P, Chapron B. Retrieval of Sea Surface Range Doppler Velocities from Envisat ASAR Doppler Centroid Measurements. IEEE Transactions on Geoscience and Remote Sensing. 2011;49:11. 
Hansen MW, Johannessen JA, Dagestad K-F, Collard F, Chapron B. Monitoring the surface inflow of Atlantic Water to the Norwegian Sea using Envisat ASAR. Journal of Geophysical Research. 2011;116:13. 
Hansen MW. Sea Surface Range Doppler Retrievals from Synthetic Aperture Radar