1. Gerard McCarthy and David Smeed National Oceanography Centre
RAPID/MOCHA/WBTS
RESULTS FROM THE 26ºN MOORING ARRAY
Gerard McCarthy and David Smeed
National Oceanography Centre UK
Thank you. It gives me great pleasure to kick off this conference with a review of the results from the 26ºN AMOC monitoring array. This moored array is a trans-Atlantic collaboration between NERC in the UK, which funds the RAPID program, NSF, which funds the MOCHA heat transport array and NOAA which maintains the Florida Current timeseries under the WBTS program. I'd like to thank my co-authors and collaborators on both sides of the Atlantic many of whose work I will present here today.
Molly Baringer, Adam Blaker, Harry Bryden, Julie Collins, Stuart Cunningham, Aurélie Duchez, Eleanor Frajka-Williams, Joel Hirschi, Will Hobbs, Bill Johns, Chris Meinen, Matt Palmer, Darren Rayner, Chris Roberts

2. INTRODUCTION

3. INTRODUCTION Why we study the AMOC: Impact on climate
Evidence of major shifts in the past associated with major climate events
Rahmstorf, S. and A. Ganopolski, Long-term global warming scenarios computed with an efficient coupled climate model. Climatic Change, : p
from Rahmstorf, S. and A. Ganopolski, Long-term global warming scenarios computed with an efficient coupled climate model. Climatic Change, : p

4. The AMOC in a changing climate
The AMOC is predicted to slow in a warming climate
Bryden, H. L., H. Longworth and S. Cunningham (2005), Slowing of the Atlantic meridional overturning at 25N, 438, Nature 30 20 10
Suggestions of a slowdown:
To understand how the AMOC varies, we need to measure it…

5. The impact of the 26ºN measurements
Mean [Sv] GS
31.8±3.1
MOC
18.1±4.3
Ekman
2.9±3.0
UMO
-16.6±3.4
Reference Cunningham 2007
AMOC timeseries and related data products are available from
Data from individual instruments are available from

6. The impact of the 26ºN measurements
Mean [Sv] GS
31.8±3.1
MOC
18.1±4.3
Ekman
2.9±3.0
UMO
-16.6±3.4
’57 ’81 ’92 ’98 ‘04
Bryden, H. L., H. Longworth and S. Cunningham (2005), Slowing of the Atlantic meridional overturning at 25N, 438, Nature
AMOC timeseries and related data products are available from
Data from individual instruments are available from

7. Seasonal Cycle Seas. Amp. GS 3 Sv MOC 6 Sv Ekman UMO 6.5 Sv
Just show the large seasonal cycle;
7 Sv
Driven by Mid-ocean contributions
Particularly the eastern boundary contribution from wind stress curl
AMOC timeseries and related data products are available from
Data from individual instruments are available from

8. Heat Transport Mean [Sv] GS 31.8±3.1 MOC 18.1±4.3 Ekman 2.9±3.0 UMO
-16.6±3.4
Heat transported north in GS is recirculated by mid-ocean and overturning circulation
Overall MHT of 1.3 PW similar to hydrographic estimates
AMOC timeseries and related data products are available from
Data from individual instruments are available from

9. Interannual Variability
Double Dip
Major Downturn
Continued increase?
Update this figure
Would you expect this type of variability? Extreme events: Gaussian etc.
Large downturn in 2009/10 driven by short term Ekman variability (3 months) and long term strengthening of the gyre/UMO (18 months)
Double dip in winter 2010/11
ERA winds is a difference from previous release

10. How we measure the AMOC at 26.5°N?
Previous estimates of OHT transport have been derived from hydrographic sections. At 26.5°N these have given estimates between 1.1 to 1.4 PW with an uncertainty of about 0.3PW.
Variability of ocean heat transport is not well known but obviously is an important factor in long-term climate variability. So one goal of the RAPID array is to provide continuous estimates of the meridional heat flux.

11. Boundary Currents and the mid-ocean Dynamic Height and Bottom Pressure Array
Johns, W. E., L. M. Beal, M. O. Baringer, J. Molina, D. Rayner, S. A. Cunningham, and T. O. Kanzow (2008), Variability of shallow and deep western boundary currents off the Bahamas during : First results from the 26°N RAPID-MOC array, J. Phys. Oceanog., 38(3),
Principle of mid-ocean array is to calculate the profile of geostrophic velocity from boundary dynamic height moorings. WB : moorings WB2, WBH2, moorings on either side of the MAR to account for any deep flows on either side of the ridge; WB a series of moorings crawling up the slope. AABW. In addition to dynamic ht moorings we have an array of bottom pressure recorders measuring.
Describe plot: Shows the annual mean meridional component of velocity measured by current meters for the period March 2004 to May 2005.
Antilles current, part of the subtropical circulation which remains outside of the Carribean. It is a subsurface intensified current.
Below about 900m the flow is southward. This is the inshore edge of the DWBC. However flow is weak in the inshore region because a submarine ridge to the north deflects the DWBC offshore.
Rayner, D., et al. (2011), Monitoring the Atlantic Meridional Overturning Circulation, Deep Sea Research II, in press.

12. The AMOC Streamfunction
red dots
LHS: Equation ; transport-per-unit-depth for each of the components
RHS : The MOC is taken as the maximum of the overturning stream function. It is the maximum northward transport shallower than 1025m.
Internal Transport:
The AMOC:

13. Mass compensation
BPR transport (black, dashed) and Mass compensation (red)
Bottom Pressure records at WB2;
Show the location of the measurements; Update
Observed Flow Compensation Associated with the MOC at 26.5°N in the Atlantic
Link with a bathymetry figure…
Mass compensation is highly correlated with BPR derived transport (r > 0.8) at the western boundary
Residual has seasonal signal indicating seasonal compensation at the east
Kanzow, T., et al. (2007), Observed Flow Compensation Associated with the MOC at 26.5°N in the Atlantic, Science

14. Basinwide transports in an eddying ocean
This plot shows dynamic height variability from moorings (crosses) and SSH variability (lines) from satelite measurements. The moorings are 25, 40 & 500km offshore from the Bahamas. The satellite periods are black for a long time series (green is it’s low pass equiv). The red and blue dashed color coded to match the mooring deployments.
We have the expected high variability offshore but near the coast there is an abrupt drop in variability which is captured by both the moorings and the satellite.
RAPID measures a variability of 3.5 Sv in UMO rather than a 16 Sv variability that would exist if we were measuring transports based on an eddying ocean (Wunsch et al., NGS)
SSH variability decreases towards close to the western boundary
We measure a small variance of 4 Sv whereas an eddy dominated signal would display a variance of around 16 Sv
Kanzow, T., H. Johnson, D. Marshall, S. A. Cunningham, J. J.-M. Hirschi, A. Mujahid, H. L. Bryden, and W. E. Johns (2009), Basin-wide integrated volume transports in an eddy-filled ocean, J. Phys. Oceanog., 39(12), 3091–3110.

15. AMOC with time-varying AABW (red) and non-varying AABW (black)
Recent Changes
Impact of time varying AABW
Moving from EOS-80 to TEOS-10
Improved extrapolation above shallowest measurement
AMOC with time-varying AABW (red) and non-varying AABW (black)
Little impact using time-varying AABW
Formerly, 2 Sv of AABW was included
Now, 1 Sv more representative of both northward flowing AABW and southward flowing DWBC below 4820 dbar, west of 72ºW
Increases AMOC by 0.2 Sv
Difference
Shear extrapolation above the shallowest instrument
TEOS-10
Frajka-Williams et al., (2011) Variability of Antarctic Bottom Water at 24.5°N in the Atlantic. JGR

16. Recent Changes Impact of time varying AABW
Moving from EOS-80 to TEOS-10
Improved extrapolation above shallowest measurement
Change of equation of state decreases the AMOC by 0.4 Sv
Change due to absolute salinity difference in deep east and west
Maximum change of 0.7 Sv at 2700 m x
Shear extrapolation above the shallowest instrument
TEOS-10
The composition of Standard Seawater and the definition of the Reference-Composition Salinity Scale
Millero et al., (2008). The composition of Standard Seawater and the definition of the Reference-Composition Salinity Scale
DSR1

17. Recent Changes Impact of time varying AABW
Ben Moat
Poster Session 1
Impact of time varying AABW
Moving from EOS-80 to TEOS-10
Improved extrapolation above shallowest measurement
Old (blue) and new (red) methods
Shallowest measurement often at 100 m depth
Linear extrapolation can miss 1 Sv in the summer
Argo based monthly extrapolation reduces to <0.2 Sv
Haines et al., (2013). Atlantic meridional heat transports in two ocean reanalyses evaluated against the RAPID array,
GRL
Transport Anomaly (Sv)
Shear extrapolation above the shallowest instrument
TEOS-10
Total reduced by 0.5 Sv
Summary: 0.2 Sv AABW change, -0.4 Sv TEOS-10 and improved seasonal extrapolation

18. SEASONAL CYCLE
Seasonal variability
Stationarity
Rossby Wave model

19. The Seasonal Cycle due to the East
±1.8 Sv,SE=1.0 Sv
±2.6 Sv, SE=0.5 Sv
SD 3.5 Sv, Range 7.0 Sv
SD 3.1 Sv, Range 6.2 Sv
Kanzow, T., et al. (2010), Seasonal variability of the Atlantic meridional overturning circulation at 26.5°N, J. Clim., 23(21), doi: /2010JCLI
TOP: Explain how isolate WB/EB contributions to MOC (fixed GS, EK and boundaries respecitvely)
Isolation of seasonal cycles of western and eastern boundary to MOC. Done by keeping the boundary density (alternatly east and west) and, ekman and gs fixed.
EB has larger and better defined seasonal cycle than the WB, so is a more important contribution to the seasonal cycle in the UMO and MOC than the WB.
Expectation is that the seasonal cycle is mainly in Ekman; Rossby waves going around the Bahamas.
Largest seasonal influence in mid-ocean transports due to the east
Recent variability hasn’t changed this

20. Wind Stress Curl drives density anomalies at the Eastern Boundary
Study density variability at the Eastern Boundary - Maria-Paz Chidichimo at Max Planck in Hamburg - using the merged density profile from moorings at the eastern boundary. Crosses are depth of a microcat CTD.
LHS : Density anomalies at the eastern boundary as a function of month and depth. The error bar is the variability in the monthly mean anomaly based on 4-years of data. Where density anomaly is positive isopycnals have been uplifted.
Strongly suspect that this seasonal cycle in density anomalies is a dynamical response to seasonal variations in wind stress curl at the eastern boundary and not by surface buoyancy forcing because of the depth over which the density variability extend.
In fact, if you revisit the hydrographic estimates and remove the seasonal cycle, then any decrease is reduced substantially.
How stationary is this seasonal cycle? Check long timeseries of WSC. At least as long as the Canary Islands exist…
Chidichimo, M. P., T. Kanzow, S. A. Cunningham, and J. Marotzke, 2010: The contribution of eastern-boundary density variations to the Atlantic meridional overturning circulation at 26.5°N. Ocean Science, 6,

21. HEAT TRANSPORT
Seasonal variability
Heat transport

22. HEAT TRANSPORT Net Heat Flux = 1.27 ± 0.30 PW (uncertainty 0.14 PW)
Overall MHT of 1.3 PW similar to hydrographic estimates
Seasonal variability is in the mid-ocean heat transport
47% variance in Ekman
Seasonal variability
Heat transport
Johns, W. et al. (2011), Continuous, Array-based Estimates of Atlantic Heat Transport at 26.5°N, J. Clim., 24, pp. 2429–2449.

23. HEAT TRANSPORT
Heat transported north in GS is recirculated by mid-ocean and overturning circulation
90% is in the overturning
Seasonal variability
Heat transport

24. INTERANNUAL VARIBILITY
Event Winter 2009/10
Heat content, budget, North Atlantic tripole
Double dip, reemergence, seasonal predictability

25. Slowdown in winter 2009/10
Ekman shift associated with negative Arctic Oscillation/NAO Only explains 3 month downturn
Full 10 day MOC and Ekman timeseries
4 month smoothed AO
Arctic Oscillation
2 s

26. Slowdown in winter 2009/10
Longer time scale changes: 18 month weakening of MOC
Anomalously southward UMO: shift from overturning to gyre circulation
A 5.5 Sv downturn persisted for 1.25 years
*Seasonal cycle was removed, and data smoothed with 180-day filter
McCarthy, G., et al. (2012), Observed Interannual Variability of the Atlantic Meridional Overturning Circulation at 26.5N, Geo. Res. Lett.

27. Expect for a balanced heat budget
Implications for Heat Content
Expect for a balanced heat budget
the thick pink curve (total predicted heat content change) would match the black curve (observed heat content change)
Cunningham talk
Today 14.30
Let’s add a map
Cunningham, Frajka-Williams, Roberts, Palmer et al., in prep

28. Double Dip: Winter 2010/11 Double Dip: Winter 2010/11 AMOC
Arctic Oscillation
Second large dip in AMOC transport in winter 2010/11 following Arctic oscillation low
This is largely explained by Ekman contributions
Ocean re-emergence of SST links the two events
Taws SL, Marsh R, Wells NC, Hirschi JJM (2011) Re-emerging ocean temperature anomalies in late-2010 associated with a repeat negative NAO. GRL

29. Historical Analogues Double Dip: Winter 2010/11 MOC Ek AO Blaker talk
Linked with positive and negative phases of the NAO
In an ensemble of NEMO runs, double dips of MOC have occurred previously in 1969/70 and 1978/79
Extreme negative Arctic Oscillation (AO) correspond with double dip analogues
Corresponds with Ekman lows
Blaker talk
Today 14.50
Blaker et al. in prep, Historical analogues of the recent extreme minima observed in the Atlantic meridional overturning circulation at 26◦N. Clim. Dyn.

30. Double Dip: Winter 2010/11 Double Dip: Winter 2010/11 Re-emerging SSTs
Reemerging SSTs are observed in 1969/70 as well as in 2010/11
These were conducive to the development of the negative NAO in winter 2010
Buchan et al. submitted, North Atlantic SST anomalies and the cold north European weather events of winter 2009/10 and December submitted to Monthly Weather Review
Evidence that this second negative is predictable due to the ocean
Winters of 2010/11 (green), 1969/70 (blue) and 1978/79 (red). Black shows mean ( ) with 1, 2 std envelopes
One-year pattern correlation function for North Atlantic SSTAs (5-65N, 80-10W) from March March 2011 (green), March 1969-March 1970 (blue), March 1978-March 1979 (red), and a 50-year ( ) average (black). Light (Dark) shading denotes the range of correlation within one (two) standard devi- ation(s) of the long-term mean signal.
Update with the Maidens reference
Winter of 2011, economic impact on UK. In the runup to Christmas estimated cost of 1.2 bn a day, culminating in a 13 bn loss.
: The first of these 2 winters saw snow in late December, around the New Year, in Eastern Scotland and England. Eastern Yorkshire saw a massive 16 inches! Mid February saw more snow, this time more to the West, with England and Wales seeing the most. Mid March saw more in the Pennines, and a TV mast fell down saw snow for Northern England, North Wales, and Scotland in mid November. Mid December saw snow for the North again. Mid February, most parts, and early March, snow in Wales and England, with the Midlands getting 12 inches.
the temperature remained at a low level all the time. and no matter which weather station you watch - from paris to stockholm on to warsaw. persistent snow cover. April and May were cold. Average temperature in spring in Hanover just 6.7 ° c. The December 1969 was in contrast to this winter also been very cold.
: Mid January, 6 foot drifts! A week later, and 4 inches fell. Mid February saw 4 inches also. Late January, heavy snow in Scotland, drifting, 28 inches falling in parts! Mid February (see above) was very snowy in the North East, East and South West. February 11th had 1 ft in Durham and Edinburgh. Feb th South West England, blizzard with huge drifts, sounds like my cup of tea! : The last really severe, snowy winter, for now anyway, and one my parents go on about! Late December falls of 6-7n inches in Southern Scotland and the North East started it off. It was very cold in parts. Mid February saw drifts of 6-7 feet on the East coast of England. Mid March had severe blizzards and drifting, in North Eastern England drifts reached a staggering 15 feet! Very snowy.
Maidens et al. in prep, The Influence of Surface Forcings on Prediction of the North Atlantic Oscillation Regime of Winter submitted to Monthly Weather Review

31. Evidence of a downturn IPCC predicts a downturn of 0.5 Sv per decade
We see a decline of 0.6 Sv per year
Even excluding the extreme of 2009, this is significant at 90% level

32. CONCLUSIONS
The 26ºN array has revolutionised our understanding of the AMOC and its variability
The variability observed in the first year gave context to the 5 previous hydrographic estimates and painted a picture of a noisy AMOC
A 30% decline in a single year 2009/10 including the brief cessation of northward flowing warm water
Re-emergent patterns driving a double dip in 2010/11
IPCC projects a 0.5 Sv per decade decline this century, we are seeing a decline of 0.6 Sv per year. Is this a long term decline? The only way to know is to measure.
The array should not only be seen as a long-term monitoring tool e.g. the potential to improve seasonal forecasts.
- sub-annual
- seasonal
- interannual
Where to next? 10 year timeseries; Zonal separation of the components; Seasonal forecasts.

33. End

34. The research leading to these results has received funding from the European Union 7th Framework Programme (FP ), under grant agreement n NACLIM