Continuous observations in the North Atlantic challenges current view about ocean circulation variability

Kevin Kilty

May 10, 2019

clip_image002Figure 1: Transect of the North Atlantic basins showing color coded salinity, and gray vertical lines showing mooring locations of OSNAP sensor arrays. (Figure from OSNAP Configuration page)
Figure 1: Transect of the North Atlantic basins showing color coded salinity, and gray vertical lines showing mooring locations of OSNAP sensor arrays. (Figure from OSNAP Configuration page)

From Physics Today (April 2019 Issue, p. 19)1:

The overturning of water in the North Atlantic depends on meridional overturning circulation (MOC) wherein warm surface waters in the tropical Atlantic move to higher latitudes losing heat and moisture to the atmosphere along the way. In the North Atlantic and Arctic this water, now saline and cold, sinks to produce north Atlantic Deep water (NADW). It completes its circulation by flowing back toward the tropics or into other ocean basins at depth, and then subsequently upwelling through a variety of mechanisms. The time scale of this overturning is 600 years or so2.

The MOC transports large amounts of heat from the tropics toward the poles, and is thought to be responsible for the relatively mild climate of northern Europe. The heat being transferred from the ocean surface back into the atmosphere at high latitudes is as large as 50W/m2, which is roughly equivalent to solar radiation reaching the surface at high latitudes during winter months2.

In order to evaluate models of ocean overturning oceanographers have relied upon hydrographic research cruises. But the time increment between successive cruises is often long, and infrequent sampling cannot measure long term trends reliably nor gauge current ocean dynamics.

To get a better handle on MOC behavior an array of sensors to continuously monitor temperature, salinity, and velocity measurements known as the Overturning in the Subpolar North Atlantic Program (OSNAP) was recently deployed across the region at multiple depths. Figure 1 shows sensor moorings in relation to the various ocean basins of the North Atlantic. Figure 2 shows data from the first 21 months of operation, and displays a rather large variability of overturning in the eastern North Atlantic between Greenland and Scotland that reaches +/-10 Sverdrup (Sv=one million cubic meters per second) monthly, and amounts to one-half the MOC’s total annual transport. Researchers had thought that such variability was only possible on time scales of decades or longer.

Figure 2: Twenty-one months of observational data showing large month to month variation in MOC flows.
Figure 2: Twenty-one months of observational data showing large month to month variation in MOC flows.

The original experimental design for sensor placement in OSNAP was predicated on much smaller variability of a few Sv per month3. The report does not address what impact this surprising level of transport variability has on validity of the experiment design; but the surprisingly large variations in flow challenge expectations derived from climate models regarding the relative amount of overturning between the Labrador Sea and the gateway to the Arctic between Greenland and Scotland.

As one oceanographer put it, the process of deep water formation and sinking of the MOC is more complex than people believed, and these results should prepare people to modify their ideas about how the oceans work. This improved data should not only help test and improve climate models, but also produce more realistic estimates of CO2 uptake and storage.


1. Alex Lopatka, Altantic water carried northward sinks farther east of previous estimates, Physics Today, 72, 4, 19(2019).

2. J. Robert Toggweiler, The Ocean’s Overturning Circulation, Physics Today, 47, 11, 45(1994).

3. Susan Lozier, Bill Johns, Fiamma Straneo, and Amy Bower, Workshop for the Design of a Subpolar North Atlantic Observing System, URL=, accessed 05/10/2019.

0 0 vote
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
Rod Evans
May 11, 2019 2:13 am

“This improved data should not only help test and improve climate models, but also produce more realistic estimates of CO2 uptake and storage”.
Here is my prediction .
If the data can be used to further complicate the story advanced as Man Made Climate Change, then the modellers will take this study on board.
If it in some way counteracts the blind acceptance that CO2 drives climate, then the modellers will continue to ignore real time data, as it does not advance their imagined climate outcome.

Reply to  Rod Evans
May 11, 2019 5:06 am

What ? Does this mean that the ” death spiral ” of Arctic sea ice is not totally caused by CO2 after all??

Maybe that is why Arctic sea ice is at the same level as it was in 2007 when everyone started screaming that it was finished, ice free Arctic etc.

Now they will have to spin this to explain this is why it has not reduced as expected while trying to avoid talking about the possibility that this is why it was melting in the first place.

Reply to  Rod Evans
May 11, 2019 6:35 am

Well, if it isn’t about CO2, why do it?

Bloke down the pub
May 11, 2019 2:31 am

Good to know the science is settled then.

May 11, 2019 2:46 am

“The heat being transferred from the ocean surface back into the atmosphere at high latitudes is as large as 50W/m2, which is roughly equivalent to solar radiation reaching the surface at high latitudes during winter months.”
Amount of the heat lost depends on the wind strength in the winter months using the atmospheric pressure as a rough guide. More heath is taken out, deeper is down-welling with somewhat reduced velocity of the return current
One part of the return current up-wells some 15 or so years in the North Atlantic’s tropics along the west coast of North Africa, the area where the N. Atlantic hurricanes are initiated. Greater the temperature differential between the up-welling and the surface currents in the area more likely is that a hurricane will be spawned.
Consequently, the temperature and arrival time of the cold up-welling current reflects the heat lost in the down-welling process which, as mentioned above, is related to the wind intensity (or the atmospheric pressure) in the region to the south of the Denmark straits.
It follows that the atmospheric pressure anomaly in the far N. Atlantic can be used as a forward estimate (or a prediction of a reasonable probability) for the ACE (Accumulated Cyclone Energy) intensity as you can see here
which to me (and hopefully to anyone else who might consider the above as credible) is known as ‘vukcevic hypothesis’ and therefore thank you for your attention.

Reply to  vukcevic
May 11, 2019 6:41 am

Interesting hypothesis. Atmospheric pressures over the tropical Atlantic and Pacific basins seem to very much determine the Atlantic hurricane season’s overall activity along with the location of the warm pools. You also have the influences of El Niño, La Niña and La Nada and where the highs and lows over the basins create or don’t create wind shear in the tropics.

I think now with a better understanding of all this, we are able to generate more accurate general forecasts, but unfortunately even if you have a slight chance of getting hit this year, it can still happen.

Rhoda R
Reply to  rbabcock
May 12, 2019 6:39 pm

Andrew was the only hurricane to make landfall in that year.

Joel O’Bryan
Reply to  vukcevic
May 11, 2019 8:14 am

The more heat that is extracted the faster a parcel of water will sink, speeding up the current. Basic thermodynamics.

Reply to  Joel O’Bryan
May 11, 2019 10:54 am

Hi J O’B
Thanks for your observation, as a consequence in order to avoid any misunderstanding, in the future I may include following paragraph:
‘Sinking current velocity has short term (in time and distance) vertical component directly proportional to its salinity. It also has by far much longer term (in time and distance) horizontal component that is inversely proportional to the depth at which current flows.’

Any further observations or criticism from any of commentators are welcome.
Dr. Curry considered that the hypothesis credibility would be enhanced by further years of observations (shown in green) which suggests that trend of the prediction is in a general agreement with trend of the observed events. This is not meant to be only, but one of number of tools to be taken into account for prediction of a decadal ACE trends.

May 11, 2019 2:54 am

So in other words, ocean overturning is much more complex than they thought but somehow they can make more accurate estimates of CO2 uptake? And why is it that any natural science article has to genuflect to the god of climate change, if only in passing?

Dodgy Geezer
Reply to  GeoNC
May 11, 2019 3:18 am

because if you don’t you
a – won’t get a grant
b- will get a reputation as a ‘denier’
c – will find that your colleagues desert you
d – will be subject to vague allegations of impropriety or ethical misdemeanour
e – will then lose your job….

Kurt Linton
May 11, 2019 3:04 am

“But the time increment between successive cruises is often long, and infrequent sampling cannot measure long term trends reliably nor gauge current ocean dynamics.” Solution: more cruises. Call Carnival. Concerned taxpayers will happily pay. Maybe a “Climate Love Boat” film/series (starring Leonardo DiCaprio as the tragic hero) could help defray the cost.

Joel O’Bryan
May 11, 2019 3:16 am

More evidence that the GCMs are not just wrong, but they producing negative learning for the entire climate science community.

Negative learning is the interference of previous learning on adapting to new learning and/or integrating new ideas to replace outdated information. This is likely the main reason new, younger entrants to a field of science can produce new ideas and theories where older scientists have become intellectually stuck or stalled unable to break through decades of “negative learning.”

Couple this 1) negative learning with 2) enforcement of consensus Groupthink by 3) rampant gatekeeping and we have before us the unholy trinity creating the global junk science enveloping the field of climate science.

Ron Long
May 11, 2019 3:38 am

Interesting scientific report and great, to-the-point comments. Why were scientists surprised that adding real data into a topic tends to produce results different from earlier models? Because the earlier models were politically correct and supported the AGW doomsday crowd? Are we waiting with great anticipation for the climate models to now be changed and show less importance of CO2? I’m not, I’m going to get a drink.

May 11, 2019 4:25 am

a few Sv per month^3

A few Sv per cubic month?

I need coffee.

Joel O’Bryan
Reply to  commieBob
May 11, 2019 5:27 am

Reference #3.

Reply to  Joel O’Bryan
May 11, 2019 6:22 am

The reference number goes after the period. link

May 11, 2019 4:47 am

Oh look! The December low in 2016 was not as low as that in 2015 – the fingerprint of global warming! (Sarc)

Joel O’Bryan
Reply to  Susan
May 11, 2019 5:32 am

drawing a linear trend would also likely show the MOC reaching zero in abor 15 years. The Day After Tomorrow scenario, becase ….science.

Tom Abbott
May 11, 2019 5:31 am

This is more evidence that “The Science” (Climate Science) is not settled.

I dare say there is more that we don’t know about the Earth’s climate than we do know. We shouldn’t waste Trillion of dollars on trying to change the Earth’s climate until we have a little better handle on everything involved.

CAGW (Catastrophic Anthropogenic Global Warming) is more a fixation than science. Those fixated on it are those who promote it for personal and/or political gain, and those who have been duped by the lies of the CAGW promoters.

Reply to  Tom Abbott
May 11, 2019 6:18 am

‘I dare say there is more that we don’t know about the Earth’s climate than we do know. ‘

When I first started writing articles about AGW some 12 years ago I likened our knowledge of climate change to an expedition to Everest and at the time I thought the ‘mountaineers’ were basically in katmandhu and had a long arduous trek in front of them before they could mount a final assault on the mountain.

Now I would say they have just met by chance in a pub in London and after a few drinks and a packet of peanuts are talking vaguely about mounting an expedition to Everest ‘one day’ and that everyone ought to meet up in Starbucks for a coffee and a chat the next time they are all up in London so please exchange telephone numbers.


Brooks Hurd
Reply to  tonyb
May 11, 2019 7:55 am

I like your analogy.

Reply to  Tom Abbott
May 11, 2019 8:44 am

Yes, of course.

I expect that what we DON’T know about how the earth’s climate regulation system works is orders of magnitude greater than what we DO know today.

Today’s climate researchers are much akin to the old Asian parable of the blind men examining an elephant with their fingers, and coming up with very different and yet equally wrong conclusions as to what it was they were sensing.

““Hey, the elephant is a pillar,” said the first man who touched his leg.

“Oh, no! it is like a rope,” argued the second after touching the tail.

“Oh, no! it is like a thick branch of a tree,” the third man spouted after touching the trunk.

“It is like a big hand fan” said the fourth man feeling the ear.

“It is like a huge wall,” sounded the fifth man who groped the belly .

“It is like a solid pipe,” Said the sixth man with the tuskin his hand.

They all fell into heated argument as to who was right in describing the big beast, all sticking to their own perception. A wise sage happened to hear the argument, stopped and asked them “What is the matter?” They said, “We cannot agree to what the elephant is like.”

The wise man then calmly said, “Each one of you is correct; and each one of you is wrong. Because each one of you had only touched a part of the elephant’s body. Thus you only have a partial view of the animal. If you put your partial views together, you will get an idea of what an elephant looks like.””

Izaak Walton
Reply to  Duane
May 11, 2019 5:26 pm

Do you have any evidence or reasoning to back up your claim about our lack of understanding
of the “earth’s climate regulation system” (whatever that actually means). Currently the best
weather models can accurately predict the weather up to about a week in advance — suggesting
that on short time scales there are no unknown physical effects that are not included in the model.
And on longer scales global climate models can predict the global temperature to within 0.1%
suggesting again that climate models are about 99.9% accurate at least when it comes to temperature. Adding in the Milankovitch forcing allows models to predict ice ages etc. etc. So there would appear to be no grounds to suggest that there are large gaps in our knowledge of the climate as you suggest.

Reply to  Izaak Walton
May 11, 2019 7:16 pm

I watch the daily predictions for the next week and see how they march forward day by day. Generally speaking, a rainfall or temperature prediction appears at 7 days out and advances unchanged to three days out then changes again but less frequently at one day out. They only change two days out of seven, that is pretty good, eh?
Accurate a week in advance? Not around here.
“And on longer scales global climate models can predict the global temperature to within 0.1%
suggesting again that climate models are about 99.9% accurate”
Nicely phrased babble.

Dudley Horscroft
Reply to  Izaak Walton
May 12, 2019 1:31 am

For a relevant description you should read “The Ascent of Rum Doodle”.

An accurate description of mountaineering by “Climate Scientists”!

Dudley Horscroft
Reply to  Izaak Walton
May 12, 2019 1:45 am

Izaak Walton May 11, 2019 at 5:26 pm said:-

“Currently the best weather models can accurately predict the weather up to about a week in advance — suggesting that on short time scales there are no unknown physical effects that are not included in the model.”

This is only a slight improvement for weather forecasting in the UK, which says that if the pressure is rising and it is summer you will have warm sunny days for the next week, and if it is winter you will have cold, possibly frosty days for the next week. But is the pressure is falling you have a 50% chance of rain, and a 25% chance of gales or strong winds for the next week.

“And on longer scales global climate models can predict the global temperature to within 0.1% suggesting again that climate models are about 99.9% accurate at least when it comes to temperature.”

But there is just one trouble with that predicted temperature, while the predicted temperature is within 0.1%, the actual temperature after 30 years differs by +- 1.0 K.

Obviously the actual temperatures must be wrong (sarc).

Of course, if you are in Bombay, you can predict that there will be a 1% chance of rain within the next week in January to April, but in June to September you can predict 100% chance of rain – and heavy rain a that – in the next week. And you will be right. Beats anything the Bureau of Meteorology can do!

Reply to  Izaak Walton
May 12, 2019 8:22 am

First of all, I wrote regarding the earth’s climate regulation system – i.e., the combination of everything, from geological to astrophysical to biological and all the other scientific engines that produce what we call “climate”.

I was not referring to “weather” which has nothing to do with climate. Weather is the result of climate .. it is not climate. Weather what you experience today, or next week in a specific area or location at a specific time. Climate is what controls weather collectively across the planet over geologic timeframes.

All of the existing climate models fail totally to predict anything, including past known climate data. That’s been well known for decades. That is why the modelers have to “hide the decline” and massage historical data until it bears no resemblance to reality in order to make their models look less ridiculously feckless in even predicting the past, let alone the future,.

It is an easy bet, however, to say that what is known -even that which is correctly known, as opposed to much of “climate science” that is baldfacedly incorrect (such as, dying polar bears, extinction crises, etc. etc.) – is vastly exceeded by what is not yet known, or understood.

Anybody who claims “the science is settled” is by definition not a scientist, but a polemicist and a politician. All real scientists know that real science is never settled. Even age old assumptions and “knowledge” is constantly proven either wrong altogether, or is misleading and incomplete at best.

After all, 66 years ago nobody knew the slightest thing about the mechanisms of genetics, until the double helix was finally discovered by Watson and Crick. And it was many decades beyond that discovery before the first sequencing of the human genome was accomplished .. and that based upon both discoveries, human medical science practitioners finally gained little more than a “learners permit” on how human DNA functions to affect everything from birth defects to disease treatment.

Or consider the existence of exo-planet. Asrtonomers and astro-physicists argued for centuries, indeed thousands of years, over whether any planets existed outside of our solar system. Yet it was not until 1988 that the first exo planet was actually found .. and it was not until 2012 that it was actually confirmed that such was indeed an exo planet. And since then, astronomers have sensed and confirmed the existence of more than 4,000 exo planets, whereas less than 7 years ago, none had ever been confirmed to exist.

I laugh at anyone who has the hubris to state that “the science is settled” on any scientific subject. All science is about the search for scientific truths, and about constantly challenging what we assume to be truth by skeptical questioning.

Another Paul
Reply to  Izaak Walton
May 16, 2019 4:17 am

Izaak says: “Do you have any evidence or reasoning to back up your claim about our lack of understanding…”

I have one simple questions, what’s the driver of ENSO. We know what it looks like, we know the effects of it, but it appears we do not have a clue as to why it happens. If we can’t explain (or predict) a short term climate event, how can any model of climate be correct? That says we might have “large gaps in our knowledge of the climate” loud and clear.

May 11, 2019 5:43 am

The publication by Lozier et al in February was more forceful:

“Last week, at the American Geophysical Union’s (AGU’s) Ocean Sciences meeting here, scientists presented the first data from an array of instruments moored in the subpolar North Atlantic. The observations reveal unexpected eddies and strong variability in the AMOC currents. They also show that the currents east of Greenland contribute the most to the total AMOC flow. Climate models, on the other hand, have emphasized the currents west of Greenland in the Labrador Sea. “We’re showing the shortcomings of climate models,” says Susan Lozier, a physical oceanographer at Duke University in Durham, North Carolina, who leads the $35-million, seven-nation project known as the Overturning in the Subpolar North Atlantic Program (OSNAP).”

Paper at

My synopsis

May 11, 2019 5:43 am

This is wrong. As someone at these pages has already mentioned, the currents are primarily driven by the wind. MOD is a secondary effect at best.'s_Mild_Climate

May 11, 2019 5:48 am

The superscripted number three is a footnote/citation reference. (A very poor way to arrange them.)

May 11, 2019 5:54 am

It is amazing how complex and variable Nature on Earth is and yet here on Earth things are so relatively stable over thousands of years.

Ron Long
Reply to  nicholas tesdorf
May 11, 2019 6:45 am

Nicholas, I can attest to “how complex and variable Nature on Earth is” from personal interaction. I am a geologist and rocks are very patient, you can pound on them with a hammer, scratch them with a knife, pour acid on them, cut them into slices and study them under a microscope or scanning electron microscope, and they don’t move or resist. Still, only about (last number I saw) 11% of mining exploration geologists are main actors in discovering a mineral deposit. This is with a substantial monetary reward for those who participated in the discovery. Complex? Without doubt! But, can you study the issue and come to correct conclusions? Absolutely, but not with political correctness hung around your neck.

Stephen Rasey
May 11, 2019 7:14 am

The Nyquist Theorem at work again.
They have spatially undersampled a field known to have vortices.

Henry Galt
May 11, 2019 7:31 am

All the weather observations in the ICOADS dataset (version3), over the period 1800-2014.

At any given moment, observations from a 72-hour period, centered on noon of the given day, are shown:-

Highlights for me? WWI and II. And … the array being placed in the Pacific from the end of the Eighties onward.

More detailed here:-

Stephen Rasey
Reply to  Henry Galt
May 11, 2019 8:57 am

A marvelous presentation of the dataset. For me, the development of the Tea Trade in the 1820’s, the whaling industry in the 1830-1855. The rise of San Francisco (Gold Rush). The US Civil War Blockade. The Building the Suez. The Rise of Japan. Panama Canal was more subtle than I figured. The rise of Seattle as a port. Yes, the Island war march in the Pacific (1942-1945). What I gather to be the fishing fleets of the North Pacific. Fascinating. Thank you.

May 11, 2019 7:35 am

The climate and its many subsystems are so complex that the idea that we can model it and get meaningful information from the models is silly. Climate science is a joke for any experimental science practitioner.

Reading the bibliography on oceanic currents it soon becomes clear that the complexity is amazing and those cartoons with arrows of the conveyor belt are a lie. The bottom water in the North Atlantic moves South at different speeds from area to area and from time to time and it is quite heterogeneous in its salinity and temperature. The Mediterranean contributes a lot of highly saline warm water that sinks and partly mixes with the North Atlantic bottom water contributing to its heterogeneity.

Andy Pattullo
May 11, 2019 7:54 am

Not sure I have enough information from the article to answer this question. The measures are described as being on a transect over the North Atlantic. If so is it possible that meanderings of the current over time with some variation in direction and preferred paths could account for the variable flow rate rather than just an absolute change in the MOC flow (i.e. large scale “turbulence” in the flow) ? In other words is a transect of measures accurately reflecting the entire MOC flow?

May 11, 2019 8:05 am

“As one oceanographer put it, the process of deep water formation and sinking of the MOC is more complex than people believed, and these results should prepare people to modify their ideas about how the oceans work.”

I’m not sure I’m ready to go there. I think first we need to consider what would cause nature to not match our models? Perhaps it’s the way we’re measuring, or the data itself that needs to be tweaked or adjusted so that it more closely matches what we believe is happening?

Robert of Texas
May 11, 2019 8:33 am

Would be nice to see other measurements with the volume of water estimates. Is the temperature and salinity varying as much? More? Less? Is it all correlated or not?

This reminds me of watching basaltic lava slowly running and cooling – you get undulating wave after wave on the surface where it is cooling. Makes me wonder if you don’t have a series of smaller overturns in the water all pushing on each other north.

I bet the computer models have this all figured out perfectly. (<– Yes, sarcasm)

Clyde Spencer
May 11, 2019 9:00 am

“13.15 +/- 3.31 Sv”

Why are two digits to the right of the decimal point being shown when the variance is in the units position of the mean value? That is, the last significant figure is in the units position. Everything to the right of the decimal point is meaningless noise, making it appear that the mean is known to greater precision than it actually is.

May 11, 2019 10:00 am

I don’t think we should really be surprised by this great variability because I believe the nature of the Gulf Stream can be highly variable and that is the input to the system. If the input is highly variable, the output will be, too. The two main things that determine the rate of flow are going to be temperature and salinity and they act in opposite ways. The warmer the water is in the Gulf Stream, the farther north it goes before it sinks. Conversely the saltier it is,the sooner it sinks. Both temperature and salinity can vary considerably. A stalled hurricane or a series of them can pull great amounts of heat out of the gulf stream if they happen to stall at the right spot. Weather in the upper midwest can bring flood waters to the Gulf of Mexico. Cloudy conditions can reduce evaporation and also act to reduce salinity.

Even various steering winds can have an impact. The failure of the Bermuda High to form reduces the general southerly winds along the coast of the western Atlantic.

One of the things I have noticed with both the “climate scientists” and even in the rhetoric they feed the public is that they seem to have this expectation that the things that drive climate are very stable and that humans are the prime factor that cause perturbations in the overall system. I believe the system *is* basically stable but not in its current state. It is most stable during the glacial state (though with even considerable variability within that state) as evidenced by the fact that the climate stays in that state roughly 85% of the time. I also have a hunch that like other natural systems, it does not transition cleanly but the state of the system “buzzes” a little back and forth between states when it is transitioning between them. As we are likely reaching the end phase of this interglacial, I would expect climate over the next few thousand years to be more unstable going forward than it was, say, 7000 years ago.

What I expect to see are more episodes like the Little Ice Age with them becoming colder and longer lasting interspersed with episodes such as the modern warm period but with those becoming shorter and not as warm. We have been in these cycles for about the past 4000-5000 years since the end of the “Holocene Climate Optimum”. We are going to get a bit lucky, though, as we will see an uptick in northern hemisphere insolation for the next few thousand years that might moderate that fall into the next glaciation somewhat.

But generally speaking, I expect to see a destabilized climate system on a millennial scale going forward. This expectation that climate is somehow stable is unfounded, in my opinion.

Reply to  crosspatch
May 11, 2019 12:03 pm

I’ll check back with you in 500 years to see how we are doing.

Joel O'Bryan
May 11, 2019 10:52 am

The THC is a major negative feedback control system on the planet’s climate, maybe the most dominant for climate on scales of decades to centuries. And the AMOC (or MOC in this article), a big one of those for its control on North America eastern seaboard climate and Northern Europe’s climate. This is the internal system variability that the climate models cannot recreate, due to both complexity and nonlinearity.

The basics of the MOC are best understood from first principle of thermodynamics.
Much like a hurricane is a cyclonic vortex heat engine in the atmosphere, the MOC is a heat engine in the ocean. Hurricanes extract heat in the form of latent heat of water vapor from the warm tropical surface waters. The wind fields have advected copious amounts of water and turned to latent heat that is then bouyant. The towering cumulo-nimbus clouds are driven by this buoyancy and sustained by the adiabatic lapse rate as water condenses out, releasing the heat, driving the parcel ever higher to reach the tropopausewhere the cumulonimbus clouds tops spread out in a large radiating surface delivering ocean heat to the stratosphere for loss to space. Hurricanes continually get “new heat” by being driven by prevailing pressure pattern winds across new fertile warm waters. A hurricane that sits in one place too long will exhaust its supply of warm water as local upwelling will bring cooler water upwards and it cannot strengthen and will likely weaken. So a hurricane’s wind field must move to live.

Of course one major difference in this hurricane heat engine analogy to the MOC is water changes phases in the atmosphere with hurricanes, it does not in the ocean MOCs. So where is the heat loss? The heat loss is along its northward legs and entry in the high latitudes where cold air blowing across warm waterr is effciently removing heat by removing water leaving the salt behind. This is where salinity differences come into play. Higher salinity = higher density and the tendency to sink. But warmer water is also more buoyant. In the heat engine of the MOC, the working fluid is the tropical warm water mass moving northward along the Eastern US Canadian coastline, losing water to evaporation and getting saltier. The wind is not driving this water as the wind can only push about the first 100 meters layer. The MOC is deeper than that. Coriolis and mass balance from the downwelling forces the current to move at depth. The MOC is driven by the downwelling in the high latitude regions when the salinity-driven density finally overcomes the warmer buoyancy reduced by the advective cooling. Coriolis keeps it moving, for its eventual upwelling where it can once again be warmed by tropical sunshine. The downwelling pushes and pulls the MOC along.

That the seasonal variation in the MOC flow is strongest when the water is warmest (Sept-Oct, see data graph above) supports the idea that higher water temps drive the MOC faster. Mass balance ensures that downwelling must be increasing if the flow is increasing. For downwelling to increase, more heat must be being shed to advection, which it is in the Fall. This is balancing negative feedback of MOC circulation responding to higher tropical water temperatures. Downwelling is the powerstroke in the heat engine.

Where a hurricane is moving water vapor-laden air between two temps vertically, T1 and T2 being the sea surface to the tropopause respectively, the MOC is moving water of differing densities between two temp regimes as well, the tropical ocean temps and the high latitude ocean temp where the top 300 meters or so of water can eventually reach a density to sink under gravity. The air temperature is not the factor in the MOC heat engine here. The air temperature response though to additional heat release though is seen as “Arctic amplification” when the additional heat from the MOC being released into the cold polar air for eventual radiation to space. It is the wind-driven removal of heat that cools the water and increases its salinity so it can sink that ultimately drives the entire overturning current system. The MOC current is not being driven by the wind fields, that would be the 1 lb tail trying to wag the 100 lb dog.

Reply to  Joel O'Bryan
May 11, 2019 12:15 pm

Great explanation! Thanks. Actually the winds oscillate in direction anyway, sometimes pushing against the surface current and sometimes with it as the high/low pressure systems move off the continent and drift eastward, so I would think the net is close to break even anyway.

I’ve been in a sailboat in the Gulf Stream with the wind pushing against the current and it is someplace you just don’t want to be. The surface waves are rough to say the least. Turn that around and you go like a rocket ship (for a sailboat anyway).

Gordon Dressler
May 11, 2019 1:18 pm

Uhhhh . . . why the big difference in the absolute salinity, and the salinity gradient, between the Labrador Basin (western North Atlantic) and the Rockall Trough (eastern North Atlantic)?

This does not appear to be explained as a simple result of the MOC model.

Reply to  Gordon Dressler
May 11, 2019 2:57 pm

Possibly a combination of freshwater sources off the eastern seaboard and Greenland and the wind driven removal of latent heat causing increased salinity (mentioned above by Joel O’Bryan) during the west to east flow. The diagram does, however, show the lower salinity below the more concentrated which is not what you might expect from Joel’s explanation unless the overturning is still in progress. Help required !!

John F. Hultquist
Reply to  Gordon Dressler
May 11, 2019 8:40 pm

Rockall is receiving water from the outflow of the Mediterranean Sea.
Outflow is warm, salty, and subsurface.
In the Atlantic, it turns north.

Gordon Dressler
Reply to  John F. Hultquist
May 12, 2019 7:39 am

Really? According to the general water circulation patterns in the North Atlantic (see ), any Mediterranean Sea outflow that is claimed to move northward would be moving contrary to the flow of the Canary current and the North Atlantic Drift.

John F. Hultquist
Reply to  Gordon Dressler
May 12, 2019 12:17 pm


Kevin Whalen
May 11, 2019 8:31 pm

Wait…. is this…. actual observation science? I didnt think anyone did this anymore

mike macray
May 12, 2019 9:09 am

Great Post!
and your explanation J O’B of the driving forces of the THC/MOC.
Are there any data on the relative power ( for lack of a better term) of the Tropical and Arctic phase change driven salinity/density changes.?
It seems as though there is evaporative cooling and increased salinity in the tropical summer while ice melt cools and lowers salinity in the Arctic then a winter reversal as ice forms in the Arctic generating cold dense sinking brine. As an old engineer I’m visualising a current with two pulses, a winter one derived from the water to ice phase change and a summer one from water to vapour phase change six months apart each producing increased salinity and sinking.

%d bloggers like this: