Stronger Overturning Circulation in the Pacific During the Last Glacial Period

Data from fossil corals points to changed circulation of ocean currents – an important finding for climate models

Peer-Reviewed Publication



Located between Australia and New Zealand, the Tasman Sea is an important but so far neglected component of the global ocean conveyor belt. Now a new study has discovered evidence that this marginal sea in the South Pacific also played an important role in the exchange of water masses between the large ocean basins during the last ice age. These findings will help to refine climate models and improve our understanding of ocean circulation and carbon storage in the sea, an international team of researchers led by geoscientist Dr Torben Struve from the University of Oldenburg reports in the journal Nature Communications.

In their study the researchers examined 62 fossil specimens of the stony coral Desmophyllum dianthus. These were collected by the underwater remotely operated vehicle JASON during a research expedition south of Tasmania at depths between 1,400 and 1,700 metres. According to dating analysis, these animals lived about 10,000 to 70,000 years ago, a period that included the peak and end of the last glacial period. “The corals grow in areas with strong currents and turbulence that inhibit the deposition of sediment,” explained Struve, who conducts research in the Marine Isotope Geochemistry group at the University of Oldenburg’s Institute for Chemistry and Biology of the Marine Environment.

Because the skeletons of these sedentary animals record the chemical fingerprint of the surrounding seawater, complex analyses can reveal the chemical composition of the ocean at the corresponding water depth during the corals’ lifetime. This in turn provided clues about which water masses flowed through the Tasman Sea at the time. “These cold-water corals are a particularly good archive for studying the chemical composition of deep ocean currents in the past,” Struve explained.

Young water flowed through the depths of the Tasman Sea

In their study the researchers focused specifically on the ratio of different variants of the trace element neodymium, some of which are produced by radioactive decay and are commonly referred to as radiogenic isotopes. The analysis showed that water from the Pacific Ocean flowed through the depths of the Tasman Sea around the peak of the ice age – as indicated by the relatively high content of radiogenic neodymium in the coral samples. The investigations also showed that this water from the Pacific had been in contact with the sea surface relatively recently compared to other water masses in the same depth range, or in other words, that it had been relatively “young”. As the team writes in their paper, the data supports a scenario in which the upper Pacific Ocean was more mixed during the last ice age than it is today – while at the same time the deepest layers were more isolated from the atmosphere, which contributed to the long-term storage of carbon dioxide and the cooler glacial climate.

According to the new study, the circulation patterns during the last glacial period would have looked like this: in the North Pacific, surface water sank to a depth of about 2,000 metres and then spread a long way southward. After flowing around the southern tip of the Australian island of Tasmania, this water could have flowed into the Indian Ocean where it joined the global “conveyor belt” of ocean currents and reinforced it. This conveyor belt plays an important role in distributing heat among the various ocean basins: the warm North Atlantic Current, for example, is responsible for the comparatively mild climate in northwestern Europe. From the North Atlantic, the circulation extends across the Antarctic Circumpolar Current and the Indian Ocean to the northern Pacific – and then back again. In today’s system, the water in the North Pacific is the oldest, meaning that the last contact with the surface occurred a very long time ago.

The historical view allocated the return flow of this conveyor belt to the Indian Ocean mainly to a relatively shallow strait north of Australia. However, recent studies suggest that the outflow of Pacific waters through the Tasman Sea is also significantly involved in the exchange of water masses between ocean basins – albeit at shallower depths than during the last glacial interval. It is possible that up to half of the water flowing northwards within the global conveyor belt in the Atlantic today originated in the area south of Australia. “Our study contributes to a better understanding of the dynamics of this global ocean circulation system under changing climatic conditions,” said Struve. Now there is evidence that there was a close link between changes in the deep Tasman outflow and circulation changes in the Pacific Ocean during the last glacial period.


Nature Communications




Observational study


Not applicable


A deep Tasman outflow of Pacific waters during the last glacial period




The authors declare no competing interests.

From EurekAlert!

4.4 7 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
July 12, 2022 2:41 am

Such oceanic overturning would have been a response to changes in convective overturning within the atmosphere.
Whenever the system starts go go out of radiative balance the atmospheric convective overturning system alters its configuration to neutralise the imbalance.
Naturally induced imbalances will always swamp any effect from the puny influences of the extent that the latter will be indiscernible.

Reply to  Stephen Wilde
July 12, 2022 6:14 am

The convective engine is self-governing. It regulates to 30C ocean surface temperature in all but a few zones near land close to the Equator. At precisely 30C, the sunlight reaching the surface is just enough to keep the convective column fully stoked. At full throttle, the convective engine averages 11W/m^2 of work in pumping up the atmosphere at a Carnot efficiency of 24%.

There are two clear break points in Earth’s temperature response to solar forcing. One at the 30C upper limit and the other at 4C where sea ice is present for part of an annual cycle. Another less evident transition is the increase in slope of the response above 26C. That is the temperature where the surface sunlight starts to reduce as cloud reflection rapidly increases.

I am not sure what you mean by this:

Whenever the system starts go go out of radiative balance the atmospheric convective overturning system alters its configuration to neutralise the imbalance.

Deep convection is best described as a self-governing heat engine with a 30C limit at the hot end being the surface. The governor is simply the persistence of the reflective cloud.

It is a temperature limiting system. It has a response time of about 25 days when at 30C. Roughly the lag between the temperature UAH measures in the lower troposphere and surface measurements.

Last edited 27 days ago by RickWill
Reply to  RickWill
July 12, 2022 8:53 am

“The governor is simply the persistence of the reflective cloud.”

What you call the “30C limit” is an outcome of the Temperature/Vapor pressure of water correlation of about 7% more water molecules per degree, at equilibrium, in the air immediately above the water……air above the water with water molecules being lighter than the dry air at high altitude, thus convecting upwards….rising air cooling as it rises by expansion….some of the air, being not as well mixed with descending dry air, meeting its dewpoint and forming fluffy clouds….which reflect sunlight into outer space…..the lack of surface heating caused by the clouds causing the ocean surface to cool again….

So yes, the reflective cloud is the governor, but the “law” is not 30 C, but the Clausius- Clapeyron equation. Also, the basic numbers show SW reflection of low Cloud cover overwhelms possible CO2 warming by about an order of magnitude.

Last edited 27 days ago by DMacKenzie
Reply to  DMacKenzie
July 12, 2022 3:32 pm

The limit of 30C is precise. You should read up on Level of Free Convection so you understand why the atmosphere partitions in a lower zone of free convection and upper dehumidifying zone.

The reason the 30C limit is consistent is that the LFC approaches freezing altitude at 30C. Clear sky conditions are reduced to around 16% of the time. That results in an average of 200Wm^2 of sunlight getting to the surface to keep the heat engine running – mostly from dispersed sunlight.

The hot end of the heat engine is the surface at 303K. The cold end is the dehumidifying zone averaging around 240K. The drying air acts as the compressor above the LFC because it has higher density than moist air that expands it after cloudburst and the LFC acts as the expansion nozzle once instability occurs; akin to the engine firing. A full cycle at 30C takes 53 hours and 45 of that is dehumidifying by ice forming rather than water condensing. Only 8 hours of clear sky. The cycle rarely builds full convective potential though.

Reply to  RickWill
July 12, 2022 4:05 pm

“…should read up…”
I’m already quite conversant with Skew-T and Tephigrams, CAPE, CAP, CIN, etc., thank you anyway…now that you mildly ticked me off…
You should read up this, especially chapter 5 so your “30 C limit” is based on accepted technical terms instead of sounding quackish….
R. Stull also has another similar text for engineers and physicists.

Reply to  DMacKenzie
July 12, 2022 6:11 pm

No mention of cloud persistence or 30C ocean surface temperature limit.

All land based as well. Nothing over tropical oceans. That is where all convective potential is generated to drive the global circulations. The relatively humidity in these regions is in excess of 80% to the LFC.

No thermal response curve like I have above to the incoming solar EMR.

All focused around land based storm prediction rather than global energy balance.

Using that book, tell me why the Southern Ocean is cooling, the Nino34 region is trendless and the oceans in the Northern Hemisphere are warming. All easily proven with reliable data:

July 12, 2022 2:44 am

More refinements to models!

And that’s what you call a huge gap in knowledge and understanding; a refinement

Ron Long
July 12, 2022 2:45 am

Amazing how new and significant data keeps getting added to “the science is settled” theme. Looks like climate is complex and chaotic and model-resistant.

July 12, 2022 3:23 am

Many articles have displayed graphs showing that sea level has risen somewhat over 100 meters since the last glacial maximum. This article talks about coral that “lived about 10,000 to 70,000 years ago” which would seem to be when sea level was at that low point. However, the fossils were collected “at depths between 1,400 and 1,700 meters”, which seems to compute to perhaps 1200 to 1500 meters depth at the time the corals were alive.

Sunlight doesn’t reach those depths. Are there corals that don’t depend on photosynthesis? That seems unlikely.

Tom Abbott
Reply to  AndyHce
July 12, 2022 5:19 am

Good questions.

Reply to  AndyHce
July 12, 2022 5:24 am

Andy: Corals are animals grouped taxonomically with jellyfish and anemones into the phylum cnidardia. They aren’t plants. The don’t actually need sunlight. But they do need food. Shallow water corals do coexist with symbiotic algae(plants), but deeper water corals survive on whatever organic they can filter out of nearby sea water.

Reply to  DonK
July 12, 2022 5:30 am


That’s phylum Cnidaria…

Reply to  AndyHce
July 12, 2022 5:37 am

I have not read the paper and I don’t know if they disclosed the species they used but, yes, there are coral species that do not need zooxanthella to survive. Therefore, they can live in medium without much light.

Reply to  AndyHce
July 12, 2022 6:07 am

The other explanation would be subsidence; meaning that the corals sampled today were not at their original depth but were brought deeper by geological processes.

Reply to  Agamemnon
July 12, 2022 12:49 pm

Subsidence seems to require certain conditions such as isostatic rebound happening because a great weight, like a huge ice sheet, having previously raised the land beyond their borders and said raised land now being let back down due to the disappearance of that great weight, or the subsiding land is on a tectonic plate that is moving under an adjoining plate. Maybe relevant but maybe doesn’t seem appropriate for ocean bottom that is a channel between two ocean basins???

Peta of Newark
July 12, 2022 4:16 am

Let’s cut the crap shall we….

What they’re saying is that the ocean > weather > climate was wilder & turbulent when the climate was colder. (than it is now for instance)

Errrrm, isn’t that a capital offence in the crazy surreal world of climate science, isn’t The Precise Opposite supposed to happen?

If nothing else it proves how crazy/surreal and what utter junk it all is.

AGW is Not Science
Reply to  Peta of Newark
July 12, 2022 10:45 am

Yes but do note the reference to CO2 being “stored” in the deep oceans via those waters being “isolated” from the surface.

By at least intimating that CO2 is somehow potentially “involved” in a “climate” event, the requisite boot licking is done.

July 12, 2022 4:42 am

I notice a chicken and egg logic conundrum in the text.

“While at the same time the deepest layers were more isolated from the atmosphere, which contributed to the long-term storage of carbon dioxide and the cooler glacial climate”

Did the CO2 reach the bottom layer because it was cold, or did the CO2 reaching the bottom layer cause the cold? They are indoctrinated to the idea of CO2 as a control knob for temperature and thus say the latter, while ice cores almost uniformly show CO2 lagging temperature which would imply the former.

Or it could just be the required unrelated genuflexion to the climate gods to continue funding.

Tom Abbott
Reply to  OweninGA
July 12, 2022 5:23 am

It didn’t seem to me that they tied their research much to CO2. They did mention CO2, but as you point out, confusingly, and they did mention improving “the models” but other than that, it looks like straightforward reporting on the circulation patterns with no conclusions as regards human-caused climate change.

Reply to  Tom Abbott
July 12, 2022 10:23 am

I was quoting the press release. It is possible the paper is good, but the “science communicators” inserted the required climate genuflexion.

AGW is Not Science
Reply to  Tom Abbott
July 12, 2022 10:59 am

Yes, but an “honorable mention” of CO2 keeps the grant money flowing.

Tom Abbott
Reply to  AGW is Not Science
July 13, 2022 5:01 am

I agree. They have to put CO2 in there somewhere.

July 12, 2022 5:53 am

No climate model has predicted the cooling trend in the Southern Ocean. No climate model has predicted the zero trend in the Nino34 region of the Pacific. They happened to align with the warming trend in the mid latitudes of the northern hemisphere but have failed to recognise the reason.

So the question is how smart is CO2? It can warm the NH. It does not do anything at the Equator and it cools the southern ocean. CO2 the fabulous magic gas.

People who believe CO2 has anything to do with Earth’s energy balance should stop smoking whatever they are smoking. CO2 does zip!

Note how the Nino34 temperature reflects Jupiter’s 11.8 year cycle of muscling the sun about.

Reply to  RickWill
July 12, 2022 6:30 am

Atmospheric CO2 ppm increase follows ocean temperature rise. It always has historically.

When the ocean mean temperature ultimately causes the ocean to outgas more CO2 annually than is sequestered annually – then the annual rise is rather dramatic!
The resulting incremental annual increase is alarming to sections of the population.

Reply to  RickWill
July 12, 2022 8:42 am

The earth is on average 93 million miles from the sun. Jupiter changes the location of the center of the sun by a couple thousand miles.
Do the math.

Reply to  MarkW
July 12, 2022 12:53 pm

PS: At the same time Jupiter is pulling on the Sun, it is also pulling on the Earth.
As a result, even though Jupiter does move the Sun around, it moves the Earth in the same direction by close to the same amount.

DD More
Reply to  MarkW
July 12, 2022 3:18 pm

Bit more than 2,000 miles.
mean radius of the sun is 432,450 miles
Based just on Jupiter,  the barycenter of Jupiter and the sun isn’t in the center of the sun. It’s actually just outside the sun’s surface! Get Saturn and Neptune on the same side and it will be further.
That’s from NASA Science, but what do they know.

Reply to  RickWill
July 12, 2022 12:08 pm

CO2 is the interhemispheric heat pirate 🏴‍☠️ !

July 12, 2022 7:11 am

One thing which continually surprised me from reading Nature and Science for 30 years was how sloppy climate models were. Every year had several new reports of how much they’d improved the climate models — they’d add or improve cloud cover or mountain ranges or water vapor; they all blurred together and I no longer remember any of the changes, just that each change seemed so basic that I wondered how they could possibly have relied on the previous models.

This just seems like more of the same. If your model of ocean currents is so drastically improved, what makes you think any of your past modeling has been even close to correct? What makes you think this new version is suddenly and forever correct and won’t get new improvements six months from now?

July 12, 2022 12:15 pm

It is possible that up to half of the water flowing northwards within the global conveyor belt in the Atlantic today originated in the area south of Australia.

It’s been traditional to talk about the downwelling in the North Atlantic / Norwegian Sea as the driver of the global THC (Thermo Haline Circulation). But it’s not surprising that this is not so: Antarctica, the dominant player generally in earth’s current climate, is bigger and colder than the Arctic and thus the most voluminous and dominant downwelling and deep water formation occur around Antarctica. This is indeed the case, including the seas south of Australia, which is probably the most important region for cold deep water formation.

Last edited 27 days ago by Phil Salmon
Reply to  Phil Salmon
July 13, 2022 9:40 am

The purple “From Pacific” line seems to be in the wrong place, there isn’t a deep water path between Java and Australia but there is one from the Coral Sea down the east coast of Australia under Tasmania into the Southern Ocean. Isn’t that what this paper says?

July 13, 2022 4:46 am

That is a lot of major conclusions from 62 specimens.

These findings will help to refine climate models and improve our understanding of ocean circulation and carbon storage in the sea”

The climate models that programmers and designers which refuse to make substantive changes in their models to reflect the real world?
Pure narcissist hubris.

%d bloggers like this: