In several other threads I’ve looked at lunar influence on ocean depth and currents and speculated that the cycles in lunar tidal forces might account for the periodic swings in our long term weather, ( Sometimes called ‘climate’, which IMHO, is not correct. The California Redwood Forest is still a redwood forest [and the California northern pines are still in a boreal forest] even if the State has a 900 year long drought. That, BTW, is a real example.) This notion of a lunar metronome on our weather and ‘climate cycles’ was picked up by Clive Best and expanded when he mixed it in with his ideas on drivers for interglacial timing. Well, what goes around comes around… Clive got me thinking about how this might reflect in the Arctic Melt, that got me looking at the ‘melt pulses’. That, then landed me on some Arctic maps. One thing leads to another, and now we have a new speculation. My speculation is just this:
During the deep glaciations, Arctic Ice goes very deep, and the ocean level drops. So much so that the Bering Strait closes, the Canadian islands and all their channels close, and the Arctic Ocean reduces to a small puddle in the middle (compared to what we have today), almost completely isolated from the warming ocean currents. As that reverses, as the ice melts some and water rises; those processes reverse at specific depths and there is a new “flushing step” as a current can now re-establish and flow under newly ungrounded ice. At particularly high tide intervals, when lunar tidal force is greatest, will be the moment when that (barely) grounded ice first lifts off the bottom, or when that (barely) closed channel first opens. Once that melt starts in a warming trend, it continues in a ‘pulse’ as the thinned bottom is no longer closed.
I’d further speculate that in a cooling cycle, the Arctic stays liquid until a very modest tide cycle, then it has time to thicken just enough to NOT lift off or melt in the next, colder, heavy tides cycle. Essentially, at the marginal state in a warming or cooling cycle, the lunar tidal metronome will kick off the exact moment of the pulse. It will flip the switch on the hysteresis of ocean currents and ice grounding.
Some links on Lunar Tidal influences:
The foundation for a lot of the ‘The Moon Did It’ speculation / thesis.
Clive Best does an interesting look at things here.
Links to the Clive Best article, but has interesting commentary.
The ‘pick up’ at WUWT.
Then a bunch of my articles in retrospective:
Essentially points out that the 18.x year cycle lines up with the same ‘face’ of the Earth on an ‘about 55’ year basis and that this is close to the ‘about 60’ year cycle of PDO et. al.
A sort of descriptive catalog of some of the lunar cycles as an attempt to put some order into the cycle-mania.
Since “orbital resonance” dominates orbital cycles, and the moon orbits, it can also have correlations with other orbital stuff. Perhaps even cometary remnants that whack into us periodically.
A look at periodic failures of the Thermohaline Circulation (and how Florida gets summery weather then ;-) along with where we are in the cycle. Next ‘Little Ice Age’ you want to be in Florida, not England. This is periodic, not random, and not driven by atmospheric gasses.
An ‘unfinished bit’ would be integrating circulation changes with depth at the Southern Ocean where I think Drake Passage is key.
Here is a nice image of the surface currents of the Arctic:
First thing to notice is that the Arctic Ocean circulates. ( I can hear someone saying “Well Duh.” Remember that it is important to notice all things, even the obvious. Like bilateral symmetry in a fish – there is a story behind that example…)
So why does circulation matter? Two reasons. First off, you can see warm water entering on the Pacific and Atlantic connections and cold water leaving via Canada and Greenland / Fram Strait. During a Glacial, that circulation stops. With a mile of ice over Canada, that exit is closed. With ocean levels 100 meters lower, folks can walk from Russia to Alaska. (Well, they do it sometimes now over the ice, but it will be easier and less seasonal during a Glacial ;-)
So look again. No Bering Sea warm intrusion. No Canadian cold drain. No Beaufort Gyre when the ice is deep, since there will be no wind driven circulation under the ice. The Asian current toward the Bering Sea will end. The entire Asian warm river drain into the Arctic likely freezes up and doesn’t happen – which raises the interesting question of where does it go then? But that is for another day. Like asking where the Alaskan rivers drain then, or are they just glaciers at that point?
In short, what is left is just the North Atlantic Drift (aka Gulf Stream for Americans) warming a small patch near Europe and some cold water near Greenland. As Scotland was under ice in the last Glacial, even that North Atlantic Drift circulation likely didn’t get very far north.
But is there more to this? What about layers of depth?
A deeper look
A bit further down, we see that the North Atlantic Drift skirts along the Asian side as is washes under the Arctic Ice. There it mixes with warm fresh water from Asian rivers and helps keep that side more melted. The Canadian side is colder, and that IMHO likely explains why the Glacial Shield tends to build up over Canada and northern USA. Less melt means more snow and precipitation needs to pile up and spread as glacial ice.
Who state it is from Jack Cook from Woods Hole Oceanographic Institute.
We can also see that at the deeper level, a bit of cold current is flowing at the bottom layers of the Bering Straight. Block that off, and block off the Labrador Current and those various Canadian circulations, all that cold water has to leave via the place where the North Atlantic Drift is trying to enter. All in the context of a lower ocean level, so less total channel volume. That will tend to block warm intrusion and drive the process toward freeze-up. Ever less circulation and flushing in the Beaufort Gyer area, stunting of the warm intrusion, more cold and ice blocking flow out via Greenland / Iceland.
Now look even deeper…
Arctic bathymetry: Notice the depths
Most of the Arctic Ocean is a shallow Europe / Asian margin or shallow Canadian shelf. When ocean levels drop, that becomes land. Ice covered land, but still, not ocean and without circulation. The “drop” in ocean depth is somewhere over 100 meters. That is the first 4 or 5 blue color bands. Essentially, the Arctic Ocean reduces to that dark pit in the center and the narrow dark channel leading into it past Greenland. The ‘North Atlantic Drift’ (or whatever is left of it) and the cold exit water must churn past each other in whatever little space remains under the ice there. IMHO, not much is going to happen in terms of ice melt in the Arctic under those conditions. Thank God for Sea Level Rise! (And pray that Iceland spreading does not close that trench or we will not exit the next Glacial after closure, IMHO.)
Now, play it backwards
At some point in the melt / deglaciation process, ocean levels rise just enough that warm water CAN start to circulate under that ice in those light blue areas. How deep is that? About 100 meters. At some point ocean levels rise just enough that warm water can start to enter via the Bering Strait. How deep is that? About 50 meters (more on that in a minute). At those times, an exceptional tide can be the trigger that lifts that ‘grounded ice’ just enough to unground it and start melting from below. Starting the bending / tidal surge / melt / ocean rise cycle and letting it run to the point of the next shelf of depth.
It looks like those depths are in meters, as the Wiki says: “The Bering Strait is approximately 82 kilometres (51 mi; 44 nmi) wide at its narrowest point, with depth varying between 98 feet (30 m) and 160 feet (49 m)” and that reflects the numbers in the image.
So at about 50 m of ocean drop, the strait no longer exists at all, but above that point, ever more Pacific Ocean (relatively) warm surface waters can go north, and ever more salt dense cold Arctic water can drain out into the deeper Oceans. In essence, the Thermohaline Circulation as we know it today can form, as can the Transpolar Drift and some of that Asian cold water draining out to the oceans and letting more of the Northern Drift into the Arctic through an ever increasing Fram Strait in the Norwegian Greenland sea. Prior to that point, the Bering Sea will have gone from ice to melt. That depth is likely the point where the melt begins, but then the opening of the Strait lets it continue on into the Arctic Ocean basin.
Melt Pulses. Notice the depth.
There are two “knees” in this chart that are of particular interest. One is at “pulse 1A” at about 110 m of depth. At that point, the Asian shelf glaciers of the Arctic starts to float and the North Atlantic Drift can scour a much larger flatter part of the Arctic Ocean margin. It is no longer constrained to nibbling at a tiny channel edge at the narrow deep bits of what is now the Fram Strait.
At about 80 and 60 meters we get the opening of the Canadian drainage (and melt of the ice over it) along with the opening of the Bering Sea and eventually the Bering Straight (and that ice melting). Once a new level of depth opens a new bottom water level, then melt from below can proceed until all the overlaying ice is gone. This process would be hindered by heavy ice above, but enhanced when there was a larger than typical tidal surge / flushing event. Once the melt of the ice above begins, the event proceeds to a new stability point at shallower bottom depth. As we reach modernity, there isn’t much more ice left to raise any sea level, and the whole process stops in that long flattish area at the end. Essentially, the Arctic contribution to sea level rise is over and the process stops.
It would be more complete to have an actual ocean bottom (global) model with computation of the water level at each point, and couple it all to a tidal model based on the long term 1800 year tidal surges; but until I get a grant to do that, someone else will need to ‘run with the idea’. Footnotes of attribution always appreciated ;-)
I would also assert that the time of greatest “risk” of a re-freeze is when ice is accumulating in the Antarctic. Why? That lowers sea level just enough that a low tidal surge ‘moment’ can let the Arctic start into a lowering / freezing feedback cycle. It can’t do it on its own (as it is melting from below due to warm water and stirred by winds). It needs an external event to get it into the refreeze cycle. That, IMHO, is a build up of ice. But where? The Polar See Saw would have that be Antarctica, then with a couple of years of low tidal flushing the Arctic ice cap starts to build a lot of ‘multi year ice’ and ocean level drop can start the feedback loop. The next half of the polar see saw builds more Arctic ice. When the next polar see saw or high tidal surge cycle arrives, it is too late to cause the melt, and we proceed directly to a long glacial period.
In short, the Millankovitch Cycle is the long slow driver of conditions, but it is the lunar tidal flushing cycle perhaps with a Polar see saw ‘kicker’ that throws the switch at each end of an interglacial. It provides the hysteresis that leads to switching.
How deep does the ice go during Arctic Glacials?
One “confounder” for this thesis is just how deep Arctic glacial ice went. It’s all well and good to say that the Bering Strait opens at 50 m of ocean depth, but what if the ice is grounded 1000 m down on each side of it? I’ve glossed over that “issue”, but it can easily explain the error bands on things like 60 m where the melt pulse starts vs 50 m for the strait depth. (or perhaps that 60 m was really the 100 m of bottom depth point, but with grounded ice needing 40 m more ‘lift’ to get off the shelf). All that is ‘in the details’ that would need a lot of work to sort out. Here’s an interesting link, though, to get the thoughts moving on that point:
The last decade of geophysical seaﬂoor mapping in the Arctic Ocean from nuclear sub-marines and icebreakers reveals a wide variety of glaciogenic geomorphic features at waterdepths reaching 1000m. These ﬁndings provide new and intriguing insights into the Quatern-ary glacial history of the Northern Hemisphere. Here we integrate multi- and single beambathymetric data, chirp sonar proﬁles and sidescan images from the Chukchi Borderlandand Lomonosov Ridge to perform a comparative morphological seaﬂoor study. This invest-igation aims to elucidate the nature and provenance of ice masses that impacted the ArcticOcean sea ﬂoor during the Quaternary. Mapped glaciogenic bedforms include iceberg keelscours, most abundant at water depths shallower than ~350–400m, ﬂutes and megascaleglacial lineations extending as deep as ~1000m below the present sea level, small drumlin-like features and morainic ridges and grounding-zone wedges. The combination of thesefeatures indicates that very large glacial ice masses extended into the central Arctic Oceanfrom surrounding North American and Eurasian ice sheets several times during the Quaternary.Ice shelves occupied large parts of the Arctic Ocean during glacial maxima and ice rises wereformed over the Chukchi Borderland and portions of the Lomonosov Ridge.
So don’t fret the 1’s to 10’s of meters. We’ve got 1000 m of ice below the present surface and who knows how much above to confound the small bits.
But what this says to me is that clearly the 100 m level around that Asian margin was grounded ice, the Bering Strait was closed and under ice, the Canadian drainage was an ice dam, and the Arctic Ocean was that deep puddle under the middle, with, at most, a very narrow channel in the depths connecting it to a cold north Atlantic, and not much circulation.
But once those ice dams get lifted and start to melt, perhaps under exceptional tides and with extra insolation / longer summers (as per Milankovitch), then the cycle of melt can proceed to a new stability point at the next blocked depth.
This needs a lot more to tie it all together. At this stage, it is just a speculation. Actual basin depths, volumes, etc. would need to be modeled. Actual (expected) ice volumes, depths, floatation ocean depths, etc. too. Then the lunar cycles matched to onset times (if available to that 1800 year precision) and the whole model set in motion to see if it had better explanatory power or predictive skill.
With that said: I do think that too little attention has been paid to the ocean bottom profile interaction with ocean depths and ice volumes. Adding in a lunar ‘kicker’ to get thing started would explain some exact timings, should they match, and at a minimum the bottom profile effects need to be added to any lunar flushing / interglacial thesis. Just I don’t have the time to do it right now. But at least I can point at it ;-)
Maybe Clive can do it. (hint hint ;-)