Now that we can simply dismiss the CO2 hypothesis as a climate driver, we must resolve the outstanding issue of the transfer of heat into the Arctic in particular.  This is also matched by a heat transfer into the South Atlantic as confirmed in a report this month in the National Geographic in their article on South Georgia.

Simply put Atlantic Ocean heat is been pushed out to the extremities.

Our clue as to why comes from a report over a decade old in which it was observed that the flow rate of the Gulf Stream had slowed.  In order to maintain balance, this implies that the volume had increased.  If the volume had increased, then obviously the amount of heat transferred had increased.

We only have the two observation points on this with the first back in 1959 or so.  The later one coincided with the discovery that sea ice had thinned by forty to sixty percent.

My point is that a slowed current moving much greater volume will warm the surface layer in the Arctic.  It is also a fairly constant injection of heat and the ice itself is insulating it from the Arctic cold.  Thus somewhere a switch got thrown and this heat injection is responsible for the eminent destruction of all multiyear ice regardless of atmospheric conditions.

No additional heat has been put into the Atlantic itself, put the normal heat content has been redistributed.  The fine details are not clear, but the consequences certainly are.  They include collapsing Arctic sea ice and a changing biology in the South Atlantic down current from the Antarctic Peninsula. 

The cause of this must be a change in the volumes handled by the global countercurrent.  Likely this same event will trigger a cooling event in the south Pacific as a natural balancer.  The last time it happened it remained stable for around seven hundred years and was called the Medieval Maxima.  Greenland famously became green.

Compared to the size of ocean currents and general volumes involved, this effect is possibly a small adjustment in the global heat machine.  We assume it is a form of climatic fine tuning.  That may be optimistic.  However following several centuries of warmth in the Northern Hemisphere with two or three centuries of little ice age seems a good balancing act since perhaps only two or so of those warm centuries were noteworthy.  Two up and two down and the rest average seems about right.

This also suggests a long pattern into which we also can fit the cold weather that ended the Roman Empire in the fifth century.  The cycle appears to be around 1200 years long.  Our minima can then be isolated as 1700AD, 500 AD, 700BCE, 1900BCE, and 3100BCE, plus or minus a century or so.  We know that the first three dates were coincident with particularly cold conditions.  I presently have no data for the earlier eras and the general protracted warmth of the Bronze Age optimum may obscure any such cycle.

What it means is that if this conjecture stands up, then we have fine climate for a full millennia during which the Arctic will be ice free in the summer.

This is all driven by the global counter current.  That in turn is affected by the Antarctica cooler that will presently continue to expand to generate cooling in the South Pacific. (not as anyone will ever notice) I suspect that at some point this becomes unstable and a small gyre is produced or destroyed altering the volume of the counter current and thus the speed.  It may simply be stretched somehow or shortened.  Both methods work nicely.  I prefer the advent and or collapse of a gyre because it explains the abrupt nature of the switching which is what has been effectively observed.

The dissipation of such a gyre after a couple of centuries also fits the gradual strengthening recovery we have experienced. (the gyre may be large enough to stretch the counter current.

A quick check with Google tells me that the climate in the Indus valley became cooler and dryer in around 1900 BCE. One down!

3200 BCE is not as clear but again climate change is strongly noted and indicated.

Of course local variations existed and it would need a lot of tree ring chains to make a really convincing argument.  So far we have no negative evidence and the actual conjecture appears to be quite robust for now.

The switch kicks in every 1200 years or so and the actual variation is under a century.  That implies that the next switch over will take place during the twenty ninth century.  In the meantime the climate will trend toward a climatic optimum as it has been.

As an aside, the ocean controls the climate rather than the climate controlling the ocean.  It makes atmospheric issues a likely sideshow at best whose effect is at best modest and usually immeasurable.