In Part Eleven we looked at the end of the last ice age. We mainly reviewed Shakun et al 2012, who provided some very interesting data on the timing of Southern Hemisphere and Northern Hemisphere temperatures, along with atmospheric CO2 values – in brief, the SH starts to heat up, then CO2 increases very close in time with SH temperatures, providing positive feedback on an initial temperature rise – and global temperatures follow SH the whole way:
This Nature paper also provided some modeling work which I had some criticism of, but it wasn’t the focus of the paper. Eric Wolff, one of the key EPICA steering committee members, also had similar criticisms of the modeling, which were published in the same Nature edition.
In this article we will look at He et al 2013, published in Nature, which is a modeling study of the same events. One of the co-authors is Jeremy Shakun, the lead author of our earlier paper. The co-authors also include Bette Otto-Bliesner, one of the lead authors of the IPCC AR5 on Paleoclimate.
For new readers, I suggest reading:
- Part Six – “Hypotheses Abound” - the many different theories that go under the same name: “Milankovitch theory”
- Part Nine – GCM III - which had a GCM modeling exercise with many common elements to this one, a “speeded up” simulation from 120kyrs BP to the present
- Past – Eleven – End of the Last Ice age - which covered the timing of important events
He et al 2013
Readers who have followed this series will see that the abstract (cited below) covers some familiar territory:
According to the Milankovitch theory, changes in summer insolation in the high-latitude Northern Hemisphere caused glacial cycles through their impact on ice-sheet mass balance.
Statistical analyses of long climate records supported this theory, but they also posed a substantial challenge by showing that changes in Southern Hemisphere climate were in phase with or led those in the north.
Although an orbitally forced Northern Hemisphere signal may have been transmitted to the Southern Hemisphere, insolation forcing can also directly influence local Southern Hemisphere climate, potentially intensified by sea-ice feedback, suggesting that the hemispheres may have responded independently to different aspects of orbital forcing.
Signal processing of climate records cannot distinguish between these conditions, however, because the proposed insolation forcings share essentially identical variability.
Here we use transient simulations with a coupled atmosphere–ocean general circulation model to identify the impacts of forcing from changes in orbits, atmospheric CO2 concentration, ice sheets and the Atlantic meridional overturning circulation (AMOC) on hemispheric temperatures during the first half of the last deglaciation (22–14.3 kyr BP).
Although based on a single model, our transient simulation with only orbital changes supports the Milankovitch theory in showing that the last deglaciation was initiated by rising insolation during spring and summer in the mid-latitude to high-latitude Northern Hemisphere and by terrestrial snow–albedo feedback.
[Emphasis added]. The abstract continues:
The simulation with all forcings best reproduces the timing and magnitude of surface temperature evolution in the Southern Hemisphere in deglacial proxy records.
This is a similar modeling result to the paper in Part Nine which had the same approach of individual “forcings” and a simulation with all “forcings” combined. I put “forcings” in quotes, because the forcings (ice sheets, GHGs & meltwater fluxes) are actually feedbacks, but GCMs are currently unable to simulate them.
AMOC changes associated with an orbitally induced retreat of Northern Hemisphere ice sheets is the most plausible explanation for the early Southern Hemisphere deglacial warming and its lead over Northern Hemisphere temperature; the ensuing rise in atmospheric CO2 concentration provided the critical feedback on global deglaciation.
In this paper the GCM simulations are:
- ORB (22–14.3 kyr BP), forced only by transient variations of orbital configuration
- GHG (22–14.3 kyr BP), forced only by transient variations of atmospheric greenhouse gas concentrations
- MOC (19–14.3 kyr BP), forced only by transient variations of meltwater fluxes from the Northern Hemisphere (NH) and Antarctic ice sheets
- ICE (19– 14.3 kyr BP), forced only by quasi-transient variations of ice-sheet orography and extent based on the ICE-5G (VM2) reconstruction.
And then there is an ALL simulation which combines all of these forcings. The GCM used is CCSM3 (we saw CCSM4, an updated version of CCSM3, used in Part Ten – GCM IV).
The idea behind the paper is to answer a few questions, one of which is why, if high latitude Northern Hemisphere (NH) solar insolation changes are the key to understanding ice ages, did the SH warm first? (This question was also addressed in Shakun et al 2012).
Another idea behind the paper is to try and simulate the actual temperature rises in both NH and SH during the last deglaciation.
Let’s take a look..
Their first figure is a little hard to get into but the essence is that blue is the model with only orbital forcing (ORB), red is the model with ALL forcings and black is the proxy reconstruction of temperature (at various locations).
We can see that orbital forcing on its own has just about no impact on any of the main temperature metrics, and we can see that Antarctica and Greenland have different temperature histories in this period.
- We can see that their ALL model did a reasonable job of reconstructing the main temperature trends.
- We can also see that it did a poor job of capturing the temperature fluctuations on the scale of centuries to 1kyr when they occurred.
For reference here is the Greenland record from NGRIP from 20k – 10kyrs BP:
Figure 3 – NGRIP data
We can see that the main warming in Greenland (at least this location in N. Greenland) took place around 15 kyrs ago, whereas Antarctica (compare figure 1 from Shakun et al) started its significant warming around 18 kyrs ago.
The paper basically demonstrates that they can capture two main temperature trends due to two separate effects:
- The “cooling” in Greenland from 19k-17k years with a warming in Antarctica over the same period – due to the MOC
- The continued warming from 17k-15k in both regions due to GHGs
Note that my NGRIP data shows a different temperature trend from their GISP data and I don’t know why.
Let’s understand what the model shows – this data is from their Supplementary data found on the Nature website.
First, 19-17 kyrs ago, Antartica (AN) has a significant warming, while Greenland (GI) has a bigger cooling:
Note that the MOC (yellow) is the simulation that produces both the GI and AN change. (SUM is the values of the individual runs added together, while ALL is the simulation with all forcings combined).
Second, 17 – 15 kyrs ago, Antartica (AN) continues its warming and Greenland (GI) also warms:
Note that GHG (pink) is the primary cause of the temperature rises now.
We can see the temperature trends over time as a better way of viewing it. I added some annotations because the layout wasn’t immediately obvious (to me):
Figure 6 – Red/blue annotations on side, and orange annotations on top
Again, as with figure 1, we can see that the main trends are quite well simulated but the model doesn’t pick shorter period variations.
The MOC in brief
A quick explanation – the MOC brings warmer surface tropical water to the high northern latitudes, warming up the high latitudes. The cold water returns at depth, making a really big circulation. When this circulation is disrupted Antarctica gets more tropical water and warms up (Antarctica has a similar large scale circulation running from the tropics on the surface and back at depth), while the northern polar region cools down.
It’s called the bipolar seesaw.
When your pour an extremely big vat of fresh water into the high latitudes it slows down, or turns off, the MOC. This is because fresh water is not as heavy as salty water, it can’t sink and it slows down the circulation.
So – if you have lots of ice melting in the high northern latitudes it flows into the ocean, slowing down the MOC, cooling the high northern latitudes and warming up Antarctica.
That’s what their model shows.
The available data on the MOC supports the idea, here is the part d from their fig 1 – the black line is the proxy reconstruction:
The units on the left are volume rates of water flowing between the tropics and high northern latitudes.
What Did Orbital Forcing do in their Model?
If we look back at their figure 1 (our figure 2) we see no change to anything as a result of simulation ORB so the abstract might seem a little confusing when their paper indicates that insolation, aka the Milankovitch theory, is what causes the whole chain of events.
In their figure 2 they show a geographical look of polar and high latitude summer temperature changes as a result of simulation ORB.
The initial increase of the mid-latitude to high-latitude NH spring–summer insolation between 22 and 19 kyr BP was about threefold that in the SH (Fig. 2a, b). Furthermore, the decrease in surface albedo from the melting of terrestrial snow cover in the NH results in additional net solar flux absorption in the NH (Supplementary Figs 12–15). Consequently, NH summers in simulation ORB warm by up to 2°C in the Arctic and by up to 4°C over Eurasia, with an area average of 0.9°C warming in mid to high latitudes in the NH (Fig. 2c, e).
In their model this doesn’t affect Greenland (for reasons I don’t understand). They claim:
Our ORB simulation thus supports the Milankovitch theory in showing that substantial summer warming occurs in the NH at the end of the Last Glacial Maximum as a result of the larger increase in high-latitude spring–summer insolation in the NH and greater sensitivity of the land-dominated northern high latitudes to insolation forcing from the snow–albedo feedback.
This orbitally induced warming probably initiated the retreat of NH ice sheets and helped sustain their retreat during the Oldest Dryas.
1. If we run the same orbital simulation at 104, 83 or 67 kyrs BP (or quite a few other times) what would we find? Here are the changes in insolation at 60°N from 130 kyrs BP to the present:
It’s not at all clear what is special about 21-18 kyr BP insolation. It’s no surprise that a GCM produces a local temperature increase when local insolation rises.
2. The meltwater pulse injected in the model is not derived from a calculation of any ice melt as a result of increased summer temperatures over ice sheets, it is an applied forcing. Given that the ice melt slows down the MOC and therefore acts to reduce the high latitude temperature, the MOC should act as a negative feedback on any ice/snow melt.
4. The Smith & Gregory 2012 paper that we look at in Part Nine maybe has different effects from the individual forcings to those found by He et al. Because 20 – 15 kyrs is a little compressed in Smith & Gregory I can’t be sure. Take, for example, the effect on (only) ice sheets during this period. Quite an effect in SG2012 over Greenland, nothing in He et al (see fig 6 above).
It’s an interesting paper, showing that changes in the large scale ocean currents between tropics and poles (the MOC) can account for a Greenland cooling and an Antarctic warming roughly in line with the proxy records. Most lines of evidence suggest that large-scale ice melt is the factor that disrupts the MOC.
Perhaps high latitude insolation changes at about 20 kyrs BP caused massive ice melt, which slowed the MOC, which warmed Antarctica, which led (by mechanisms unknown) to large increases in CO2, which created positive feedback on the temperature rise and terminated the last ice age.
Perhaps CO2 increased at the same time as Antarctic temperature (see the brief section on Parrenin et al 2013 in Part Eleven), therefore raising questions about where the cause and effect lies.
To make sense of climate we need to understand why:
a) previous higher insolation in the high latitudes didn’t set off the same chain of events
b) previous temperature changes in Antarctica didn’t set off the same chain of events
c) whether the temperature changes produced in simulation ORB can account for enough ice melt seen in the MOC changes (and what feedback effect that has)
And of course, we need to understand why CO2 increased so sharply at the end of the last ice age.
Articles in the Series
Part One - An introduction
Part Two – Lorenz - one point of view from the exceptional E.N. Lorenz
Part Three – Hays, Imbrie & Shackleton - how everyone got onto the Milankovitch theory
Part Four – Understanding Orbits, Seasons and Stuff - how the wobbles and movements of the earth’s orbit affect incoming solar radiation
Part Five – Obliquity & Precession Changes - and in a bit more detail
Part Six – “Hypotheses Abound” - lots of different theories that confusingly go by the same name
Part Seven – GCM I - early work with climate models to try and get “perennial snow cover” at high latitudes to start an ice age around 116,000 years ago
Part Seven and a Half – Mindmap - my mind map at that time, with many of the papers I have been reviewing and categorizing plus key extracts from those papers
Part Eight – GCM II - more recent work from the “noughties” – GCM results plus EMIC (earth models of intermediate complexity) again trying to produce perennial snow cover
Part Nine – GCM III - very recent work from 2012, a full GCM, with reduced spatial resolution and speeding up external forcings by a factors of 10, modeling the last 120 kyrs
Part Ten – GCM IV - very recent work from 2012, a high resolution GCM called CCSM4, producing glacial inception at 115 kyrs
Pop Quiz: End of An Ice Age - a chance for people to test their ideas about whether solar insolation is the factor that ended the last ice age
Eleven – End of the Last Ice age - latest data showing relationship between Southern Hemisphere temperatures, global temperatures and CO2
Thirteen – Terminator II - looking at the date of Termination II, the end of the penultimate ice age – and implications for the cause of Termination II
Northern Hemisphere forcing of Southern Hemisphere climate during the last deglaciation, He, Shakun, Clark, Carlson, Liu, Otto-Bliesner & Kutzbach, Nature (2013) – free paper (there is considerable supplementary information probably only available on the Nature website)