Earth's response to CO2 underestimated

Global warming resulting from slowly changing Earth systems could be up to 50% greater than previously thought, according to research by UK and US scientists. The study reinforces the notion that certain poorly understood systems such as ice sheets or vegetation are integral to accurately predicting future temperatures. It also paints an ever-bleaker outlook for our planet at a critical time when world leaders are gathering for a United Nations conference in Copenhagen to discuss practicable ways of mitigating climate change.
Last edited by a moderator:


The Middle Pliocene was a period in earth’s history between 2.85 to 3.15 Million years ago.
It was a significantly warmer period lasting about 300,000 years. The continents were within a hundred miles of where they are now with plants and animals similar to today.
There was less ice in the Antarctic and a minimal ice cap on Greenland.
What is now tundra was covered with forest. There were also fewer deserts as they were
replaced by tropical shrubs and savanna. Sea levels were about 35 meters higher.

However, CO2 levels were only slightly higher than they are now. Much of the additional
warmth was a consequence of less ice at both poles and additional vegetation that absorbed more of the sun’s energy. Precipitation patterns were also altered as were the ocean currents.

At 0.016C/year, Middle Pliocene like temperatures could be achieved within 225 years.
At 3mm/year, Middle Pliocene sea levels could be obtained within 12,000 years.
Obviously, there could be some acceleration of the rate at which ice sheets are melting.
However, these time periods illustrate the scale over which full realization of the warming could be felt. Even with some acceleration, it’s going to take a long time to melt much of Greenland and for trees to grow on the exposed soils.
There are other sets of CO2 estimates covering this period and the general trend of CO2 over the past 20 million years is for CO2 at around 280 ppm (+/- 40 ppm).

Individual estimates for individual time periods can be cherrypicked but looking at the overall trend over time and the variability that exists in the various estimates, a good average for CO2 at 3 million years ago would be 280 ppm. [Broken]
Last edited by a moderator:


Nice chart Bill;

Do you have a link to the source?

The Middle Pliocene occurred around 3 million years ago and may line up with a spike on your chart that briefly rises to about 1100 ppm. The values I've seen are in the range of 360 to 440 ppm.

During that period, CO2 values are estimated to have reached 360–440 ppmv (e.g. Raymo et al. 1996) and global mean annual temperatures (MATs) were approximately 3°C higher than today (Chandler et al. 1994; Sloan et al. 1996; Haywood et al. 2000).

and this from Raymo:

Three million years ago, prior to the onset of northern hemisphere glaciation, global mean temperatures may have been as much as 3.5 °C warmer than at present. We present evidence, based on the carbon isotopic composition of marine organic matter, that atmospheric CO2 levels at this time were on average only about 35% higher than the preindustrial value of 280 ppm. We also present carbon isotopic evidence for stronger thermohaline circulation in the Atlantic Ocean during the warmest intervals and propose that the North Atlantic “conveyor belt” may act as a positive feedback to global warming by enhancing sea ice retreat and decreasing high latitude albedo. Based on our results, it seems unlikely that the mid Pliocene warm period was a doubled CO2 world.

135% of 280ppm is 378ppm.
Current CO2 levels are around 385ppm.

However, Raymo is also pointing to a stronger thermohaline circulation as a positive feedback and of course it will take a long time to melt ice and for plants to re-establish themselves.
One can also cherrypick temperature estimates as well.

Raymo's newest dO18 isotope stack would put the temperatures 3 million years ago at just a little above today (not quite +1.0C from today would match up with the other dO18 isotope temperature estimates). [Broken]
Last edited by a moderator:


Don't cherry pick!

Keep in mind that dO18 percentages are not 1:1 equivalent to temperature. Also, there is no indication of where the data for the above chart was collected. Was it a tropical location? Anyhow, this is why it's always good to include a link to a peer reviewed paper for what ever data or chart is being presented. It helps everyone from being mislead.

Here is a 2009 paper from Raymo and others that has found North Atlantic sea surface temperatures during the early pliocene about 6C warmer than present. [Broken]

We find that while ocean surface temperatures were significantly warmer (6C) than modern temperatures during the early Pliocene, they were also as variable as those during the late Pleistocene, a surprising result in light of the subdued variance of oxygen isotopic time series during the interval of 3–5 Ma.
Last edited by a moderator:
All these isotopes need to be calibrated to temperatures we are reasonably sure about.

The dO18 isotope data does not end at 4 million years ago. It goes back to 540 million years and earlier.

If one is going to put a +3.0C or +6.0C on the values of 3 million years ago, then 15 million years ago, we are up to +9.0C to +18.0C. The Eocene Maximum goes to +20.0C or more. The Cretaceous Hothouse is closing in on +30.0C.

Then one has to match up to the ice age temperatures of -4.0C to -5.0C.

If we are comparing CO2 sensitivity to temperature, then the global temperature estimates should be used rather than the polar equivalent temperatures (which vary by twice as much as the global temperature) (site 983 in the North Atlantic is probably a little less than twice and I don't know much about this Uk37 isotope).

This page shows the global temperature numbers based on the dO18 isotopes (I have all this data as well).


I agree Bill; but I'm not sure if that Wikipedia graph has been peer reviewed.
There appear to be some discontinuities.

UK'37 is an index of the di-unsaturated to tri-unsaturated alkenones found in Coccolithophores.
It is measured using a gas chromatograph.

UK′37=[C37:2]/[C37:2+ C37:3]

As far as I know, it's a good method for measuring sea temperatures from sediments up to 100 million years old.

T (οC) = (UK′37 - 0.039) / 0.034

Prahl, F.G. & Wakeham, S.G. (1987) Calibration of unsaturation patterns in long-chain ketone compositions for palaeotemperature assessment, Nature 330, 367-369
I found the data on the Lawrence-Raymo 2009 paper using the Uk37 isotope data and it seems to be calibrated okay to the global temperature estimates (might be high by 0.5C or so and the age estimates might be off a little, it doesn't quite match the Greenland ice core temps but close enough) ...

What this data does let us do, however, is look at the last several Ice Ages in the Northern Hemisphere and compare that to the Milankovitch Cycles. [The Greenland ice core data doesn't really make it to the height of the last interglacial or we are not sure about the ages at the bottom of the core because there is so much ice distortion].

Previously we only had the Antarctic ice core temperature data which seems to match the Northern summer solar insolation much better than it matches the Southern summer solar insolation data.

So, as one can see, we still do not get a good match to the Northern summer solar insolation either (I used the data at 75N rather than 65N because I am more interested in the Albedo and sea ice impacts but 75N is almost identical to that at 65N). [Broken]
Last edited by a moderator:

The Physics Forums Way

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving