Global deep ocean temperature for past 60 million years
Global deep ocean temperature for past 60 million years
The paleoclimate record provides evidence that Earth's temperature started cooling about 50 million years ago and that about 34 million years ago it had dropped by about 5 °C when Antarctica began to glaciate. Since then Antarctica has remained glaciated. The temperature 52 million years ago was about 14°C higher than the last ice age 20,000 years ago. CO2 concentration, which was about 180 ppm in glacial times, was much higher, about 1400 ppm, at the beginning of the cooling period leading to the glaciation of Antarctica. This is in qualitative agreement with evidence from a wide range of sources that has revealed a close correlation between temperature and atmospheric CO2 concentration. Assuming that CO2 does drive temperature, the IPCC has estimated a quantity called climate sensitivity. When the atmospheric CO2 concentration doubles, the Earth's surface temperature is projected to increase by an amount between 1.5–4.5 °C in the short term. (As the amount of snow and ice covering the Earth's surface decreases, its reflectivity or albedo decreases and the surface absorbs more solar energy. In the longer term when changes in the Earth's albedo are taken into account the temperature increase could be double that amount.)

CO2 concentration for the past 65 million years
CO2 concentration for the past 65 million years
Symbols with error bars Observed CO2 concentration for the past 65 million years
long-dashed grey line Present (2012) CO2 concentrations
short-dashed grey line Pre-industrial CO2 concentrations
light blue shading Uncertainty band

The paleoclimate history of atmospheric CO2 during the period 50 to 34 million years ago provides a long term perspective for assessing the impact of varying atmospheric CO2 levels on the Earth's climate. For example, it has been argued that decreasing CO2 was the main cause of the cooling trend that began 50 million years ago. Support for this conjecture has been limited by two issues. Knowledge of atmospheric CO2 levels has been limited by widely varying observations except for recent ice core data covering the last 800,000 years. This study addresses this issue by providing reliable estimates of atmospheric CO2 levels during this period. Secondly, it has not been possible to determine whether temperature or CO2 has led during periods of deglaciation or whether CO2 has acted to amplify temperature changes. It has only recently become possible to measure the paleoclimate record during the last deglaciation with sufficient resolution to determine whether rising temperature or increasing CO2 drove climate change.

It has been argued that a major source of CO2 during this period may have resulted from the subduction (one continental plate sliding under another) of carbon-rich crust when the Indian subcontinent collided with the Asian continent. Between 60 and 50 million years ago the Indian subcontinent moved north rapidly, averaging about 18-20 cm/year, through a region that long had been a center for deposition of carbonate and organic sediments. It has been suggested that the subduction of carbon-rich crust may have provided a large source of CO2 outgassing. The warming climate peaked 50 million years ago when the Indian subcontinent collided with Asia. It is expected that CO2 then dropped because there was less subduction. As well there was increased weathering (which removes CO2 from the air) as the Himalayan/Tibetan plateau rose. Since then, the Indian and Atlantic Oceans have been major centers for the deposition of carbon. The subduction of carbon-rich crust has been limited to small regions near Indonesia and Central America. Thus atmospheric CO2 would be expected to decline following the Indo-Asian collision.


In this paper recent analytical and methodological developments have been used to generate a new high-fidelity record of CO2 concentrations using the boron isotope (delta-boron-11) composition of well preserved plankton shells from the Tanzania Drilling Project. Boron isotopes (delta-boron-11) in marine carbonates are a well understood proxy of seawater pH, allowing for high-fidelity reconstructions of atmospheric CO2.

CO2 concentration 50-30 million years ago
CO2 concentration 50-30 million years ago
Comparison of CO2 concentrations with delta-boron-11 and delta-oxygen-18 ratios
a - delta-boron-11 of shallower foraminifera, asterisks represent the average of those for each time slice. Blue squares and orange circles represent the warmest species
b - Atmospheric CO2 using the warmest species of each time slice
c - delta-oxygen-18 from bottom-dwelling foraminifera


The new data confirms that atmospheric CO2 was probably (at a 95% confidence interval) greater than 1,000 p.p.m. around 50 million years ago. The data reveals that atmospheric CO2 concentration decreased from 1,400 ± 470 p.p.m. around 52 million years ago reaching a minimum of 550 ± 190 p.p.m. around 34 million years ago. If CO2 did drive this cooling, climate sensitivity during this period when the Earth was 10 °C warmer than at present was estimated to be 3.8 °C with an uncertainty interval of 2.6–4.6 °C (66% confidence) which is within the IPCC estimated range of 1.5–4.5 °C.


Changing atmospheric CO2 concentration was the primary driver of early Cenozoic climate, Eleni Anagnostou, Eleanor H. John, Kirsty M. Edgar, Gavin L. Foster, Andy Ridgwell, Gordon N. Inglis, Richard D. Pancost, Daniel J. Lunt & Paul N. Pearson, Nature 533, 380–384 (19 May 2016) doi:10.1038/nature17423

Are long term effects of elevated CO2 really what we should be worried about ?

CO2 drove, but did not trigger, warming during last deglaciation

CO2 lagged rising temperature in the Southern Hemisphere during last deglaciation

IPCC AR5: Earth surface temperatures, greenhouse gas concentrations, sea level rise and ice loss from the paleoclimate record