EGU2020-13972, updated on 21 Apr 2021
EGU General Assembly 2020
© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Atmospheric CO2 during the Late Miocene Cooling

Thomas Tanner, José Guitián, Iván Hernández-Almeida, and Heather Stoll
Thomas Tanner et al.
  • ETH Zurich, Geological Institute, Earth Science, Zurich, Switzerland (

Alkenone sea surface temperature records recently observed suggest a substantial long-term and large-magnitude ocean surface cooling during the Late Miocene. At the same time, starting about seven million years ago, both hemispheres on Earth witnessed synchronous cooling and large areas of the continents experienced drying and enhanced seasonality. Coinciding with this climatic shift were significant changes in ecology, including the rise of C4-photosynthesizing terrestrial plants and the emergence of so-called "vital effects" in oceanic coccolithophores. These changes are collectively hypothesized to be induced by declining atmospheric CO2. However, the sparse proxy data available for this time interval limits our understanding of the link between these changes and atmospheric greenhouse gas fluctuations and has let people to propose a "climate-CO2 decoupling".
In this study, the alkenone based pCO2 proxy is used to reconstruct atmospheric CO2 for the time interval between 4.5 and 8.5 Ma. Estimations are based on the carbon isotopic fractionation during photosynthesis (εp) and a new statistical multilinear regression model based on an analysis of culture and sediment data. Past coccolithophore growth rates are reconstructed using foraminiferal isotopic-based proxies, related to water column structure which favour or limit nutrient supply to the photic zone. A thorough sensitivity analysis of modern and past  εp values and its influencing factors in the Southern Ocean yield to a new, high resolution pCO2 record. Estimated pCO2 concentrations synchronously decline with the observed long-term cooling (5°C) from 6.8 to 5.9 Ma, periodically decreasing to sufficiently low values of <200 ppm, potentially inducing ephemeral Northern Hemisphere glaciation. CO2 concentrations during the Late Miocene Cooling Event are thus successfully reproduced in this study and allow a reasonable interpretation of past conditions as has not yet been previously achieved in the relevant literature. 

How to cite: Tanner, T., Guitián, J., Hernández-Almeida, I., and Stoll, H.: Atmospheric CO2 during the Late Miocene Cooling, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-13972,, 2020.


Display file

Comments on the display

AC: Author Comment | CC: Community Comment | Report abuse

displays version 1 – uploaded on 01 May 2020
  • CC1: Comment on EGU2020-13972, User deleted account, 04 May 2020

    Hi Thomas, Nice work. Very usefull for us modellers :).

    To reply on your comment of: "Thomas Tanner ETHZ (author) (14:50) @David: Thank you. I would guess but might be a very strong biological pump that shuts off during ice-sheet advances."

    I don's these strong changes back in the benthic d18O record. I would think it's only Antarctica at this time or perhaps small ice caps in the NH, but not significant ice growth.. well don't have sea-level data to support this though.. Also, how certain are you on the low values of pCO2?

    • AC1: Reply to CC1, Thomas Tanner, 06 May 2020

      Hi Bas, Thank you.

      Sorry for not getting back to you sooner. It is difficult to show the confidence we have in the lower (or upper) end of the error bar. In the paper, we show multiple sensitivity studies that show how and how much different factors influence our final CO2 values. It is also important to note that the statistical model we use (Stoll et al. 2019) has some limitations as well (e.g. limited amount of culture data). This is one of the sources of the quite large error envelope.

      • AC2: Reply to AC1, Thomas Tanner, 06 May 2020

        Additionally, about your comment of the ice sheet:

        We see a very strong freshening of the surface waters (see display slide 4) in the d18O of up to 1.5 permil. We think that this shows a strong northward movement of the polar front in the Southern Ocean, potentially together with a large expansion of sea ice. Maybe thats why it's not visible in the sea-level reconstruction...

  • CC2: Comment on EGU2020-13972, Charlotte Beasley, 04 May 2020

    Hi Thomas, really interesting study - looking forward to reading the paper!

    It's interesting that the decreases in the biological pump (Dd13C) aren't synchronous with the lows in the eccentricity and CO2 records (or perhaps appear to lag these changes?), but are synchronous with the end of the LMC. Do you have any ideas why these offsets may be occurring?


    • AC3: Reply to CC2, Thomas Tanner, 06 May 2020

      Hi Charlotte, sorry for not answering earlier.

      We suspect that the biological pump might react on longer timescales. That might explain why it does not express a 400kyr-cycle but might trigger (by stopping/slowing down) the release of stored carbon back into the atmosphere and quickly reverse the long-trend cooling.

      Honestly, here I struggle to find a good explanation...