I have been trying to find a dataset for the earth's long-term historical temperature record.

I have seen a few charts like this one from wikipedia.

Is there an open dataset with this information? If it exists, where can it be found?

  • Original art w/ description, and reference list : globalwarmingart.com/wiki/File:65_Myr_Climate_Change_Rev_png
    – Joe
    Jun 20, 2013 at 1:41
  • Thanks, ya. I have been able to track down two of the three datasets listed in his sources. The third is this article from the magazine "Science". I can't get into it without paying a subscription fee. The other two sources contain information only for the most recent 5 million years. Jun 20, 2013 at 2:45
  • Hmm ... you might try contacting the authors, but I'll go and ping one of the folks I know at globalchange.gov .
    – Joe
    Jun 20, 2013 at 3:32
  • Thank you this is excellent progress. Re: contacting the authors: I haven't actually read the article, so I'm not sure what I would say. Also, I tend to believe that they are protecting the information. Do you think I should try emailing James Zachos? I fear I will be rejected, or even worse, misunderstood, or both! Jun 20, 2013 at 19:57
  • 2
    It's generally better to read the paper before asking people for their data. In some cases, it might be available as an attachment to the article, or have an 'extended methods' document.
    – Joe
    Jun 21, 2013 at 3:37

1 Answer 1


The Zachos et al. article has supplemental data section, but there are no actual datasets in this section, only sources for ΔO-18 and ΔC-13 isotopic data. The article itself says that the temperature estimates are given for an ice-free world ocean. If you want to cite the data from Figure 2 in the article without all the caveats (and then some) of the original, your readers will be either fooled or offended.

A sample caveat from the supplement:

Sampling Biases: One of the limitations on reconstructing long-term secular variations is the highly uneven distribution of deep-sea stable isotope data in both space and in time. The global signal for some key intervals is based on data from just a few records (emphasis mine - DH). In general, these spatial biases increase with age, moving toward the Atlantic, and shallower water depths. In other words, the Pacific, and abyssal portions of the oceans tend to be under-represented in existing stable isotope records. These biases do not pose a problem for our temperature/ice-volume reconstruction of the late Neogene oceans which were thermally homogeneous. Such biases, however, are a concern for establishing the mean climate-state of some "warmer" time intervals when the thermal gradients within the deep-sea were greater.

  • Thank you @deer-hunter and Joe. Based on this answer, I decided to bite the bullet and pay to see the article... until I realized that it costs $20 to view the article for one day! (ouch! I cannot pay this much!) DeerHunter I promise I will include the caveat you provided in any reconstructions I create. Can you provide the sources for the ΔO-18 and ΔC-13 isotopic data? Also, does the article provide a description of the algorithm used to extract temperature estimates from the isotopic data? If you are able to provide any of this it would be awesome. Jun 22, 2013 at 20:27
  • @MattLemmon - possibly later, haven't got it ATM. However, you may be able to get hold of the paper copy of Science at your local library. Jun 22, 2013 at 20:54
  • Ya, I live in Honduras, where the libraries are, um, different. Nonetheless, it's worth trying. I have nothing to lose... If I was in the states, I would have already found the information at the library... perhaps I can get a friend in the states to check it out for me... Well, this is all excellent progress. Onward and upward. Jun 22, 2013 at 22:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.