An article about ocean temperature history/modeling has this plot:

Purty colors

Where the units are given as cK.

cK does not seem to be a standard temperature unit and Google searches return millions of results that don't seem relevant.

While it might be "fun" if the temperatures shown were Kelvin, centi-Kelvins, or even Celsius, that seems unlikely as we are all still alive.

So, what is cK as shown on the graphic, and how do you convert the temperatures shown to degrees Celsius?

If the axis is really some delta temperature, what is the zero point?


1 Answer 1


It actually is the common centi- prefix, but applied to temperature changes in units of Kelvins rather than absolute temperatures. The research as quoted by the Daily Mail obscures this by simplifying the results for a general audience.

The analyses and results are much more clearly described in the original refereed journal article, which clearly refers to changes in temperature - see in particular the discussion of their Fig. 1 (reproduced below), which explicitly refers to rates of temperature change of 0.1 and 0.02 °C at depths below 2000 m over the previous century, corresponding to cK contours of ~10 and near zero in the corresponding section of that figure.

So, the contours in the figure quoted by the Daily Mail represent changes in temperature between measurements made in the 1870s and the early 2000s, in units of 0.01 K (or equivalently, 0.01 °C).

Fig 1 Figure 1 from Gebbie and Huybers (2019), https://doi.org/10.1126/science.aar8413 - reproduced here for non-commercial use.

  • $\begingroup$ The article you linked says that the question plot ("Fig. 2") is not over the last century, it shows changes "between the 1870s and 2000s". Add that to the answer and I will accept it. ... Also, should the axis be labeled something like "Change (cK)" or "𝚫 cK" etc? $\endgroup$ Commented Jan 11, 2019 at 20:01
  • $\begingroup$ That's correct, I edited my answer to clarify both the timescale shown in the Daily Mail's figure and point out the differing timescale I referred to for the figure I quoted. I would label the figure "𝚫T (cK)" personally - not to mention adding a request for more descriptive labels overall if I had been a reviewer :) $\endgroup$
    – dplmmr
    Commented Jan 11, 2019 at 20:10

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.