Talk:Geologic temperature record

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Geology (Rated Start-class, Mid-importance)
WikiProject icon Geologic temperature record is part of WikiProject Geology, an attempt at creating a standardized, informative, comprehensive and easy-to-use geology resource. If you would like to participate, you can choose to edit this article, or visit the project page for more information.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.


I recently created this page to give a deep time perspective to fluctuations in Earth's temperature history. Right now it is quite qualitative and not at all well documented. I intend to continue to improving this page over time, but it may be a while before I have ample opportunity to do so. Dragons flight 06:59, Mar 3, 2005 (UTC)


I believe you should merge this with Paleoclimatology, as that is the term for old climate. You apparently created this page by thinking of "temperature record" over "time scales used in geology". However, this page may be suitable for the paleoclimate records which have been extracted from rock and geologic deposits. — SEWilco 19:12, 4 Mar 2005 (UTC)
(previously posted reply from Talk:temperature record) I had forgotten about that page. I'm probably not impartial, but the paleoclimate page is rather icky right now with things that don't belong and material that should be there which isn't. In my opinion, wikipedia could tolerate both a well-written page on paleoclimatology (including methodology, forcing factors, and an outline of past climate) and a more focused page, such as I have tried to start, that just goes back and tries to describe climate changes through geologic time. However, I am willing to entertain other opinions, such as merging the two, or whatever. Dragons flight 08:54, Mar 3, 2005 (UTC)
It looks like a lot of this material fits in the existing structure of Paleo. Merge them, then deal with restructuring. Some subgrouping usually becomes apparent during a merge. — SEWilco 22:11, 4 Mar 2005 (UTC)
I will keep paleoclimatology in mind, but I think I am probably going to make additional refinements to this page in its current location before deciding what to do long term. Dragons flight 22:37, Mar 4, 2005 (UTC)

Processes affecting global temperatures[edit]

Re Connolley's {{cn}}, entirely agree. Actually, the much-cited Royer et al says precisely the opposite. This whole section is loaded with pretty arguable stuff. Needs re-wording for NPOV. --Gergyl (talk) 00:57, 31 January 2008 (UTC)

The 70M-year Geologic Interval Section has been removed because? All data presented was verified by the references which are still included in the article. It appeared that this entry was a competing assemblage of verifiable facts, properly referenced. Because this was presented in this article, and has been deleted, violates NPOV. MOShaver —Preceding unsigned comment added by (talkcontribs)
I cut it [1] because it had no refs. Are you Shaver? If so, which papers are you referring to? As to the ice age stuff at the end... what did you have in mind? William M. Connolley (talk) 20:17, 2 February 2008 (UTC)
On second thoughts, I hacked out even more [2] William M. Connolley (talk) 20:20, 2 February 2008 (UTC)

Yes you did cut out all competing discussions. The reference of P.C.Frisch "Solar Journey: The Significance of our Galactic Environment for the Heliosphere and Earth" ISBN-10 1-4020-X(HB) has reference to GMC ice age causal explicity indicating a 1000/cm3 and 20KM/sec penetration threshold for the Sol Earth Heliosphere (see Chapter 11 Fahr and Yeghikyan). Within this same book, reference to the Galactic Gas ouflow rate of 26KM/s is made in reference to published papers. User Morbas, 2008 February 2. —Preceding unsigned comment added by (talk) 20:38, 2 February 2008 (UTC)

OK, lets try to make sense of this...
  1. Alternately, Shaver has recently correlated.... Who is Shaver? You signed yourself MOShaver, so I thought it might be you. Or it might be a misspelling of Shaviv. Which?
  2. P.C.Frisch, Fahr and Yehikyan, have presented papers that interstellar material (ISM) density of 1000cm-3 at 20KM/sec will eliminate the Sol Heliosphere at 1AU, causing O2 depletion, Ozone depletion and an Ice Age. This is garbled. What does it mean? Is it a prediction of the future, or an explanation of the past? What is the source...? You seem to be now saying its F&Y, so why add F?
William M. Connolley (talk) 20:48, 2 February 2008 (UTC)

1) I am M.O.Shaver, alias morbas both here and on There is another Shaver,an editor, you can correct me on this.

So when you wrote Alternately, Shaver has recently correlated.... you were referring to your own unpublished work? You're not allowed to: see WP:OR William M. Connolley (talk) 10:57, 3 February 2008 (UTC)
I somehow question the fair play nature of your WP:OR arguement. (talk) 03:23, 4 February 2008 (UTC)
I published it on the web. It was open to review and I had advertised this to many scientists in the same field. I am writing a professional level paper, being audited/reviewed/edited by an anonymous PhD. I am presently waiting for a very relevant paper to be published, so I am taking my sweet time.Morbas (talk) 16:35, 3 February 2008 (UTC)

2) a) Honorable P.C.Frisch indicate that Interstellar Clouds impinging on the Solar System with this intensity will cause "Accretion of Interstellar Material into the Heliosphere and onto Earth". As a consquence the hydrogen impact will purge the O2, the Ozone, and cause an Ice Age temperature depression. Chapter (11) is an analysis made by H.Fahr and A.Yeghikyan that evaluates this penetration of the Heliosphere. In this anaylsis they also attempt to evaluate the probability of such an intersection and come to a few hundred Million years frequency. This is in the above referenced book by P.C.Frisch. I presented the 70M-year interval to suggest a strong correlation, and also included the 220M-year (Kvet) to show all periods were accounted for. I was not picking a few among all the Periods. Which brings us to the Central Bar being aligned to this 210M-year Nemesis and the Oligicene Ice Age. I added F as a consequence, the present Central Bar +28 degree position, adding about one degree per M-year.

So the FF&Y stuff is for the past, not the future? What does "purge the O2 mean"? William M. Connolley (talk) 10:57, 3 February 2008 (UTC)
The hydrogen inflow combines with the O2 and O3 (per P.C.Frisch). The future is apparent because Sol will pass across three arms cycling into the next zero degree central bar alignment. The past 210M-years will re-cycle, however Sol main sequence will favor more GW.Morbas (talk) 16:27, 3 February 2008 (UTC)

2 b) Also Frisch, P.C., 2006, Solar Journey: The Significance of our environment for the Heliosphere and Earth, pp55 et al. Witt M (2004) on Ulysses GAS HE flow.ISBN-10 1-4020-4397-X indicated measured intergalactic gas flow rates at 26KM/sec vectored from the Galactic Core (relative to Sol).

Morbas (talk) 23:47, 2 February 2008 (UTC) The excellent references have been removed, as they have no relevance to your highly edited discussion. Good Luck...Morbas (talk) 14:42, 4 February 2008 (UTC)

wikiproject Earth[edit]

Hello i have recently proposed the Wikiproject Earth. This Wikiproject`s scope includes this article. This wikiproject will overview the continents, oceans, atsmophere and global warming Please Voice your opinion by clicking anywhere on this comment except for my name. --IwilledituTalk :)Contributions —Preceding comment was added at 15:38, 30 March 2008 (UTC)

{{cn}} tags[edit]

Most of the things you're {{cn}}'ing are uncontroversial, and discussed at length in the references given on the graph image pages. Why not transfer the citations from the graphs, rather than just littering tags everywhere?

"Ma" unit[edit]

Regardless of ISO/SI, the edit effectively makes the article unreadable for most. The sense is clear with the unabbreviated form ("million years ago"). Why destroy it with a frankly obscure abbreviation ("Ma")? --Gergyl (talk) 02:57, 14 May 2009 (UTC)

Defined and wikilinked the first occurrence. Awickert (talk) 04:59, 14 May 2009 (UTC) (signed late)
Still not happy. Regardless of standards, your abbreviation is not current common usage. I cannot see how it is superior to the unabbreviated form that was used here. Anyway, how is this english, " age that began about 40 Ma with the glaciation of Antarctica"? "Ma" means "million years" not "million years ago"; so s.r. "40 Ma ago".--Gergyl (talk) 04:49, 14 May 2009 (UTC)
Sorry - forgot to sign earlier - I actually didn't change it to Ma; it's one of those weird things in that it is typically used for years ago. I didn't actually change it to Ma in the article; I'm not sure about common-common usage, but it is common scientific usage, so perhaps it will be useful (if properly linked and defined) to teach people what it means. Awickert (talk) 04:59, 14 May 2009 (UTC)
WikiProject Geology suggests the use of 'million years ago' / 'million years' in full if practical, which I think it is in this instance. I've implemented this with the {{Ma}} template. Martin (Smith609 – Talk) 13:16, 14 May 2009 (UTC)

Logarithmic scale seriously misleading.[edit]

The "Temperature of Planet Earth" graph is seriously misleading in a very subtle way, in that it has a logarithmic scale. I think most people would not notice it (I missed it myself at first glance), with the resulting impression that those most recent wiggles on the right end are minor compared to the big swings on the left (when the earth's crust was still molten). There are times when log scales are useful, but they are too sophisticated for general use.

It may be interesting to see the long-term decline in temperature of the planet, but this should be on a constant scale. Expansion of a segment to see finer detail should be done in the usual way. - J. Johnson (JJ) (talk) 23:01, 15 January 2010 (UTC)

Hi JJ. The scale isn't logarithmic (actually that would be reverse logarithmic - it's compressed to the left). What there is is just a change of linear scale at each vertical grid line. Those are the places where the graphs from the various sources are joined. The x-axis labels show that, though I'd agree it might not be completely obvious on first viewing. Maybe needs an explanatory note?
This sort of age-compressed plotting is common in geology. It reflects the fact that we know less and less about longer and longer ago (the data resolution is lower), so progressively less time-detail is generally shown. We also tend to be more interested in the more recent, so the scale is expanded. You can see that in the way the geological timescale is defined and plotted.
If the temperature record plot was same-scale linear for the whole ~540 My of the Phanerozoic, all of the interesting ice age detail (Pleistocene - 2 My) would be so compressed that it'd be invisble. The recent temps (Holocene - 12 ky) would all plot as one point. Zachos, for example, plots just the 65 My of the Cenozoic same-scale linear, and the recent detail is invisible. --Gergyl (talk) 22:25, 17 January 2010 (UTC)
The Holocene scale is non-linear, close to logarithmic. (Well, strictly speaking it is semi-logarithmic, the vertical scale being linear, or if you wish, reverse semi-log.) On all other points you are correct, yet the chart is misleading. Geologists (and engineers) understand about variable (e.g., semi-log) scales, but most readers generally do not, and the image powerfully projects the misleading notion of a wildly variable history only recently settling down to stability. Unless they serve a particular purpose, and are communicated to and understood by the intended reader, changes of scale are so tricky to handle that they are properly suspect.
What I would suggest is a stack of horizontal charts, each descending chart being a segment of the prior chart at an expanded scale, but each chart having a strictly linear scale. So the topmost chart could represent four billion years across the width of the page, the next chart, also across the width of the page, could represent, say, the last five million years, then another chart showing, say, ten thousand years, and each chart with the lines connecting it with the segment of the previous chart. (I regret I don't have a sample right at hand, but I'm sure you know the kind of thing I'm talking about.) This kind of arrangement clearly shows where each sub-part fits into the higher level, and, if properly done, graphically communicates that there are different scales. - J. Johnson (JJ) (talk) 20:49, 19 January 2010 (UTC)
With respect, you are incorrect. No logarithmic scales are used in this plot.--Gergyl (talk) 22:14, 24 January 2010 (UTC)
For sure, the scale of the leftmost chart ("Holocene") may not be precisely logarithmic, but it is non-linear. They key point is that what we have here is essentially five different charts, with different scales, cobbled together as one. And these differences of scale convey a misleading impression to everyone that does not pay careful attention. - J. Johnson (JJ) (talk) 22:50, 24 January 2010 (UTC)

Logarithmic scales are good.[edit]

The figure used in this article All palaeotemps.png is absolutely horrible. It hints at a logarithmic scale and then just starts using a slightly different scale along the horizontal axis. This whole article is surprisingly bad given everyone is trying to change the world based on this data. There needs to be a scientist in the field who has access to the data to create a good plot to replace this plot. Mjspe1 (talk) 04:31, 4 March 2010 (UTC)palaeotemps.png

(Yeah, just what I was saying! :-) There may already be suitable charts, but they have to be located. The big problem is that nearly all the scientific journals where the reputable charts would be found are not "free". Which can be dealt with, but first the charts (and underlying data?) have to be located. Also, has anyone checked all the IPCC material for any suitable charts? - J. Johnson (JJ) (talk) 22:51, 5 March 2010 (UTC)

For the record, even though I made several of the individual panels, I consider that stacked plot (made by someone else) to be really icky. Even beyond the awkward aesthetics, I think the stitch between the first and second panels is probably significantly in error. I have at various times tried to make a logarithmic (or compressed) plot of long-term temperature, but I've never managed to make anything I felt comfortable showing. It is very easy to be very misleading when you are comparing multi-million year averages to the temperature last month. Dragons flight (talk) 23:08, 5 March 2010 (UTC)

Don't get me wrong, the individual plots are good. The stitch is weird. If you have access to the data and plotted the originals (that were stitched) PLEASE create a real logarithmic plot. No matter who people are they can look at the horizontal axis and see that the time is 300M years ago, or whatever. If you have to, make it a linear horizontal axis with the same data so that dolts can see what is happening. This article is a shambles without a proper plot of the temperature record. Mjspe1 (talk) 05:27, 8 March 2010 (UTC)
"Icky", well, yes. Aesthetics are your specialty, so why not fix it, as I suggested long ago.
I don't see what's wrong with the first stitch. The Royer et al scale at the left is reproduced exactly from the reference. It's meant to represent global temps AFAICT. Your two scales (Vostok and Polar Ocean) are polar, so an amplification factor is expected, and x2 (approx. as used) is a common choice. The splice can only be approximate due to the ice volume effect on the Zachos et al temps, but at least the temperature traces roughly meet at the join.
The OP contended "everyone is trying to change the world based on this data" (really?), so I'd be just a little cautious in assuming good faith here Rob. --Gergyl (talk) 02:35, 6 March 2010 (UTC)
Look, I am not trying to suppress the data or anything. I am acting in good faith. No matter where you are coming from, the horizontal axis on this plot is awful. Mjspe1 (talk) 05:27, 8 March 2010 (UTC)
My apologies then. I'll certainly grant you that the axis scale is ugly; but awful, no. The original purpose of this plot was to combine all the stuff that Dragons flight had so usefully plotted together on one page, so we can see the whole 500 My history of the living planet at a glance. This data amounts to one of the finest achievements of modern geology (and all in the last 20 years). That is actual proxy measurements of temperature going back 540 My, not just the traditional geological inference.
As discused above, a uniform linear scale is unsuitable for the purpose because it loses most of the interesting recent detail. We are progressively more interested in more recent detail; and we do know progressively more about it.
Unfortunately DF says he has found that a (reverse) logarithmic scale is also unsuitable. That's not really surprising, because:
  • Time is not a naturally logarithmic variable. It's rarely plotted on a log scale, and almost never in geology.
  • Our data has a zero (this year's temperature), and a zero cannot be plotted on a log scale.
  • A log scale tends to over-expand really recent data. For example, on a log scale the decade 2000-2009 would occupy about the same width on the graph as the whole of the Tertiary. The century 1900-1999 would occupy the same width again. Try it and you'll see straight away what I mean.
So I still think the best choice is a piecewise linear expanding scale, which is what is used here (and more generally in geological timescale plotting). Could it be improved? Absolutely. It sure seems to need more and better explanation.
(Could you point us to some examples? - J. Johnson (JJ) (talk) 23:20, 12 March 2010 (UTC) )
Um, I pointed you to one above JJ. Maybe try Google: eg here, or here.--Gergyl (talk) 06:04, 13 March 2010 (UTC)
What I started out trying to do was to make the "5 to 5" intervals roughly the same width across the graph. So ~500 to ~50 Myr gets one chunk, ~50 to ~5 Myr a similar chunk, ~5 MYr to ~0.5 Myr another. So far those happen to fit pretty well with the source plots, so it was just a matter of making them equal widths (yes, I know, not quite). After that the whole thing gets less manageable. Like a log scale, the expansion is just too strong if the rule is continued up to the present. And the more recent sources are more disparate, making it logical to break the scale at the source change rather than at even intervals.
So why not make some useful and practical suggestions here? I'm prepared to try again, or if we can persuad DF to do it, all good. But he seems less interested these days.--Gergyl (talk) 08:16, 10 March 2010 (UTC)
You make some good points, thankyou. I am not saying the data is bad, just the plot of the data. At the very least, there should be small gaps between each of the seperate time scales so that it is immediately obvious that the horizontal scale is discontinuous. What do the words 'climatic optimum' refer to? My other practical suggestion you can find below. If the data was easily accessible then we wouldn't even be discussing this. I could have had a go at plotting it up myself. Mjspe1 (talk) 22:33, 11 March 2010 (UTC)
Most of this data is available in public databases or in the supplementaries published with the source papers - eg. Zachos et al is still here. There's about a dozen sources though, so it's no small task. And if one were replotting the more recent stuff there'd be several newer sources you'd want to add. Also it's not all simple. A degree of specialist understanding is sometimes required to plot things sensibly, which DF has so far supplied. Eg. the Zachos temps are from marine oxygen isotope ratios, which depend on the global ice volume as well as temperature. That's why DF used two temperature scales - one for the early part where there's no ice, and another for the last bit where the baseline is reasonably understood (the bit in the middle stays in the too hard basket, but isn't totally meaningless).--Gergyl (talk) 06:04, 13 March 2010 (UTC)

Plot idea: Rate of change in temperature[edit]

It may be useful to show a 'rate of change' plot. One where the vertical axis is in `deg/year' and the horizontal axis is `year' (logarithmic). It is this which will emphasize the drastic events of recent times, since the dramatic change in temperature in such a short period of time is unprecedented. Perhaps Dragons flight could create this plot?? Mjspe1 (talk) 05:35, 8 March 2010 (UTC)

I think the objections to the current chart are essentially of two kinds. First, the "stitching together" of disparate scales. That really isn't good, so having a uniform (single) scale would be an improvement. The second objection is to having a non-linear (log, semi-log, or whatever) scale. Yes, these are often useful, but generally for experts; many general readers are too easily misled by them.
What I would suggest is one of these – I don't what they are called, stepped? stacked? – charts. Like a stack of horizontal charts, lower ones being at a larger scale, with lines showing the segment it represents of the larger chart. (Sorry I don't have any examples right at hand.) - J. Johnson (JJ) (talk) 23:00, 8 March 2010 (UTC)
I understand what you mean, but such a stacked chart makes it very difficult to compare the temperature over time. Defeating the whole purpose of creating the chart/charts in the first place. No matter what you think about the debate, you must agree that people should have access to clearly presented data. —Preceding unsigned comment added by (talk) 23:39, 8 March 2010 (UTC)
For sure, data should be presented clearly. (Which is why non-linear scales for non-experts is problematical.) But a critical aspect of this issue is the proper interval of "time" over which temperature changes are to be considered. E.g., do you want to show the range of temperatures over which Earth has supported life? (Which has varied considerably, and implies that global warming is no big deal.) Or do you want to show the drastic impact of anthropogenic warming, which has some very serious implications for the current environmental balance (like species extinctions), and which is evident in only the past century? Both of these approaches is misleading, depending on the intended purpose. The advantage of a stacked chart is being able to show both long-term and short-term trends, and (!!) show them in a context that relates them ("this one is a subset of that one"). - J. Johnson (JJ) (talk) 23:08, 9 March 2010 (UTC)
Nobody has yet addressed the suggestion under this discussion heading. That the plot show the rate of change in temperature. This is the really dangerous quantity which should emphasise the recent temperature changes. If only someone could tell me where to access the data I would do it myself!Mjspe1 (talk) 21:26, 10 March 2010 (UTC)

You are quite right. While all the above discussion is applicable either way, we have overlooked your idea of plotting not the changes of temperature, but the rate of change. I think this is actually quite a good idea, as the key problem of "global warming" is not whether life will continue at a higher temperature, but whether the changes will come faster than existing life (with all of its many ecological adaptations) can cope with. So I have to wonder: why haven't I seen this done? (I may have slept in that morning.) As to getting such a chart, my first approach would be to search, especially on Google Scholar. If nothing turns up, then start e-mailing some of the authors of various articles. (They are often quite helpful.) If no charts are available, some one might still have some data. But this starts to run into a problem: if the data is not published in some way then we run afoul of original research. But take a look around, see what you can find. - J. Johnson (JJ) (talk) 22:54, 10 March 2010 (UTC)

General review and improvements needed.[edit]

Looking around a bit I see two other Wikipedia articles (Temperature record and Temperature record of the past 1000 years) not even mentioned here, and similar enough that I think the relationship between these articles ought to be reviewed. There is also the Paleoclimatology article (discussed above), which I think sets a standard this article has yet to attain. Not that these should necessarily be combined – there is something to be said (as was said above) for this article being the narrower, more specifically focused article. I think it would be to the betterment of the entire topic to have a broad review of just what each of these articles should cover, and how they can work together. With that in place this article could work on various improvements. - J. Johnson (JJ) (talk) 20:54, 13 March 2010 (UTC)


A little Googling found some possible good sources (and a whole slew of unreliable sources). I would recommend, which has various examples of how this data can be presented, including an example of the stacked chart I suggested above. The NOAA site ( site has links to a lot of data for anyone that wants to try constructing a chart, though I wonder if it might be better to stick to charts already in the scientific literature. - J. Johnson (JJ) (talk) 21:03, 13 March 2010 (UTC)

Re Baez, you mean that (vertical) plot with the piece-wise linear expanding scale? Or the (horizontal) one of the last 1.35 My, also with a piece-wise linear expanding scale? BTW, the Saltzman text ref'd there was written before publication of some key sources used here. Plots in the lit. are generally copyright.--Gergyl (talk) 22:09, 13 March 2010 (UTC)
Baez has a horizontal stacked chart further below ("Temperatures over the last 1 million, ...") such as I have in mind, though it is not the best example as it is not as clear as could be in showing the expansion. I recommend Baez as a source of inspiration, not for specific charts. For actual data and charts for sure I would recommend starting from the NOAA site.
Baez' vertical chart is an interesting case. It is a standard from used in many fields, with a rather standard compression of the geologic time scale (those older geologic ages tend to be so tedious) familiar to experts. But same criticism as above: to the lay reader not familiar with non-linear scales the compression of one axis over-emphasizes the other axis; it makes the recent squiggles seem inconsequential. Which could lead to an interesting editorial consideration: should the article explain – educate the reader – why some charts are "better" than others? (But discussion of this should be in a different section.) - J. Johnson (JJ) (talk) 19:12, 14 March 2010 (UTC)


The article says multi-million to billion (109) year time scales. But the only ref in the "Evidence for past temperatures" section is for the recent ice-age cycles, ie the 100 kyr scale. Which isn't really geological. Does d-O-18 apply more than 1Myr ago? William M. Connolley (talk) 23:22, 25 November 2011 (UTC)

δ18O does get used as a proxy for climate change over long times, though in those cases you are talking about the oxygen isotope content of either bulk sea sediments or fossil corals / shells. Over long times, what you are mostly measuring is bulk changes in the ocean isotopic concentration as a proxy for the presence or absence of large ice sheets on land. Dragons flight (talk) 23:52, 25 November 2011 (UTC)