What can global temperature data tell us?

2015-12-03 at 7:17 pm 33 comments

Debates about anthropogenic climate change often centre around data on changes in global temperatures over the last few decades. There are good scientific reasons to look at this data, but it also plays a prominent role in political advocacy, sometimes fairly, sometimes not so fairly. This is the first in a series of posts in which I’ll discuss what this data can and cannot tell us, and examine some recent papers concerning whether or not there has been a “pause” in global warming over the last 10 to 20 years, and if so, what it might mean.

I will focus on anthropogenic warming that results, via the mis-named `greenhouse effect’, from CO2 produced by burning fossil fuels. There are other human-generated `greenhouse gasses’, and other human influences on climate, such as changes in land use, but the usual estimates of their effects are smaller than that of CO2, and in any case, they would call for different policy responses than reducing fossil fuel consumption.  Other possible anthropogenic influences are, however, a possible complication when trying to determine the effects of CO2 by looking at temperature data.

What I’ll call the `warmer’ view of the effect of CO2 is what is accepted (at least verbally) by most governments, and is more-or-less found in the reports of the Intergovernmental Panel on Climate Change (IPCC) — that burning of fossil fuels increases CO2 in the atmosphere, resulting in a global increase in temperatures large enough to have quite substantial harmful effects on humans and the environment.  The contrasting `no-warmer’ view is that increases in CO2 cause little or no warming, either (implausibly) because CO2 has no warming effect, or (somewhat more plausibly) because strong negative feedbacks limit its effects. In between is the `lukewarmer’ view — CO2 has some warming effect, but it is not large enough to be a major cause for worry, and does not warrant imposition of costly policies aimed at reducing fossil fuel consumption. This is the predominant view at some `skeptical’ web sites such as Watts Up With That.

There is also the `extreme-warmer’ view, that the effects of CO2 will be so large as to `fry the planet’, leading to the extinction of humans, and perhaps all life, which is surprisingly common among the general public, despite being utterly implausible. Of course, they are encouraged in this belief by alarmist papers such as `Mathematical Modelling of Plankton–Oxygen Dynamics Under the Climate Change‘ by Sekerci and Petrovskii, who apparently don’t understand that any arbitrary system of differential equations has a good chance of producing unstable behaviour, and that calling such a system a `model of a coupled plankton–oxygen dynamics’ does not make it a good model. It is very, very unlikely that life on earth would have lasted for over three billion years if the global ecosystem were really as unstable as is suggested in this paper.

The `warmer’ and `lukewarmer’ views are sufficiently plausible that it’s worth asking whether global temperature data has anything to say about which is closer to the truth.  An alternative source of evidence is physical theory, embodied in computer simulations.  Unfortunately, earth’s climate system is too complex to be simulated without various simplifications and approximations being made, so simulation cannot provide definitive answers, and must ultimately be checked against observations. Observations also have a rhetorical role, being potentially convincing to those who may put no trust in theory and simulation, but who naively think that measuring global temperature is a simple matter of reading thermometers.

Unfortunately, measuring global temperature is not so simple.  Earth is a big place, with few observing stations, and every observing station is subject to biases from factors such as changes in the nature of its surroundings and in the time of day when observations are made. Measurements of temperature from space are indirect, and have potential biases from factors such as decaying satellite orbits.  All time series of global temperatures are therefore the result of complex processing of raw data, whose appropriateness can be questioned.

It should come as no surprise to those aware of the political nature of this debate that supporters of the `warmer’ and `lukewarmer’ views tend to favour different global temperature datasets, which show different temperature trends in recent years.  A favourite of the warmers is NASA’s GISS data, whose land-ocean version combines land temperature observations with sea surface temperature data. This data set was recently revised, with the new version showing a larger upward trend in temperature in recent years. The lukewarmers tend to favour the UAH data from satellite observations, also recently revised, with the new version showing a lower trend than before.

One should note that these two data sets are not measuring the same thing, or even trying to.  GISS measures an ill-defined combination of water temperature near the top of the ocean and air temperature a few feet above the ground, in some variety of surroundings. UAH measures temperature in the lower part of the atmosphere, up to about 8000 metres above the surface. So it’s conceivable that the different trends in these two data sets both accurately reflect reality, though if so it’s hard to see how these different trends could continue indefinitely.

I’ll first show the monthly GISS global land-ocean temperatures (retrieved 2015-11-30) from 1880 to the end of 2014. (That’s when some other data I’ll be looking at ends; 2015 is so far mostly warmer than 2014.)  These temperatures are expressed as `anomalies’ (in degrees Celsius) with respect to a base period (separately for each month of the year), since absolute values are meaningless given the arbitrary nature of what GISS is measuring. Here they are:

This graph is often portrayed (to the public) as convincing evidence that CO2 causes global warming. Look at that upward trend from about 1910!  However, the rise from 1910 to 1940 can’t really be due to CO2. The direct warming effect of CO2 is generally accepted to be proportional to the logarithm of its concentration, with a doubling of CO2 producing roughly one degree Celsius of warming, which might be amplified (or diminished) by feedbacks. Here is a plot of the log base 2 of CO2 over the period above (data from here):

The increase from 1910 to 1940 is only about 0.05, which even with a generous factor of four allowance for positive feedback would give only 0.2 degrees Celsius of warming, compared to the warming of about 0.5 degrees in the GISS data. And if the 1910-1940 warming was really due to CO2, the warming from 1970-2000 should have been even greater than it was.  Furthermore, part of the effect of CO2 is expected to be delayed by decades, making it an even less likely explanation of the 1910-1940 warming, since CO2 is thought to have been more-or-less constant before 1880.

Clearly, there are other influences on temperature than CO2. Once one realizes this, the upward temperature trend from 1970 to 2000 becomes less convincing as evidence of a warming effect of CO2.  Furthermore, since CO2 has been increasing pretty much monotonically for over a hundred years, it is highly confounded with everything else that has been increasing over that period, as well as with long-period cycles.  So any really persuasive argument regarding the effect of CO2 must be based on physical theory and on more detailed measurements that can confirm the effects of CO2 at a greater level of detail than a simple global average of temperature. This is the subject of `attribution’ studies, the critique of which is beyond the scope of this blog post (and beyond my expertise).

Nevertheless, there seems to be value in trying to better understand the global temperature data, partly as a `sanity check’ on claims based on more complex, and perhaps more questionable, analyses, and also to see whether there is any evidence of the data being wrong.

To lukewarmers, an aspect of the data that provides evidence of other factors being comparable in importance to CO2 is the `pause’ in warming (or at least a `slowdown’) that one can visually see in the plot above from about 2002.  For a closer look, here is the same GISS data, but going back only to 1979:

The UAH satellite temperature data starts in 1979, so we can now compare with it (version 6.0beta4, downloaded 2015-11-30):

The base period for the anomalies in the UAH plot is different from GISS, so only the changes are comparable.  (I’ve made the vertical scales match in that respect.)

Both data sets seem visually to show a slowdown or `pause’ around 2002, with this being more prominent in the UAH data (in which one might see the pause as going back as far as 1995).  To lukewarmers, the significance of this pause is not that global warming has stopped, showing that CO2 has no effect, since they think that CO2 does have at least some small effect.  Rather, they see it as evidence that other effects are large, sometimes large enough to cancel any underlying warming trend from CO2, and sometimes making any such trend appear larger than it actually is — and hence the warming in the 1970-2000 period cannot be taken as indicative of the magnitude of the warming due to CO2, or of what to expect in future.

As alluded to above, simple linear least squares fits to the GISS and UAH data for 1979-2014 show a greater trend for GISS (1.59 degrees C per century) than for UAH (1.12 degrees C per century).  But if there is actually a change around 2002, a single trend line is of course largely meaningless.

Reactions to the `pause’ (or `hiatus’) from the warmer camp have taken several forms:

  1. Claims that the pause is an artifact of poorly adjusted temperature measurements, that disappears when adjustments are done properly.
  2. Claims that the visual appearance of a pause is deceiving — that the `pause’ is just chance variation, which the human eye overinterprets.
  3. Claims that if one subtracts changes due to known effects, such as volcanic eruptions, the pause disappears, showing that the underlying trend due to CO2 continues unabated.  (Note that depending on the size of the underlying trend that is revealed, this would not necessarily be contrary to lukewarmer views.)
  4. Claims that warming from CO2 continues at a substantial rate, but that the heat is going somewhere that escapes measurement in global temperature data sets.

I will leave claims in category (4) for others to critique.

Claims in category (3) include a blog post by `tamino’.  I plan to present my own analysis of this sort in a future blog post, and compare to that of `tamino’.

Two recent papers making claims in category (2) are `Debunking the climate hiatus‘, by Rajaratnam, Romano, Tsiang, and Diffenbaugh, and `On the definition and identifiability of the alleged “hiatus” in global warming‘, by Lewandowsky, Risbey, and Oreskes. Both of these papers look at (or say they look at) the GISS land-ocean temperature data, displayed above, but before the recent revision.  I plan to comment on these papers in my next blog post.

Regarding (1), the GISS temperatures displayed above show a less prominent `pause’ than the version of GISS land-ocean temperatures distributed prior to July 2015 (obtained from the wayback machine’s version of 2015-04-18, stored here), which is shown below:

The revision results in a greater upward trend during the `pause’ period, as shown by the following plot of differences (with enlarged vertical scale):

To tell whether or not this revision was justified, one would need to examine in depth the temperature adjustments done for the GISS data set, which I haven’t done.

However, it’s not too hard to see some interesting things by examining the GISS land-ocean temperature data in more detail.  I’ll look only at the most recent version (accessed 2015-11-30) .

First, one can look separately at the Northern Hemisphere:

and Southern Hemisphere:

The difference is rather striking. One would expect some overall difference due to the greater amount of ocean in the Southern Hemisphere, and the different nature of the polar regions. But that doesn’t explain the abrupt increase in the scatter of  Southern Hemisphere data points after about 1955.

We can also look at each month of the year separately.  Here’s the Northern Hemisphere:


And here’s the Southern Hemisphere:


In the Northern Hemisphere, variability is obviously greater in winter than in summer.  The variability in the Southern Hemisphere winter seems slightly greater than in summer, but much less so than in the Northern Hemisphere. These are differences that I’ll take account of when modeling this data later.

I’ve marked 1955 by a short line at the bottom. In the Northern Hemisphere, the dip in January temperatures from 1955 to 1975 seems odd, since it doesn’t show up in December and February, but it’s hard to be sure that it’s not a real climatic effect.  Something does happen around 1955 in the Southern Hemisphere plots, which increases the variance in May and August, and maybe June, July, and September.  This can be confirmed by looking at plots for each of the 12 months of the year that show the difference of the anomaly for that month from the average anomaly for that month in the three preceding and three following years:


May through September seem to have higher variability in the years after 1955, and this is very clear for at least May and August. In contrast, similar plots for the Northern Hemisphere show no change in variance, or perhaps a slight decline after 1955 for May and June. It’s hard to see how this Southern Hemisphere variance change can reflect a real change in climate, given its abrupt onset, and that it does not appear in the Northern Hemisphere. More likely, it is an artifact of how the data is processed. A rapid improvement in quality of measurements after World War II might also be a possible explanation (though one would expect that to lead to less variability, rather than more).

Whatever the reason, it seems that relying on GISS data before 1955 might be unwise. In my later analyses, I will look at data only from 1959, since that is when some other related data sets begin, or from 1979 when comparing to the UAH data.

I note that obtaining all but the most recent GISS data is difficult.  Some versions can be accessed at the wayback machine, but many versions apparently saved there produce an ‘access denied’ error. UAH has an extensive archive, but even it seems not to have all the versions that were distributed. GISS distributes the programs they use, but only the current version.  I can’t find any programs at the UAH website.  Both GISS and UAH ought to have a public repository that uses a source-code control system such as git, which would allow all versions of programs, raw data, and processed data to be accessed, with documentation of all changes.

To reproduce the results in this post, you will first need to download the data using this shell script (which downloads other data too, that I will use for later blog posts), or manually download from the URLs it lists if you don’t have wget. You then need to download my R script for reading these files, and my R script for making the plots (and rename them to .r from the .doc that wordpress requires).  Finally, run the second script in R as described in its opening comments.

Entry filed under: R Programming, Science, Society, Statistics, Statistics - Nontechnical. Tags: .

Exact computation of sums and means Has there been a ‘pause’ in global warming?

33 Comments Add your own

  • 1. Wayne  |  2015-12-04 at 8:16 am

    As far as 1955 and the southern hemisphere, you might want to consider when the first permanent Antarctic stations was established (1957) and how GISS interpolates temperatures across vast distances.

  • 2. Ken  |  2015-12-05 at 5:17 pm

    There are some parallels here with economics. In both areas there is very little knowledge about the underlying cycles that may take decades to conclude. So in economics we had the Great Moderation followed by a crash, resulting from increased risk in the financial system. I think we are still part of an even larger cycle as debt and therefore risk is still increasing. For the climate we have shifts in large volumes of sea water and the generation of unstable patterns which will cause cycles of varying length.

  • 3. Nick Emblow  |  2015-12-05 at 6:23 pm

    Obviously there are other factors that affect climate, in particular and probably most significantly would be the Milenkovitch cycles, and perhaps to a lesser degree the sunspot cycles.

  • 4. Robert Young  |  2015-12-05 at 10:52 pm

    — There are some parallels here with economics.

    Well, if so, it’s all coincidence. Physical phenomena are ruled by God’s Rules (or Mother Nature’s, as you prefer). Economics is ruled by humans’ fungible rules. Physical phenomena have no control over the environment and processes which determine their value. Economic phenomena are directly determined by prior decisions of humans, mostly by small cabals which profit from changing/bending/breaking some existing set of rules. Which is how we got the Great Recession; it wasn’t the result of any analog to thermodynamics or Brownian Motion.

    There are two external forces over which such cabals have limited control: burgeoning population and diminishing natural resources. Or as one wag, whose name I don’t remember, put it: “we don’t need 7 billion people to make stuff”. IOW, economic growth doesn’t require or need more people, just more leveler distribution of wealth. Not that the Social Darwinists are interested in that vector.

    • 5. Radford Neal  |  2015-12-06 at 8:16 am

      The last paragraph above is off topic. So no more comments along those lines, please, regardless of how tempting a rebutal might be.

  • 6. hypergeometric  |  2015-12-07 at 9:59 pm

    Scientific argument in favor of anthropogenic climate disruption does not come from observing global temperatures and finding an explanation. Rather, it comes from really fundamental physics (atmospheric radiation + blackbody radiation thermodynamics + law of conservation of energy). The outcome of substantially increasing CO2 emissions were appreciated in the 19th century. The atmospheric lifetime of CO2 is a late 20th century contribution.

    Accordingly, while, as mentioned, there are many things which can alter global temperatures, including bursts of volcanic particulates, and atmospheric aerosols (and less so, solar variations, even historically: query Faint Young Sun paradox), the long term homeostatic control involves CO2 fluctuations. Plucking a trend out of a temperature series, then, seems to require a pretty strong prior that there is a trend, lest much of independently verified physics lie in a rubbish heap.

    There have been several attempts to independently verify the major trend finding sources, each vying with one another, with the BEST project being the most pronounced. Experimental science like this is tricky, whether speaking of satellite sensors or (innocently) misprogrammed ARGO floats, and the struggled to reconcile observations with ab initio modeling is always a challenge.

    Still, I think the geophysicists have come a long way, even if they are better then they use some of the better statistical methods. Two guys who have contributed a lot, in my uneducated book, are Mark Berliner of OSU and Nathan Urban of LANL.

    My own take on this question has been informally reported at:
    http://johncarlosbaez.wordpress.com/2014/05/29/warming-slowdown-2/ and http://johncarlosbaez.wordpress.com/2014/06/05/warming-slowdown-part-2/.

  • 7. ...and Then There's Physics  |  2015-12-16 at 4:36 am

    Can I ask if you are aware of the ocean heat content data and the significance of that data? Anthropogenic Global Warming (AGW) is really a consequence of adding radiatively active gases to the atmosphere. These gases change the planetary energy balance. A vast majority (>90%) of the energy associated with AGW goes into the oceans. Only a small amount (a few percent) is associated with warming of the surface. So, assessing AGW, and the role of CO2, using surface temperature datasets only, means that you’re assessing this using data that represents only a small part of the overall system.

    However, returning to energy balance requires warming of the surface, so understanding the evolution of the surface temperatures is clearly important and understanding how much the surface will warm for a given change in atmospheric CO2 is also important, but focussing only on surface temperatues does ignore a good deal of relevant information.

    • 8. Radford Neal  |  2015-12-16 at 7:32 am

      I’m aware that such data exists, but I haven’t examined it in any detail. Off-hand, I’m skeptical that one can actually measure ocean temperature to the very high accuracy, with comprehensive coverage, needed to track heat content, though as I say, I haven’t looked at how they claim to be able to do this.

      • 9. ...and Then There's Physics  |  2015-12-16 at 8:03 am

        Technically, there are two ways to measure ocean heat content. One is using in-situ measurements (ARGO floats are the ones being used at the moment). The other is sea level rise, which is partly a consequence of thermal expansion. As I understand it, these are largely consistent.

        My broader point, though, was that focusing on only surface temperatures ignores a great deal of other relevant information and can lead one to infer things that may not be consistent with the overall evidence.

      • 10. dikranmarsupial  |  2015-12-16 at 8:05 am

        It is something that needs to be taken into account if inferences are drawn from GMST/MSU observations regarding climate sensitivity, particularly because changes in ocean circulation have a significant effect by redistributing heat between the oceans and atmosphere.

      • 11. hypergeometric  |  2015-12-16 at 9:40 pm

        Links. Argo is a remarkable system.





        Some of these floats are very creatively placed. One, for instance, in the eastern, northern Atlantic, hitches a ride on a surface current in one direct, which it rides for 100 km (approx), then descends to where there is a current flowing in the reverse direction, where it rides back to the vicinity of the area it left. It then ascends and repeats.

        I have looked at the profiles and series of some of these floats individually.

        The general picture is the one recorded at:

        recorded from https://johncarlosbaez.wordpress.com/2014/05/29/warming-slowdown-2/

      • 12. Glenn Tamblyn  |  2015-12-16 at 11:06 pm

        See my comment below about ARGO…

  • 13. Zeke Hausfather  |  2015-12-16 at 11:06 am

    A few quick comments:

    As others have mentioned, there is no data for Antarctica prior to 1955, which might help explain the changing nature of variability after that point. A separate temperature series by Berkeley Earth explicitly tries to account for variations in spatial coverage in their uncertainty envelope: http://static.berkeleyearth.org/papers/Methods-GIGS-1-103.pdf

    Second, comparisons between temperature and CO2 are useful, but CO2 is not the only forcing influencing the climate system. If you include our best estimate of all forcings, models do a reasonable job simulating the 1930-era temperatures: http://postimg.org/image/5vukh7h4z/

    • 14. Radford Neal  |  2015-12-16 at 2:55 pm

      Yes, perhaps one can explain 1930s temperatures by looking at various forcings. That’s more-or-less my point, though – that just looking at CO2 and temperature isn’t going to be adequate. This may seem obvious, but I think it isn’t obvious to the public, and is ignored in some of the “evidence” for CO2 causing warming that is offered to the public.

    • 15. Glenn Tamblyn  |  2015-12-16 at 11:03 pm

      I would add also, going back to the period 1910 to 1930, that a similar process was occurring in the Arctic. Many more stations were added, particularly by the Soviet Union. So similar issues of changes in coverage apply then.

      And there is also a well documented bias in the Sea Surface Temperature records during the late 30’s/40’s due to a significant change in the mix of ships sampling water temperatures using buckets vs sampling the engine cooling water inlet. This change in the mix occurred due to WWII and the change in the mix of nationalities operating ships. Different datasets have included a correction for this bias to differing degrees.

    • 16. Radford Neal  |  2015-12-16 at 11:32 pm

      Regarding the increase in variance in 1955, new stations in Antarctica do seem to be the most likely explanation. In an email, Reto Ruedy pointed me to this paper, in which Fig. 2 shows that 3/4 of Antarctica started to be covered only in 1955-1957.

    • 17. manicbeancounter  |  2016-03-10 at 3:58 pm

      Zeke is quite right about there being no data for Antarctica prior to 1955. However, Gistemp still has data for this region right back until 1880. There must be some proxy data used.
      For the period up until 1955 the data for 90S to 64S very closely resembles that of Base Orcadas located at 60.8 S 44.7 W.
      However, Base Orcadas has some highly erratic temperature anomalies, and shows cooling between 1880 and 1930.

  • 18. dikranmarsupial  |  2015-12-16 at 11:50 am

    “Furthermore, since CO2 has been increasing pretty much monotonically for over a hundred years, it is highly confounded with everything else that has been increasing over that period, as well as with long-period cycles. ”

    Can you give some information on what long-period cycles you had in mind? As far as I can see the confounding tends to operate more in the other direction, with likely effect of the combined actions of the forcings being misdiagnosed as a long-period cycle (with perhaps only one or two cycles present in the temperature record). There are good physical reasons to support the forcings having an effect on climate, the evidence for some of the purported long-term cycles have very little physical basis.

    • 19. Radford Neal  |  2015-12-16 at 2:58 pm

      This was a general point. I don’t have any particular cycles in mind, though others may have such ideas. As I say in the next sentence, “any really persuasive argument regarding the effect of CO2 must be based on physical theory”, which seems to be what you’re saying too.

  • 20. Peter Jacobs  |  2015-12-16 at 1:49 pm

    Hi Dr. Neal,

    This seems like an interesting first step in what I hope will be a nice series of posts and mutually beneficial interaction with commentors. I would like to offer a few constructive comments:

    On the classification of “warmer” types, I would strongly suggest that this is largely irrelevant to any sort of statistical questions you’re likely to ask, and presents a potentially biasing effect on how things may proceed. The classifications you’ve presented have little to do with anything based on science and function basically as in-group and out-group identifiers for the purposes of partisan squabbles. If you look at the physical science literature itself, these labels are essentially non-existent. And indeed this should not be surprising. One could easily believe both that Equilibrium Climate Sensitivity (ECS) is on the low end but that the biosphere and human systems are fragile, or one could believe that ECS is on the very high end but that the profit of fossil fuel corporations and distribution of that wealth are an acceptable trade for the effective loss large portions of the developing world’s health and economy on a purely cost-benefit basis. So aside from whether or not one rejects basic physics relating to radiative forcing and planetary energy balance, these labels are irrelevant from a scientific perspective.

    In terms of causation (or perhaps more appropriately attribution), looking at the global surface temperature response (or the beleaguered lower tropospheric microwave-oxygen-brightness product) and CO2 concentrations alone are going to tell one very little.

    I would strongly object to this characterization “This graph is often portrayed (to the public) as convincing evidence that CO2 causes global warming.” Plots of the surface temperature record are often portrayed as convincing evidence that the planet is warming. Our understanding of the drivers of climate tell us about causation, and in combination we are confident that warming is occurring and being driven largely by human changes in radiative forcings (principally but not exclusively the combustion of fossil fuels and resulting release of CO2). I have found that people who under the mistaken impression that a plot of surface temperatures is supposed to be evidence of human causation of warming are unaware that there are people who deny that warming ongoing, of the evidence behind the attribution of warming, or both.

    Without consideration of internal climatic variability (e.g. phasing of Tropical Pacific or North Atlantic variability on high and low frequencies) and non-anthropogenic “external” forcings (such as changes in solar irradiance and volcanic aerosol loading), as well as understanding expected spatio-temporal changes to these through the deep ocean to the upper atmosphere, very little is going to be accomplished in investigating attribution.

    Alternatively, if this is the start of a comparison of the relative merits of the surface instrumental record vs. the microwave products, I would encourage a deep dive into the methodologies of producing both, as well as their reproducibility, before passing judgment. I have seen a number of politicians and climate skeptics make claims of superiority on behalf of the microwave product that have almost no correspondence to the actual qualities of said product.

    I am very much looking forward to your future posts.

    Peter Jacobs

  • 21. Radford Neal  |  2015-12-16 at 10:25 pm

    I’m hoping to be ready to put up my subsequent posts in the near future, so I won’t make any long reply to this.

    But I am puzzled by your objection to my saying that the graph of temperatures since 1880 is often presented to the public as convincing evidence that CO2 causes warming. I see this all the time. For instance, I think there was recently a version of this graph on the White House web site (or some such, I unfortunately didn’t save the URL), albeit with the labels on the two curves accidently reversed.

    The implication is usually that it’s warming, and there is nothing else that could have caused the warming than CO2. As I note in the post, this is particularly misleading if the 1910-1940 warming is shown, without comment.

    • 22. Peter Jacobs  |  2015-12-17 at 12:51 am

      Hi Dr. Neal,

      My initial objection was that surface temperature plots alone were not “often portrayed (to the public) as convincing evidence that CO2 causes global warming” to the best of my knowledge.

      You’re responding that a plot of temperatures along with CO2 serves as an example. I would suggest that the addition of CO2 to the chart is suggesting a relationship, and that this is not what I originally objected to. Nor do I think that what this other plot is claiming is “there is nothing else that could have caused the warming than CO2”- the figure is from a National Climate Assessment and neither the caption for it nor the surrounding material make this claim based solely on the two increasing at the same time. Indeed attribution is discussed in terms of discriminating between possible natural and anthropogenic drivers.

      Perhaps this seems nit-picky to you, and if so, I apologize. As someone who tries to reduce misunderstanding about climate change as a side academic interest, I try to make sure all parties agree on their assumptions going into a discussion (at least as much as possible) to preclude avoidable misunderstandings later. Almost if not all materials I have seen targeted at pretty much every age group of the public don’t simply rely on the two variables increasing at the same time to claim causation (the physics are mentioned at varying levels of sophistication), and non-CO2 drivers of temperature change are mentioned as well. Indeed the caption of the second figure you mention makes clear reference to shorter term excursions relating to non-CO2 drivers. This is not to say that you might not be able to find a counter example that does not mention the greenhouse effect or non-CO2 drivers, but I am quite dubious that you will find that surface temperature plots alone are “often portrayed (to the public) as convincing evidence that CO2 causes global warming” if you look around.

      Again, looking forward to your future posts.


    • 23. dikranmarsupial  |  2015-12-17 at 4:38 am

      I don’t think it is reasonable to expect politicians to spell out the caveats everytime they present a graph in the same way that a scientist would. The IPCC reports make it quite clear that there are natural forcings as well as anthropogenic ones and that these are taken into account in attribution studies. If someone wants to find out the scientific details, then the IPCC reports are freely available and there are many good primers on climate change aimed at the general public. I rather doubt that politicians spell out the caveats when discussing e.g. economics or foreign policy, so I’m not sure why we should expect anything different on climate change (although it would be a good thing IMHO).

  • 24. Glenn Tamblyn  |  2015-12-16 at 10:56 pm

    Adding to some of the comments here, I would like to mention this statement you make:

    “Unfortunately, measuring global temperature is not so simple”.

    None of the surface data sets measure global temperature! They measure global temperature ANOMALY.

    The surface temperature datasets are based on taking a baseline average for each individual temperature station. Then taking the difference between any individual reading for a station and that stations individual baseline average to produce a temperature anomaly for that station. These temperature anomalies are then averaged, using various area weighting schemes to account for non-equidistant spacing of the measurement stations, to produce regional and global averages of these anomalies. At no time is a global temperature calculated or used.

    By always using anomalies and averaging them biases in individual readings such as inaccuracies in the individual instruments are cancelled out. This would not be the case if averaging of absolute temperatures were applied.

    Bear in mind also that it is climate change that is being measured, not weather. So the methods used need to be evaluated based on the degree to which climates across large distances are correlated. Consider the example of the capital of Chile, Santiago. It is close to sea level but the Andes nearby tower over the city.

    Santiago and those peaks obviously have very different climates, but their climates are correlated; the same weather systems pass over both. Thus any change in their climates will be correlated as well; they experience the same changes in the patterns of weather systems. Thus uncovering the underlying changes in climates is the focus of the design of how these datasets are calculated.

    If you have not yet done so it would be worth reading Hansen & Lebedeff 1987 which provides the basis for the GISS algorithm. It includes an analysis of these correlation of climates issues.

    Regarding measuring OHC again anomalies are used, not absolute temperatures, to calculate heat content changes. The thermistors used on the ARGO floats have an accuracy of +/- 0.002 Deg C. And an important factor in this is the question of how much ocean temperatures vary over distances. Basic sampling theory says that sampling interval, whether in the time domain or the spatial domain, needs to be based on the underlying variability of the signal being measured. A rapidly varying signal needs greater sampling density. The oceans do not vary much in temperature over shorter distances so the sampling distances needed to adequately characterise them can be quite large.

    I look forward to your future posts.

    • 25. Radford Neal  |  2015-12-16 at 11:43 pm

      Yes, I know that GISS data is fundamentally on anomalies, not meaningful absolute temperatures. That’s why I said “These temperatures are expressed as `anomalies’ (in degrees Celsius) with respect to a base period…, since absolute values are meaningless given the arbitrary nature of what GISS is measuring”. For instance, an absolute temperature would be sensitive to how many stations had trees growing near them.

      For UAH data, although I don’t know the details, I think absolute values ought be meaningful, though they aren’t usually presented. I think, for instance, that one could at least in theory hope to match climate simulation outputs with absolute temperatures derived from satellite data, though I’m not sure whether this is at all realistic at present.

  • 26. Kevin Cowtan  |  2015-12-17 at 7:37 am

    Dear Dr Neal

    I’ve got a little experience with temperature records which may be relevant. I’m mostly interested in communicating and demystifying the science, so a lot of my focus has been on finding the simplest possible methods for calculating temperature records.

    In the case of the surface temperature data, the simplest methods are indeed very simple. To produce a land surface air temperature record from weather stations takes about 60 lines of code. A simple sea surface temperature record from the most reliable buoy data is about 100 lines of code. Automated homogenization of weather station records can be performed for the densely sampled parts of the network (e.g. the US and Europe) in about 150 lines of code. So these are not complex calculations, and by virtue of simplicity we can comparatively easily analyse for bias and uncertainty. I’d recommend doing the exercise for yourself – it’s a very good way to get a grip on the real uncertainties in the data. And the experience gained from the simple analyses has been enough to lead to a few useful papers.

    Of course a more complete analysis with uncertainties is more complex, but not excessively so. I’d recommend the Berkeley analysis as methodologically the best, however the Hadley analysis is better from the point of view of quantifying uncertainty.

    The satellite data are in a whole different league I’m afraid. A few of us have taken a look at the basics of reproducing the satellite record, but it looks to me to be a different order of complexity. For a brief overview, look at Figure 2 in this paper:
    Note, for example, that the RSS data do not produce direct temperature estimates, indeed they need to be calibrated against climate models to account for diurnal temperature variation.

    As a result, I find it almost impossible to say anything useful about the uncertainties in the satellite data. I have done some analysis of the satellite data against the surface data, which shows for example that the UAH 5.6 data show significant geographical correlation with the surface temperature over land, but not over oceans. The RSS data and UAH 6.0 are sensing higher in the atmosphere and thus show less spatial correlation with the surface air temperatures. However this doesn’t tell us much about bias or uncertainty in the time domain. Estimating either is complicated by the fact that both the climate signal and the internal variability change significantly as you move up through the troposphere, and in particular it is vital to account for the El Nino signal when trying to understand the differences.

    The complexity of the satellite record means that, for now, we are dependent on the providers for estimates of the uncertainties and biases in the data, something which has only been addressed in general terms up to now. However Carl Mears of RSS has a very interesting abstract at AGU this year, here:
    (In fact the talk is today.) It looks as though RSS are going to produce an ensemble of temperature reconstructions encompassing both uncertainties in the observations and methodological choices in the processing algorithm, similar to what Hadley have done for the surface temperature record. I am hoping that this will provide us, as users of temperature data, with our first opportunity to understand the relative uncertainties in the trends in surface and satellite temperature records.


  • 27. ralfellis  |  2016-01-12 at 12:06 pm

    Global temperatures from ice cores can tell us that albedo is the primary modulator of temperature, not CO2.

    Prof Clive Best has a nice review of the dust-ice-albedo theory on his blogsite. And the bottom line is that ice ages are forced by precessional insolation, but initiated and modulated by dust-ice-albedo.


  • 28. Steve McIntyre  |  2016-01-12 at 12:52 pm

    Nice to see such thoughtful comments from a fellow Torontonian.

    Around 2008, I looked at GISS and inventoried quite a few different GISS versions, including ones back to the 1990s. You’re welcome to it if you’re interested.

    One of the oddities of GISS at the time (and I don’t know whether they still do it) was a “two legged adjustment”. in which, as I recall, they fitted each station series with two trends and a changepoint, from which they calculated their anomaly. This led to peculiar results, to say the least. It wasn’t clear whether this led to a bias, as opposed to being a completely pointless exercise. I compared it to the meaningless toy in the Peter, Paul and Mary children’s song – it goes zip when it moves, bop when it stops and whirr when it’s standing still. See http://climateaudit.org/2008/06/22/nasa-step-2-another-iteration/.

    • 29. Radford Neal  |  2016-01-12 at 7:57 pm

      Thanks. I’ve been amazed at the amount of careful, detailed analysis you’ve done on these and other topics.

      I’ll keep your archive in mind if I need it. Do you know what the copyright status of this data is? Could one just set up a public archive? Or include the data actually used in supplementary information for a paper?

      Of course, it would be best if GISS set up an archive, as I suggest at the end of the post. As it is, they say one should include the date of access when citing their data, but it’s not clear how that is supposed to help when you can’t get past versions…

  • 30. Steve McIntyre  |  2016-01-12 at 11:05 pm

    As I understand it, as long as the originator of the GISS data was acknowledged, academic use is fair use. I’m sure that you could put it into SI and I would encourage this. If you wanted belt-and-braces, you could ask them for permission. Gavin Schmidt is well aware of the bad publicity that arises out of data refusal and I can’t imagine that he’d object.

  • 31. johnmarshall  |  2016-01-13 at 6:10 am

    CO2 has zero effect on temperature. Ice core data clearly shows that temperature change precedes CO2 changes not the reverse as the alarmists claim. The laws of thermodynamics also show zero effect and debunk the AGW hypothesis. in fact there are no papers to date that show that the GHE in fact works.

    • 32. Radford Neal  |  2016-01-13 at 9:39 am

      Causation clearly goes both ways: more CO2 leads to warming, and warming leads to more CO2 (released from the ocean). So presenting the ice core data as evidence that CO2 leads to warming without discussing the issue, as is sometimes done, is misleading. But I also don’t think one can untangle the effects in the two directions just by looking at the timing.

      Regarding thermodynamics, would your argument also show that adding another blanket to your bed won’t keep you any warmer at night? If so, there’s something wrong with it. (Of course, the atmosphere is more complicated than a blanket, since it expands when warmed.)

      I think the only viable argument that CO2 has no effect in the end would be that there are strong, negative, non-linear feedbacks that essentially act as a thermostat, keeping temperature within narrow limits. An analogy would be that warm outside temperature has a causal effect on indoor temperature, but nevertheless doesn’t lead to an increase in indoor temperature if the house has a thermostatically-controlled air conditioner.

  • 33. Joffa  |  2016-01-13 at 5:25 pm

    There is a presumption here that temperature data is “random” and accurate. Neither is true. We know about the urban heat phenomenon (where most temperature data is collected from) and the confusing introduction of “homogenised data” and fiddling the dat procedures. We also know about “selective data collection by the meteorology departments globally.

    No way are the official “data collecting stations” – random. They are deliberately selected or “deleted” to satisfy the authorities. Plenty of work was done by “Watts Up With That.which clearly demonstrate how biased the land data is collected and processed.

    We should know the differences there are between “global urban temperatures” and global rural temperatures. That might give us a more accurate picture than the methods scientists currently work with. In my view most of the above post and commentary is based on “pseudo information”. Building nonsense onto nonsense.

    Why is this data continually used as if it is “dinkum”?


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed


December 2015
« May   Jan »

Most Recent Posts