Warming Stopped in 1998

This is just one of dozens of responses to common climate change denial arguments, which can all be found at How to Talk to a Climate Sceptic.


Global temperatures have been trending down since 1998. Global Warming is over.


At the time, 1998 was a record high year in both the CRU and the NASA GISS analysis.  In fact, it was not just a record year, it blew away the previous record by .2oC. (That previous record went all the way back to 1997, by the way!) According to NASA, it was elevated far above the trend line because 1998 was the year of the strongest El Nino of the century. Choosing that year as a starting point is a classic cherry pick and demonstrates why it is necessary to remove the very chaotic year to year variability that exists (aka: weather) by smoothing out the data. Looking at the CRU’s graph below, you can see the result of that smoothing in black.

Clearly 1998 is an anomaly and the trend has not reversed.  (Even the apparent levelling at the end is not the real smoothing.  The smoothed trend in 2005 depends on all of its surrounding years, including a few years still in the future.)  By the way, choosing the CRU analysis is also a cherry pick because NASA has 2005 breaking the 1998 record, though by very little.

Now this is an excusable mistake for average folks who do not need the rigors of statistical analysis in their day jobs, but any scientist in pretty much any field knows that you can not extract any meaningful information about trends in noisy data from single-year end points. This is why it is hard to hear a scientist make this argument and still believe that they are a voice of integrity in this debate, rather it appears more to be an abuse of the trust people would like to place in them as scientists.  Bob Carter is such a voice and was the first to trot out this argument in an article in the Daily Telegraph.  Since then it has echoed far and wide and has been used by Richard Lindzen as well as a host of sceptic websites.  

Interestingly, Bob Carter seems to know what he is doing as he tries to pre-empt objections in his article by basically insinuating that any choice of starting point, (such as 1978), will just be a cherry pick with the opposite motive! But cherry picking is about choosing data for the sole purpose of supporting a pre-conceived conclusion, it is not the simple act of choosing at all, as one must choose some starting point. In the case of his example year, 1978, this is often chosen simply because it is the first year that satellite records of tropospheric temperatures were available.

So what choices are there, what are the reasons for those choices and what are the conclusions we can draw from them?

  1. As just mentioned above, one could chose to examine the last 30 years because that is the period of time where both surface and tropospheric readings were available. We have been experiencing warming of approximately .2oC/decade during this time.  It would take a couple of decades trending down before we could say the recent warming did in fact end in 1998.
  2. You could choose 1970 in the NASA GISS analysis as this was the start of the late 20th century warming and as such it is a significant feature of the temperature record. The surface temperature over this period shows .6oC warming.
  3. You could choose 1965 in the CRU analysis as this is when the recent warming started in their record. This record shows around .5oC warming of the smoothed trend line.
  4. You could choose 1880 in the NASA record. This shows .8oC warming.
  5. You could choose 1855 in the CRU record. This shows .8oC warming.  Again, with this trend and the above we can not say it is over without many decades more data all indicating cooling.
  6. You could choose to look at the last 500 years in the bore hole record analysis because that is its entire length. This puts today about 1oC above the temperatures in the first 3 centuries of that record.  The record of today’s trend in that kind of analysis will be hidden from view for many more decades.
  7. You could choose to look at the last one thousand years, because that is as far back as the dendrochronology studies reliably go. Then the conclusion is:

    Although each of the temperature reconstructions are different (due to differing calibration methods and data used), they all show some similar patterns of temperature change over the last several centuries. Most striking is the fact that each record reveals that the 20th century is the warmest of the entire record, and that warming was most dramatic after 1920.

  8. You could choose to look at the entire period of time since the end of the last ice age, around 10kyrs ago. Then the conclusion is that GHG warming has reversed a very long and stable period with a very slight downward trend and we are now at a global temperature not experienced in the history of human civilisation, the entire Holocene.  Such a long view applied to today will take many centuries to clear up.  The situation is a bit more urgent than that!

I think that about covers any periods of time relevant to today’s society. Clearly, "it has stopped warming" is only supported by taking a single specific year out of context and using a 7 year window to look at multi-decadal trends in climate. That is a classic cherry pick.

This is just one of dozens of responses to common climate change denial arguments, which can all be found at How to Talk to a Climate Sceptic.

“Warming Stopped in 1998” was first published here, where you can still find the original comment thread. This updated version is also posted on the Grist website, where additional comments can be found, though the author, Coby Beck, does not monitor or respond there.

80 thoughts on “Warming Stopped in 1998

  1. SemiChemE: the temperatures may appear to be down, but I think if you take any recent ten year period of climate data, from any source, and actually calculate the regressions you will find that it has never been negative. The high value for 1998 gives a powerful optical illusion.


  2. Can you help me understand why there is a correlation between global temperature and CO2?

    What really confuses me is the long term data that goes back hundreds of millions of years back to the Cambrian Period when CO2 levels peaked at 7,000 PPM to modern day where they are only a few hundred. The CO2 levels have no correlation to global temperatures whatsoever. Temperature and CO2 levels vacillated all over the place, but there is clearly no correlation.
    However, during the last few hundred years of Earth’s evolution the claim is that there suddenly is a correlation.


  3. According to a recent opinion survey, man-made global warming is viewed as settled science among the majority of computer software developers in the United States.


  4. 15

    Tim, the graph is from the Hadley center, which is unaffiliated with Hansen. He had nothing to do with that graph.

    If you can’t tell the difference between Hadley and GISS, how can you claim to be informed here?

    Posted by: Brian D | September 19, 2008 10:37 AM


    Well, we now know that there WAS connection to Hansen, as Hadley-CRU was committing fraud and fudging data, and all kinds of graphic and math tricks to make the data conform to what they PRE-DETERMINED the result should be!

    Are you ready to admit that the Hansen/Hadley-CRU/IPCC connection is what has been driving temps up, and not anthropogenic CO2?


  5. I sometimes teach statistics, so the recent media/blogosphere misinterpretations of Chris Jones’ remarks about statistically nonsignificant short-term “trends” looked like a teaching opportunity — for my students, if not for less educable pundits.

    Anyway, I wrote a short program to calculate trends in GISTEMP global temperatures over the period 1880-2009, then 1881-2009, 1882-2009, and so forth up to 2007-2009 (a three-year “trend”). The trends (regression slopes) all are positive and statistically significant for starting years up through 1996. They remain positive but nonsignificant through 2001. After 2002 the slopes turn negative but with confidence intervals that go far above zero as well as below.

    Here’s a graph of the whole series, GISTEMP 1880-2009:

    Here’s how it looks if you cherry-pick 2002 as your start date:

    The “trend” goes slightly positive if we start just one year earlier, in 2001:

    And becomes steeply positive if we start in 2000:

    But of course short-term trends are missing the point with respect not only to climate, but to statistics as well.

    Here’s a graph showing all the trends (slopes) versus start year:

    And here’s a similar graph visualizing how confidence bands explode as you shorten the time window:


  6. Loren: Do you know the difference between absolute and relative warming? Good. Now, the period preceding the Cambrian was exceptionally cold. Almost the entire surface of Earth was frozen over. Clearly, other climate factors caused this – there was no higher animal life, there was no terrestrial vegetation. The main suspect is plate tectonics – land warms and cools more rapidly than ocean water, so it’s continental movements to or away from the poles (with their polar night/day) which will cause a switch between warm ages and ice ages.

    The Cambrian represents a period of rising CO2 and drastic global warming, *compared to the time preceding it*. Or *relative to the tectonic situation*. How can you say “there is clearly no correlation”?

    Also consider later periods: CO2 was usually above today’s levels, climate was warmer (in the mid-Mesozoic it was drastically warmer), with the usual consequences: Europe was mostly covered by a coral sea, as were the central USA. Suzch is the world in a period of exceptionally high CO2 levels, and even though we did have the usual Milankovic variations – cool and warm periods -, as it was a warm age back then, “cool period” did not mean “ice all over Central Europe” as we had it just 15000 years ago, but merely “a few mountains in Antarctica had glaciers”.

    Nowadays we are in an ice age. Many contrarians are sloppy with their language, and use “ice age” for what is actually a “glacial”, a period of ice advance, no matter whether in an ice age or warm age. We are in an interglacial in the current ice age (which started in earnest a few million years ago). That’s why Scandinavia is not covered in glaciers entirely (as it would be if we would be in a glacial period), but there are some glaciers (there would be none if we were *not* in an ice age).

    Then, there is the known property of CO2 (and other greenhouse gases) of capturing heat radiation. If this radiation is travelling in a specific direction (in our case, it is always travelling away from Earth), the effect of having GHG in the atmosphere is to block part of the surplus heat we receive from the Sun. The consequence is that increasing GHG content of the atmosphere must cause *some* amount of global warming, *relative* to the temperature immediately preceding it. Either that, or the laws of thermodynamics are dead wrong.

    We have been unable to refute the laws of thermodynamics since 150 years (when the first studies by Avogadro, Arrhenius et al found that there were indeed such laws, even though they lacked the maths to correctly express them at that time). We have also seen that temperature rise may cause atmospheric CO2 levels to increase, but that peaks in atmospheric temperature *always* appear to occur immediately following peaks in atmospheric CO2 (this last part is often omitted by contrarians. If you see someone mentioning only the “CO2 follows temperature” part, you know that person is dishonest). Also, extremely high global temperatures are apparently impossible to achieve *without* extremely high GHG levels.

    So there is and always has been a quite pronounced correlation. Plate tectonics is of course the stronger factor and can – as in the Precambrian-Cambrian case – overrule CO2. That’s why one speaks of climate “forcings” – a specific climate factor need not have an *absolute* influence, but it *always* forces the trend in a particular direction. GHG force it up. Having all continents stuck together at the South Pole would force it down, down, down.

    The bottom line is that:
    * the entire history of human civilization has taken place exclusively within a single interglacial in an ice age.
    * you would NOT want a switch from ice age to warm age. It would necessitate more people dying than anyone could stomach, because sea levels would rise so grossly as to flood much of the interior of all continents: the Solnhofen archipelago, the Turgai Strait, the Oceans of Kansdas, the Erromanga Sea; you might want to look these up to see what it’s like in a warm age.

    From a purely “humans-first” perspective, we would want the present interglacial to continue indefinitely, with a very, *very* slight warming trend (to increase global precipitation and temperatures).

    Problem is: the extra CO2 we added is – purely based on thermodynamic considerations – already good for about half the difference between a glacial and an interglacial. At least. This does not mean that we will see exactly this effect – other forcings also come into play – but it means that CO2’s effect on climate is probably more pronounced today than it ever was since the early Permian.

    As an evolutionary biologist, one cannot help but be disgusted by people like Lomborg with their “global warming is GOOD” stance. Their ignorance of good science permeates everything they say – but many of these people, including Lomborg, are economists, and as such they couldn’t foresee the global recession we’re in now. That is not science; a proper scientist should be able to make predictions about the future that are well-reasoned and turn out to be accurate. In any case, humans are not reptiles. Meaning that we cannot be expected to thrive in a warm age. Indeed, it was only the turn from the Mesozoic-Paleogene warm age to the current ice age that brought about the mammal-dominated world of today; had this not happened, we would probably have a civilization of intelligent birds by now (birds are better able to cope with very high temperatures than mammals).

    So, you do not want pronounced global warming – in the short run, you do not want it for your real estate property values, in the long run, you do not want it for your species. We thrive best in an interglacial within an ice age, and our ability to adapt to anything else has not yet been tested but given the extinction rates every time such climate shifts happen, it is not necessarily high even with all our inventiveness. At the least, we won’t be able to cope without forced resettling of many 100s of millions of people. I for one do not want this.

    (I have noted that most of the inability of many people to grasp the problem is a problem of size. The amounts of energy pertinent to the issue are mind-bogglingly large, and though air weighs next to nothing, so is the total mass of the atmosphere.

    What does help is to grab
    * a good record of atmospheric CO2 levels
    * a good record of atmospheric temperature
    * a Google Earth paleomap overlay, such as http://bbs.keyhole.com/ubb/ubbthreads.php?ubb=showflat&Number=624709&page=1

    and compare them. If one accounts for plate tectonics, the correlation between atmospheric CO2 and global temperature/habitable landmass should be strikingly obvious.


  7. Isn’t it increase like 1% ever 50 years or so. Is this really something to worry about? I dunno enough about the issue to comment heavily but I just feel it’s not the end of the world.


  8. Isn’t it increase like 1% ever 50 years or so.

    CO2 or temperature?

    CO2 is currently increasing at over 0.5% per year.
    If you are going to talk about percentage increase in temperature, it only makes (some) sense to use Kelvin, in which case a 1% increase is about 2.9C (more than 5F), not a trivial matter when it comes to global averages.


  9. “Why is it that we are not to accept that warming peaked and we have begun cooling?”

    Because it isn’t shown to be cooling?

    That would be a good reason for not accepting warming peaked.

    Ask yourself: why would this be a peak? Temperatures don’t act like a struck jelly, you know.

    “The AGW period was from 1978 to 1998”

    Why? What happened over that period that we did that we’re not doing any more?

    And why, if the “AGW period” ended 1998, have there been at least three and probably five years that have been as warm or warmer?

    Do you know what “peak” means?


  10. I agree with Coby in that no matter what time period you use it can always be construed as a cherry pick. So where does that leave us?

    We can argue about whether warming has stopped or not but neither side can claim victory.


  11. You can use statistical analysis to find out what period you need to take data over to display signal more than noise.

    Guess what. WMO did that. They figured 30 years.

    So go ahead and pick a 30 year running mean and calculate the trend for that 30 year period, year by year.

    And a cherry pick isn’t picking a year. A cherry pick is picking a year to get a desired result. You can deduce a cherry pick by picking several nearby numbers and seeing if the result changes significantly. If it does, it was a cherry pick.


  12. Prett sure Coby didn’t say that no matter what time period you use it can be ccnstrued as a cherry pick.

    I think polly’s reading skills are right up there with his/her ability to do statistics and science.


  13. Hi. I notice nothing has been posted in almost 3 years on this topic, but the sceptis still make this argument.

    To me, the data measurements must be the final determinant in the theory of AGW. My basic question I’m looking for an answer for is this:

    Assuming no volcanoes, and that ENSO events are appropriately considered, after how many years of low temperature measurements relative to models will the theory be in doubt.?

    Has this been discussed somewhere? Please point it to me, or tell me.



  14. Vonhayek,

    to add to what mandas said, note that they used one specific temperature record (RSS) in this latest piece. If you look at other temperature records, including the earlier favorite UAH (also a satellite-based temperature record), you will see those don’t give such a long time period of apparent (and it really is apparent) lack of increase.

    Mandas already points out that the surface record does fit with models, but perhaps more important is that there is no evidence of any stall/pause in the ocean heat content. Since that amounts to about 90-95% of the total heat capacity of the earth, focusing on the trends of the atmospheric temperature can be quite misleading.


  15. And note that those who screech “GIGO! GIGO!” at models when it comes to AGW will readily accept the data that is entirely dependent on a computer model to turn a radiant intensity into a surface temperature record.

    We have thermometers, and they don’t require computer models to interpret.

    For some, however, they don’t give the “right answer” therefore must be ignored.

    By this fact you can tell that they are not “honest brokers” in their claims.


  16. V –
    It’s worse than we thought.
    The warming continues at record levels.
    Many areas will see climate depart from normal ranges as early as 2025 to 2040.
    We have seen the beginning of the 6th great extinction.

    The human population may dwindle by half as early as 2050.

    You may want to review what is presented at Skeptical Science.
    Click on Climate Myths or Newcomers Start Here.


  17. Wow, stop! Thermometers tell you that the “global temperature” (which makes no physical sense) has not increased in the past 15 years. Why do you show problems in accepting reality?


  18. Global temperature does make sense, it’s just you that doesn’t, denierbot.

    Temperature of your body? 98F.
    Temperature of the Sun? 5600K.
    Temperature of the Earth? 17C.

    Just because you can’t work it out doesn’t mean it isn’t possible, dearie.


  19. Actually, freddy, in most of the canonical series the temps have increased a bit under .1 degree C over the time period 2000.5 to 2014.5–the 15 years you refer to. RSS is the only outlier and it is barely negative.

    You should check before you talk.


  20. JGarland: your lies about non-existing temperature increases speak for itself. Learn that even the IPCC, your authority, accepts the hiatus. You should accept your authority and not invent unbased alarmistic lies.


  21. remember, JG, there are people brainwashed into thinking that if only you believe hard enough, then it doesn’t matter what the facts are, if they are against you.

    kaitrollbot is just such a genetic mistake.


  22. kaitroll, so you admit that a temperature can be given for a body and “makes sense”.

    So why not the earth?

    Because you don’t want it to?


  23. Only in the mind of the deluded is arctic sea ice extent more than 1-sigma below the 1981-2010 average “big”.

    This is also below IPCC projections. Looks like the IPCC is rather conservative.


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