Global cooling bites the dust: Hottest January followed by second hottest February. Now March is busting out.

Last month, NOAA reported the world experienced the warmest January in both satellite records.  And NOAA just reported (here) that it was the second warmest February on record in both satellite records.  Now the UAH satellite data shows record-smashing temperatures in the first half of March:

UAH 3-10

The yellow line is the 20-year average temperature, the purple line is of the 20-year “record highs,” and the green line is the 2010 temperature [make your own chart here — I have a more complete, though messier, graph at the end].

Other temperature datasets show slightly different results.  For NASA, January and February were tied for the second hottest on record.

Of course, there never was any global cooling — see Must-read AP story: Statisticians reject global cooling; Caldeira “” “To talk about global cooling at the end of the hottest decade the planet has experienced in many thousands of years is ridiculous.” The vast majority of the warming went right where scientists had predicted — into the oceans (see “How we know global warming is happening” and below).

In fact, 2005 was the hottest year on record in both NOAA’s and NASA’s dataset — and in every dataset, the 200os were the hottest decade on record.  But the anti-science crowd loves their much-vaunted satellite data.  Why?

First, the satellite data shows the warmest year on record to be the uber-Ni±o year of 1998, allowing the disinformers to ignore the long-term trend and keep repeating the mantra, no warming since 1998.

Second, I think many in the anti-science crowd still operate under the mis-impression that the satellite data doesn’t show any significant long-term warming — a mis-impression creating by some mis-analysis by anti-scientists John Christy and Roy Spencer, mis-analysis that just happened to bias the data in the direction of their beliefs (see “Should you believe anything John Christy and Roy Spencer say?“).  Go figure!

In fact, NOAA points out that both satellite data sets show about the same amount of warming as the land-based record, “which increased at a rate near 0.16°C/decade (0.29°F/decade) during the same 30-year period” — once you remove the expected stratospheric cooling from the satellite records (see NOAA discussion here).

This is from Spencer’s blog on March 5:


It’s clearly warming, and Spencer himself says the “trends since 11/78 [are] +0.132 deg. C [+0.234 F] per decade.”

It’s also worth noting that this current El Ni±o is puny compared to the one in 1998.  This is from the weekly update, “ENSO Cycle: Recent Evolution, Current Status and Predictions” from NOAA’s National Climatic Data Center:

Nino 3-10

The current El Ni±o is already winding down, although the temperature anomaly in the key region of the tropical Pacific has stayed steady in the last few week around 1.2°C.

Finally, the record temperatures we’re seeing now are especially impressive because we’ve been in “the deepest solar minimum in nearly a century.” It now appears to be over.

But, of course, it’s hard to stop the upward march of human-caused global warming — other than by a sharp reductions in greenhouse gas emissions.

Anyway, the media and anti-science crowd probably only have a few more months to push the “global cooling” meme, because it’s looking increasingly likely that — barring a very sharp drop into a La Ni±a like we saw 2 years ago (or a major volcano) — this will be the hottest year on record in every temperature dataset.

Memo to media:  If you are going to try to squeeze in one more global cooling story before the end of the year, please do remember that 80% of anthropogenic warming was always expected by scientists to go into the oceans, which have seen a pretty steady increase in heat content:

Figure 1: “Total Earth Heat Content [anomaly] from 1950 (Murphy et al. 2009). Ocean data taken from Domingues et al 2008.”

And from another JGR article, “Global hydrographic variability patterns during 2003-2008” (subs. req’d, draft here [big PDF])

Figure 2: Time series of global mean heat storage (from 0 to 1.24 miles).

So perhaps we are near the end of the global cooling nonsense, at least until the next big La Ni±a or volcano….

UPDATE:  Their are flaws in Spencer’s graphing system.  Dr. Danny Braswell, NSSTC, an author of the graphing page, writes me, “The period used to compute the 20yr record highs ended several years ago.  Gaps are because of missing data.”  He also writes, “The yearly plots are computed with temperature data from AMSU. The first AMSU was launched in May 1998 on NOAA-15 and data became available later that year.  The 20 year records are only there for a few channels and are based on data from the older MSU.”

That means if you want a plot that includes the first half of 1998, which set many records, you need to use Channel 5 and include the “20-year record highs.”  But if you want the true record highs, you need to include all recent years (although that makes the graph messy).  That’s what I’ve done here:

UAH 3-10 Full

So March 2010 still looks like it will blow out the records.

Related Posts:


40 Responses to Global cooling bites the dust: Hottest January followed by second hottest February. Now March is busting out.

  1. Kevin Johnstone says:

    Sorry Joe, this meme will never die. A new record high will just give them a new starting point for their cooling graphs. In a few years it will be ” if global warming is real how come the world has been cooling since 2010?”.

  2. paulm says:

    This year looks like it might be the reign of climate extreme terror!

    (and we haven’t even warmed by 1C yet!)

  3. Steve Bloom says:

    Joe, there’s something wrong with that 20 year record trend line on the UAH graph, as plugging in all of the available years shows a number of points in time when it’s been exceeded. Although this month is still running much warmer than any of the available individual years, since those only start partway into 1998 we can’t tell for sure that this month is the sharp record the graph makes it appear to be.

    [JR: You are right. There is something wrong. I had noticed that, but as you can see even when you put in all the years, March 2010 is still a stand out. Indeed, for this to be the hottest year, each month obviously does not have to be the hottest month of all time. It already seems likely that a large fraction of the month this year will be the hottest or second hottest on record.

    I have emailed them and see if they can explain/fix this. Frankly, what is most annoying to me about this graphing system is that they don’t post the 1998 data for the first half of that year. I probably should have complained to them earlier. If they get around to fixing it, I will repost a better graph.

    Note that Spencer’s name is on this flawed graphing system as one of 2 page authors!]

  4. zj 77 says:

    “But you can’t show anthropogenic causation!!!” they’ll howl.

    Might you drive the correlation home by including at the end of this grave posting the most recent graph showing atmospheric GHG concentrations over time?

    Because anti-science deniers get embarrassingly tangled trying to exclaim that “it is purely coincidental!” that temperatures are increasing right along with GHG concentrations, albeit with a (terrifying) 30-year lag.

  5. Michael T says:

    This NOAA/NCDC graph is very impressive in showing the rise in CO2 and global temperature.

  6. MapleLeaf says:


    I could be wrong, but I think the reason they only show data for the second half of 1998 on the site in question is b/c that is when the new AMSU sensors came online. Of course they do have MSU data before that, so maybe they can splice/merge them into a continuous time series on that site.

    Yes, I too have noticed the error with the max. line. Regardless, as you say March is going to be the warmest, or very close to the warmest, on record

  7. Tim A says:

    The funniest thing about this post is;
    If you look at the UAH’s trend since 1995 it’s almost 0.20C/decade.

    I would assume that is significant at the 95% confidence level?

    I predict Spencer’s graphs will be become less and less relevant over time.

  8. MarkB says:

    “So perhaps we are near the end of the global cooling nonsense, at least until the next big La Niña or volcano…”

    Absolutely not. While Canada had its warmest winter on record, it was below average in the U.S. southeast. Below-normal temperatures in some location will always exist in any given month or season, regardless of where global mean temperature is at. Deniers will be sure to point out those locations and use it as evidence that the temperature record is fraudulent. Just point out that Roy Spencer is a fervent contrarian and watch them run away in confusion, scrambling for new talking points.

  9. Steve L says:

    Joe, thanks for providing links so we can make our own graphs. What channels or elevations or whatever should we pay the most attention to? When deniers cite UAH data, to which are they referring?

    [JR: Channel 5 is what is most commonly used. The graphing system is flawed.]

  10. Wit's End says:

    Why can’t we just admit reality?
    The seasons are screwed up, the plants are careening out of season.
    And let’s face it, without plants, humans aren’t going to find much to eat.

    Anybody who imagines that climate change isn’t going to alter their (YOUR and YOUR CHILDREN”S) ability to obtain food is just as pathetic a denier as Beck, Morono, and BilloReilly.

  11. Mark S says:

    Joe (or anyone who can explain) a quick question: why do you use the 14.4k layer instead of the near surface layer?

  12. Dave says:

    I think you’re misinterpreting the graph. The 20-year record high, 20-year record low, and 20-year normal are all based on the period 1979-1998. The 20 year record high at the beginning of the year is from 1998. Some more recent years have had occasional periods above that – which is why the record high isn’t the warmest at all times. Of course, that’s to be expected – we’re still experiencing global warming. In any case, while the graph doesn’t exactly follow the reported numbers, it’s usually a good proxy. So a record warm March is quite likely.

  13. Dave says:


    I think Channel 5 is used because the satellites measure the lower tropospheric temperature (i.e. not just the near surface).

  14. MapleLeaf says:

    MarkS @ 11,

    The RSS MSU product is much better explained, nice graphics too. See here,

  15. MapleLeaf says:

    Dave @12, you appear to be correct about how they calc. the 20-yr record values. That is a little odd,a nd they need to communicate that better.

    Anyhow, thus far in March 2010 the near 600 mb temps. have been higher than any values observed during the satellite record for the month of March. So it would appear highly likely that March 2010 will be the warmest since at least 1979.

    Global SSTs are also running quite a bit higher than data since 2003 (from same AMSU site).

  16. steve says:

    I looked at the site that updates the temperature everyday from satellites and it gives a variety of choices of altitude to choose from. It seems for this graph, you chose 14,000 ft up so I’m wondering why that was chosen as opposed to higher or lower? I’m really curious because I would like to start checking it myself but have no idea which altitude to choose and why. Thanks so much in advance!


  17. MapleLeaf says:

    Steve @16,

    Depends what you wish to do with the data. If you are interested in monitoring near surface temperatures, then the TLT data are best (for RSS). The mid-tropospheric data (RSS, TMT) are comparable with those from the global upper-air network (GUAN), i.e., weather balloon network. Those data are processed and used to generate a product called RATPAC. As for the UAH AMSU data I would recommend following the near surface (ch04) or 600 mb (ch05).

    The contrarians like to use the mid-tropospheric data b/c the long term rate of warring is slightly lower. IMO, the TLT data are best correlated with the global surface air temperatures though, and thus most indicative of what is happening in that portion of the atmosphere which is most relevant to us.

  18. chrism says:

    personally i don’t think there is necessarily any trend in the satellite data alone, not the kind you can use to predict. it’s too short a period. I mean, on the left half of the graph, it kind of lopes along at one level, then after 1998 it seems to have been bumped up, but who’s to say what’s going to happen in the future? I mean, looking only at that I would think it’s just as likely to go back down tomorrow where it was in ’79-’98.

    in my opinion in order to identify a new trend in a series of data, there has to be a long-established status-quo that is clearly being broken. 1979-2010 just doesn’t offer enough of a track record, given the multiple cycles to which we understand the climate is subject.

  19. Mark S says:

    mapleleaf, thanks for the link, very nice.

    steve @16, I think the reason why Joe is using the 14.4k layer (channel 5) is that it is the only channel programmed to draw a graph with 20 year average and 20 year record high. It’s probably easier for him to display to his audience. I just started tracking channel 4 (near surface layer) about a month ago because I thought it would be the one that most closely correlates with surface temp records. Hopefully someone with some knowledge on the subject can respond and we can both learn if that is the best channel to monitor.

  20. joe says:

    chrism: Are these records long enough?
    (credit: Bart Verheggen)

    And in case you’re wondering how the satellite records compare with the surface temperature records, the answer is “very well”:

  21. B McC says:

    Another in a line of great posts – and very nice use of irony. Keep it up!

  22. Rune Jensen says:

    Everyone really has to start doing what they can to save the environment. I love the article btw! :)

    Also just found this project which can CO2 neutralize your Facebook profile for a pretty reasonable donation to the environment:

  23. fj2 says:

    There is a huge war brewing: A battle of unprecedented benevolence for a planet where human civilization continues to thrive and advance into states increasingly utopian.

    Unfortunately, we often are our own worst enemies, petty, tenacious, formidable.

  24. Turboblocke says:

    Are you aware of Spencer’s (and McKitrick’s)participation in the Cornwall Alliance:

    “The ECI’s “Call to Action” rests on the following four assumptions:
    • Human emissions of carbon dioxide and other greenhouse gases into the atmosphere as we burn
    fuels for energy are the main cause of global warming.
    • Global warming is not only real (which we do not contest) but is almost certainly going to be
    catastrophic in its consequences for humanity–especially the poor.
    • Reducing carbon dioxide emissions would so curtail global warming as to significantly reduce
    its anticipated harmful effects.
    • Mandatory carbon dioxide emissions reductions would achieve that end with overall effects that
    would be more beneficial than harmful to humanity and the rest of the world’s inhabitants.
    All of these assumptions, we shall argue below, are false, probably false, or exaggerated.”

    Scary stuff.

  25. dhogaza says:

    I looked at the site that updates the temperature everyday from satellites and it gives a variety of choices of altitude to choose from. It seems for this graph, you chose 14,000 ft up so I’m wondering why that was chosen as opposed to higher or lower? I’m really curious because I would like to start checking it myself but have no idea which altitude to choose and why. Thanks so much in advance!

    The channel looking at the upper level of the troposphere gets a bit of the stratosphere, too, as I remember.

    Also, one channel on two of the older MSU sensor was bad on a couple of the early satellites, so there’s not reliable data all the way back to 1978. The channel observing centered at the 14K level (4 on AMSU, IIRC 3 on the older MSU sensor but I’m too lazy too look) has good data going back to 1978.

  26. dhogaza says:

    Eh, I really need my coffee, hopefully that’s not so fractured as to be incomprehensible …

  27. Chris Dudley says:


    Just a reminder that I like these monthly posts very much. I’d give a little more emphasis to the GISS surface data than to other sets in each post myself but keeping tabs in this way is great.

    [JR: Thanks. Usually I do separate posts, and I agree about GISS but they provide so little background, I tend to combine them with NOAA/NCDC. This month I’ve been busy and the same will be true of next month.]

    I cleaned up trash along the road to our house yesterday. This is the window when the poison ivy is not yet out nor the mosquitoes or the spider webs but the leaf litter is largely decayed so you can see bottles and cans and such in the woods. It went well, but for our St. Patrick’s day picnic in the evening, a mosquito was buzzing me. A very rapid transition to spring here in the DC area.

  28. J Bowers says:

    @ 24 turboblocke,

    Yes, I’m certainly aware of it. Check out the vow made by McKitrick, Spencer and Ball, amongst others.

  29. Chris Winter says:

    Wow, that Cornwall Alliance is scary stuff indeed. There are so many ways that 20-page document runs off the rails. I’ll just pick on one, from page 16:

    “The Cornwall Alliance agrees with the opinions of the Asia-Pacific Partnership on Clean Development and Climate…”

    That would be the organization established as an alternative to Bush administration participation in the Kyoto Protocol? The organization that John McCain called a nice little public relations exercise?

    Right. More confirmation, as if any were needed, that this is another attempt to justify business as usual.

    [JR: Read the stuff on the biblical climate change.]

  30. Paul Pierett says:

    Using the same web site NOAA:

    You will find that the average winter temperatures have dropped in the last 10 years by 6 degrees.

    If you go to SIDC and build your own sunspot charts from scratch you will have something to correlate your temperatures to.

    Or, you can go to October Newsletter and Under my name, I did that all for you.

    Most Sincerely,

    Paul Pierett

  31. KeenOn350 says:


    I am curious as to why this post of yours can say
    “Hottest January followed by second hottest February. Now March is busting out” – with what appears to be valid reference to UAH atmospheric dataset on atmospheric temps,

    while at the NOAA site we also get this headline:
    “NOAA: Sixth Warmest February in Combined Global Surface Temperature, Fifth Warmest December-February”

    I have followed climate crisis info quite a lot – but find this a bit confusing unfortunately. I note that NOAA article refers to “20th century average” as baseline for comparison – believe UAH sets “hottest” scales against 1961-90 avg or some such.

    Is there need for some single website, somewhere, that would rationalize figures etc. to a common base period, explain how different rankings are defined.

    It doesn’t really create a great problem for me, but we can’t afford to give the “deNihilists” any extra ammo for their arguments on “it’s hottest – it isn’t – ’tis too!!”

    Thx for great blog, great work.

    [JR: I should have linked to the NOAA report. Sorry about that. They all use slightly different schemes and of course the satellite data typically reported is not surface temperature, so you wouldn’t expect it to match up exactly with those who are trying to measure surface temperature.]

  32. fj2 says:

    Paul Krugman
    “Hot Stuff”, March 18, 2010, 2:10 pm

    “Hmm. When the Northeast had an unusually snowy winter, it proved that global warming was a myth. But when temperatures are much higher than usual for mid-March . . . crickets chirping.

    “Meanwhile, here’s global temperatures so far this year. The yellow line shows 2005, the warmest year to date.”

  33. MapleLeaf says:


    Good points. Yes, it can all be rather confusing at times, and is complicated by the fact that the MSU satellite data only go back to 1979. Generally the different data compare very well, and rankings can be a little misleading b/c sometimes rankings are separated by 100th of a degree (in C), so slight changes can move the ranking around.

    Bart Verheggen’s site has some nice graphics:


    The latter site is great for generating your own graphics.

    Regarding the MSU satellite data, it should be noted that the temps in the troposphere are more sensitive to El Nino and L a Nina events than are global surface air temperatures. So that could be why the satellite data (incl. RSS) are ranking higher than the global surface data.

    Either way, November 2009, January 2010 and probably March 2010 will be the warmest on record for the UAH data. So when making these claims one has to be very careful to site the source and place the numbers in context (base period of anomalies etc.), and I think Joe has done that.

    The climate system is in a positive energy imbalance and is continuing to accumulate heat on account of higher GHG levels (most of which are from anthro activities)– see work by Murphy et al. (2009, JGR-A)

  34. asterisk says:

    This is steve@16 (changed my name because I see another steve here)

    First of all thanks for the replies. Second, could someone please tell me if I am reading too much into this…

    My reading is that not only is this looking like the hottest years ever, but it seems miles above the others so far (including as the previous post points out, 2005).

    Considering that the permafrost was just 2 weeks ago reported to be leaking ‘record’ levels of methane that shocked scientists, and that was with LAST years temps (which were cooler).

    I mean, seriously are we talking about major MAJOR MAJOR problems in the next year or two (a la James Locklove)?

    I guess I’m asking because I just saw the movie a REALLY inconvenienct truth by Dan Miller (see it if you haven’t…on Youtube) and it was bleak as hell. In it, he said the number 1 thing he fears is the melting permafrost. Then a week later I saw the headlines “permafrost leaking methane at record levels), now I see that this year is SIGNIFICANTLY hotter than all previous ones.

    I mean, seriously guys, is it me or is the situation REALLY NOT LOOKING GOOD??!!

  35. MapleLeaf says:

    Steve @ 34,

    It is not clear yet whether or not the ‘leaking methane’ is related to warming. In facts, they do not know. Anyhow, so while that may be true, it does not mean that further warming will not release even more methane from the hydrates. The melting of terrestrial permafrost is a very valid concern, and will escalate should Eurasia and N. Canada continue to warm– that would not be good in terms of positive feedbacks.

    Yes, the AMSU data for March (up until now anyways) are way above normal. That said, there is a 5 or so month delay between the peak of the ENSO and global air temperatures. So the impact of this El Nino event on global air temps will probably peak sometime between March and May. What will be really interesting is what happens after this current El Nino has dissipated (current indications/guidance are that that will happen in June-July).

    The demise of the El Nino could also make for an interesting N. Atlantic hurricane seasons, especially given that SSTs in the tropical N. Atlantic are way above normal, and expected to remain so until the hurricane season starts:

  36. asterisk says:

    Damn, there is some serious high IQ going on in this forum! :-/ lol

  37. John Hirsch says:

    I apologise if I missed a reference in the article – but is it not relevant to note the increase in CO2 in the atmosphere in these updates? Mauna Laua reports year over year increase in CO2 of 2.5 ppm Feb 2010 over Feb 2009- a dramatic acceleration over previous months. (389.91 vs. 387.41.I guess the recession is over and/or all the mitigation efforts we have been hearing about are just not real.

  38. Ray says:

    Can someone help me? I was impressed with the first graph here (of the UAH satellite data – showing the 20 year average [yellow], the record high temps [purple], and the 14,000 ft temps [green] for 2010). I wanted to use it as powerful evidence for global warming in a presentation I was giving today, but found there were aspects of the graph that I did not understand and which seemed to be counter intuitive.

    1. It is labelled as a daily global average temperature, but the yellow (20 yr average) and purple (daily 20 year high temps) are much higher in the northern hemisphere summer months (June, July, August). How can that be? Are these really northern hemisphere averages? Wouldn’t the southern hemisphere winter temperatures approximately balance out the northern summer temps and give a roughly straight line? I have a science background but not in climate science. Can anyone help me understand this apparent anomaly?

    2. It is a surprise (even at 14,000 ft) to learn that the global average temperatures vary between about minus 21.25 degrees celsius and minus 19.5 degrees celsius. That is MUCH colder than I would have expected. Is there some sampling error here? Is it weighted toward the poles and less sampling in the equatorial regions? Can a climate scientist explain this to me?

    Help anyone?

  39. asterisk says:

    good questions Ray….looking forward to reading the answers myself!

  40. Ray says:

    On Q1 in my post:

    I have drawn my own graphs using the excellent tool you offered. It is stranger than I first thought. If you go to sea level, there is no big peak in the northern hemisphere summer months. Also, if you go to 25,000 ft the peak in the northern hemisphere summer months is much smaller than at 14,000 ft.

    I’m still hoping someone can explain this.