Thursday, January 15, 2015

Temperatures 2014 summary

I headed the last post on 2014 "Prospects for surface temperatures 2014 final". In my town, the evening paper used to come in three editions, announced by many newsboys - Final, Late Final, and Late Final Extra. So this is Late Final - my excuse is that GISS is dragging its feet (and NOAA hasn't even posted its November MLOST file).

I ran the TempLS Grid version, and it showed a considerable rise for December - from 0.518°C to 0.638°C. That actually makes December the warmest month of 2014. TempLS Mesh is also showing a greater rise with extra data, now from 0.59°C to 0.655°C. So I think it is time to make predictions (while we wait):

2014 Jan-Dec2010 Jan-Dec
GISS Land/Ocean0.670.66
NOAA L/O0.680.65
HADCRUT 40.5630.556

This is on the basis that GISS agrees with TempLS mesh, and NOAA/HADCRUT with TempLS grid. As you see, HADCRUT and GISS narrowly reach a record, NOAA with more to spare. Actually, my GISS estimate came to 0.675, so 0,68 is equally likely.

Update: GISS and NOAA have now released their results with a  joint press release. GISS gave 0.68°C as their 2014 value; NOAA announced 0.69°C (re 20th Cen ave, it's worse than I thought ;)).

Update. There is an active plot of the historic record years of all major indices (and also both TempLS) in this later post.


  1. I just want to remind the world that when December was looking ice cold, I stayed with my prediction that it was going to be pretty warm!

  2. Thanks for adding all the bells and whistles, Nick. On to 2015. Note that NCEP has gone warm.

    1. Thanks JCH. Yes, your faith in December was right. NCEP has been very up and down. The average for Jan is still fairly low. But it hasn't been a good quantitative predictor lately.

  3. Berkeley Earth reported the warmest year too:

    1. The global surface temperature average (land and sea) for 2014 was nominally
    the warmest since the global instrumental record began in 1850; however, within the margin of error, it is tied with 2005 and 2010 and so we can’t be certain it set a new record.
    2. For the land, 2014 was nominally the 4th warmest year since 1753 (when the land surface temperature record began)
    3. For the sea, 2014 was the warmest year on record since 1850
    4. For the contiguous United States, 2014 ranked nominally as the 38th warmest
    year on record since 1850.

  4. 38th warmest? - members of the contiguous lower 48 Tea Party must have left their refrigerator doors open all year long. Memo never made it Alaska.

  5. Yes, well predicted, Nick. You remarked the other day that it's four years since The Great Climate Change Debate Reconciliation Conference in LIsbon. One of the main talking points then was "the pause". I see Judith Curry is not letting go of the issue quietly, despite the news from NOAA and GISS, indeed supporting a US Senator whose response has been to say there has still been no temperature rise since 2000. I've posted a response at Hotwhopper, citing data from your temperature trend plots: thank you for providing this resource.

    1. Bill Hartree: I see Judith Curry is not letting go of the issue quietly, despite the news from NOAA and GISS, indeed supporting a US Senator whose response has been to say there has still been no temperature rise since 2000.

      I'd dispute that there's been no temperature rise (positive trend) since 2000 in any case, since this is true only if you cherry pick the data seta.

      But in any case, a single year, especially one that is not even statistically warmer than the competing "warmest years", does not a temperature trend make. Pointing to this year by itself as evidence of anything is pretty much the same as treating weather as climate.

    2. Judith's post at

    3. Carrick, Senator Cruz claims that despite the data for 2014 that there is STILL no increase since 2000. Juidth argues that he is "broadly correct" in so doing. For further info see discussion at as well as the link to Judith's blog

    4. Bill thanks, I just looked at it.

      I don't agree at all with this statement from Judith:

      So, what is wrong with Cruz’s statement? Well, assuming that by ‘recorded warming’, he means the satellite-derived lower atmospheric surface temperatures his statement is absolutely correct, since at best the evidence that there was "no warming" is equivocal. RSS shows a cool down, but I think there is a good reason to select UAH over them for that period.

      UAH of course shows one of the fastest warming rates.

    5. Bill,
      Yes, Judith has been totally erratic on surface temp. I see Sou has a new post setting all that out.

  6. One thing that will be interesting at the end of the February 2015 will be to look at the 12 months March 2014 through February 2015. It could be the warmest 12 months ever, and it could be by a pretty wide margin. On GISS, dropping a 68 for Jan 2014 and a 43 for February, and replacing them with low-to-mid 70s, which is possible, makes quite a difference. After that, just depends on how the winds blow.

  7. There is a very useful slide in the NOAA/NASA presentation which estimates the probability of each of the highest years in the series actually being the highest given statistical vagaries.

    Go look at the slide (~5th), but the take home is that there is only a ~5% chance that it was 1998, and all the other years were in the 2000s. The two warmest by a longshot were 2010 and 2014 and of the two 2014 was about 2.5 times more likely to be warmest in NOAA and 1.5 times more likely in GISS.

    1. I get something similar by asking the question "is 2014 the warmest year on record"? As I wrote on Lucia's blog:

      Well here’s the top three temperatures for GISTEMP:

      2014 0.68
      2010 0.66
      2005 0.65
      2007 0.62

      But of course these measurements have uncertainty associated with them.

      This figure shows the GISTEMP estimate of annual uncertainty, which appears to be take from Global Surface Temperature Change, Hansen et al. 2010. Table 1 of that paper gives the uncertainty from 1960-2008 as ±0.05°C (2σ).
      So looking at the difference in temperatures (for which the uncertainty is approximately $latex \sqrt{2} * 0.05^\circ \approx 0.07^\circ$), I get:

      Range,DeltaT,p-value (1-sided)

      So 2010 and 2005 are both statistically indistinguishable from 2014, based on the GISTEMP data and analysis products.

      But remember, Sou assures us:

      "It wasn’t an “almost warmest year”, it was the warmest year. No almost about it".


    2. I'd be interested in seeing GISTEMP with the Cowtan and Way correction added to it. I seem to remember Kevin Cowtan saying something about their correction leading to a bit lower number after correction.

      But either way, it'd be interesting to look at.

    3. I mean to say "I seem to remember Kevin Cowtan saying something about their correction leading to a bit lower number after correction for 2014"

    4. Carrick,
      "I get something similar by asking the question "is 2014 the warmest year on record"?"

      So what is your answer? Is it 2014? 2010? Can't say? Then was there ever a warmest year? Have we been wrong when we spoke of 2010, or earlier 1998?

      What does a year have to do to be the warmest? Be 0.05 C ahead of any other? 0.07?

    5. What I said on Lucia's blog was: What matters is what measurement science says, and if measurement science says two measurements aren’t sufficiently distinguishable within measurement uncertainty so that we can’t make a conclusion that one was larger than the other, then we can’t make a conclusion that one was larger than the other.
      That’s how science works.
      In any case, there really isn’t anything particularly exciting about an individual year being the warmest by itself, so I’m not going to break out in a sweat if we can’t say, within measurement error, that any particular year was the warmest.
      It’s much more important to me whether this year is part of a pattern versus whether it’s a “one off”. Single datums do not a trend make.
      But if you reduce your error (presumably by fixing the holes where there’s not data, like near the poles), your measurement error goes down. So yes, probably there will come a day when we can say ‘this year was the warmest’.

  8. Don't worry Carrick, when 2015 becomes the warmest year with no doubt, 2014 will get its comeuppance

    1. Using Carrick's reasoning, every sporting event or political election that has required refereeing or officiating has uncertainty with respect to the outcome. Every refereeing decision has uncertainty associated with it so that the winning outcome should not be finalized to 100% certainty.

      The likelihood that the last Bush was not actually elected president was likely well above 50%, yet he was president according to the history books.

    2. The photo finish is tossed on the trash bin of history. The nose/bumper out front is no longer a win. Secretariat is now just another horse.

    3. One of the most shameful appellations that the nasty WUWT hordes apply is "racehorse".

      JCH, let's symbolically shove this in their faces.

  9. JCH, a horse race involves direct measurements in a single point in time. Even then there are dead heats, races too close to call.

    However, the quantity reported by GISTEMP for example: annually averaged global mean temperature anomaly series has little in common with a horse race. It is what t I would refer to as an "indirect measurement":

    It relies on a series of direct measurements (voltages that are converted into temperature using calibration values) at different points in time and different places on the Earth and uses a series of algorithms to combine these numbers to produce a single quantity.

    The number that has been reported for 2014 has measurement uncertainty with it, but in addition the value almost certainly will change in time as more data arrives, and will change over time as the algorithm gets tweaked.

    This is easily seen by looking at how previous annual temperatures have changed for GISTEMP:


    The data series used to produce this figure are located here.

    As I mention on Lucia's blog, I view the temporal variation of these quantities as estimates of the uncertainty in the "recorded value". That is, the recorded value at any point in time is based on incomplete data and algorithms that will get tweaked and improved in time. We'd expect that given enough time (probably a decade) all of the data sources used by GHCN will have reported, if they're ever going to report.

    And the algorithms will evolve as researchers understand the problem better. I could see an update like Cowtan & Way using some type of medium resolution weather model to reconstruct the field in regions where there is missing data for example.

    Anyway, if you look at the Secretariat Race, the video remains unchanged between then and now. If you look at the temperature record for a city, that doesn't change either.

    But thinking that the result from a horse race has anything in common with the quantity reported in a "annually averaged global mean temperature anomaly series" is I'm afraid is incredibly naive on your part.

    I think we can learn a lot by seeing how the authors of a particular work report their own results. Gavin Schmidt does not report which year has the largest value, since he knows these will probably change over time in any case. Instead he quotes the probability that a given year is truly warmest temperature on record.

  10. I do not like lucia and rarely read her blog. For awhile I participated in guessing the UAH anomaly, which was fun. I won that a couple of times, but just being there creeped me out so I stopped.

    I think there are pretty stout limits on what you can learn about somebody based upon something as trivial as the report on the warmest year of these years:

    1998 - .63C
    2005 - .65C
    2010 - .65C
    2014 - .69C

    Over the last day on Climate Etc. I've seen them called liars. Using the ward "liar" tells me a lot about the person making the accusation. It tells me they are wild-eyed, agenda-ridden radical.