Data Comparisons

HadCrut December: Down from November.

25 Comments

HadCrut December reported an monthly surface temperature anomalyof 0.407C; this is down 0.039C from their November anomaly. However, their update page also includes a number of provisos. I’ve plotted surface temperature anomalies reported by both GISS and HadCrut along with the multi-model mean of A1B simulations since 1975 below; december values are highlighted:Comparison of Projections and Observations of Earth's Surface Temperatures

Next week, I’ll comment more, but for now note that despite presence of El Nino,

  1. Observed HadCrut temperature anomalies (blue) remain below the multi-model mean projection (purple) and the trend computed since 1975 is lower. The HadCrut least squares trend computed starting Jan 2001 remains negative (-0.08C/decade) but if positive if computed beginning in Jan 2000 (+0.03C/decade).
  2. The recent Hadley anomalies have been falling below the trend line computed starting 1975. Mathematically, this causes that computed trend since 1975 to decline slightly as more data arrive.
  3. The overall trend since 1975 is positive for both observational sets. With respect to testing models, the question is: is the excess trend exhibited by models statistically significant?

As many are aware, I’ve been wondering if the weather noise induced by El Nino was going to be able to drive any individual monthly temperature anomalies above the the multi-model mean projections. So far, it has not. Still, surface temperatures do lag El Nino, and El Nino was ripping in Nov. and early Dec. Satellite readings indicate January was hot. Maybe in January, we’ll see weather noise manage to break through the multi-model mean projections. Or not.

I guess I should set up a HadCRUT bet!

25 thoughts on “HadCrut December: Down from November.”

  2. of 0.407C; this is down 0.039C from their November anomaly

    Edit – oops, I misread the sentence.

  3. TerryMN–
    I commit typos often enough. So I checked– but I see you did too. Yes, it’s down 0.039C. That’s not a lot. It could easily pop up quite a bit this month!

  4. I probably missed the discussion of this, but is there a known explanation for the sudden dip in the model mean about 93/94? Volcanoes?

  6. Should climate sceptics be giving this sort of data credibility by reporting and analysing it? If it is, as I suspect, largely bogus, what is the point? Auditing has proved a wonderfully fruitful activity, but that’s rather different, isn’t it?

  8. The provisos make interesting reading, it seems they are pulling their fingers out. It would be interesting to know if the corrections are due to the MET taking an interest, or whether CRU are blowing out some cobwebs. Either way I can but suspect it is due to additional oversight, but I welcome it irrespective of its cause.

    Alex

  9. dearime,
    .
    Lucia has stated that she is a lukewarmist. She believes in radiative physics, as should we all. The details do get interesting though!!!
    .
    I believe her position is similar to Steve McIntyre’s.

  10. dearime,
    .
    One other little issue, it is hilarious to watch warmists freak out over Lucia’s detailed analysis of the temp data, that has been wholly modified and controlled by the IPCC and James Hansen types, as it currently resists supporting the Holy Models and Holy Theory!!!!

  11. Well, I still consider it as a twelve year flat trend; and I still consider GISS to be an outlier.

    the denialist blogosphere still is in decline. how could a person with such an opinion be allowed to post a guest post on wattsUp?

    simply beyond me.

    ———————–

    on topic:

    let me apply denialist theory to this one.

    thermometers are concentrated in the north, and are covered with snow at the moment. this should eliminate a lot of the micro site issues, as documented by Anthony Watts.

    looking at his famous list of station quality, we should find a temperature that is more than 5°C below the usual one.

    for some weird reason, we don t find such an effect….

  12. Sod

    looking at his famous list of station quality, we should find a temperature that is more than 5°C below the usual one.

    Huh? Why should we expect to find one of these because of snow?

  14. Huh? Why should we expect to find one of these because of snow?

    the majority of micro site effect, is grass being turned into pavement. (i am ignoring the stupid barbecue and air condition units for a moment)

    according top Anthony, the majority of stations is type 4 or 5 with “error >2°C” or “error>5°C”.

    snow will remove the microsite impact of pavement. we would expect a dramatic effect on temperature. it does not happen.

  15. Lucia,

    Why does the annual series differ when averaging monthly data? I’ve been tracking the HadCRUT monthly reports and although they’ve been changed a lot during the year, the annual series is way off compared to averaging the monthly.

    http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/

    Averaging monthly series places 2009 in 6th position with 2009 at .438 and 2004 at .447, yet the annual series is 2009/.440 and 2004/.431

    What am I missing?

  16. Re: sod (Jan 31 13:05),

    I believe you are conflating error with bias or offset to create a strawman. Microsite issues do not cause a constant offset 24/7 which would disappear with snow cover.

  17. DG– I have no idea. . .
    I also have no idea why HadCRUT had a NH/SH average (for ordinary purposes), a simple area average and a variety of other products all of which differ. Did you at least compare the annual NH/SH average to the monthly NH/SH? Have you tried adjusting for number of days in a month?

  18. sod–
    I’ll admit not to follow these microsite issues in details.

    Still, why would you expect the grass/pavement error issue to mater much in winter? Grass does have transpiration cooling in summer, making grass cooler than pavement on hot supper days. Is the effect large in winter when plants are mostly dormant? Seems to me it wouldn’t be, but I don’t actually know.

    If the microsite problem is a summer problem, why would snow make any difference in winter?

    Plus, if the thermometer is over a concrete pad, what makes you think the sidewalk and paths aren’t shoveled in winter? If so, we’d still have concrete, lava rock or asphalt rather than natural surface under the thermometer.

    I’m not really understanding what you think would be easily detectable in the data. Right now, there is snow and it is cold. Thermometer are probably reporting cold temperatures in snow covered places right now because it is cold.

  19. Sod,

    You’re right! Urban Heat Island effects should only be operable in the summer!

    I can look out the window and see where snow has melted and not melted. It has not melted on the grass. It has melted on the pavement, where it was not first cleared off. So, your comment is cute, but sadly doesn’t pass the smell test.

  21. Sod,

    Now if the thermometers were covered by snow in the summer, that would tell us something about climate.

  23. I don’t know if anybody will find this interesting, but this is a fit I just obtained to the HadCrut annual data for the past 130 years:

    1.85 Log[CO2(t)] – 10.8 + 0.00219 t + 0.108 Cos[0.101t]

    where t is the time in years since 1942, and CO2(t) the CO2 concentration in ppm. (If anybody wants to do a quick plot without looking CO2(t) up, a good fit to it for the past 50 years, that is, for t>17 or so, is CO2(t) = 304 + 0.436 t + 0.012 t^2.)

    The logarithmic term is pretty much what one expects for greenhouse warming in the absence of any extra feedbacks. It predicts a change in temperature of about 1.3 C if the CO2 concentration is doubled, which I believe is about right. The linear term is, I think, consistent with pre-industrial age warming trends. The sinusoidal term is a fudge; it may or may not represent a real oscillation, but it gives an idea of the natural variability of the system over a decadal time span.

    For the 1975-2010 period a linear regression to this fit would give a slope of 0.0178, which is within 5% of what your graph shows. The current (“instantaneous”) slope is much smaller (0.006), though still positive.

    The warming for the 21st century predicted by this fit is about 1.2 C (+/- 0.1 C, because of the oscillation). I have no idea if this is right or not, but at the moment it seems to be doing better than the fancy climate models…

  24. Do we get a post on what everyone predicted for Jan Uah? Or do we wait until the actual Uah figure is released?

Comments are closed.