More huge errors in HadCRUT3 gridded temperature data

The brave prediction, “Sydney’s climate to ‘become like Brisbane’s'” for 2100 by staff at James Cook University means that Sydney Airport will warm by ~2.75 degrees C relative to Brisbane Airport (based on 1961-1990 averages) for this prediction to come true. It is obvious that the Sydney Urban Heat Island (UHI) has already notionally moved Sydney north but we will all be departed when it falls due to adjudicate on this claim by JCU staff in 90 odd years time.

However the review paper that this prediction was extracted from is titled, “Expansion of the tropics”. The authors do not seem to present evidence directly themselves, preferring to cherry pick quotes from a wide range of IPCC compliant literature.

I just want to point out that whatever merits this concept of the “expanding tropics” might have, the tropics are only warming slightly. According to 30 years of temperature trends in the lower troposphere generated by NASA satellites and calculated by the University of Alabama at Huntsville, the tropics are warming at about 0.05 C per decade. That trend is partly driven by cooling due to volcanoes early in the 30 year period then warming from the huge El Nino in 1998 – cooling early in the 30 years and warming late in the 30 years forms a couple which to some extent inserts a warming trend into the data.

However one hopes that none of the papers reviewed and relied upon by our JCU academics are quoting the authoritative (much IPCC quoted) HadCRUT3 land sea gridded temperature data compiled by the University of Norwich, Climate Research Unit, Dr. P. D. Jones and the UK Met. Office Hadley Centre.

HadCRUT3 errors in tropical Africa

This graphic shows that for a huge region of tropical Africa the HadCRUT3 data has errors of about 0.8 degrees C over the 30 odd years.
And world leaders are discussing huge changes to our economies assuming all the science is settled.
I should have said I got my data from the The Royal Netherlands Meteorological Institute useful website Climate Explorer. Follow the Monthly observations link on the right.

20 thoughts on “More huge errors in HadCRUT3 gridded temperature data”

  1. To Warwick or anyone who might know,
    I recently looked at the Reference Climate Station network for NSW. Do you know what the data from these stations is used for, or to whom it is sent (ie WMO)?
    Having checked out all the NSW stations, I find that some seem not to meet the BOM’s own guidelines in order to qualify as an RCS.

  2. You’ll find the same warming with the GISS (with a lot of “holes” in the time series). And with the third surface temp. database at NCDC/NOAA, the gridded data for 20N20S/10-40E don’t exist for most of the 1979-2009 period!
    No wonder why Jones doesn’t want to disclose his data and method: the warming in his data is pulled out of his hat. It’s not error, it’s scientific malfeasance.

  3. I am not sure that the RCS means much to the BoM now – it looks to me like a “good idea” from a decade long ago that is just hanging there.
    A list of WMO numbers for Australia or any selected country can be made at this US site.
    If you use “Display All Stations In a Country” – I think that is the WMO list.
    Geoff Sherrington has recently assembled a lot of the RCS data – most compare OK with satellites but there were a couple of clangers.
    I think Jones et al started with the WMO list in the 80’s when they compiled “global warming” – we are talking about circa 2000 stations USED for NH and ~300 for SH.
    Then in the early 90’s the US (NOAA-NCDC) compiled the GHCN – a total of ~6000 temperature stations.
    They had the same WMO list plus whatever else they could find – they did things their way and GHCN is distinct from Jones CRU UKMO HadCRUT data. GHCNV2 found even more stations – I have a file of 7280. GISS took the GHCN and did their own thing on again.
    So GISS data has a completely different lineage to Jones-CRU-UKMO. GISS (homogenization) has changed over the decades too – my impression is they are more prone to end up with warming trends.

  4. Sorry if this is a stupid question: The graph shows a discrepancy between the Hadcrut and UAH satellite data – why are you so sure that the Hadcrut data is incorrect and the UAH data is correct?

  5. Thanks for a good question Annabelle. We are fortunate in that there is a competing organization RSS – compiling the NASA satellite MSU data. The pioneering UAH work has been subjected to intense scrutiny over the decades and at present RSS finds a very small quota of warming more than UAH. IMHO there is a case that RSS needs to look at some cooling adjustments to their data.
    So I think there is a negligible chance that UAH data could be in error to this extent.
    If you look at the RSS Figure 3. Color coded map of decadal trends in MSU channel TLT (1979 – 2008)
    You can see for that region the anomaly colour coding would need to be dark red at the far right end of the scale for UAH to be wrong.
    Note also the RSS Fig 2 showing the altitudes of the various data products – see that TLT (temperature lower troposphere) is at an average altitude way below 5kms. So with our chosen grid box being 50 degrees E-W, at the equator this is over 5500 kms across. So the very thin skin of atmosphere measured by the MSU’s is circa one thousandth of its width. I think there is a miniscule possibility that surface trends could vary by such a large amount over the 30 years.
    Taking another approach, the Jones et al hemispheric compilations in 1986 incorporated urban heat island (UHI) warming right from day one. See Dr Fred Wood’s critique.
    My city pages and grid cell pages specify many instances where Jones et al city data warms compared to surrounding more rural stations.
    My USSR pages also demonstrate also how Jones et al are great truncators of early data that does not fit.
    In 1991 I wrote to the South African Weeburo asking if they had a greater coverage of temperature stations than I could find in public sources. As you know the nation was in some turmoil then but my letter ended up with a helpful person who sent me two diskettes of max and min data from 1960-1990 which included data far more rural then that used by Jones et al. This lead to the 1996 paper which showed up the urban bias in Jones et al trends across South Africa.
    On my blog I have many posts highlighting urban bias in Jones et al data, check out Jones et al – Surface Record and Urban Heat Islands in my Categories list down the left side.
    I will just pick out two more examples.
    I have explained EXACTLY how Jones et al incorporate urban errors, see this case from Puerto Rico which is the subject of published papers – and was picked out in the Wood critique above.
    In 2008 Chinese climate scientists finally finally got Jones et al to sign on to a paper demonstrating the very large urban signal over eastern China.
    History made as Jones et al 2008 paper admits huge urban warming in IPCC flagship CRUT3 gridded data over China
    Followed up by a closer look at the Jones et al 2008 paper.
    Will the UKMO Jones gridded data be adjusted ? They have had over a year.
    Sorry to have gone on Annabelle but the entire issue is complex and goes back over 2 decades now.

  6. Thanks Warwick

    The RCS network I found is at
    and seems to be updated to Oct 2007. It doesn’t reflect the link you gave me to the WMO which appears to be the weather stations they use to calculate Aust temp averages. As you suggest, I good idea at the time.
    Anyway, I ran some checks on the NSW stations and nearly 40% do not meet the criteria they give for an RCS. Two have even closed.

  7. Annabelle Says:
    July 11th, 2009 at 2:13 am
    Thanks for taking the time to write such a detailed and thorough reply.

    On the other hand, Annabelle, the GISS, HadCrut, and RSS (satellite) global trends are at ~0.16 deg per decade remarkably similar over the past 30 years. UAH is slightly less at ~0.13 deg per decade. However, since 1992 the trends of all 4 datasets are virtually identical (at ~0.2 deg per decade), which suggests that UAH was “running a little warm” in it’s first decade of operation.

    The Urban Heat issue is not the big problem some believe. For a start 70% of the earth’s surface is covered in water. Across the 48 contiguous US states where Anthony Watts has highlighted station siting problems the GISS (surface) and UAH (satellite) trends are the same,i.e. 0.25 deg per decade.

    Finally, Warwick makes a point about the USSR. If there is a problem with former Soviet Union data, why aren’t Russia making more of a fuss. The last thing the Russians want is for the world to get bogged down in energy regulation. They’re one of the few countries who have argued against AGW. The Russian Academy of Science has rejected the science behind AGW. Why haven’t they made more of a fuss about “dodgy data” from their own country.

  8. SST were adjusted to better fit land (UHI affected) trends – so the UHI lives on.
    Russian rural data is gappy like much rural data – I just can not recall now how it rates with other places – I recall Canada did not rate well either.
    scroll way down to link;
    Report comparing record quality in national datasets above 50 degrees north. Downloadable zipped Word 97 file .
    I think it is a red herring to lay it on the Russians to rebut Western bad science. They have had quite enough to do for 2 or 3 decades copiing with the greatest case of a failed State in history.
    The main issue with global T trends is the data Jones chose to use or reject.
    I think I show case after case why that is so in my USSR study.

  9. For Ian George,

    Might you please be so kind as to email ne a list of what you consider to be te most recent BOM RC stations? If you have gone to the trouble of making annual averages, that would be wonderful.

    Thanks Geoff.

    P.S. Here are a couple of graphs illustrating how different agencies with different adjustments have made different results. I do not know which one is the best to use for serious work, bit they cannot all be right. I’m intereted in how CRU will fit in these – maybe there will be several versions and Life’s a Lotto.

  10. I can’t think of any CRU gridbox that has had its trend confirmed upon examination by someone outside of CRU. Maybe Warwick can correct me.

    The CRU raw data is simply monthly means of the daily averages. If CRU had been doing its job properly, they would have similar files for daily maxima and daily minima. Trends in daily minima seem to be about 3 times larger than trends in daily maxima everywhere they have been examined. Thus, it is likely that the 0.6 C/century global trend in the daily averages is really 0.9 C/century in the minima and 0.3 C/century in the maxima. The difference in trends arises from UHI and microsite issues, so the real warming of the 20th century was probably no larger than 0.3 C.

  11. I tried to post some graphs but they did not come through.

    The limited rural Australian adata I have examined in Tmax and Tmin terms shows about all combinations of convergence, divergence and parallel, whatever these mean within the noise envelope. I have a developing suspicion that UHI is greater at night, but I am working mostly with truly rural stations. I’m not at the stage of working with grid boxes because coming from a background of interpolation of chemical analyses between grid points, in 2D and 3D, even-spaced and otherwise, I can see a whole chapter that will be some time in the future.

    A problem with station data is that you need a lot of them combined to bring out reliable trends. I do not have the maths packs to do this well. But, ask me if I have seen an effect and the chances are i can show you a station that displays it. The whole country might not. Signal/noise yes, but if GW is global, one needs a lot of specific excuses to explain the many exceptions.

  12. I think you would find close agreement with satellites over the USA Doug – which I take to be due to intense grooming of each gridbox – I have a post on that back a while. Then there is Eastern China with huge discrepancies – where Jones was shot down in flames by the 2008 paper. I would welcome a global gridbox map of HadCRUT3 minus UAH. Come on someone.

  13. Can anyone tell me why the Hadcrut3 data for September is about a week late. I’ve never seen it more than a few days difference to the 20th of the month, and now the month is almost over and I’m seriously wondering what legitimate reason they could have for delaying.

  14. What happened to the 2009 temps?

    All months have been deleted back to Mar 2009.

    First September was 2 weeks late.

    Then the email release.

    Now temps are deleted.

    What’s going on?

  15. njcons, mike and others; normally by this time of the year their minds are working on the usual pre-Xmas puff that 200x was the nth hottest year stuff. Maybe this year they are occupied with other matters. Some things we can say for certain – whatever is going on with pulling and revising their global HadCRUT3 or CRUT3 is happening in utter secrecy and we will likely never know the truth.

  16. as ever, Wikipedia has an informative article about the satellite temperature discrepancies which addresses a lot of interesting points on both sides of the argument — e.g. pointing out that tropospheric temperature should rise at a greater rate than surface temp, which was not the case in the tropical dataset provided — while explaining potential issues with the satellite readings.

    Satellites do not measure temperature as such. They measure radiances in various wavelength bands, which must then be mathematically inverted to obtain indirect inferences of temperature.[1][2] The resulting temperature profiles depend on details of the methods that are used to obtain temperatures from radiances. As a result, different groups that have analyzed the satellite data to calculate temperature trends have obtained a range of values. Among these groups are Remote Sensing Systems (RSS) and the University of Alabama in Huntsville (UAH). Furthermore the satellite series is not fully homogeneous – it is constructed from a series of satellites with similar but not identical instrumentation. The sensors are subject to fade over time, and corrections are necessary for satellite drift in orbit. Particularly large differences between reconstructed temperature series occur at the few times when there is little temporal overlap between successive satellites, making intercalibration difficult.

    In the late 1990s the disagreement between the surface temperature record and the satellite records was a subject of research and debate. The lack of warming then seen in the records was noted.[21] A report by the National Research Council that reviewed upper air temperature trends stated:
    “Data collected by satellites and balloon-borne instruments since 1979 indicate little if any warming of the low- to mid-troposphere—the atmospheric layer extending up to about 5 miles from the Earth’s surface. Climate models generally predict that temperatures should increase in the upper air as well as at the surface if increased concentrations of greenhouse gases are causing the warming.”[22]
    The same panel then concluded that
    “the warming trend in global-mean surface temperature observations during the past 20 years is undoubtedly real and is substantially greater than the average rate of warming during the twentieth century. The disparity between surface and upper air trends in no way invalidates the conclusion that surface temperature has been rising.”[23][24]
    As noted earlier, these temperature data, misinterpreted from the satellite data, are now known to have been too low.
    An important critique of the satellite record is its shortness—adding a few years on to the record or picking a particular time frame can change the trends considerably.

    The process of constructing a temperature record from a radiance record is difficult. One widely reported satellite temperature record, developed by Roy Spencer and John Christy at the University of Alabama in Huntsville (UAH), is currently version 5.2 which corrects previous errors in their analysis for orbital drift and other factors. The record comes from a succession of different satellites and problems with inter-calibration between the satellites are important, especially NOAA-9, which accounts for most of the difference between the RSS and UAH analyses [17]. NOAA-11 played a significant role in a 2005 study by Mears et al. identifying an error in the diurnal correction that leads to the 40% jump in Spencer and Christy’s trend from version 5.1 to 5.2.[18]
    For some time, the UAH satellite data’s chief significance was that they appeared to contradict a wide range of surface temperature data measurements and analyses showing warming in line with that estimated by climate models. In April 2002, for example, an analysis of the satellite temperature data showed warming of only 0.04 °C per decade, compared with surface measurements showing 0.17 ± 0.06 °C per decade. The correction of errors in the analysis of the satellite data, as noted above, have brought the two data sets more closely in line with each other.

    note particularly this graph which compares readings from all four global temperature sources:

    While there may be individual variation for whatever reason in measurements and also regionally, surely it’s the overall trend which is important — and the patterns in this graph show all four readings following a remarkably similar pattern.

    Also, looking at the readings in the Surface Temperature Anomaly Map e.g. today on the HadCRUT3 website:
    you can see that overall temperatures are higher compared to 1961-1990, though there are areas where there is no variation, notably a large swathe of Africa.

    Another useful source is the UAH temperature graph which currently shows a clear upward trend though this becomes less clear when other years are added:

Leave a Reply

Your email address will not be published. Required fields are marked *