How the Australian Bureau of Meteorology can get vital issues wrong

I have been looking at the BoM’s new iteration of adjusted Australian temperature data – ACORN SAT. Taking a look at the pdf report 049 “Techniques involved in developing the Australian Climate Observations Reference Network – Surface Air Temperature (ACORN-SAT) dataset”
p 86 of 104 Table 8 – Urban classification of ACORN-SAT locations.
– we see

My eye is drawn to to Canberra Airport 070014 classified non-urban. It just seems delusional for the BoM to fail to recognize that UHI affected air from the steady growth of Canberra has been steadily impacting airport instruments at least since WWII. – Not to mention the booming Airport Brindabella and Majura Business Parks – the suburb of Fyshwick and town of Queanbeyan. Not to mention warm air wafting around from the tons of burnt jet fuel and avgas – all of the above effects increasing over the decades.
A photo series here shows the growing business parks development at the airport itself.
image 6 of 16 shows the instruments in the top left – I have marked a copy looking east – business building to south – Fairbairn RAAF to the far left – BoM instruments new site is just above the red “BoM”. Frames 1, 2 & 4 have good views of the business park.
I am working on station by station comparisons of the various BoM adjusted versions over the last 2 decades. If anybody has ideas examples to use, please email them in or let me know.
A 2010 post on UHI at Canberra Airport.
the Google map images seem to have reverted to some odd old version ?

13 thoughts on “How the Australian Bureau of Meteorology can get vital issues wrong”

  1. Warwick, your comments about Canberra make sense. I would suggest the same for Dubbo. Some time since I have been to Dubbo but the airport then was at the edge of town, 5-8 minutes drive from the centre and was a hub for Qantaslink with flights out to Cobar, Bourke and other towns. Dubbo I think has about 40,000 people (maybe more now as it was growing rapidly)and there is housing and industry around the airport. It certainly should not be called non-urban, maybe fringe is justified but it could still be regarded as Urban.

  2. It seems things at the Canberra site changed after 24 November 2010 with a move elsewhere having occurred prior to 6 May 2011.
    Gidday Bob, Yes the old site now called Canberra Airport Comparison 70014 ended in Nov 2010 and was located north of the new site quite near the RAAF base in the upper left of photo 4 of 16 above.
    The new site Canberra Airport 70351 started Sep 2008 and is the site (white specks) just above the red “BoM” in the photo 4 of 16 above. So it was not much of a move. The story of the reason for the move is all covered on my site –
    “ACT (Canberra) Commissioner for Sustainability and the Environment chides BoM re urbanization near Canberra Airport observations site” – May 9th, 2009
    In a later post I demonstrate how the minimum UHI can be measured against a station in far south Canberra.
    “Evidence for a strong urbanization signal; 0.3 degrees C per decade in Canberra Airport temperature data 1997-2009”
    January 12th, 2010 – linked above.
    I find it bizarre that the BoM and the ACT Comm Env make a big deal about the old site being compromised by a car park and med rise building at the RAAF base- while the BoM contrives to ignore the growth of Canberra to 300K pop – the recent rapid growth of the airport business parks – increased traffic burning more fuel etc etc – which has been adding UHI contamination to the data for half a century. A classic case of ignoring the elephant in the room.

  3. As to the UHI effect for ACT/Canberra, under some weather conditions it does exist. Anyone driving in or out of the region with a outside temperature indicator on their vehicle will probably observe a temperature change. On two occasions during mild conditions (16 – 22 C) a 2C rise was noted. On both occasions we drove (from the coast) in from Bungendore direction, with the temperature rise from east of Queanbeyan; Our destination was the Garran area and left less than two hours later, travelling to Cooma, noting a return to the original temperature well out of the Tuggeranong Valley towards Williamsdale.

    An interest comparison would be to compare temperature time-series records for Cooma, Canberra (the several sets) and Goulburn. However, the recent Cooma data come from the Airport AWS some 15km SW of the town, convenient for the BoM, but not close enough for the towns folk. The use of airport AWS data sliced into long term data undoubtedly distorts the result.

    On a global scale it is difficult to understand with the addition and removal of data collection locations being at differing altitudes could not alter the determined “mean” global temperature.

    The BoM have been wary of including early data (pre-1910) lest a Stevenson Screen was used, while changing to the smaller plastic louvred devices closer to buildings, etc., connected by cable to recorders. [Some BoM devices appear to be closer to the standard screen, but many, such in the U.S. are not.] Some newer devices now have a wireless link so can be located further away.

  4. Paul80
    ALL australian state capitals were using Stevenson screens before 1900. Even WA, the last to start using them, had at least 3 by 1901.
    Melbourne had at least 30 years of data using Stevenson screens which the BOM doesn’t want to show.

    Anyway they are busy adjusting data from 1910 onwards. Strangely the adjustments all seem to increase the “warming” effect.

  5. Re Graeme Inkster’s comment: Yes, I am well aware of Warwick’s work on this topic and his “25 Towns” data and I discovered back in 2008 that Moruya Heads NSW had a warm period, 1881-1910, almost as warm as 1971-2000. Can hiding data be considered as a lack of integrity?

  6. One possible source of bias in temperatures is irrigation, called the Oasis Effect. I’ve been to quite a few regional and remote airports in WA and they typically have irrigated grass around the terminal building.

    I picked Carnarvon airport at random and found a picture of the stevenson screen. Where it is located there is green grass, absent from the surrounding area.

    I don’t know when this picture was taken, but my suspicion is that in the past steveson screens were located in irrigated areas, until the BoM woke up to the fact this wasn’t such a great idea and turned the irrigation off, introducing a warm bias.

    Note that in pretty much the whole of WA, grass that isn’t irrigated dies. So when you see green grass you can be certain it is irrigated.

  7. I noticed when looking up Gunnedah (NSW) on the ACORN data set, that it starts at 1948. The BOM data goes back to 1878 (even the daily data goes back prior to 1900 – so why not ACORN. After checking the long-term data, this might be the reason.

    OR the HQ site graph since 1900.

    Much warmer prior to 1940. Unless there was a change of thermometer (or siting), why wouldn’t you use the full data.

  8. Jo Nova and Ken Stewart have a post up about ACORN now

    there’s a link there to WAClimate with a table – below are the percentages of minimum and maximum whole .0 Fahrenheit recordings before 1972 and whole .0 Celsius recordings since 1972 at each of the 201 weather stations in the ACORN-SAT dataset
    Station Minima .0F %
    before 1972 Maxima .0F %
    before 1972 Minima .0C %
    after 1972 Maxima .0C %
    after 1972
    Canberra Airport Comparison 70014 16.5% 13.9% 12.3% 12.2%
    Canberra Aero Comparison 70338 Opened after 1972 Opened after 1972 11.4% 13.1%
    Canberra 70351 Opened after 1972 Opened after 1972 8.3% 10.4%

    sorry the table doesn’t copy properly

    In relation to the adjustments I’ve made a comment on Jo’s post that BOM agreed to disclose reasons for its adjustments and to make those publicly available which I don’t think has happened yet (but it may have and I’m not expert enough to find them)

    There is also a comment by RichardC from NZ querying as to the reason for very recent adjustments

    Curiouser and Curiouser!

Leave a Reply

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