The Australian Temperature Record: Part 2- Northern Territory

Ken Stewart, May 2010

“All thermometers are equal, but some thermometers are more equal than others.”

(Apologies to George Orwell)

Australia’s Northern territory in the remote outback is home to the tourist meccas of Ayers Rock and Kakadu, spectacular scenery, miles of desert and savannah, and a population of 219,800.  It covers 1,349,129 square kilometres, or roughly 17.5%  of the Australian landmass.

This vast area, like the rest of Australia, has a network of weather stations managed by the Bureau of Meteorology (BOM).  BOM produces climate information for the public and the government.  The official climate record is presented as trend maps and time series graphs for Australia as a whole, various regions, and each State and territory.  This official record is produced from the High Quality Data Sites- 100 non-urban sites across Australia.

As I showed in 

the HQ series has some inherent problems:

  • It is based on data that has been subjectively and manually adjusted.
  • It makes no allowance for Urban Heat island (UHI) effects.
  • In Queensland, it produces a trend in mean temperatures that is 0.2 degrees Celsius greater than the raw data does.

 Now, how many HQ stations were used to build the Northern Territory’s climate record?  2.  That’s right, TWO.  Alice Springs Airport and Tennant Creek Airport.  Darwin isn’t used because it is urban.

These two climate records, and the thermometers that produce them, are the most powerful in the nation.   Here’s a graph showing the area of each State per HQ climate site.

Here is the official climate record for the Territory.  The trend map:

Note the big red section around Alice Springs!

And the time series:The Northern Territory has a warming trend of 1.1 degrees Celsius per 100 years, so a lot depends on those two sites, because as you can see from the table below, the trend has been increased from 0.98 degrees to 1.13 degrees per 100 years.

Round all figures to tenths, 1 degree C to 1.1C.  No, no other sites were used.  A 10% warming bias- and all due to Alice Springs, as you can see.

When you look at the individual records, the arbitrary nature of the adjustments becomes apparent. 

Tennant Creek raw data- has a Post Office site  to 1969 and the Airport:

I spliced the two, as did BOM.  Here’s the comparison:

BOM has warmed the past to decrease the trend.  But at The Alice:

There were actually a number of sites used at Alice Springs.  The Old Telegraph Station from 1878 to 1932; the town Post Office from 1932 to 1989; the old airport (Connellan’s), the new Airport (The Seven Mile) from 1941 but with several changes in position there.  Note the cooling trend to 1950, and the similarity (though slightly warmer) of Airport to PO 1943-1953.  By the way, the mid 70s dip is reflected in several other sites.  Compare the splice with HQ:

Warming increased from 0.95 to 1.65.  No nearby stations have a complete record.  Only one (Charlotte Waters) covers the early decades, and it shows cooling.  So the adjustment must be based on records from many hundreds of kilometres away- like Tennant Creek, which shows less warming!  The nearest long term station in Queensland showing similar warming is Boulia (raw trend 1.1 degrees, adjusted to >1.6 degrees!) Further south, Oodnadatta has records only from 1941, showing considerable warming since then, but there is no evidence that this warming started in 1910.  Therefore, the HQ record for Alice Springs, subjectively and manually adjusted, is arguably highly suspect.  It follows that the time series for the Northern Territory is also incorrect.

By the way, Darwin Airport, which is urban and not included, is an interesting site:  The Post Office records to 1940, and the Airport from 1941.  The Post Office measurements apparently were based on dodgy practices, like no Stevenson screen.  Raw data:

Note the airport shows a warming of about 0.8C- if the trend is extrapolated back to 1910, which is not good practice.

Now the adjusted data:

From 0.8 to 1.3 degrees per 100 years!  There are no long term stations within hundreds of kilometres of Darwin.  Daly Waters is from 1926, and surprise, does NOT support such a warming trend.

Now here’s a plot of a splice of Darwin Airport constructed by reducing PO values by 0.6C, and Daly Waters by 0.5C (to approximately match the AWS):

Darwin’s trend: 0.5C, and a good mirror of Daly Waters.  Note the 1970s dip, common throughout the inland.  BOM’s adjustment does not stand up.

Here’s the Northern Territory showing the trends at each site:


The climate record for the Northern Territory is based on very limited data.  Very few stations have long records, much data is missing, and the Trend Maps and Time Series Graphs thus depend on only two sites.  Both of them showed about 1 degree of warming, but Alice Springs has been manually adjusted to give extra warming which cannot be justified.  Similarly, Darwin’s extra warming (though not used in the record) cannot be justified.

The official BOM record for the Northern Territory is guesswork.  BOM would have been better off leaving the record as is, and saying “We don’t know enough to show any trends for the Northern Territory.”

Progress Report on the HQ series:

BOM assertion:  “On the issue of adjustments you find that these have a near zero impact on the all Australian temperature because these tend to be equally positive and negative across the network (as would be expected given they are adjustments for random station changes).”

After 30 stations (30%): Raw trend: +0.79C  HQ trend: +0.96C Warming bias 21.5% (in Celsius terms)

About these ads

23 Responses to “The Australian Temperature Record: Part 2- Northern Territory”

  1. val majkus Says:

    Hi Ken; great work; I’ve sent the link to a number of friends and sceptic colleagues including Warwick Hughes whose work I admire also; great to see you on Jo Nova’s site as well

    • kenskingdom Says:

      Thanks Val- nice to hear from you. We’ve had a 6 hour power outage today so I’m slow to catch up. It’ll take some time to complete this but we’re on a roll. Next state is Western Australia- Jo’s home. Thanks for your support.

  2. FrankK Says:


    Looking at the raw data from Tenant Ck and Darwin airport. Particularly from about 1988 to the end of record it looks to me to be a decreasing temp trend does it not?? What does the Excel trend line say for this period at those two sites ?

    • kenskingdom Says:

      Gday FranK:
      Yes, it looks to be dropping since 1988. 4th order polynomial shows this clearly. But trends are funny things- can go anywhere. Which is why we shouldn’t use them to forecast the future.

  3. Sinan Unur Says:

    This is good stuff. One point about that BOM assertion though: I can see no reason to expect adjustments to add to zero given that these are not random and independent changes to station environments.

    • kenskingdom Says:

      I agree. But the highest in BOM say otherwise- purely random and independent adjustments! You would think that a quality control check would have found and corrected this if it was it was random- so either there was no quality control or else it is deliberate. Am I being unfair? We’ll see after I have finished the 100 sites.

  4. Tweets that mention The Australian Temperature Record: Part 2- Northern Territory « kenskingdom -- Says:

    [...] This post was mentioned on Twitter by Maurice H Rich, Alice on Todd. Alice on Todd said: The Australian Temperature Record: Part 2- Northern Territory …: There were actually a number of sites used at A… [...]

  5. Ian Beale Says:


    From a comment well down in

    there is this one for Australia

    (For the record in case you missed it)

  6. FrankK Says:

    “But trends are funny things- can go anywhere. Which is why we shouldn’t use them to forecast the future.

    Its interesting that the graphs on pages 20 to 22 in the pdf download at this site.

    also show more recent decreasing temp trends.

  7. val majkus Says:

    Ken I had an idea (but I may be completely wrong) that there was someone doing in Western Australia what you are currently doing for Qld and NT; might be worth asking Warwick Hughes if he is aware as to whether or not that is the case;

  8. FrankK Says:

    “Time will tell. 100 years is even too short to see trends.

    Well, some have indicated that the Pacific Decadal cyle is some 60 years in duration and that quite definite trends are evident over the period throughout 1880 to 2010. See:

    Re F

    • kenskingdom Says:

      Yes I do agree- I was talking about the times since Roman WP and Mediaeval WP, warming since little Ice Age. AGW is focussed on far too short time frames. It is possible to see 30 year and 60 year trends in quite a few sites and they probably have a lot to do with ENSO, PDO, Indian Ocean Dipole. As we learn more we will unravel these mysteries. BOM’s climate department doesn’t help with bogus warming.

  9. Geoff Sherrington Says:


    Here is the graph of Tav for Alice Springs from the data in Jones’ early work. This ends in 1991. I have infilled an occasional month by averagie of those around it. Data from Warwick Hughes.

    When you add later efforts, you get the customary spaghetti as shown here:

    Legends for series. 1=Jones as before. 2=BOM online. 3=BOM CD. 4=GISS homogenised from 1904. 5=KNMI using GHCN ver 2 adjusted, finishes 1992.

    It seems like a case of the deeper you dig, the worse it gets.

    • kenskingdom Says:

      Gday Geoff
      Thanks! And Torok and Nicholls say Alice Springs is an urban site ,therefore should not be included. But BOM still uses it- even though for part of the time it was located at the post office in the middle of town.
      As you say, the deeper you dig…. Just wait till you see the WA stuff.
      Someone did some work on WA sites that I saw a while ago but I can’t find it…was it you?

  10. Geoff Sherrington Says:

    Ken, I have done detailed work on Broome, Meekatharra, Carnarvon, Esperance, Forrest, Giles, Hall’s Ck, Learmonth.

    These and other rural stations are described on Niche Modeling at

    BTW, there is also a graph of Tennant Creek from a BoM CD shown below.

    Darwin has been shown before, e.g.

    In the Darwin spag graph, I have subtracted the findings of various adjusters from the BoM online data. Even the eralier BoM data do not agree with the online data.

    A little-shown aspect of Darwin seems to be a 7-year comparison of the old Post Office site versus the airport site. I am not certain of which actual sites were used. The airport site has been changed a number of times. It appears to put to rest the need for a major adjustment when the main station was shifted from the PO to the Airport. A couple of winter months are cooler in Tmin each year. But then, hot jetwash is less of a problem in the light traffic of predawn temp minima.

    It is quite hard to work out what has happened to NT temperatures. (I spent 9 years travelling to Darwin almost monthly, with 3 of them as President of the Northern Territory Chamber of Mines and Energy. Mining/energy was by far the biggest NT industry. My employer company built the town of Jabiru and the mine at Ranger, plus several mines at Tennant Creek. I first visited the NT in 1960 and I can say, tongue in cheek, that I have not noticed catastrophic climate change. Not even at Kakadu, which I visited over 50 times. Boring).

    • kenskingdom Says:

      Gday Geoff!
      It was you! As you know I’m nearly half way through checking the entire 100 site HQ network which I don’t think any one has done yet. So I will have a good look at your work this morning. I’ve compared raw and adjusted at all of the above sites except Forrest and Learmonth so will read with interest. One thing I have noticed is that some of the adjustments appear fairly skilfull, and others seem really heavy handed, as if two different people did them and never checked. I’d be interested in your thoughts!
      I have to go and do some work in the garden now- back to it later.

  11. Geoff Sherrington Says:


    There is a little value in numerical methods to indicate (but not prove much about) if a string of figures has been worked over or not. In the stations I mentioned above, it is possible to take a term like a a year of days Tmax and look at the number of times the least significant digit appears. 0, 1, 2, 3, etc should all have equal probability. At some of the above stations, there are some years that pass this elementary test and others that show a decided preference for some digits. It’s not as simple as conversion from F to C, it seems to happen sometimes when thermometers were replaced by MMTS or whaterver name you want to use for thermocouple or thermistor devices. This was mainly in the 1988-1996 period for the Reference Climate set (change from daily to half hour recording, in general) then later to recording at even shorter intervals, as you can see if you go looking for metadata sheets on the BoM website. At most Australian stations, the least significant Temp figure reported is usually 0.1 deg C, but at some stations for some periods the temps are reported in round degrees.

    • kenskingdom Says:

      That is another indicator of adjustment. I haven’t had time to look closely at “raw” min and max figures as I’m busy enough comparing the raw with the HQ, but I have noticed distinct preferences (.4 and .6 ?) in some of the HQ data. The plot shows where and when a site has had a major adjustment. I’ll be posting WA in a couple of days.

  12. Warwick Hughes Says:

    In my 1995 paper on the introduction of the Stevenson Screen to Australia,
    there is a photo from the Darwin library dated 1890 showing a SC at the Darwin PO. See the 4 page paper scanned here

    Chris Gillham from Perth has much info online re WA T records

    There is also a visitor to my blog that I only know as Ripper – who puts up pages on WA stations. There must be some central page for Rippers work.

  13. kenskingdom Says:

    Gday Warwick.
    Thanks for the links. All good work. Food for thought. I notice John Daly wrote about the poor observations/ screen at Darwin- but your study appears to show widespread use of Stevenson Screens, including Darwin. Was it kept in good order I wonder? Even so, BOM’s splice and adjustment has no link with reality.
    I have seen Chris Gillham’s excellent site but there doesn’t appear to be any way to comment or contact him. Watch later this pm for my WA post!

  14. adjusted temperatures kill tree … ABC reports | pindanpost Says:

    […] Or adjusted by climatologists,  Raw data: […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 35 other followers

%d bloggers like this: