The Australian Temperature Record- An Update

Ken Stewart, March 2011

Readers familiar with my analysis of the official climate record of Australia may be wondering what has happened since Part 8 posted in July last year.

Not much.

I have had ongoing correspondence (on my part) with BOM, and a reply on behalf of the Minister, Tony Burke MP.  I posted an Open Letter in reply (see below).

I have also performed a sort of internal monitoring check on my own results.  I have been begging BOM- or anyone- to check my analysis but to no avail.  I therefore have gone through every one of the 100 High Quality sites and checked my splices, calculations, and plots, for the third time.  (All of these are based on publicly available data from the BOM website.)

I have made some changes to some sites- notably in New South Wales- and reassessed some trend figures in Victoria and Cape Bruny in Tasmania.  I have gone back to my original assessment of Bowen in Queensland.  A point that needs emphasising here is that anyone going through the data could come up with a slightly different result each time- as the subjective, manual method used by BOM is not reproducible.  Another is that many of these sites should not be used at all because of missing or short data, urban influence, or lack of good overlap.

The changes?   While I have reduced NSW’s overall warming bias of raw vs BOM’s published figure from 60% to 45%, this has involved a reduction in the average adjustment of only 0.02 degrees C; Victoria’s warming bias of over 140% is unchanged.  The Australia-wide average amount of adjustment per site is +0.22 degrees C per 100 years- down from 0.23.  The median adjustment is 0.25C (a change of 0.025 down from 0.275C, or one place in the ranked list of stations.)  And the Australia wide graph is unchanged, with the difference in trends of 0.25C exactly the same.

Once again, the results speak for themselves: when the results of the homogeneity adjustments are compared with the raw data, there is a discrepancy of over 40%.

My next project is to plot the trends in maxima to 2010, just checking how hot 2010 was.    Just to whet your appetite, here’s BOM’s own plot for Tenterfield NSW- I can’t be accused of not doing my sums right or cherry picking.

No, one station does not prove anything, but watch this space.

6 Responses to “The Australian Temperature Record- An Update”

  1. Geoff Sherrington Says:

    Ken, Please do not think that you are alone and unappreciated. Several others are working away on similar projects. There are over 1,200 sites in Aust that could be compared between one set of workers and another, but the chanes are slim that we have all chosen the same set. I’m sure there is no reluctance to swap sites in common.

    It might help if you again clearly state the properties of what you describe as raw data, by saying for example that it does contain homogeneity adjustments according to Torok or whomever, that it does not contain adjacent site interpolations, UHI corrections, Stevenson screen presence/absence adjustments, Tobs adjustments and so on. In order to compare, we have to have apples with apples. And, to be sure, it would help to have a written statement from BOM confirming the presence/absence of adjustments in the raw.

    At the base of it all is the meta data. I can sympathise with the BOM because the original intent of data collection could not be foreseen in the global warming context and it’s understandable that they have fallen behind, if that’s what has happened. You can see that groups comparing data need to compare data of identical origin and description.

    A short note on weather for Melbourne people follows:
    At the BOM Regional Office in LaTrobe Street, the highest January maximum temperature on record was 31.0 degrees in 1908. The average maximum January temperature over 155 years, 1856 to 2010, is 26.2 degrees.

    The January 2011 average maximum temperature was 26.5 degrees, close to the 155 year record.

    The highest monthly February maximum on record was 30.2 deg C in 1898. The average maximum February temperature over 155 years, 1856-2010, is 26.1 degrees.

    The February 2011 average maximum temperature was 25.3 degrees, somewhat below the 155 year average.


    Figures from the Bureau of Meteorology site at and related.

    • kenskingdom Says:

      Hi Geoff, nice to see you! I’m well aware that I’m not alone or unappreciated.
      HQ data are the figures as downloaded from the HQ means for each site on the BOM HQ network. I use means because that’s what Della-Marta et al quote.
      The raw data I show are all calculations of Tmax +Tmin/2. The max and min are downloaded from BOM Climate Data for all open and closed stations within about 20km of the HQ site. Occasionally I go further but rarely use them. I have to rely on the integrity of this data knowing it may contain errors and adjustments already. For the vast majority of sites there is no one record long enough so 2 or 3 records have to be spliced to form a composite record. This can only be done if the records overlap by a few years! If not, we (and BOM) can only guess. I calculate the average difference between the records and adjust one (or two) to fit the other. You can see from my posts that the vast majority of my splices are virtually identical to BOM’s. The difference arises from adjustments.
      Thanks for your interest.

  2. Niche Modeling » BoMs Climate Hockey Stick Says:

    […] results of the Australian Temperature Record analyzed by kenskingdom posted as a series from July 2010 show an antipodean hockey […]

  3. Niche Modeling » Global Warming Trends – Gimme Some Truth Says:

    […] results of this audit corroborate the results of Ken Stewart’s audit of the Australian temperature […]

  4. Wally Webster Says:


    So what was 2009’s Black Saturday 47 C — a block of flats? Methinks there’s a bit of sloppy book-keeping here, and that what’s being talked about is in fact the highest AVERAGE January / February maximum temperature …

  5. Wally Webster Says:

    Sorry about that omission — I was singling out the following quote:

    At the BOM Regional Office in LaTrobe Street, the highest January maximum temperature on record was 31.0 degrees in 1908. … The highest monthly February maximum on record was 30.2 deg C in 1898.

Comments are closed.

%d bloggers like this: