Homogenisation – 3 examples

Updated 11 March

Here we look at some subtle changes in trends that result from homogenisation. It is often the case for long time spans that a station has either been moved  or else  two different stations in the same location have been combined.  In this case  a small offset in temperature values is to be expected.  The objective is where possible to create a continuous temperature series from 1910 to the present time. However in every case I have looked at the homogenisation procedure itself has extended way beyond the join and always increased the apparent  warming. Here are three examples.

  1. Launceston, Tasmania. The ACORN time series is actually  a combination of 3 nearby sites in the city. a) Launceston Pumping Station from 1910-1946. b) Launceston Airport (original site) from 1939-2009 and c) Launceston Airport (current site) from 2004 onwards. The animation below consists of 3 frames. Frame 1 is the raw average temperatures (monthly and annual) from the 3 sites is the middle trace in 3 colours – red, green and pink. Above is the maximum annual recorded temperature and the bottom trace is the minimum recorded temperature. Frame 3 is the same thing for the homogenised data – single red colour for monthly, blue for annual, purple for maximum and light blue for minimum. Frame 2 is a mix of the two.



The joining of the 3 bands at first sight looks to be fine, but closer inspection shows that the maximum and minimum temperatures in the central section are differentially being  shifted so as to produce a linear rising temperature trend where there was none apparent before. There are no obvious kinks in the raw data to justify this.

2. Alice Springs

A  similar animation is displayed below for Alice Springs


Alice Springs consists of a merge between the Post Office station (1910-1953) and the Airport since 1953. Note how the animation shows an increase in minimum temperatures at the airport resulting in a linearisation of the trend, neither of which has any direct connection with the merge.

3 Darwin

Darwin is tropical and shows little warming since 1910, but here again we see small adjustments extending up to 1980 from a merge in 1940.


Darwin is a merge of 3 different sites, the post office before 1942 and two airport sites. This should be a straightforward linear shift between the two stations but again the shape of the early data is completely changed producing a small linear warming trend.


There is evidence of  warming in the raw temperature measurement data. However, I strongly suspect this has then been boosted in ACORN-SAT by their  ‘homogenisation’ process. My guess is probably by about 33%. This must  also apply to GHCN, CRUTEM and BEST, since they all basically use the same algorithm.

Posted in AGW, Australia, Climate Change | 21 Comments

Raw data shows no warming in Australia

I have the raw measurement data for all 112 stations used by ACORN-SAT.  So I processed the daily minimum and maximum temperatures in exactly the same way as for the homogenised data. Here is the result.

The raw result shows no warming after ~1990. The apparent warming of ACORN-SAT by ~1C is therefore a result of the homogenisation process.  How can that be?

Several ACORN stations are not autonomous but instead are merges of nearby stations. For example Launceston Airport is a merging of the current station, a previous airport location and one in the centre of town. Choices of how to align such offsets between different locations are called homogenisation. Launceston is rejected from the raw Anomaly calculations simply because the current airport site has no coverage between 1961-1990.

Homogenisation consists in looking either manually or statistically for kinks or offsets in temperature data, and then correcting for  them. These can be caused by recorded site moves, instrumentation changes or often for unknown reasons. The corrections are supposed to be determined by near neighbours which in Australia may be 100s of km away. Automated changes are designed to make signals smoother, but this results in a tendency to force all stations to end up following the same trend. That is what homogenisation means. Often these adjustments are based on differences in derived monthly anomalies, which after homogenisation then get converted back into ‘measured’ temperatures. This all seems rather incestuous to me, so I took decided to take a deeper look.

This is Richmond, Queensland which shows no shift in raw temperature anomalies, yet it gets subtly modified by homogenisation thereby increasing a small warming trend. 

The second example is Wilcania which shows a flat anomaly distribution in the raw data but which after ‘homogenisation’ shows a positive warming rrend.

Here is a comparison of the resulting temperature anomalies. The red curve are the annual average anomalies

It is true that several stations indeed show shifts in raw anomalies, mainly due to station re-location, but these should be corrected with a simple offset. The same argument applies in the case that different stations are combined within the same town eg Launceston. However, I now have no doubt that automated ‘pairwise’ algorithms will always exaggerate warming. We see the same bias in GISS, in GHCN, in CRUTEM and especially in Berkeley Earth.

The question is by how much?

Posted in AGW, Australia, Climate Change | Tagged | 9 Comments

ACORN-SAT analysis

The Australian Climate Observations Reference Network (ACORN) Surface Air Temperature (SAT) is BOM’s ‘homogenised’ temperatures from 112 stations around Australia. These consist of daily maximum and minimum temperatures at each station from 1910 to 2017. I downloaded all the data and then calculated the daily, monthly, and annual averages and temperature anomalies for each station relative to a 1961-1990 baseline. I then made an area weighted average of all stations to produce annual anomalies based on a Lat, Lon triangulation. The result is shown below and compared to GHCN V3C.

Comparison of ACORN-SAT and GHCNV3C annual anomalies.

The result  agrees almost exactly with GHCNV3C for their Australian stations. I therefore suspect that the station ‘homogenisation’ calculations are the same. The real  surprise for me though was that the locations of the 112 stations are totally different to those from the hourly data described previously. Here are the ACORN stations.

There are over 400 Australian stations contained in CRUTEM4 and GHCN but the extra stations seem to make very little  difference.  I repeated the same calculation first for coastal stations only, and then for inland stations. Here is the comparison between anomalies averaged over Inland stations compared to those averaged over Marine Stations.

Inland anomalies compared to Marine anomalies. Essentially no difference

There is no difference. However, this result looks strange to me as I would have expected that stations on the coast would have followed more closely SST anomalies. Even the off-shore stations based on islands and peninsulars show the same warming trends as everywhere else.

Anomalies for 2 offshore stations.

I have to admit that I find this a little bit strange, and it implies to me that the homogenisation has probably gone a little bit too far. This is how the averaged Marine station anomalies compare to the ocean temperature anomalies (HADSST3) in the Southern Hemisphere.

The Marine stations appear to be warming about twice as fast as the surrounding ocean!

Diurnal Range

ACORN-SAT provides just the daily minimum and maximum temperatures to work with. The diurnal range is then simply Max-Min for each day. To reduce the daily variability I have taken a 30 day  average, and a 365 day  average. Missing data then leaves gaps in the traces. Here are some examples.

There is some evidence of a drop in Diurnal range, implying that minimum temperatures are rising slightly quicker than maximum temperatures. The UHI effect can also reduces diurnal range.  However it is not a consistent story and several stations show no discernible trend.


ACORN-SAT is the official BOM temperature series for climate change studies, and contains maximum and minimum daily temperatures from 112 stations. The data has been corrected for any apparent shifts in temperatures and ‘homogenised’ so that trends have been made consistent with near neighbours. The data show a net warming on land of about 1C since 1960. The trend is identical with that calculated from GHCNV3C which includes over 400 stations.

There is no evidence of any difference in temperature response between Marine stations and inland stations. This is surprising because it implies that the coast is warming twice as fast as the surrounding ocean. There is some evidence that the diurnal temperature range  is reducing although the effect is not as large as seen in the hourly data.

Posted in AGW, Australia, Climate Change | Tagged | 5 Comments