Temperatures

I HAVE already detailed my concerns with the claim by the Bureau of Meteorology that 2013 was the hottest year on record. Thank you for your many comments in the threads that followed the two blog posts. I am taking many of these comments into consideration in the development of a further set of questions.

In the meantime Bob Fernley-Jones, a retired engineer based in Melbourne, suggests that even if one accepts the truncating and adjusting that the Bureau has undertaken to arrive at the 2013 record annaul average temperature, given the actual temperature statistics published by the Bureau, David Jones was not justified in making the headline-grabbing statements that he did on January 3, in particular that there is a general warming trend and that it is Australia-wide.
Mr Fernley-Jones makes the following points, with the charts and supporting references available for download here:

1. One data point does not make a trend. There has been considerable variability over recent years in both the annual mean temperature anomaly and also the annual maximum temperature anomaly, Chart 1.

2. When regional variability is considered it is evident that only one state, South Australia, was signficiantly warmer in 2013, Chart 2.

3. The seasonal distribution of temperature is important, for instance warmer winters might arguably be a good thing for South Australia, while hotter summer are generally not. When the season mean temperatures for South Australia are considered for the period 1990 to 2013, the summer of 2013 was not particularly hot, Chart 3. In fact not a single season was hotter in 2013.

4. While much has been made of January 2013 being exceptionally hot, when the mean January temperature anomaly is plotted for South Australia back to 1910, Chart 4, it is apparent that 2013 was not an exceptionally hot January with hotter January’ occuring in the 1930s.

5. According the Bureau’s own time series data based on truncated and adjusted data series, mean January maximum temperatures have been flat since 1934 for the Northern Territory, and since 1947 for Queensland. New South Wales and the Australian Capital Territory have hotter January’s going back to 1938. South Australia, the hottest region for 2013, has eight hotter January’s going back to 1933. Tasmania has about 25 hotter January’s going back to 1917. Victoria has about 20 hotter January’s going back to 1939.

Mr Fernley-Jones’ critic can be read in more detail by downloading this document BobF-J_BlogVer2.

THE Australian Bureau of Meteorology, BOM, doesn’t have a set of temperature thermometers regularly positioned across the landmass of Australia from which it might derive an average annual temperature. Rather many more of it’s approximately 750 temperature recording stations are in south eastern Australia, some have records that date back to the 1850s, while others only started recording last year.
How do you derive an average annual temperature for Australia from such a mix of measurements?

The Bureau’s solution is to select a subset of about 112 stations from the 750. Some of the stations in the subset started recording temperatures in the 1850s, others not until the 1960s. Then the bureau truncates the longer temperature records, in some cases by discarding over 50 years of data, indeed everything before the somewhat arbitrary date of 1910. Data before this date is not considered reliable, but then the Bureau applies corrections to some of the rest of the data even though it should be reliable. Then the adjusted and truncated values from the subset are fitted to a grid to generate an area-weighted average of the data.

There is no single document that describes this methodology. Rather in a letter from the Bureau dated 24th January, responding to my request for information, I was directed to a mix of Bureau reports, peer-reviewed papers and also a PhD thesis by way of justification, methodology and for a list of stations.

In short, there is no straightforward way to verify the claim that last year, 2013, was the hottest calendar year on record. This is the claim the Bureau made in a media release on 3rd January 2014.

When individual stations with long temperature records are examined the late 1890 and 1930s appear to be as hot as recent years.

Ever since Climategate, deeply disturbing questions have been asked about the way climate science is conducted and also the state of the climate data. The Australian Bureau of Meteorology’s own databases feature in the leaked emails with the infamous ‘Harry Read Me’ file complaining about jumbled values and incorrect start dates for particular stations.

The mainstream climate science community has a vested interest in the average temperature for Australia increasing year-on-year because it has embraced the theory of anthropogenic global warming and invested heavily in research that assumes this theory.

It is time that the Bureau was more transparent in how it calculated its average annual values, and that it developed a method for benchmarking these annual average values. The benchmarking could be against satellite data and also against individual stations in Australia for which there are long temperature records.

**************

Click here to download the Bureau’s reply to my letter of 9th January requesting information to enable verification of the claim that 2013 was the hottest year on record.

Click on the image to see a chart of the Bureau’s annual mean temperature anomaly for Australia, this value is derived from the annual mean.

Dr David Jones
Manager of Climate Monitoring and Predictions
Australian Bureau of Meteorology

Dear Dr Jones

Re: Request Verification of 2013 Temperature Record

I am writing to request information be made publicly available to myself and others so we may have the opportunity to verify the claim made by you on behalf of the Australian Bureau of Meteorology that 2013 was the hottest year on record in Australia. In particular it is claimed that the average temperature was 1.20°C above the long-term average of 21.8°C, breaking the previous record set in 2005 by 0.17°C.

This claim is being extensively quoted, including in a report authored by Professor Will Steffen of the Climate Council, where he calls for the Australian government to commit to further deep reductions in greenhouse gas emissions because of this “record-breaking year”. Accurate climate records are not only of political interest, but are also of importance to those of us who rely on historical temperature data for research purposes. For example, the skill of the medium-term rainfall forecasts detailed in my recent peer-reviewed publications with John Abbot, have been influenced by the reliability of the historical temperature data that we inputted. From a very practical perspective, businesses will adjust their plans and operations based on climate data, and ordinary Australians worry and plan for the future based on anticipated climate trends.

Further, I note that you said in a radio interview on January 3, 2014, following your “hottest year on record” press release that, “We know every place across Australia is getting hotter, and very similarly almost every place on this planet. So, you know, we know it is getting hotter and we know it will continue to get hotter. It’s a reality, and something we will be living with for the rest of this century.”

The Australian Bureau of Meteorology is the custodian of an extensive data network and over a long period now, questions have been asked about the legitimacy of the methodology used to make adjustments to the raw data in the development of the Australian Climate Observations Reference Network – Surface Air Temperatures (ACORN-SAT). Furthermore, questions have been asked about why particular stations that are subject to bias through the Urban Heat Island (UHI) effect continue to be included in ACORN-SAT. In particular why is ‘Melbourne Regional Office’, a station at the corner of Victoria Parade and Latrobe Street (Melbourne CBD) still included in the ACORN-SAT network when this station is known to have become sheltered from previously cooling southerly winds following construction of office towers.

I understand ACORN-SAT was used to calculate the statistics indicating 2013 was the hottest year on record, but it is unclear specifically which stations from this network were used and how data may have been further adjusted in the development of the record breaking temperature anomaly.

Rockhampton-based blogger Ken Stewart, for example, has suggested that in the calculation of the annual average temperature for Australia, the eight sites acknowledged as having anomalous warming due to the UHI would not have been included. Is this the case? I had assumed that the Bureau used all 112 ACORN-SAT locations, and thus that the record hot temperature anomaly announced by you, actually includes a UHI bias.

Radio presenter Michael Smith has given some publicity to claims made by blogger Samuel Gordon-Stewart that the Bureau has overestimated the average Australian temperature by about 4 degrees. Mr Gordon-Stewart calculated average temperatures and temperature anomalies from data from all the weather stations listed by Weatherzone.

Furthermore, given many ACORN-SAT stations have continuous temperature records extending back to the mid-late 1800s and many stations were fitted with Stevenson screens by 1900, why does the Bureau only use data after 1909, all the while claiming that 2013 is the hottest year on record? Indeed it is well documented that the 1890s and early 1900s, years corresponding to the Federation drought, were exceptionally hot.

In summary, given the importance of the historical temperature record, and the claim that 2013 is the hottest year on record, could you please provide details concerning:
1. The specific stations used to calculate this statistic;
2. The specific databases and time intervals used for each of these stations;
3. The history of the use of Stevenson screens at each of these station;
4. How the yearly average temperature is defined; and
5. Clarify what if any interpolation, area weighting, and/or adjustments for UHI bias, may have been applied to the data in the calculation of the annual mean values.

Kind regards

Dr Jennifer Marohasy
Scientist, Blogger, Columnist
PO Box 692
NOOSA HEADS Q. 4567
Tel Mobile: 041 887 32 22

***
The above text was sent as an email to Dr David Jones earlier this evening. The photograph is of a sunset taken in northern Australia by Neil Hewitt of Cooper Creek Wilderness.