Official reply to my queries at last!

Here is the complete text of  the reply to my letter of 26 October 2010 to Tony Burke MP, the Minister responsible for the Bureau of Meteorology amongst other things.  Dr Ayers also included a hard copy of the Jones et al. paper referenced below.

I also include below for the public record previous correspondence with BOM.

Readers may make up their own minds about the adequacy of this official letter, but I will leave my comments to a reply post shortly.

The reply:

Dr Greg Ayers

Director of  Meteorology

GPO Box 1289

Melbourne VIC 3001

Dear Mr Stewart

Thank you for your letter of  26 October 2010 to the Minister for Sustainability, Environment, Water, Population and Communities, the Hon Tony Burke MP, titled “Formal complaint and request for Information- Bureau of Meteorology”.  I have been asked to reply on the Minister’s behalf.

Monitoring Australia’s climate is a high priority for the Bureau of M eteorology and we have developed a reference network of stations that have little or no urban influence to their climate records.  The data from this network are available free of charge at http://www.bom.gov.au/climate/change/.  The data which underpin this network are being continuously improved and added to through the digitisation of historic paper-based records and the collection of new observations.  We also committed (sic) to improving our data  analysis methods and therefore continually review these (see later) and adjust them where appropriate.

When assembling these data, we take great care to ensure that the science and methodologies underpinning our work are documented in the peer review scientific literature and thus made available to the public.  The website noted above provides links to scientific papers that describe in detail the data, processes and analysis methods that have been employed.  I urge you to familiarise yourself with this material as it is quite rich and authoritative.

We welcome examination and testing of our climate data and the analysis methods we use.  The Bureau receives much correspondence detailing findings, theories, observations and other matters.  We are unable to respond to all of these requests in detail, as to do so would distract us from critical operational duties.  The appropriate forum for review of our work is the scientific peer review journals.  A basic tenet of science is that critiques of scientific arguments should be subject to the same level of expert scrutiny as the original arguments.  Bureau scientists are more than happy to respond in detail to critiques of their work when they are published in this forum.

Specific to the queries set out in your letter, I offer the following comments:

1.  What explanation is there for the large discrepancy between the 100-year trend of the homogenised data from the 100 high quality sites, and the trend of the raw data from these and nearby sites?

The Bureau of Meteorology has assessed this issue by a comparison of the high quality datasets developed by Torok and Nicholls (1996) with raw climate data from the Bureau of Meteorology climate database.  In a recent analysis we have spatially interpolated temperature data across the continent to assess the differences between raw and homogenised data trends on Australia’s average temperature.  Doing so removes bias inherent in having monitoring stations clustered in particular areas and sparse in others.  I refer you to Figure 12 in Jones et al. (2009) which compares spatially interpolated raw and homogenised mean temperatures aggregated over Australia for the period of the historical warming from 1950 to the present.  The series labelled as AWAP (shown in green) is the average temperature for Australia using all raw monthly temperature data.  The series labeled (sic) as Torok and Nicholls (shown in black) shows the average temperature for Australia based on the annually homogenised high quality dataset.  Comparing these two series demonstrates that the warming trends are effectively the same for the raw and homogenised data.

2. Why, if adjustments “tend to be equally positive and negative across the network”, have the great majority of sites had adjustments that have resulted in an increase (sic) warming trend?

Details on the homogenisation process are available in two references that I understand have been provided to you by Bureau staff previously:  Della-Marta et al. (2004) and Torok and Nicholls (1996).  In addition the PhD thesis by Torok (1996) contains graphs of every adjustment applied in the analysis.

As noted earlier, the Bureau continually reviews the data analysis methods used and I assure you that an evaluation of potential bias in warming trends, either positive or negative, is firmly within the compass of that work.  We will publish our findings on this matter later in the year and we will forward you a copy of the journal paper as soon as it is available.

3. What explanation is there for the inclusion in the High Quality series of 15 sites previously excluded because of urban influences?

It is not clear to which 15 sites you are referring.  The current criterion for inclusion of a station in the data set is set out in the publication by Della-Marta et al. (2004), with operational updates made in latter years.  The decision to classify a station as urban or non-urban is made on the basis of evidence for an urban warming signal and the immediate environment of the station.  It also depends on the period of observation.  For example, the Darwin composite series is classified as urban if the pre-1941 data are used, and non-urban if data are limited to the post 1941 period.

4.  Why, in supposedly “High Quality” records has so much low quality data been included?

All data included in the data series are deemed to be of high quality on the basis of length of record and careful examination of the station records and meteorological recordings.  As emphasised in the peer-reviewed scientific publications cited above, the data series included have been chosen because they provide the longest and highest quality records in a particular region.  Further detailed descriptions of the high quality stations are available in the PhD thesis by Torok (1996).

5.  Have there been any further adjustments made to the Australian temperature data apart from those documented at the BoM website?  If so, kindly supply details of all reviews and adjustments to the Australian raw temperature including, but not limited to, Torok and Nicholls and Della-Marta et al.,

The three papers described above provide details about how the high quality station data have been prepared, with the thesis by Torok providing detailed accounts of adjustments at stations.  An updated summary of operational adjustments is under preparation and will be available later this year.  I will arrange for a copy to be sent to you after this is published in the scientific peer reviewed literature.

6.  Can you please provide details of the quality control checks on the homogenisation adjustments?

The homogeneity adjustments are part of improving the quality of climate data for climate change analysis.  The adjustments are described in the peer-reviewed scientific literature (most notably in the three papers cited above) and all the high quality data is available via the Bureau’s website, as previously noted.  The homogenisation procedure involves careful metadata checks and a number of robust statistical tests.  Uncertainties in calculating the homogeneity adjustments are minimised, for example, by having a large number of comparison sites available, very good metadata and a year or more of parallel observations for the same station.

7.  Can you please provide details of the peer reviews of the two papers referenced by Dr Jones?

Scientific peer reviews are undertaken in confidence, with only the internal and external reviewers, the authors, and the journal editors having access to the reviews.  I can assure you that all publications from the Bureau of Meteorology are subject to careful review and quality control.  Publishing our work in respected scientific journals assures that the work has been independently reviewed by experts and found to be sound.

8.  Please provide complete details for additional sites.

I have asked our Climate Data Services Section to provide you with the specific data you have requested in your letter.  I understand that in the past this Section has made a range of data available to you and also pointed you to the wide variety of data which are available on the Bureau’s website.

I trust my considered reply to your letter will address your concerns.  Should you have any further concerns about the validity of our data analysis methods, I urge you to submit your arguments to scientific journals that mandate independent peer review.  This will ensure that we can engage in a reasonable and orderly debate over scientific facts and further improve our understanding of Australian climate trends.

Yours sincerely

(DR GREG AYERS)

DIRECTOR OF METEOROLOGY

10 February 2011

References

Della-Marta P., Collins C. And Braganza K. (2004), Updating Australia’s high quality annual temperature dataset, Australian Meteorological Magazine 53, 75-93.

Jones D.A., Wang W. And Fawcett R. (2009), High-quality spatial climate data-sets for Australia, Australian Meteorological Magazine 58, 233-248.

Torok, S. (1996), The development of a high quality historical temperature data base for Australia, PhD Thesis, School of Earth Sciences, University of Melbourne.

Torok S.J. and Nicholls N. (1996), A historical annual temperature dataset for Australia, Australian Meteorological Magazine 45, 251-260.

My letter to Tony Burke MP

26 October 2010

The Hon Tony Burke MP

Minister for Sustainability, Environment, Water, Population and Communities

PO Box 6022
House of Representatives
Parliament House
Canberra ACT 2600

Dear Sir,

Formal Complaint and Request for Information- Bureau of Meteorology

Given the importance of the consequences of a changing climate, the government and the public is justifiably interested in getting the most accurate picture of Australia’s temperature record.  For some months I have been comparing the Australian High Quality Temperature Sites’ adjusted data with raw data from those sites and nearby open and closed sites.  Data have been obtained from the Bureau of Meteorology’s excellent website.  The 100 High Quality (HQ) sites are used by BOM in their analyses of Australia’s climate.

I have published the results online at http://www.kenskingdom.wordpress.com .

My analysis has revealed apparent significant discrepancies between the HQ record and the raw data on which it is based.

This is problematic as Dr David Jones, Head of Climate Monitoring and Prediction, assured me in April that

“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).”

Despite numerous requests via email and letter, Dr Jones has not responded to the issues which I have since outlined to him, nor has he answered three simple questions I put to him.  The only response was an anonymous email from Webclim suggesting that I publish a scientific paper before I get any answer.  Previous correspondence is enclosed.

In view of Dr Jones (and BOM’s) apparent refusal to answer my queries, I wish to make a formal complaint, and ask you to investigate this unwillingness to comply with standards of openness explicit in the APS Values and Code of Conduct:

“Openness is at the core of Australia’s modern system of government. It is essential in a healthy democracy that members of the public have the opportunity to contribute to policy development and decision-making, and that there is public scrutiny and accountability of government. Public access to information in the possession of government agencies helps to make this possible.”

Further, I am asking for answers, through you, to the following questions:

1.  What explanation is there for the large discrepancy between the 100-year trend of the homogenised data from the 100 High Quality sites, and the trend of the raw data from these and nearby sites?

2. Why, if the adjustments “tend to be equally positive and negative across the network”,  have the great majority of sites had adjustments that have resulted in an increased warming trend?

3. What explanation is there for the inclusion in the High Quality series of 15 sites previously excluded because of urban influence?

4. Why, in a supposedly “High Quality” record, has so much low quality data been included- stations with short records (e.g. Woomera, Giles), large gaps infilled with estimates (e.g. Wilcannia, Cape Borda) or with data from sites many kilometres away (e.g. Newman), and records constructed by combining data from stations with no overlap at all (e.g. Port Hedland, Bourke, Cashmore)?

5.  Have there been any other adjustments made to the Australian temperature data apart from those documented at the BOM website?  If so, kindly supply details of all reviews and adjustments to the Australian raw temperature data including, but not limited to, Torok and Nichols, and Della-Marta et al.

6.  Can you please provide details (including dates, personnel, methods, results) of the quality control checks on the homogenisation adjustments.

7.  Can you please provide details of the peer reviews of the two papers referenced by Dr Jones, namely

Della-Marta, P., Collins, D., Braganza, K. “Updating Australia’s high-quality annual temperature dataset” Australian Meteorological Magazine Vol. 53, no. 2, June 2004

and

Torok, S.J. and Nicholls, N. 1996. A historical annual temperature dataset for Australia. Australian Meteorological Magazine, 45, 251-260.

8. Please provide complete details, including station metadata, of the reasons for the large adjustments to the temperature records of the following sites:

Omeo

Deniliquin Post office

Nhill

Wagga Wagga AMO

Kellerberrin

and of the following Urban sites (not used in climate analyses but adjusted):

Wangaratta Aero

Echuca Aerodrome

Benalla Shadford St

Dubbo Airport AWS.

I look forward to your reply.

Yours sincerely

Ken Stewart

Copy of email and hard copy sent to Dr Jones 08 October 2010

08 October 2010

Dr David Jones

Head of Climate Monitoring and Prediction
National Climate Centre
Bureau of Meteorology
GPO Box 1289, Melbourne
Victoria 3001

Dear Sir,

Attached copy of email sent to Webclim and yourself today.

Yours sincerely

Ken Stewart

Dear Dr Jones

I refer to my previous correspondence with you:

1. My emails dated 19/04/2010; 24/04/2010 and replies of 23/04/2010 and 25/04/2010.

2. My email dated 27/07/2010 (copy included below).

3. My letter dated 31/08/2010 (copy included below).

and 4. Email from Webclim (below) dated 09/09/2010.

As it is now at least one month since my letter (“3.”) should have been received, I am disappointed that neither you nor anyone else has been able to respond to the issues raised at https://kenskingdom.wordpress.com/ , and in my letter, or to the specific questions I asked.

I am disappointed that the only response I have had to my email of 27/07/2010 and my letter of 31/08/2010 has been an anonymous, possibly automated, email from Webclim.

I am disappointed that these issues and questions apparently will not receive serious consideration unless published in a scientific journal.  I do not intend to write a scientific paper, but as an Australian taxpayer and concerned citizen I have noticed some apparent discrepancies and I would like my specific questions answered by an identifiable human being.

I ask again:

1.  In view of your statement,

“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).” ,

can you please explain the apparent discrepancy between the raw and High Quality data?

2.  Can you please provide details (including dates, personnel, methods, results) of the quality control checks on the homogenisation adjustments?

3.  Have there been any other adjustments made to the Australian temperature data apart from those documented at the BOM website?  If so, kindly supply details of all reviews and adjustments to the Australian raw temperature data including, but not limited to, Torok and Nichols, and Della-Marta et al.

I look forward to your reply within 14 days.

A copy of this follows by surface mail.

Yours sincerely

Copy of my email of 27/07/2010:

To: David Jones

Sent: Tuesday, July 27, 2010 1:29 PM

Subject: High quality temperature record

Dear Dr Jones

You may remember answering my queries in April regarding the HQ record for Queensland.  Thank you for your responses.  Since then I have completed a comparison of the HQ data with the raw station data for all 100 non-urban sites.

The results are published at https://kenskingdom.wordpress.com/2010/07/27/the-australian-temperature-record-part-8-the-big-picture/

I would be interested in your response to the issues raised, especially in view of your claim that

“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). ”

I will be glad to publish your response in full.

Yours sincerely

Ken Stewart

(Copy of my letter of 31/08/2010)

Dr David Jones
Head of Climate Monitoring and Prediction
National Climate Centre
Bureau of Meteorology
GPO Box 1289, Melbourne
Victoria 3001

Dear Sir

Given the importance of the consequences of a changing climate, the government and the public is justifiably interested in getting the most accurate picture of Australia’s temperature record.

I have recently completed a comparison of the Australian High Quality Temperature Sites’ adjusted data with raw data from those sites and nearby open and closed sites.

I have published the results online at http://www.kenskingdom.wordpress.com

In an email to me on 25 April 2010 you stated:

“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).”

My comparison shows that:

• Of the 100 High Quality sites, 22 have been cooled, 22 have had essentially no change to their trends, and 56 have been warmed by the adjustment process.
• The average difference between raw and adjusted trends for the period 1910 – 2009 across the 100 non-urban sites is about 0.25 degree Celsius per 100 years (mean: 0.23; median: 0.275).
• The average temperature trend has been increased from + 0.6 degree C to + 0.85 C for the period 1910 – 2009.
• The published Time Series graph shows a warming trend for Australia of 0.1 C per decade (1.0 C per 100 years) which is 0.4 C per 100 years greater than the anomaly trend of the raw data.

• 15 of the 100 sites were originally listed as Urban by Torok and Nichols, but have been included in the record.
• Although a long period of continuous records is desirable for a High Quality site, and Torok and Nichols regarded 80 years as being desirable, many of the sites have much less than 80 years of data, and large gaps in the record.  These missing years have been infilled with estimated data or with data from sites from many kilometres away.
• Although Della-Marta et al consider that 5 years of overlap provides excellent data comparison, and 2 years can be useful if good quality, many records are constructed by combining data from stations with no overlap at all, or from different climatic areas many kilometres away.

In view of your above mentioned statement, can you please explain the apparent discrepancy between the raw and High Quality data?

Can you please provide details (including dates, personnel, methods, results) of the quality control checks on the homogenisation adjustments?

Have there been any other adjustments made to the Australian temperature data apart from those documented at the BOM website?  If so, kindly supply details of all reviews and adjustments to the Australian raw temperature data including, but not limited to, Torok and Nichols, and Della-Marta et al.

I look forward to your response to these questions and your comments on the points raised.

Yours sincerely

(Copy of email from Webclim)

—– Original Message —–

From: Web Climate requests

To:

Sent: Thursday, September 09, 2010 10:12 AM

Subject: Email response [SEC=UNCLASSIFIED]

Thank you for your analysis/hypothesis

As part of the scientific process you may wish to prepare and submit your work to a scientific journal. This will then mean that your work will be given serious scientific consideration.

There are a number of scientific journals worth considering. See for example: http://gcmd.nasa.gov/Resources/pointers/news_journ.html

It is difficult for the Bureau of Meteorology to assess work that has not been published. For example, the pure and applied (research) work carried out by scientists at the Bureau have to comply with the publication process before further assessment of the applicability of the work (analyses) is carried out. This conforms to standard scientific process and procedure.

Good luck with your work.

regards,

Webclim

National Climate Centre-Climate Information Services

Email:   webclim@bom.gov.au

Copy of emails with BOM- April 2010:

Ken,

the high-quality Australian data is all available at http://www.bom.gov.au/climate/change/hqsites/ . A description of how these data are assembled is provided at http://www.bom.gov.au/climate/change/hqsites/about-hq-site-data.shtml .

Regards,

David

I have read the 1996 paper by Torok and Nicholls.  Where can I find the details of any other adjustments?

The update to Torok and Nicholls is described in

Updating Australia’s high-quality annual temperature dataset /  

Della-Marta, P., Collins, D., Braganza, K. 2004

Australian meteorological magazine. Vol. 53, no. 2, June 2004. pp. 75-93

 

1. I note that Innisfail is listed on the HQ page as “Urban”.  When did Innisfail’s population exceed 10,000?

Urbanisation does not simply relate to population. It also can relate to the nature of the site’s surrounds.

1. The Torok and Nicholls paper lists Cairns AMO, Rockhampton AMO and Gladstone as Urban, but they are shown as Rural on the HQ page for Queensland.  Why is this?

Each of these is well outside of the town centre. We find no evidence in our analyses urban warming at these sites – Cairns and Rockhampton are at the airports and Gladstone is on a high hill outside of the urban influence..

1. Cooktown’s raw data for the 3 stations shows practically no warming at all over 100 years, yet the High Quality data (for Cooktown Mission Strip) shows a warming of roughly 0.9 degrees.  Yet Port Douglas shows only about 0.25 degree warming.  Can you please comment on this.

You are merging data from different stations. That is why your trends are inconsistent with the data.

1. Similarly, Sandy Cape Lighthouse shows about 0.45 degrees trend, but the HQ data shows a trend of approximately 1.2 degrees/100 years. Nearby sites- Maryborough shows approximately 0.3 degrees, Childers about 0.4, and Gympie 0.3 degrees.  Can you please comment on this.

I cannot see the data you are describing.

1. Similarly, Roma’s warming trend has been increased from 0.6 to about 1.5 degrees.  Miles shows about 0.4, Dalby about 0.55, Mitchell about 1.0, and Charleville’s HQ data shows about 0.6.  Can you please explain how the Roma data was developed.

Here is the raw and corrected data for Roma. The site has moved a number of times and the screen has changed.

Here is the list of changes to the site which are related to the corrections. The main change relates to a major site move which accounts for a large part of the inhomogeneity.

 

METADATA Summary

Roma,043091,043030
10/1897: Stevenson screen supplied.
02/1908: First correspondence.
12/1908: Screen moved due to building.
11/1912: New large screen.
10/1916: Poor observations during the 1910s.
11/1929: Screen moved 50 feet due to new shed and incinerator.
10/1941: Screen needs to be repaired.
10/1962: Site has deteriorated due to bitumen and buildings.
01/1971: Temporary site during building.
09/1972: Move to new PO.
09/1983: Site moved 50 m south to better site.
01/1992: Move to composite site.
04/1994: Bubbles in Min Thermometer
02/1997: AWS (Almos) Installed (28/02/1997)
02/1997: SITE MOVE- Short distance West (@8/02/1997)
1997-2002: Routine AWS upgrades and maintenance.

Please consider the environment before printing

 

Dr David Jones
Head of Climate Monitoring and Prediction
National Climate Centre
Bureau of Meteorology
GPO Box 1289, Melbourne
Victoria 3001, Australia

---

From: Cathy Toby On Behalf Of Web Climate requests
Sent: Monday, 19 April 2010 14:23
To: David Jones
Subject: FW: Questions re high quality data sites [SEC=UNCLASSIFIED]

Hi David

I’m not sure who in CAS is best able to answer this one – could you forward it on please?

thanks

Cathy

Webclim

Climate Information Services
National Climate Centre

GPO Box 1289 Melbourne Australia 3001

Phone +61 3 9669 4082,  Fax +61 3 9669 4515

Email: webclim@bom.gov.au

 

---

From:
Sent: Monday, 19 April 2010 13:50
To: Web Climate requests
Subject: Questions re high quality data sites

Dear Sir/Madam,

I have been looking at the Australian High Quality Climate Site Data and comparing them with annual means calculated from the data for these stations at the Climate Data Online page.

I have some questions regarding the Queensland sites.  Could you please assist with these:

1. I have read the 1996 paper by Torok and Nicholls.  Where can I find the details of any other adjustments?
2. I note that Innisfail is listed on the HQ page as “Urban”.  When did Innisfail’s population exceed 10,000?
3. The Torok and Nicholls paper lists Cairns AMO, Rockhampton AMO and Gladstone as Urban, but they are shown as Rural on the HQ page for Queensland.  Why is this?
4. Cooktown’s raw data for the 3 stations shows practically no warming at all over 100 years, yet the High Quality data (for Cooktown Mission Strip) shows a warming of roughly 0.9 degrees.  Yet Port Douglas shows only about 0.25 degree warming.  Can you please comment on this.
5. Similarly, Sandy Cape Lighthouse shows about 0.45 degrees trend, but the HQ data shows a trend of approximately 1.2 degrees/100 years. Nearby sites- Maryborough shows approximately 0.3 degrees, Childers about 0.4, and Gympie 0.3 degrees.  Can you please comment on this.
6. Similarly, Roma’s warming trend has been increased from 0.6 to about 1.5 degrees.  Miles shows about 0.4, Dalby about 0.55, Mitchell about 1.0, and Charleville’s HQ data shows about 0.6.  Can you please explain how the Roma data was developed.

Thank you for your assistance.

Yours sincerely

Ken Stewart

Final reply from Dr Jones: 25 April 2010

The regions are described at http://www.bom.gov.au/climate/change/about/temp_timeseries.shtml .

Most of the metadata only exists on paper files as it can date back more than a century. If you contact your local regional office they should be able to make access to the paper copies available to you – this is very time consuming.

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).

Regards,

David

________________________________________
From:
Sent: Saturday, 24 April 2010 8:02 AM
To: David Jones
Subject: Re: Questions re high quality data sites [SEC=UNCLASSIFIED]

Dear Dr Jones,

Thank you for your prompt reply.  I have read with interest the Della-Marta et al paper.

Thank you especially for the metadata re Roma.  Is metadata available electronically for all HQ stations, and if so, where?

Another question- what are the boundaries for the different regions in Australia e.g. North Australia, South East Australia, etc.?

Many thanks

Ken