Manipulated temperature records ?
WHEN raging floodwaters swept through Brisbane in January 2011 they
submerged a much-loved red Corvette sports car in the basement car park
of a unit in the riverside suburb of St Lucia.
On the scale of the billions of dollars worth of damage done to the nation’s third largest city in the man-made flood, the loss of a sports car may not seem like much.
But the loss has been the catalyst for an escalating row that raises questions about the competence and integrity of Australia’s premier weather agency, the Bureau of Meteorology, stretching well beyond the summer storms.
It goes to heart of the climate change debate — in particular, whether computer models are better than real data and whether temperature records are being manipulated in a bid to make each year hotter than the last.
With farmer parents, researcher Jennifer Marohasy says she has always had a fascination with rainfall and drought-flood cycles. So, in a show of solidarity with her husband and his sodden Corvette, Marohasy began researching the temperature records noted in historic logs that date back through the Federation drought of the late 19th century.
Specifically, she was keen to try forecasting Brisbane floods using historical data and the latest statistical modelling techniques.
Marohasy’s research has put her in dispute with BoM over a paper she published with John Abbot at Central Queensland University in the journal Atmospheric Research concerning the best data to use for rainfall forecasting. (She is a biologist and a sceptic of the thesis that human activity is bringing about global warming.) BoM challenged the findings of the Marohasy-Abbot paper, but the international journal rejected the BoM rebuttal, which had been prepared by some of the bureau’s top scientists.
This has led to an escalating dispute over the way in which Australia’s historical temperature records are “improved” through homogenisation, which is proving more difficult to resolve. If Marohasy is right, contrary to widely published claims, last year cannot be called the hottest year on record. BoM insists it is using world’s best practice to determine temperature trend and its methods are in accordance with those of its international peers.
But in furious correspondence with BoM, Marohasy argues the computer “homogenisation” of the records is being undertaken in a way that is at odds with its original intent.
“In (George Orwell’s) Nineteen Eighty-Four Winston Smith knows that, ‘He who controls the present controls the past’. Certainly the bureau appears intent on improving the historical temperature record by changing it,” Marohasy says.
There has been correspondence between Marohasy and BoM involving federal MP Dennis Jensen and the parliamentary secretary responsible for the bureau, Simon Birmingham.
After taking up the issue for Jensen, Birmingham says he is “confident that the bureau’s methods are robust’’. On its website, BoM says it is “improving Australia’s temperature record” by carefully analysing records “to find and address spurious artefacts in the data, thus developing a consistent — or homogeneous — record of daily temperatures over the last 100 years”.
BoM says historic high temperatures are unreliable, some having been collected by thermometers housed in a beer crate on an outback veranda.
In response to questions from Inquirer, BoM says “the premise that data homogenisation results in diminished physical veracity — in any particular climate data set — is unsupported in the bulk of the scientific literature’’.
But Marohasy is not convinced.
“Repetition is a propaganda technique,’’ she wrote back to Birmingham. “The deletion of information from records, and the use of exaggeration and half-truths, are others.
“The Bureau of Meteorology uses all these techniques, while wilfully ignoring evidence that contradicts its own propaganda.’’
Marohasy has analysed the physical temperature records from more than 30 stations included in the BoM set that determines the official national temperature record.
And she remains disturbed by a pattern whereby homogenisation exaggerates, or even produces, a record of steady warming against a steady or cooling trend in the raw data.
Marohasy says the clearly stated intent of homogenisation is to correct for changes in equipment, siting, and/or other factors that conceivably can introduce non-climatic factors into the temperature record.
“The bureau, however, is applying the algorithms subjectively and without supporting metadata, in such a way that the temperature record is completely altered, despite the absence of evidence that there were any changes in siting, equipment, or any other factor which could have conceivably introduced a non-climatic discontinuity,’’ she says.
Marohasy says the “corruption” of the data was of no practical consequence to climate scientists at BoM because they do not use historical data for forecasting either rainfall or temperature — they use simulation models that attempt to recreate the climate based on assumed physical processes.
But she says the remodelling is “of considerable political value to them, because the remodelled data better accords with the theory of anthropogenic global warming’’.
Marohasy presented a paper on her research to the Sydney Institute earlier this year. She has since expanded the number of physical temperature records analysed and says the results have only added weight to her concerns.
At Amberley, in Queensland, temperatures have been collected at the same well-maintained site within the perimeter of the air force base since 1941.
But through the homogenisation process BoM has changed what was a cooling trend in the minimum temperature of 1.0C per century into a warming trend of 2.5C per century.
“Homogenisation has not resulted in some small change to the data set, but rather a change in the temperature trend from one of cooling to dramatic warming,’’ Marohasy says.
This has been achieved by jumping-up the minimum temperatures twice through the homogenisation process: once in about 1980 and then around 1996 to achieve a combined temperature increase of more than 1.5C. NASA’s Goddard Institute for Space Studies, based in New York, also applies a jump-up to the Amberley series in 1980, and makes other changes, so that the annual average temperature for Amberley increases from 1941 to 2012 by about 2C.
In correspondence, Marohasy was told by NASA the Amberley data was adjusted to take account of historic temperature records at nearby stations.
But these 310 “nearby” stations stretched to a radius of 974km and include Frederick Reef in the Coral Sea, Quilpie post office in southwestern Queensland and even Bourke post office in NSW.
Considering the unhomogenised data for the nearest weather station, BoM’s jump-up for Amberley creates an increase for the official temperature trend of 0.75C per century. Temperatures at old Brisbane aero, the closest station that is also part of the national temperature network, also shows a long-term cooling trend.
“Perhaps the cooling at Amberley is real,’’ Marohasy says.
“Why not consider this, particularly in the absence of real physical evidence to the contrary?”
Another example is Rutherglen, a small town in a winegrowing region of northeast Victoria, where temperatures have been measured at a research station since November 1912.
There have been no documented site moves but an automatic weather station was installed on January 29, 1998.
Temperatures measured at the Rutherglen weather station also form part of the national temperature network (ACORN-Sat), so the information from this station is homogenised before inclusion in the official record that is used to calculate temperature trends for Victoria and also Australia.
Marohasy says the unhomogenised/raw mean annual minimum temperature trend for Rutherglen for the 100-year period from January 1913 through to December last year shows a slight cooling trend of 0.35C per 100 years.
After homogenisation there is a warming trend of 1.73C per 100 years. Marohasy says this warming trend essentially was achieved by progressively dropping down the temperatures from 1973 back through to 1913. For the year of 1913 the difference between the raw temperature and the ACORN-Sat temperature is 1.8C.
BoM is adamant the purpose of homogenisation is to remove non-climatic disconuities. But Marohasy says because there have been no site changes or equipment changes at Rutherglen, but very large adjustments made to the data, it is perhaps reasonable to assume that the bureau has changed the record for Rutherglen because it is very different to the record for the neighbouring stations.
According to a technical manual written by Blair Trewin from BoM, changes can be made where discontinuities are apparent when the trend at a site, for example Rutherglen, is compared with up to 40 neighbouring sites.
But Marohasy says analysis of nearby sites finds temperature trends show almost no warming during the past 100 years.
Marohasy says the changes to the minimum temperatures for Rutherglen are broadly consistent with many other changes made to temperature records for eastern Australia, which make the trends consistent with the theory of anthropogenic global warming.
But these changes are not consistent with the overriding principle of homogenisation, which is that changes should only be made to correct for non-climatic factors.
In the case of Rutherglen, she says, the changes do not even appear consistent with a principle in the bureau’s own technical manual, which is that changes should be consistent with trends at neighbouring weather stations.
At Burke, in western NSW, BoM deleted the first 40 years of data because temperatures before 1908 were apparently not recorded in a Stevenson screen, the agreed modern method.
Marohasy says this could have been easily accounted for with an accepted algorithm, which would not have changed the fact that it was obviously much hotter in the early 20th century than for any period since. Instead, the early record is deleted, and the post-1910 data homogenised.
“Rather than searching for a real physical explanation for the early 20th century cooling at Bourke since the hot years of the late 1800s, the Bureau has created a warming trend,’’ Marohasy says.
“This homogenisation, and the addition of data recorded after 1996 from the airport, means that the official record shows an overall warming trend of 0.01C per century and 2013 becomes about the hottest year on record for Bourke.’’
BOM says major adjustments at Bourke related to site moves as well as comparisons with neighbouring areas, while the Amberley and Rutherglen adjustments also were made after comparison with neighbouring stations.
And the bureau says an extensive study has found homogeneity adjustments have little impact on national trends and changes in temperature extremes. source
On the scale of the billions of dollars worth of damage done to the nation’s third largest city in the man-made flood, the loss of a sports car may not seem like much.
But the loss has been the catalyst for an escalating row that raises questions about the competence and integrity of Australia’s premier weather agency, the Bureau of Meteorology, stretching well beyond the summer storms.
It goes to heart of the climate change debate — in particular, whether computer models are better than real data and whether temperature records are being manipulated in a bid to make each year hotter than the last.
With farmer parents, researcher Jennifer Marohasy says she has always had a fascination with rainfall and drought-flood cycles. So, in a show of solidarity with her husband and his sodden Corvette, Marohasy began researching the temperature records noted in historic logs that date back through the Federation drought of the late 19th century.
Specifically, she was keen to try forecasting Brisbane floods using historical data and the latest statistical modelling techniques.
Marohasy’s research has put her in dispute with BoM over a paper she published with John Abbot at Central Queensland University in the journal Atmospheric Research concerning the best data to use for rainfall forecasting. (She is a biologist and a sceptic of the thesis that human activity is bringing about global warming.) BoM challenged the findings of the Marohasy-Abbot paper, but the international journal rejected the BoM rebuttal, which had been prepared by some of the bureau’s top scientists.
This has led to an escalating dispute over the way in which Australia’s historical temperature records are “improved” through homogenisation, which is proving more difficult to resolve. If Marohasy is right, contrary to widely published claims, last year cannot be called the hottest year on record. BoM insists it is using world’s best practice to determine temperature trend and its methods are in accordance with those of its international peers.
But in furious correspondence with BoM, Marohasy argues the computer “homogenisation” of the records is being undertaken in a way that is at odds with its original intent.
“In (George Orwell’s) Nineteen Eighty-Four Winston Smith knows that, ‘He who controls the present controls the past’. Certainly the bureau appears intent on improving the historical temperature record by changing it,” Marohasy says.
There has been correspondence between Marohasy and BoM involving federal MP Dennis Jensen and the parliamentary secretary responsible for the bureau, Simon Birmingham.
After taking up the issue for Jensen, Birmingham says he is “confident that the bureau’s methods are robust’’. On its website, BoM says it is “improving Australia’s temperature record” by carefully analysing records “to find and address spurious artefacts in the data, thus developing a consistent — or homogeneous — record of daily temperatures over the last 100 years”.
BoM says historic high temperatures are unreliable, some having been collected by thermometers housed in a beer crate on an outback veranda.
In response to questions from Inquirer, BoM says “the premise that data homogenisation results in diminished physical veracity — in any particular climate data set — is unsupported in the bulk of the scientific literature’’.
But Marohasy is not convinced.
“Repetition is a propaganda technique,’’ she wrote back to Birmingham. “The deletion of information from records, and the use of exaggeration and half-truths, are others.
“The Bureau of Meteorology uses all these techniques, while wilfully ignoring evidence that contradicts its own propaganda.’’
Marohasy has analysed the physical temperature records from more than 30 stations included in the BoM set that determines the official national temperature record.
And she remains disturbed by a pattern whereby homogenisation exaggerates, or even produces, a record of steady warming against a steady or cooling trend in the raw data.
Marohasy says the clearly stated intent of homogenisation is to correct for changes in equipment, siting, and/or other factors that conceivably can introduce non-climatic factors into the temperature record.
“The bureau, however, is applying the algorithms subjectively and without supporting metadata, in such a way that the temperature record is completely altered, despite the absence of evidence that there were any changes in siting, equipment, or any other factor which could have conceivably introduced a non-climatic discontinuity,’’ she says.
Marohasy says the “corruption” of the data was of no practical consequence to climate scientists at BoM because they do not use historical data for forecasting either rainfall or temperature — they use simulation models that attempt to recreate the climate based on assumed physical processes.
But she says the remodelling is “of considerable political value to them, because the remodelled data better accords with the theory of anthropogenic global warming’’.
Marohasy presented a paper on her research to the Sydney Institute earlier this year. She has since expanded the number of physical temperature records analysed and says the results have only added weight to her concerns.
At Amberley, in Queensland, temperatures have been collected at the same well-maintained site within the perimeter of the air force base since 1941.
But through the homogenisation process BoM has changed what was a cooling trend in the minimum temperature of 1.0C per century into a warming trend of 2.5C per century.
“Homogenisation has not resulted in some small change to the data set, but rather a change in the temperature trend from one of cooling to dramatic warming,’’ Marohasy says.
This has been achieved by jumping-up the minimum temperatures twice through the homogenisation process: once in about 1980 and then around 1996 to achieve a combined temperature increase of more than 1.5C. NASA’s Goddard Institute for Space Studies, based in New York, also applies a jump-up to the Amberley series in 1980, and makes other changes, so that the annual average temperature for Amberley increases from 1941 to 2012 by about 2C.
In correspondence, Marohasy was told by NASA the Amberley data was adjusted to take account of historic temperature records at nearby stations.
But these 310 “nearby” stations stretched to a radius of 974km and include Frederick Reef in the Coral Sea, Quilpie post office in southwestern Queensland and even Bourke post office in NSW.
Considering the unhomogenised data for the nearest weather station, BoM’s jump-up for Amberley creates an increase for the official temperature trend of 0.75C per century. Temperatures at old Brisbane aero, the closest station that is also part of the national temperature network, also shows a long-term cooling trend.
“Perhaps the cooling at Amberley is real,’’ Marohasy says.
“Why not consider this, particularly in the absence of real physical evidence to the contrary?”
Another example is Rutherglen, a small town in a winegrowing region of northeast Victoria, where temperatures have been measured at a research station since November 1912.
There have been no documented site moves but an automatic weather station was installed on January 29, 1998.
Temperatures measured at the Rutherglen weather station also form part of the national temperature network (ACORN-Sat), so the information from this station is homogenised before inclusion in the official record that is used to calculate temperature trends for Victoria and also Australia.
Marohasy says the unhomogenised/raw mean annual minimum temperature trend for Rutherglen for the 100-year period from January 1913 through to December last year shows a slight cooling trend of 0.35C per 100 years.
After homogenisation there is a warming trend of 1.73C per 100 years. Marohasy says this warming trend essentially was achieved by progressively dropping down the temperatures from 1973 back through to 1913. For the year of 1913 the difference between the raw temperature and the ACORN-Sat temperature is 1.8C.
BoM is adamant the purpose of homogenisation is to remove non-climatic disconuities. But Marohasy says because there have been no site changes or equipment changes at Rutherglen, but very large adjustments made to the data, it is perhaps reasonable to assume that the bureau has changed the record for Rutherglen because it is very different to the record for the neighbouring stations.
According to a technical manual written by Blair Trewin from BoM, changes can be made where discontinuities are apparent when the trend at a site, for example Rutherglen, is compared with up to 40 neighbouring sites.
But Marohasy says analysis of nearby sites finds temperature trends show almost no warming during the past 100 years.
Marohasy says the changes to the minimum temperatures for Rutherglen are broadly consistent with many other changes made to temperature records for eastern Australia, which make the trends consistent with the theory of anthropogenic global warming.
But these changes are not consistent with the overriding principle of homogenisation, which is that changes should only be made to correct for non-climatic factors.
In the case of Rutherglen, she says, the changes do not even appear consistent with a principle in the bureau’s own technical manual, which is that changes should be consistent with trends at neighbouring weather stations.
At Burke, in western NSW, BoM deleted the first 40 years of data because temperatures before 1908 were apparently not recorded in a Stevenson screen, the agreed modern method.
Marohasy says this could have been easily accounted for with an accepted algorithm, which would not have changed the fact that it was obviously much hotter in the early 20th century than for any period since. Instead, the early record is deleted, and the post-1910 data homogenised.
“Rather than searching for a real physical explanation for the early 20th century cooling at Bourke since the hot years of the late 1800s, the Bureau has created a warming trend,’’ Marohasy says.
“This homogenisation, and the addition of data recorded after 1996 from the airport, means that the official record shows an overall warming trend of 0.01C per century and 2013 becomes about the hottest year on record for Bourke.’’
BOM says major adjustments at Bourke related to site moves as well as comparisons with neighbouring areas, while the Amberley and Rutherglen adjustments also were made after comparison with neighbouring stations.
And the bureau says an extensive study has found homogeneity adjustments have little impact on national trends and changes in temperature extremes. source