Letter to Bob Baldwin re Acorn Inquiry Terms of Reference

March 13, 2015

Dear Mr Baldwin

I am shocked, horrified, and angry after reading the Terms of Reference and Format of the Technical Advisory Forum.
The proposal that this august body can investigate the many flaws in ACORN-SAT in a one day sitting once a year is beyond ridicule, beyond parody.  It is sillier than the funniest episodes of “Yes, Minister”.
It is apparent that it has been set up not to investigate but to verify the Bureau’s claims.  It is a political solution to a perceived embarrassment.  It has been set up to appease the Bureau’s critics by seeming to examine their concerns without actually doing so, and without upsetting the Bureau or the ALP or Greens.  It will be a whitewash.
The Technical Advisory Forum needs real teeth.  It should be able to call witnesses, examine documents and analyse data.  Instead it will listen to platitudes from the Bureau in the morning and have a discussion in the afternoon.  What a joke!
Meanwhile, ACORN-SAT has been updated to include data to 31 December 2014.  It has not been improved.  It still has thousands of days of missing data, 954 days when minimum temperature exceeds maximum, still has gross and obvious errors anyone can recognise, still over-adjusts raw data so that ‘homogenised’ data is LESS comparable with neighbouring sites, and still results in Australia wide trends that are much greater than those shown by raw data.
I assume that you have acted on advice from your department.  You have been hoodwinked.  And unless you swiftly act to rewrite the Terms of Reference and Format for the Technical Advisory Forum to give it some real investigative power, this will be added to the list of bad decisions of a gutless government, more concerned about the opinions of the ABC, the press gallery, and the cross benches than its once loyal and hopeful supporters.

Yours in disgust
Ken Stewart

Limited Blogging for a While

March 6, 2015

Due to family reasons, I can’t give much attention to this blog for a while.

I’ll leave you with two graphs to contrast climate alarmism about heatwaves with reality.

Dr Sarah Perkins of the Climate Change Research Centre and Dr Karl Braganza were in the news this week claiming that

“we’re definitely seeing more heatwaves across Australia”

and

“the number of days that belong in a heatwave each season — has been increasing since the 1950s.”

and

“Brisbane has been getting heatwaves in spring”

Let’s look at Brisbane.  Dr Perkins, a heatwave expert,  defines a heatwave as “three days in a row of temperatures in the top 10 per cent”.  The top 10% of days are those with a temperature above 29.5 degrees Celsius.  Yep, 29.5 C is apparently an extremely hot day at Brisbane Airport!  This shows consecutive days above 29.5 C- the red line is the 3 day heatwave benchmark.  Note the linear trend.

spring consec 29.5 brisbane

This shows the highest temperatures reached in 3 day heatwaves:

spring temps 29.5 brisbane

Just two graphs refute all of the above claims about Brisbane:

definitely NOT more heatwaves

definitely NOT more hot days each season since the 1950s

and Brisbane has ALWAYS had heatwaves in spring, but not so many nowadays.

There has been a slight increase in the number of standalone days above 29.5 C (0.17 days in 66 years), but ‘heatwaves’ have been getting COOLER.

Enough said.

How Hot is it in Boulia? (Hotter Than News Limited Can Remember)

March 3, 2015

Yesterday I posted about the ABC’s short term memory about heat at Longreach.  Today it’s Boulia’s turn, this time due to a News Limited article, “Relentless, endless summer. The Queensland town where it’s been 40 degrees pretty much forever”. 

Time for a reality check.

Using the ACORN dataset and current data for Boulia up to 2 March, here is a 180 day running count of days above 40 degrees in Boulia.  (This captures days in one summer, but not two).

Fig. 1: 180 day running count of days above 40 C at Boulia.

boulia 40

Even with 43 missing days of data in December and January, this summer is not unusual.

The next graph shows the length of 40+ heatwaves.

Fig. 2: Consecutive days above 40 C at Boulia.

boulia 40 consec

This is definitely NOT the longest ever heatwave.  This one is very mild- 31 days in 1973 is going to be hard to beat.  Note that there is NO TREND in the length of heatwaves.

Finally, this graph shows the number of days above 45 degrees, which is hot in anyone’s language.  The missing data may have had some 45+ days, but there were many days with rain in this period.

Fig. 3: 180 day running count of days above 45 C at Boulia.

 boulia 45

It’s hot in Boulia, but not unusually so.

Tomorrow I will look at the latest claims about Australia getting hotter and cities getting more heatwaves.

How Hot Is It at Longreach? (Hotter than the ABC can remember)

March 2, 2015

Today on ABC Radio’s Country Hour, there was an interview by Lydia Burton with a grazier from the Longreach district in Central Western Queensland, which has been having some pretty hot weather this summer, as has most of Western Queensland.

Peter Whip is from Royston near Longreach and says over the last 50 years he has noticed an increase in the amount of days over 35 degrees and that is a real concern for all landholders.

“In the last couple of years we have had some really long hot dry spells when the feed quality has been terrible and we have seen cattle conditions slip really quickly.

“That is certainly something that is changing over time and….it is something that in 20 years time it will be a significant impact.”

The gist of the segment was that cattle were struggling in the heat, with so many days above 35 degrees Celsius, and especially with days over 40 degrees, let alone 45 degrees.

As I have said before, I have close links to agriculture.  My parents, grandparents, and great-grandparents were farmers; one brother still farms, another is retired, and my brother-in-law is a farmer also.  As well, several nephews and nieces are on the land or in agri-business.  I have great sympathy for those on the land at all times, especially in drought conditions.

However, is there any evidence that Longreach is having an unusually hot spell, or that Longreach is getting hotter, or that it will get hotter in the future?  And how could an intrepid rural reporter check the facts against one grazier’s anecdote?

The very simplest check would be on Longreach’s mean maximum temperatures.  From November to February, the mean monthly maximum is never below 35 C, so we might expect many days to be well above 35 C.  Longreach Aero’s highest ever maximum was back on 28 January, 1990, at 47.3 C.

Let’s have a look at the number of days above 35 C, 40 C, and 45 C.  I’ll do this for 180 day periods, which will capture days in a single summer, but not two.  I use the official ACORN dataset, plus daily data right up to yesterday, 1st March, straight from the Bureau of Meteorology, so there can be no accusation of cherry picking or of using “unreliable” raw data.

Fig. 1:  Number of days in summer above 35 C at Longreach.

longreach 35

And for the last 50 years….

Fig. 2:  Number of days in summer above 35 C at Longreach since the summer of 1964 -65.

longreach 35 since 6465

What about the length of heatwaves?  The next plot shows the number of days in a row where the temperature has exceeded 35 C.

Fig. 3:  Number of consecutive days above 35 C at Longreach.

longreach 35 consec

Mr Whip may be a little mistaken.  Since 1965, and since 1910, there appears to be no trend in the number of days above 35 C in Longreach, and there is no trend in the length of heatwaves.

But the ABC is always talking about extremes- what about days over 40 C?

Fig. 4:  Number of days in summer above 40 C at Longreach.

longreach 40

Oops!

Fig. 5:  Number of days in summer above 45 C at Longreach.

longreach 45

Oops again.

Now Lydia Burton is a good rural reporter, I frequently listen to her on the Country Hour and the Rural Report, but if only she had checked the facts, she could have saved her listeners a lot of angst.  Instead, “our” ABC continues to peddle nonsense about dangerous climate change.

It’s no wonder that people have lost faith in the ABC.

An “Eye Witness” Account of TC Marcia

February 28, 2015

A week after the event, now that our power has been restored, internet seems back to normal, and our cleanup is mostly complete, I can turn to an analysis of Tropical Cyclone Marcia.  Much initial analysis has been done, especially by Jen Marohasy at http://jennifermarohasy.com/jenns-blog/  who was in contact with me soon after the cyclone, and also by the Cyclone Testing Station of James Cook University.

However, perhaps I can add to the discussion with some additional suggestions and some personal observations.  I was there.  The eye passed over our house for about 25 to 30 minutes.  I have felt the impact in Rockhampton, experienced the electricity supply loss for 6 days and 6 hours, and cleaned up tree damage to our own 1.5 acre block, the local Heritage Village and a nursing home where a family member is resident.  Close family members were also without power.  I can also relate the experience to previous cyclones when I lived in the Mackay area.

Tropical Cyclone Marcia was a very small, intense, and short lived Complex Cyclone, with maximum winds about 30 – 40 km away from the centre, and well outside the eye wall.  At all times the Bureau of Meteorology was reporting its strength as at least one Category (more likely two) above what ground observations showed.  There is no surface evidence to show it was ever Category 5.  The only surface observation available indicates Category 3.

I watched on various websites the approach of TC Marcia throughout the day on Thursday 19th, and every couple of hours that night (few people got much sleep), and for as long as I had landline and then mobile internet connection (and power in my laptop) on Friday 20th.

Marcia rapidly intensified during Thursday.  By 5.30pm it was near Creal Reef and about 175 km  northeast of Mackay with a clearly visible eye on weather radar.

marci 17301902

The Tropical Cyclone Technical Bulletin reported it as having central pressure 958 hPa, 10 minute winds of 155 kmh and gusts to 220 kmh.  By 10.22pm it had central pressure 940hPa, winds 195 kmh with gusts to 270 kmh, and had slowed to 7 kmh but heading southwest, straight for Mackay.

By 3.20 am on Friday it was just to the east of the Percy Islands, where Middle Percy Island recorded winds from the south at 145 kmh gusting to 180 kmh.  The maximum winds here were reached between 4.00 and 4.30 am when winds reached 156 kmh gusting to 208 kmh.  Central pressure was 971.6 hPa.  It is interesting to note that the apparent temperature was 1.3 degrees C- it is very cold in a cyclone if exposed to the wind.

marci 03002002

Strongest winds and heaviest rainfall are usually in the eyewall.  Notice that the heaviest rain in the above image (yellow) is to the west of the eye- in the vicinity of Middle Percy Island.  This is what made me strongly doubt the forecast strength.  The Technical Bulletin reported it as having central pressure 929 hPa, winds 205 kmh gusting to 285 kmh, moving south at 13 kmh- straight towards Rockhampton.

The next image is from 6.00 am, about two hours before landfall, travelling south at 30 kmh.

 marci 06000220

Again note the heaviest rain outside and to the west of the eye wall.  The eye is about 25 km across.

Here is an interesting model of winds around Marcia when close to crossing the coast at 7.00 am on http://earth.nullschool.net/#current/wind/isobaric/1000hPa/orthographic=143.66,-20.38,1821.

 winds 07002002

Strongest winds are shaded red.  Interestingly, they are to the northwest and northeast of the eye.  (Note also the deep trough from the Sunshine Coast to the border.)

After this I was busily checking final preparations, having breakfast, and making and receiving phone calls, as well as watching the websites.  By 10.00 am the wind was becoming much stronger, and the Bureau was still saying it was Cat 5 and heading straight for Rocky.

marcia forecast 10000220

Here is the radar image:

marci 10100220

Note the yellow area of heavy rain (and presumably strongest winds) south of Samuel Hill and close to Byfield.

The wind strengthened and was at its strongest by about 12.30 pm, consistently from the east.  There was one final prolonged gust, the strongest, at about 12.35 and then from about 12.50 the wind died away to nothing by 12.55 pm.  Like many others I went outside and had a look around, but didn’t go far.  It was raining lightly, heavily overcast, with puffs of wind from different directions.  At 1.25 pm we could hear the wind coming, and it picked up again over about five minutes, blowing directly from the west, but didn’t reach the same strength as before.  At a speed of 22 kmh, the eye was about 11 km across at this stage.  This screenshot shows the eye right over us at about 1.10 pm.

marci 12500220

The eye is just visible, and strongest rain and winds are to the south and east.

But this one shows it just 10 minutes later:

 marci 13000220 eye collapse

Note the eye has disappeared- it collapsed while over us.

Rockhampton Airport is about 12km south-southwest, so 30 minutes later the eye should have reached there.  But it didn’t.  Here is a screenshot of observations at Rocky Aero from 12.41 to 4.30 pm.

rocky eye

I have circled key information.  At 12.50 pm the wind was ESE at 72 kmh, with a gust to 113 kmh.  Between 1.13 and 1.52 (arrowed) the wind swung from SE to WSW, but never dropped below 24 kmh.  Lowest pressure was 975.7 hPa.  The wind steadily picked up and was strongest around 2.45 to 3.00 pm, reaching 82 kmh with a gust to 113 kmh.  This was different from my experience, where the wind was consistently from the east, then very light and variable, then directly from the west, but not as a strong.  The eye had collapsed before the centre was near the airport, and the remnants passed a couple of kilometres to the east.  The cyclone was rapidly degrading.

My last screenshot before the internet died and my battery went flat, shows the BOM forecast at 2.12pm:

marci forecast 14120220

It shows Marcia still Cat 3 at 2.00 pm.  But Rocky Aero data show winds were no more than Category 1 at this time.  At our place, the winds at this time were from the west and still damaging trees but not nearly as strong as before the eye passed.  By 3.00pm the winds were light and we were out in the street talking with the neighbours.

On eastern slopes or hills, and closer to the coast, winds would have been much higher.  I estimate gusts at our place, funnelled up a gully on an east facing slope, were around 120 – 130 kmh- similar to our experience of TC Ului at Sarina some years ago.

The Cyclone Testing Station of James Cook University has released a preliminary report at https://www.dropbox.com/s/9y8hdb74ycgemfi/TC%20Marcia%20Rapid%20Assessment%20Report%2020_02_2015.pdf?dl=0 .  They will produce more definitive analysis in a few weeks.

To summarise, this report states Yeppoon experienced Category 2 winds, and Rockhampton Category 1.  The report goes on to state:

“A community that receives an over-represented wind speed report may have potential for complacency in preparation or building standards in the future. Further research is required to determine if this is an issue.”

But was TC Marcia a Complex Cyclone, one in which maximum winds are concentrated away from the eye?  All radar images show heaviest rain, and therefore probably winds, away from the eye wall.

Byfield community is 30 km northwest of Yeppoon, half way between Yeppoon and Samuel Hill.  There are reports of every tree stripped for kilometres, which appears to indicate strong Cat 3 to Cat 4 winds.  Radar image at 10.10 am shows heaviest rain just to the west of Byfield and just to the south of Samuel Hill.  Yet Samuel Hill did not experience winds of this strength.

And here’s a photo of part of our block taken at 3.30 on the afternoon Marcia passed.

OLYMPUS DIGITAL CAMERA

Note every tree has at least 50% of its leaves stripped and several branches off, and several were completely uprooted, and these are tough old ironbarks.  So allowing for 30 km closer to the sea and two hours earlier, I wouldn’t be surprised if winds were 150 – 160 kmh at Byfield- but not 200 kmh.

Until JCU can do their complete assessment, we will not know exactly what happened.  No doubt their engineers will also visit Stanage Bay, where Marcia crossed the coast.  Tropical Cyclone Marcia was a very small, intense, and short lived Complex Cyclone, with maximum winds about 30 – 40 km away from the centre, and well outside the eye wall.  At no time was it Category 5.  Despite its relative weakness compared to some other cyclones, Marcia did tremendous damage in Central Queensland.

The main thing is no one was hurt.

My Submission to the BOM Review Panel

February 9, 2015

The Hon. Bob Baldwin MHR

Parliamentary Secretary to the Minister for the Environment

PO Box 6022

House of Representatives

Parliament House

CANBERRA ACT 2600

Dear Mr Baldwin

Re: Recommendations for the Review Panel appointed to review official national temperature records

As a “citizen scientist” who has been researching Australia’s climate, and the ACORN-SAT record in particular, over the past several years, I am concerned about errors in the work done by the Australian Bureau of Meteorology, in particular warming bias introduced through homogenisation and the apparent general lack of quality control.   I draw your panel’s attention to these issues listed under four categories in the following submission.  More information, including supporting charts and tables, is provided in numbered attachments.

 1Adjustment Issues

 1.1 Homogenisation distorts temperature records causing warming bias at most locations, indicating the methods as stated in the CAWCR Technical Reports (see CTR-049) are not followed or do not work as designed. Homogenisation should lead to candidate sites having trends in temperature anomalies that are more like their neighbours’. However in many cases this does not occur, and homogenisation has resulted in wide disparities.  This is obvious from a simple visual inspection of a plot of ACORN data at candidate sites versus raw data of the listed neighbours (Attachment 1.1).

 1.2 A better but still simple method of comparison involves differencing. Differencing (anomaly data of candidate site minus data of reference sites) should show improved results following homogenising, with differences closer to zero.  Importantly, even if the differences fluctuate, there should be zero trend in differences. Yet at a number of sites, homogenising has produced worse results. (Attachments 1.2a, 1.2b, 1.2c).

 1.3 The most extreme examples of homogenising arising from warming adjustments, result in trends of candidate sites greater than the homogenised trends at neighbouring ACORN sites. (Attachment 1.3a). The converse applies where extreme cooling adjustments result in trends less than the homogenised trends of ACORN neighbours (Attachment 1.3b). This indicates over correction, resulting in the creation of artificial trends warmer or cooler than the neighbours’.

1.4  Data are homogenised by reference to up to 10 best correlated neighbours. Some of these neighbours may be hundreds of kilometres away, and with completely different climates.  Deleting the two most distant neighbours greatly improves data comparison between Mackay and its remaining neighbours. (Attachment 1.4a, 1.4b).

1.5 The Bureau has belatedly tried to explain adjustments with the release of the 28 page PDF file of all adjustments ( http://www.bom.gov.au/climate/change/acorn-sat/documents/ACORN-SAT-Station-adjustment-summary.pdf) , and has also provided Summaries of Adjustments for six sites as a further explanation. ( http://www.bom.gov.au/climate/change/acorn-sat/#tabs=Adjustments ). Incredibly these Summaries don’t agree with the adjustments in the 28 page document. (Attachment 1.5).

1.6 The Bureau claims that sites exhibiting Urban Heat Island (UHI) effect are “excluded from downstream products such as the calculation of national and regional temperature anomalies for the analysis of large scale climate change” (CTR-049, pp.71-73.) These sites include Townsville, Rockhampton, Laverton RAAF, Richmond NSW, Sydney, Melbourne, Adelaide, and Hobart.  Unfortunately they certainly are used as comparison sites when making adjustments.  Several Queensland sites, including Cairns, Charters Towers, Mackay, and Bundaberg, have Townsville and/or Rockhampton listed as neighbours used for making adjustments.  If a site’s temperatures are suspect due to UHI to the extent that they cannot be used for regional or national anomalies, it seems illogical that they can be suitable for comparison with neighbours.  This apparent contradiction needs explanation.

 2. Impact on trends

2.1 The Bureau of Meteorology has reportedly claimed “an extensive study has found homogeneity adjustments have little impact on national trends and changes in temperature extremes.”  (Weekend Australian, August 23-24, 2014). In support of this, the Bureau displays a plot on the adjustments tab at the ACORN-SAT web page purporting to show “temperature trends since 1910 from the unadjusted temperatures from more than 700 locations (AWAP), together with those that have been carefully curated, quality controlled and corrected for artificially induced biases at 112 locations (ACORN-SAT)” (http://www.bom.gov.au/climate/change/acorn-sat/#tabs=Adjustments). However, the AWAP (Australian Water Availability Project) network is not “unadjusted”, but according to CTR-050 p.41, ‘the generation of stable climatologies implicit in the AWAP … ( analysis) … goes part of the way towards removing the temporal inhomogeneities implicit in the raw data without the explicit application of temporal-inhomogeneity adjustments. … Hence  it is reasonable to describe the AWAP … (analysis) as “partially homogenised” rather than unhomogenised.’  It is therefore misleading to describe the above-mentioned plot of ACORN vs AWAP as being a comparison with “unadjusted temperatures”.

2.2 Moreover, the Bureau has made no attempt to compare ACORN data with minimally adjusted raw data (that is, adjusted only to combine two incomplete records into one through examination of overlapping data.) My comparison of annual ACORN data (1910-2012) with raw records (corrected only for overlap) at 83 sites for minima and 84 for maxima shows the increase in trend of ACORN over raw is 66% and 13% respectively (Attachments 2.2a, 2.2b).  (The remaining sites had no suitable overlap between discontinued and new stations).  Nearly two thirds of the sites analysed had trends increased (warmed).

2.3 A few very remote sites especially in Northern and Central Australia have an enormous impact on the ACORN record. This is demonstrated by differencing area averaged means (the official national annual means) and straight averaged means of the 104 ACORN sites (Attachment 2.3).  As 7 – 10% of the national climate signal is due to Alice Springs alone (International Review Panel Report, September 2011 p. 12 http://www.bom.gov.au/climate/change/acorn-sat/documents/ACORN-SAT_IPR_Panel_Report_WEB.pdf), both the influence of remote sites, and the area averaging algorithm, need to be investigated.

2.4 The South East portion of Australia (the area south and east of the median of ACORN stations’ latitudes and longitudes) has the greatest number of sites, and also the greatest change in trend in minima from raw to ACORN- 232% (Attachment 2.4a). The trend increase for New South Wales is 245% (Attachment 2.4b) and for Victoria is 350% (Attachment 2.4c).  Homogenisation in the most heavily populated areas of the country cannot be described as having “little impact”.

2.5 I have also compared network wide ACORN data with AWAP data (1911 – 2013), for annual, seasonal, and monthly analyses (the last two the Bureau has not yet completed.) The results (Attachments 2.5a, 2.5b, 2.5c, 2.5d, 2.5e) are staggering and need urgent investigation.  In particular, the 200% increase in trends for Summer Maxima is relevant to claims of increasing summer heat, especially in light of the recent Climate Institute report.

3. The effect of rounding on trends, and uncertainties in neighbouring stations’ data used for homogenising

3.1 The Bureau admits that rounding of temperatures to whole degrees in the Fahrenheit era may have led to an artificial breakpoint of +0.1C in 1972, but claims this is lost in the noisy signals of the 1970s (CTR-049 p.70). This, however, is disputable and needs thorough investigation, as an audit by several colleagues and myself of daily minima and maxima data in the Fahrenheit and Celsius eras (474 data records and 8,580,583 daily observations) found evidence that indicates the impact on trends could be between +0.1C and +0.4C. (Attachment 3.1).

3.2 Our study also found homogenising is based on records with large amounts of uncertainty (Attachment 3.2). Significantly, the first of the 2011 International Review Panel’s Recommendations (A1) was “Reduce the formal inspection tolerance on ACORN-SAT temperature sensors significantly below the present ±0.5 °C …“ (See  http://www.bom.gov.au/climate/change/acorn-sat/documents/ACORN-SAT_IPR_Panel_Report_WEB.pdf8).  It is inexcusable that no error bounds are given for ACORN data.

4. Lack of Quality Assurance

The apparent lack of quality assurance means ACORN-SAT is not fit for the purpose of serious climate analysis including the calculation of annual temperature trends, identifying hottest or coldest days on record, analysing the intensity, duration, and frequency of heatwaves, matching rainfall with temperature, calculating monthly means or medians, and calculating diurnal temperature range.

4.1 ACORN-SAT daily data (and consequently, monthly and annual means) provide many obstacles to rigorous analysis. Days of data are missing, slabs of data are offset by one day (daily data being assigned to the wrong date, usually one day early), and many adjustments show obvious glaring errors (Attachment 4.1).

4.2 Another glaring error gives Australia a new hottest day on record (Attachment 4.2).

4.3 Other researchers have reported at least 917 days where minimum temperature exceeds maximum (Attachment 4.3). Although a specific check for errors in recording maxima and minima was conducted before homogenising, this check could not have been done with the homogenised data. It might be claimed that this feature is normal and due to a cold change arriving after 9.00 a.m.  This would be especially evident in winter at high altitudes such as Cabramurra, with 212 occurrences.  However, there are no instances of maximum less than minimum in the raw data for Cabramurra.  All instances occur in the adjusted data before February 1999.  Further, despite the Bureau being aware of the problem since at least I July 2013 when Blair Trewin, lead author of ACORN, assured readers of the blog Open Mind at https://tamino.wordpress.com/2013/06/29/a-clue-for-willis/#more-6693 that “in the next version of the data set (later this year), in cases where the adjusted max < adjusted min, we’ll set both the max and min equal to the mean of the two” (which merely hides the fault caused by adjustments), the problem still exists- 212 occurrences are still in the ACORN record for Cabramurra.

In conclusion, ACORN-SAT is not reliable and should be scrapped.  ACORN-SAT shows adjustments that distort the temperature record and do not follow the stated procedures in the Bureau’s own Technical Papers, generating warming biases at a large number of sites, thus greatly increasing the network wide trends.  Furthermore, the Bureau does not take account of uncertainty, and the data are generally riddled with errors indicating poor quality assurance.  Finally, its authors have not followed up on most undertakings made more than three years ago to permit replication and improve transparency (http://www.bom.gov.au/climate/change/acorn-sat/documents/ACORN-SAT_Bureau_Response_WEB.pdf ).

I am delighted with the formation of the Review Panel.  I hope that this review will bring about much needed improvements at the Bureau of Meteorology in the way the Bureau collates, audits, analyses and reports on national temperature data.

Yours sincerely

Ken Stewart

Attachments

 Attachment 1.1:  One example of many – comparison of Acorn anomalies (black) with neighbours at Carnarvon

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/07/24/carnarvon-a-closer-look/ )

Attachment 1.2a:  Differencing: Rutherglen minus neighbours

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/09/08/homogenisation-a-test-for-validity/ ) 

Attachment 1.2b:  One example of many- Differences after homogenisation show worse results

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/09/08/homogenisation-a-test-for-validity/ )

Attachment 1.2c:  As for 1.2b, showing the mean of candidate data minus neighbours. 

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/09/08/homogenisation-a-test-for-validity/ )

 

There are many other sites with greater differences after homogenisation in minima or maxima.  I have also checked Deniliquin, Bourke, Amberley, Carnarvon, Williamtown, Mackay, Kalgoorlie-Boulder and Wilcannia, which all show this problem.  Some sites may show improved differences.  An audit of all sites is essential.

Attachment 1.3a: One example of many of warming adjustments over correcting-  Amberley Acorn vs nearest Acorn neighbours’ (mean of nearest Acorn neighbours’ homogenised data).

(For further information and full explanation see  https://kenskingdom.wordpress.com/2014/07/16/the-australian-temperature-record-revisited-part-4-outliers/ )

  

Attachment 1.3b:  An example of cooling adjustments over correcting- Acorn Tarcoola shows decreased trend compared with nearest Acorn neighbours’.

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/07/28/tarcoola-a-cooling-outlier/)

 

 Attachment 1.4a: Mackay maxima differencing including all listed neighbours

(For further information and full explanation see  https://kenskingdom.wordpress.com/2014/11/03/the-bizarre-case-of-mackay-33119/ )

 

Attachment 1.4b: Mackay differencing with 2 most distant neighbours excluded, showing improved differences.

(For further information and full explanation see  https://kenskingdom.wordpress.com/2014/11/03/the-bizarre-case-of-mackay-33119/ )

  

Attachment 1.5: The Bureau’s lists of adjustments at six stations are different- except one at Orbost.

(For further information and full explanation see  https://kenskingdom.wordpress.com/2014/11/05/more-bizarre-adjustments/)

 

Attachment 2.2a: Mean of Tmin annual anomalies at 83 sites- minimally adjusted raw data vs Acorn, 1910 – 2012 data. 

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/05/16/the-australian-temperature-record-revisited-a-question-of-balance/  )

 

Attachment 2.2b: Mean of Tmax annual anomalies at 84 sites- minimally adjusted raw data vs Acorn, 1910 – 2012 data. 

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/05/16/the-australian-temperature-record-revisited-a-question-of-balance/  )

Attachment 2.3: Differencing shows the effect of area averaging using very remote sites, 1910 – 2012 data.

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/05/16/the-australian-temperature-record-revisited-a-question-of-balance/  )

  

Attachment 2.4a: Tmin increase in trend in different regions, 1910 – 2012 data. 

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/06/01/the-australian-temperature-record-revisited-part-2-regional-effects/ ).

  Median network position map adj results

Attachment 2.4b: Increase in Tmin warming in NSW, 1910 – 2012 data.

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/06/01/the-australian-temperature-record-revisited-part-2-regional-effects/ ).

  

Attachment 2.4c: Increase in Tmin warming in Victoria, 1910 – 2012 data.

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/06/01/the-australian-temperature-record-revisited-part-2-regional-effects/ ).

  Vic chart

Attachment 2.5a: ACORN vs AWAP comparison- by month (1911 – 2013 data)

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/10/04/adjustments-grossly-exaggerate-monthly-and-seasonal-warming/).

 

Attachment 2.5b: ACORN vs AWAP comparison- annual and seasonal (1911 – 2013 data)

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/10/04/adjustments-grossly-exaggerate-monthly-and-seasonal-warming/).

summary table seasons

Attachment 2.5c: ACORN vs AWAP comparison- by season- Tmean (1911 – 2013 data)

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/10/04/adjustments-grossly-exaggerate-monthly-and-seasonal-warming/).

mean table seasons

Attachment 2.5d: ACORN vs AWAP comparison- by season- Tmin (1911 – 2013 data)

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/10/04/adjustments-grossly-exaggerate-monthly-and-seasonal-warming/).

  min table seasons

Attachment 2.5e: ACORN vs AWAP comparison- by season- Tmax (1911 – 2013 data)

(For further information and full explanation see https://kenskingdom.wordpress.com/2014/10/04/adjustments-grossly-exaggerate-monthly-and-seasonal-warming/).

tmax table seasons

Attachment 3.1: Comparison of percentage of observations recorded in values from whole (rounded to .0) to 0.9 in the Fahrenheit era, with that of the Celsius era at continuing sites, with irrelevant sites deleted, indicating suitable conditions for creation of artificial warming.  (For a full discussion see https://kenskingdom.wordpress.com/2012/03/13/near-enough-for-a-sheep-station/ ).

Attachment 3.2:  Analysis of the impact of rounding on trends and uncertainties.  (For a full discussion see https://kenskingdom.wordpress.com/2012/03/13/near-enough-for-a-sheep-station/ ).

Our study concluded:

“As more than half of all sites in Australia had rounding probably greater than 50%, truncating at significant levels (33%, 50%, or 100%) before September 1972 would cause artificial warming of between +0.1C and +0.4C per 100 years.”

“Many …. sites have recorded large amounts of data in recent  years that may be in error by up to 0.50Celsius, being rounded to whole degrees, and more than half of the sample studied have recorded erroneous data at some time in the past 40 years.”

“As well, the vast majority of sites … inaccurately recorded observations in the Fahrenheit era by recording in whole degrees. For nearly half of all sites, this amounts to at least 50% of their total observations. It is probable that more than 50% of all Australian observations were rounded. This alone means that temperatures before 1972 may be inaccurate by up to 0.250 C.”

“The large amount of uncertainty in the records of so many sites means that homogenisation as practised by BOM researchers must be in question, and with it all analyses of Australia’s temperature trends.”

Attachment 4.1: Obvious errors indicate poor quality assurance.

There are numerous glaring errors for individual days at many sites.  The following graphic shows Rutherglen maxima at Climate Data Online for September to November 1926 compared with ACORN-SAT maxima from 30/09/1926 – 05/11/1926.

cdo v acorn max ruth oct26

Also on 13/10/1926, Acorn minima records -1.2 (adjusted down from +6.9).

Data for days such as 13/10/1926 with an obvious error, possibly the result of a missing leading digit, are not unusual and are found in the records of many stations.

The Acorn record for Rutherglen has some other peculiarities as well.  There are several separate periods where Acorn’s maxima record frequently does not match with data from Climate Data Online, and is one day too early.  These are:

1/11/1920 – 19/3/1940,

1/12/1940 – 31/10/1944,

1/5/1946 – 31/10/1947, and

1/12/1947 – 31/1/1948.

Attachment 4.2: Another glaring error- one of many

Australia’s hottest temperature is supposed to be 50.7C recorded at Oodnadatta on 02/01/1960, but ACORN-SAT has a temperature of 51.2C at Albany on 08/02/1933.  Many days have been adjusted by more than +6 degrees C, resulting in this ludicrous figure which has passed quality assurance.

  albany max 1933

Attachment 4.3: List of 69 stations with ACORN minima exceeding maxima.

Station, Number of days with minimum temperature exceeding the maximum temperature.

Adelaide, 1. Albany, 2. Alice Springs, 36. Birdsville, 1. Bourke, 12. Burketown, 6. Cabramurra, 212. Cairns, 2. Canberra, 4. Cape Borda, 4. Cape Leeuwin, 2. Cape Otway Lighthouse, 63. Charleville, 30. Charters Towers, 8. Dubbo, 8. Esperance, 1. Eucla, 5. Forrest, 1. Gabo Island, 1. Gayndah, 3. Georgetown, 15. Giles, 3. Grove, 1. Halls Creek, 21. Hobart, 7. Inverell, 11. Kalgoorlie-Boulder, 11. Kalumburu, 1. Katanning, 1. Kerang, 1. Kyancutta, 2. Larapuna (Eddystone Point), 4. Longreach, 24. Low Head, 39. Mackay, 61. Marble Bar, 11. Marree, 2. Meekatharra, 12. Melbourne Regional Office, 7. Merredin, 1. Mildura, 1. Miles, 5. Morawa, 7. Moree, 3. Mount Gambier, 12. Nhill, 4. Normanton, 3. Nowra, 2. Orbost, 48. Palmerville, 1. Port Hedland, 2. Port Lincoln, 8. Rabbit Flat, 3. Richmond (NSW), 1. Richmond (Qld), 9. Robe, 2. St George, 2. Sydney, 12. Tarcoola, 4. Tennant Creek, 40. Thargomindah, 5. Tibooburra, 15. Wagga Wagga, 1. Walgett, 3. Wilcannia, 1. Wilsons Promontory, 79. Wittenoom, 4. Wyalong, 2. Yamba, 1.

(From Willis Eschenbach at http://wattsupwiththat.com/2013/06/28/australia-and-acorn-sat/ .  Another study https://tamino.wordpress.com/2013/06/29/a-clue-for-willis/  claims a total of 954 days.)

Extreme La Nina events – an alternative view

January 28, 2015

Yesterday the ABC hyped up their climate alarmism to another new level with their uncritical and unabashed reporting of a claim by the CSIRO that Extreme La Niña events … will almost double in frequency as the climate warms”.

“Lead author Dr Wenju Cai, chief scientist at Australia’s CSIRO Oceans and Atmosphere Flagship, says their work shows La Niña events will occur every 13 years compared with a past frequency of one every 23 years.”

This is the paper:

Increased frequency of extreme La Niña events under greenhouse warming, by Wenju Cai et al., published yesterday.

Time for a reality check.

The authors say they used climate data from 1900 to 2005, and 21 climate models to predict conditions for 2006-2099, and that an extreme La Nina is defined by Central Pacific sea surface temperature anomalies of more than 1.5C below normal.  They claim that an increase in severe El Ninos will lead to an increase in following extreme La Ninas.

In the paywalled article I suspect the Central Pacific region they use is actually the Nino 4 region.  In this analysis I use data from the Nino3.4 region, which is the overlap between Nino 3 and Nino 4, covering Latitudes 5 degrees South- 5 North and Longitudes 170 degrees West- 120 West.  This is the most common data region used.   I downloaded data from http://www.esrl.noaa.gov/psd/gcos_wgsp/Timeseries/Nino34/ and calculated monthly anomalies from the 1961-1990 means.  There are data from 1870, however I chose to use data from 1876 to match Southern Oscillation Index (SOI) data.

Here are the results:

Fig.1: Nino 3.4 anomalies.  Note 1900 & 2005 limits, and +/- 1.5C thresholds.

nino34

By screening for events of +/- 1.5 or more, we remove the clutter and identify extreme events:

Fig.2: Nino 3.4 data exceeding +/- 1.5C

extreme enso events

The paper claims that the incidence of extreme La Ninas will increase from one per 23 years to one per 13 years.  While there are more extreme La Ninas in the last 45 years, I count seven La Ninas from 1900 to 1999, which is one per 14 years.  There were three very high El Nino peaks since 1970, but there are clusters of extreme El Ninos in the first and last thirds of the record.  So possibly the claim for increased La Nina frequency was for an increase in the frequency of abrupt swings from El Nino to La Nina.

Fig.3:  12 monthly change in Nino 3.4 anomalies. +/- 3C is the threshold for swings from extreme El Nino to extreme La Nina.

12m enso chg

Fig.4: Removing the clutter, change exceeding +/- 3C.

extreme enso change

There we have it.  The extreme changes since 1900 have all been in the last 45 years.  Is this due to Greenhouse warming or natural climate change? Could it have anything to do with the Interdecadal Pacific Oscillation? Or is it an artefact of my arbitrary choice of extreme threshold?

More importantly, does the Southern Oscillation Index (SOI) tell the same story?

SOI data are from the BOM website.

Fig.5:  12 month running mean of the SOI inverted.  Threshold is +/- 8.  Note the historical rises and falls.

 12m soi

Fig.6: Nino 3.4 and 12 month inverted SOI match fairly well, although SOI values lag by up to 2 years.

soi v nino34

Fig.7:  El Nino and La Nina conditions per SOI criteria (+/- 8).  An extreme ENSO event might be +/- 16, although I have not seen that mentioned anywhere.

12m soi tests

Again note the clusters of El Ninos, and the spread of La Ninas, in small groups with large gaps between.

Fig.8:  12 month SOI change exceeding +/- 16.  Horizontal lines indicate the threshold for an annual swing of +/- 24 units, which is associated with some dramatic weather events.

extreme  soi change

I left all of the changes >16, to show the historical spread.  Note there were three extreme La Nina (< -24) changes from 1876- 1916, and three from 1960- 2000, and four from 1973- 2014.  There is no unusual trend.

How does this correspond with the observed rainfall record, especially for South East Australia, which is predicted to receive more extremes of rain and drought due to greenhouse warming?

Fig. 9:  Number of months of severe deficiency.

SE Oz severe droughts

Fig. 10:  Number of very wet months.

SE Oz ext wets

Not very alarming.

Queensland is especially susceptible to ENSO events.

Fig. 11:  The match for Queensland wet years is better.

Qld ext wets

Fig. 12:  But not for droughts!

Qld ext dry

Where are the extreme El Ninos?  Call me underwhelmed.

Depending on the index used, the criteria used, and the length of the record used, you can say we’ve had an increase in extreme ENSO swings, or no noticeable change other than a long period (70 to 90 years?) cycle of more and less extreme changes.

My money’s on the latter, but Time will tell.

BOM Admits “Inaccuracy”!

January 21, 2015

At last- 15 days after their widely publicised initial claims (through Mr Jeff Sabburg) that in 2014 Queensland had rain deficiencies not seen since the 1927-1929 drought, and that 37.3% of the State had the lowest rainfall on record- the Bureau has replied to my complaint.

After the Bureau’s initial perfunctory response on Monday 12 January, I asked for straight answers to whether the claims were correct, and would the Bureau correct them in a Media Release.

This morning, 21 January, I received this email:

Dear Ken,
Further to our correspondence we can confirm that media statements made to the ABC by a Bureau employee on 6 January 2014 did not accurately reflect the relative severity of the current Queensland rainfall deficiencies. Unfortunately the Bureau spokesperson misinterpreted some of the information. We have advised the ABC of the inaccuracy and asked them for an opportunity to update the story, if possible.
 
Regards,
Climate Analysis Section

So the answers to those questions were “No”, “No”, and “No, a media release would be too embarrassing and out of the question, so we’ll just advise the ABC and trust this will be buried as old news”.

I will therefore contact other media outlets (e.g. the Queensland Country Life which also ran the story), with a reminder to ABC Queensland Editor Genevieve Hussey, hoping that people across Queensland are made aware that the drought conditions (terrible as they were and still are for many) were not as bad as reported.

While this is a satisfying outcome for me, it is over two weeks since these “inaccuracies” were widely reported across Queensland in the three media that rural people mostly turn to: ABC Radio Country Hour, ABC TV 7.00 p.m. News, and the Queensland Country Life.  I cannot believe that no one in the Bureau saw these reports and did anything about the falsehood, but I’m not surprised.

If you make a mistake, own up quickly, or trust will dry up faster than the rain.

 

How not to admit a mistake

January 12, 2015

Well that explains everything I guess.  Not.

Here is the reply to my query to the Bureau regarding the claim that 2014 saw “the worst drought in 80 years.”

Dear Ken,
Thank you for your email.
The Bureau’s official publicly available Drought Statement is online at http://www.bom.gov.au/climate/drought/
 
Regards,
Climate Analysis Section
National Climate Centre – Bureau of Meteorology 

Short, sharp, but not exactly to the point.

The relevant paragraphs from the Drought Statement read as follows (my bolding):

It is generally difficult to compare one drought to another, since each drought differs in the seasonality, location, spatial extent and duration of the associated rainfall deficiencies. Additionally, each drought is accompanied by varying temperatures and soil moisture deficits.

The current drought in Queensland is comparable to the 2002–2003 drought, which was perhaps more severe in terms of rainfall deficiencies that occurred at times over a very large area. Historical data shows that the current drought is perhaps a one in ten or twenty year event over a significant part of inland eastern Australia (see for example the 24-month deciles map for 2013–2014), but very severe in some places. For example, some location in central Queensland the present deficiencies are the most severe on record, and in addition have been accompanied by record high temperatures. See: 24-month maximum temperature deciles map for 2013–2014.

Nowhere can I find any reference to “the worst drought in 80 years” or rainfall deficiencies not seen “across Queensland at least since the 1927- 1929 depression drought” or “37.3% of the state… covered by the lowest rainfall on record”.  Also included in the Drought Statement is this map showing 27 month rainfall deficiencies:

qld drought 27m

Perhaps 37% is covered by serious deficiency, but not the lowest on record.So are they admitting the reports were wrong?  I don’t think so.

I will email them again asking for a specific reply, preferably Yes or No, to the questions:

Was Mr Jeff Sabburg correct in saying “In terms of rainfall deficiencies the comparison is we haven’t seen this across Queensland at least since the 1927- 1929 depression drought”?

Was Mr Sabburg correct in saying “37.3% of the state was covered by the lowest rainfall on record”?

If the answer to either of these is “No”, will the Bureau immediately issue a correction in a media release?

I live in hope.

Not the third hottest year either

January 11, 2015

According to the Bureau’s surface temperature record, 2014 was the 3rd hottest year on record.  The satellite derived Lower Troposphere data from UAH (University of Alabama- Huntsville) show a different picture.

uah aust 2014

If rankings are important to you, 2014 at +0.40C was in equal seventh place with 2006, and cooler than 1980, and warmer than 1988 by 0.01C.

2013 0.71 1
2009 0.64 2
1998 0.63 3
2005 0.51 4
2007 0.50 5
1980 0.49 6
2014 0.40 7
2006 0.40 8
1988 0.39 9
2002 0.23 10
1991 0.22 11
2010 0.22 12
1996 0.17 13
2008 0.16 14
2012 0.14 15
2011 0.10 16
1990 0.09 17
2004 0.02 18
1981 -0.01 19
1995 -0.04 20
2003 -0.05 21
1982 -0.12 22
1979 -0.13 23
1999 -0.15 24
1985 -0.22 25
1989 -0.22 26
1987 -0.22 27
1997 -0.22 28
2000 -0.24 29
2001 -0.29 30
1986 -0.29 31
1993 -0.29 32
1983 -0.36 33
1994 -0.38 34
1992 -0.56 35
1984 -0.62 36

But don’t expect to find this reported by the ABC.


Follow

Get every new post delivered to your Inbox.

Join 55 other followers