Warming Never Completely Negated Earlier Cooling: Some Thoughts on NSW Temperature Trends

Last night I presented a paper for the Sydney Institute entitled ‘Modelling Australian and global temperatures: what’s wrong? Bourke and Amberley as case studies’. The paper written with Ken Stewart, Dennis Jensen and John Abbot concludes with some thoughts on the New South Wales annual mean maximum temperature record. In particular we suggest that the record can be thought of as showing a cooling trend of -1.95 degree per century to 1960, and a warming trend of 2.48 degree per century since, Table 1.

Table 1

Taking these values and thinking about them in terms of a rate of cooling for approximately 70 years (1890 to 1960) and then a rate of warming for 54 year (1960 to 2013), we can estimate how much New South Wales cooled and then warmed. The sum of these values gives us the net change, a value that if we then multiple by a total period of 124 years, gives an indication of the rate of cooling per century.

This trend of -0.021 degree C per century suggests that the earlier cooling was almost, but not completely, negated by the later warming. To be clear, the overall temperature trend for New South Wales, calculated from the unhomogenized maximum temperatures as recorded at locations with long series, is one of cooling.

107 Responses to Warming Never Completely Negated Earlier Cooling: Some Thoughts on NSW Temperature Trends

  1. jaycee June 26, 2014 at 8:27 am #

    Good morning Jen’…Just looking over your numbers there, the old ‘tradie’ in me sees a suspicion of cherry-picking…I say; “suspicion” as I am from SA. and not familiar with many of those locations…and I would presume they would be representative of not just locations that experience extremes both high or low temp’s more favourable to a certain outcome (now I know YOU, Jen…wouldn’t do THAT, but there are those…?)….after all, NSW. has a very large land area, stretching from the Alps to the western deserts…so to claim ANY average for such a broad spectrum of temp’ measurements could be seen as a tad ambitious..(?).

  2. bazza June 26, 2014 at 9:40 am #

    and I thought Clive Palmer was devious!

  3. Debbie June 26, 2014 at 9:43 am #

    So Jaycee?
    Why are we therefore being lectured by the CAGW crowd about GLOBAL averages and a perceived alarming human induced (or maybe cherry picked?) trend?
    I think you have perhaps just proved one of Jen’s major points about using means/medians/averages. .. albeit from a different attitude/perspective.
    How did it go Jen?
    Good crowd with good questions?

  4. sp June 26, 2014 at 10:40 am #

    Jaycee – so what if you are not familiar with those locations – what has that got to do with anything?

  5. gavin June 26, 2014 at 11:10 am #

    Given Berkley@Sydney,1.92 ± 0.24 C/100y since 1960, I probed other NSW stations in their data, and found an interesting method in their analysis of our dodgy data. In many cases they give an acceleration for AGW but best of all is their skeptics guide to climate change




  6. clipe June 26, 2014 at 11:59 am #

    berkley shmerkley


  7. Beth Cooper June 26, 2014 at 12:12 pm #

    Wish I could have been present, Jennifer.

    Deviousness in Climate Change activists, bazza, is as prevalent
    on the ground as acorns that lie beneath oak trees in Autumn.

  8. clipe June 26, 2014 at 12:42 pm #

    Carbon dioxide?


  9. jaycee June 26, 2014 at 1:21 pm #

    Well..sp., Deb…like yourselves, I am not a techo person..so I cannot vouch for either location favouring nor temp averaging….but!…I do know that temp’ like a shout, does not stop at a fence-line, so if one was to extrapolate on the finding that NSW. has been cooling over this last century..and taking into account that NSW. touches on four other states, then logic would plead the case that sections of those four states also would be cooling…for like the pebble dropped into the pond, those average temps’ would lap into the other states…and indeed!..given that a large slice of NSW. touches on the Pacific Ocean…how legitimate (in a scientific manner) would it be to merely stop the presumption at the shoreline?….given that those coastal locations are influenced by ocean currents and winds….
    If I was to come to Jennifer with such a premise that only included such data for such a prescribed location and claimed accurate conclusions with such data…what would she say to me in regards it’s scientific legitimacy?
    Perhaps I haven’t pleaded my case with great eloquence…but I think you get my drift.

  10. jaycee June 26, 2014 at 1:26 pm #

    ” Deviousness in Climate Change activists, bazza, is as prevalent
    on the ground as acorns that lie beneath oak trees in Autumn.”
    How very prescient of you Beth…considering the “negotiations” proceeding at this very moment between Mr. Palmer and the PM. in Canberra!

  11. Emily Fraser June 26, 2014 at 1:36 pm #

    Dr Marohasy’s methodology is clear. It’s up to Bazza and Jaycee to find the locations that fit the criteria that have not been included. I can only find Sydney’s Observatory Hill and this station is excluded by BOM from calculation of annual average because of UHI effect. Perhaps why Dr Marohasy did not include it.

  12. sp June 26, 2014 at 2:08 pm #

    Jaycee – You have not pleaded your case with great eloquence and I do not get your drift. I think you have written a load of ruibbish.

    And the question still remans – so what if you are not familiar with those locations – what has that got to do with anything?

    You have a unique ability to never make a point.

  13. jaycee June 26, 2014 at 2:54 pm #

    Well, sd….for your sake…..: Why I asked about those locations, is because if..I..wanted a particular result for an “average”..I would select locations that would “tilt” the average in my favour….(are we getting the picture?)…For instance, if I wanted a “hotter average” I would select locations that were averaging a higher temp over the selected period…and vice versa…now having said that I hereby DENY any accusation against “ol’ Marrowbones” that she would pull such a swifty…but I am imagining it could be done…perhaps I am wrong, as I do not know those places of which you are more familiar.
    Now!…are we up to speed?

  14. jaycee June 26, 2014 at 2:55 pm #

    Oh, sorry..: “sp”.

  15. Debbie June 26, 2014 at 2:55 pm #

    I think you are confused Jaycee.
    To a large extent. . .that is the actual point.
    Averaging averages of averages while quite interesting from a statistical perspective, is not particularly useful in this space.
    Emily is correct, if you don’t like what Marohasy, Abbot et al have presented. . .then the onus is on you to explain the fault in the methodology.
    Because this type of work uses start and end points. . .I guess it’s easy to accuse all of the climate/weather researchers of ‘cherry picking’.
    That accusation is not particularly useful or informative either.
    There is also a difference between being a super computer techie and understanding stats and methodologies.

  16. sp June 26, 2014 at 3:20 pm #

    Jaycee – you presented “a suspicion of cherry-picking” without presenting any evidence as to why.

    As Debbie said – the onus is on you to explain the fault in the methodology.

    It seems to me your aim is to transmit doubt rather than provide explanation.

    I also think you are obnoxious – reference to “ol’ Marrowbones” is crass and unecessary – it does not increase knowledge of climate.

  17. Mr Koala June 26, 2014 at 3:23 pm #

    By including the older temperature measurements as well as the recent stuff from the 1950s or 1960s onwards its much more obvious there was a downward trend followed by warming. Why would anyone throw away the earlier data? Hide the decline?

  18. jaycee June 26, 2014 at 3:54 pm #

    sp…..I feel you are being a little precious…the nick-name of ‘ol’ marrowbones” was meant in an endearing way and in no way offensive…not in my book anyway!!…but I tell you what…I don’t think I’ve met a more grumpy mob of bloggers…and if you say you’re not…I say you’re all in denial !!

    ” Jaycee – you presented “a suspicion of cherry-picking” without presenting any evidence as to why.”……are you trying to be deliberately obtuse?

  19. Debbie June 26, 2014 at 5:42 pm #

    What do you say we’re all in denial about Jaycee?

  20. Luke June 26, 2014 at 7:01 pm #

    Ah yes Amberley “.The temperatures have been
    recorded at the same well-maintained site within the perimeter of the air force base
    since 1941.” Maybe the guard dogs piddled on the met station. Check the pic of Amberley climate station in the ACORN catalogue – what a laugh – nothing like a bit of black – http://www.bom.gov.au/climate/change/acorn-sat/documents/ACORN-SAT-Station-Catalogue-2012-WEB.pdf and have a gander the station from the air – note all the scraping and associated albedo changes from how much runway terraforming over the years – and why is the enclosure a different albedo https://www.google.com.au/maps/place/Amberley+QLD+4306/@-27.6293494,152.7108664,265m/data=!3m1!1e3!4m2!3m1!1s0x6b96b646c8bb44af:0x502a35af3de7e70

    Could have cherry-picked Cape Moreton Lighthouse too but that wouldn’t suit. Don’t look you’ll go blind. Or Nanango.

    And please don’t look at the minima either – seems to be warming UHI effects everywhere – everyone has got some pavers from Bunnings

    And think of all those so dedicated artisans selfishly slaving away at Bourke dutifully noting the temperatures. So dedicated like the dwarfs in Snow White. As the trees and garden got higher and higher over the years. A little paradise in the heat. Ah yes downtown Bourke – https://www.google.com.au/maps/place/Bourke+NSW+2840/@-30.0898226,145.9364024,283m/data=!3m1!1e3!4m2!3m1!1s0x6b00671dd4029055:0x40609b4904403e0 hot one day and hotter the next

    5 mins googling zzzzzzzzz

  21. jennifer June 26, 2014 at 8:08 pm #

    Anyone want to tell us which locations are missing from Table 1 that should be there?

    There must be some more locations in NSW that have continuous Tmax. records from either pre-1900 to 1960 and/or post 1960 to 2013?

    This is a work in progress.

  22. Luke June 26, 2014 at 9:04 pm #

    After relaxing at Bourke it’s always nice to warm up a bit at Mungindi. http://www.bom.gov.au/jsp/ncc/cdio/weatherData/av?p_display_type=dataGraph&p_stn_num=052020&p_nccObsCode=36&p_month=13

  23. Debbie June 27, 2014 at 8:21 am #

    Precicely Mr Koala.
    By using a longer time frame it appears there is a different story about the trends.
    Luke. . .
    If the ‘trend’ is not manifesting itself in local/regional data/observations where is ‘IT’ happening and where are we all supposed to DO something about ‘IT’???

  24. gavin June 27, 2014 at 8:31 am #

    Jennifer; It’s been my view that during any stable weather pattern and in particular, extended highs, we can throw a blanket over large regions and get the same BoM temp within 1C at any given time of day.

    This is an ideal calibration point for remote instruments, but I don’t expect anybody here to go checking through old weather station data for max min points and other daily conditions including atmospheric pressure, prevailing winds etc. to prove me right or wrong about error hunting early records because somebody in major climate research organisations have already done much of the correcting via various computing methods.

    What we can do is accept my observations and go straight into identifying the likely cosy relationships between older station records (pre 1960) by simply drawing a line between say Broken Hill elevation 315 m and St George @205m then consider all the river towns, Menindee @61m, Bourke @107m, Walgett@130m, then add Lightning Ridge etc. with appropriate adjustments for their elevation. Any period of continuous daily measurement pre 1960 from the chosen region should be included after consideration of possible bias due to terrain.

  25. Debbie June 27, 2014 at 8:47 am #

    Bias due to terrain?
    You’re kidding ????

  26. Ian George June 27, 2014 at 9:37 am #

    Checked out Berkeley for Casino. Raw temps show downward trend (even after quality control) but after adjustments, shows large warming trend.
    So I checked Berkeley’s Australian temps. It clearly shows 2009 warmer than 1998 whereas the BoM annual summaries shows the opposite.
    I wouldn’t put too much faith in Berkeley’s data.
    Berkeley started in 2010 – they should have adjusted it by now.

  27. sp June 27, 2014 at 11:08 am #

    Gavin – “What we can do is accept my observations …..”

    So humble.

  28. Debbie June 27, 2014 at 12:53 pm #

    Why would we consider ignoring the NATURAL terrain when recording weather?
    What purpose would that serve?

  29. Mr Koala June 27, 2014 at 2:03 pm #

    It looks to me like you already have most of the locations in NSW. I have also got the bears to look separately at Victoria and Queensland. The story is very similar. Cooling pre-1900 to 1960 then a warming trend, which has taken a hiatus in recent years.

    We don’t homogenize anything because we think the physical readings speak for themselves.

  30. Debbie June 27, 2014 at 4:52 pm #

    Yes Mr Koala.
    Of course the physical readings are important!

  31. Luke June 27, 2014 at 7:50 pm #

    Well K Bear – cherry picking is easier. And don’t worry about site recon and history (like Bourke or Amberley – still laughing) or the minima. And a Glaisher here or there – who cares.

    The propagandists path is build a big pile of cherries.

    It’s issues like pan evaporimeters showing negative evaporation after Christmas parties that are fun anomalies too. Or fiddling the numbers to get those teachers to come to town.

    Hey why I am talking to a stuffed toy anyway? You’re an inanimate object.

  32. Mr Koala June 27, 2014 at 7:59 pm #

    There is no cherry picking . All possible candidate sites are included. Better to talk to stuffed toys than uneducated, overpaid, arrogant government employees. Find a useful role for yourself in society.

  33. Luke June 27, 2014 at 8:27 pm #

    “All possible candidate sites are included” LOL ! Due diligence = zero

  34. jennifer June 27, 2014 at 8:56 pm #

    Via Anne Henderson at the Sydney Institute…


    speaking on
    Modelling Global Temperatures – What’s Wrong. Bourke & Amberley – as Case Studies
    The talk will be broadcast on A-Pac Channel 648

    on Saturday 28 June 2014

    at 8.15am and 9.30pm

    Also podcast on The Sydney Institute website


    And the complete text with charts can be downloaded


  35. Luke June 27, 2014 at 9:21 pm #

    The great thing about uneducated, overpaid, arrogant government employees is that occasionally they get off their bums and develop neat open data access to groovy things like
    http://globe.six.nsw.gov.au/ – so what ya do is you open the NSW kml in Google Earth (developed by greenie lefist carbon conscious rent seeking Google workers) and have a good look in the old post office back yard at Bourke, What a great place to cool off. Nice trees. Nice whopper comms tower too. And great historic building where the incredibly diligent seven dwarfs at great personal expense recorded the temperature without error every day. They loved it and had a great sense of duty.

    It’s a great way to check up on sceptic try-ons. As is Bing Maps (often higher res than that leftist Google Earth), and http://www.dnrm.qld.gov.au/mapping-data/queensland-globe (some banana bending high res) but shhhh don’t tell the sceptics or they’ll start doing reccy missions

  36. Debbie June 27, 2014 at 10:23 pm #

    The stuffed toy is making more sense.
    So Bourke has cooled & then warmed & then plateaued.
    The stuffed toy is just commenting on the observations. . . Luke is blaming the trees?
    Where does the neat open access say that those shifting trends is explained by trees. . .or for that matter. . .that the earlier recorders who were indeed diligent. . . wouldn’t know that it’s cooler under trees?
    EVERYBODY knows that it’s cooler in the shade!

  37. Luke June 27, 2014 at 10:34 pm #

    Multiple degree genius Debs. I wonder if dem trees grew. You can hear Debs cogs working. The whole site is a changing compromised unknown Debs. Like Amberley – did you check the pic in the catalogue. Of course not. There is no site due diligence. Like BoM do. Personally I prefer the backyerd of the PO in Mungindi. NSW side of course.

  38. Debbie June 28, 2014 at 7:25 am #

    No need to wonder Luke.
    Of course the trees grew 🙂
    That’s what trees do.
    EVERYBODY knows that too.

  39. gavin June 28, 2014 at 7:49 am #

    Deb; while I searched yesterday for a simple illustration of temp v altitude and some figures for c / m in Aus data, I failed to get exactly what I’m after although there was much on international balloon studies at high altitudes. So I suggest you have a go in your free time and give us the reason all BoM sites are referenced to altitude up front.

    BTW; some retirees have a busy life between blogs. Yesterday it was dry enough to tackle the rest, a savage hand pruning of my boundary ornamentals. Because of recent winds, wires and neighbours power interruptions, the authorities will be back soon with their ultimatums for everybody. Pruned my thumb too when my hook saw encountered one too many in the thousands of whippy plum water shoots headed for the sky and that was between very exciting on line hand tool auctions I had to have an eye on.

  40. Debbie June 28, 2014 at 8:13 am #

    A bit like Luke’s trees. . .EVERYBODY knows that temps vary at varying altitudes.
    Despite all these variables (aka excuses) this alarming GLOBAL trend that is apparently subject to unquestionable due diligence has not manifested itself where it was supposed to be.
    That’s why the stuffed toy is making more sense at the moment.
    There could be infinite reasons why the modelling is not projecting correctly. . .but I suspect that the majority of those reasons may be human based variables. . . and the real environment/weather/climate is not particularly interested.

  41. Luke June 28, 2014 at 10:25 am #

    Just waffing piffle Debbie. The beloved Wattsup blog made its fame questioning site integrity. And now it doesn’t matter for you?

    So here we have some of the worst analysis I’ve seen in ages.

    (1) cherry pick two stations with no site research – and none of you looked yourselves or questioned it. Even accepted the gilding the lily about how diligent the recording technicians were – why – do you personally know they even did a good job?
    (2) cherry unpick stations with the opposite story – shhhh
    (3) misrepresent what BoM are saying
    (4) don’t discuss trends in minima

    I cherry pick Mungindi and Cape Moreton Lighthouse to tell the opposite story – took a whole 5 minutes Debbie.

    Let’s pick Georgetown Qld for a very big warming story – I hear the recorders were really diligent too. LOL. http://www.bom.gov.au/jsp/ncc/cdio/weatherData/av?p_display_type=dataGraph&p_stn_num=030018&p_nccObsCode=38&p_month=13


    Let’s pick Richmond down the road – same story – http://www.bom.gov.au/jsp/ncc/cdio/weatherData/av?p_display_type=dataGraph&p_stn_num=030045&p_nccObsCode=38&p_month=13

    BTW I just left out the max story as it didn’t suit me.

    But let’s pick 040082 with a different story to nearby Amberley with its sometimes blackened enclosure.. Or beautiful Nanango up the road 04015

    So the game is Debs to pick what suits your argument. Add a layer of suitable rhetoric to introduce your topic and call it science.

    If you’re convinced by this sort of tomfoolery well good on ya. Hurry on the Royal Commission.

    What would be very relaxing of course is to do an Aussie Wattsup-up and travel around all these remote and lost post offices and check out the sites and interview the codgers who dutifully, diligently and selfishly measure those temperatures. Wouldn’t that be a lovely holiday. And may have know that a bloody big tree eventually grew and shaded the whole site (or not) or whether the diligent dwarf was on the grog and hated doing the met obs..

    And gee I do like Lady Elliot – great trend in the max 039059 – all you have to do fun and games Debs is enter 36 or 38 or the station number in the web link and a whole new analysis. Such fun. Oh and I do like Bundaberg minima 039015 – that suits me …. and look at Birdsville go 038002

    – and a nice little trend at Giles too 013017

    BTW have you seen the minima at Nullagine WA – 004027; but not sure about Macquarie Island 30004 – but I did appreciate Lake Victoria Storage’s effort 047016

  42. gavin June 28, 2014 at 11:29 am #

    Aah Luke G’day; re our site deliberations for Deb, I always go for the soft option in blaming our pesky instruments in all their glory rather than the PM’S at their old PO or those poor buggers on their way back from the local airport tower checking something out on the perimeter fence in the wind and rain after a couple of tricky commercial landings before modern electronic guides were installed.

  43. Debbie June 28, 2014 at 1:40 pm #

    Luke. . . along with Jaycee upthread. . .I think you are confused.
    Most of your argument is precicely the point. . .albeit from a different perspective.
    Real time data and the progress of real time is judging all of it very harshly.
    But please go ahead with all the excuses that BOTH SIDES of this particular debate use. . .as each season progresses. . .people like me are becoming more and more amused and less and less impressed.

  44. Luke June 28, 2014 at 3:56 pm #

    Deb changes the subject. So Debs given the evidence I have presented here do you think Bourke and Amberley are stable reference sites. Or are the trends really due to changing local site conditions. And how do you explain wildly different trends in nearby stations which Jen just seems to have not mentioned?

    So want to do this task properly –

    lets’s see all the stations.
    Let’s see max and min.
    And let’s see some serious due diligence on the recording sites.
    Let’s see some serious breakpoint statistical analysis

    Now Debs – why hasn’t this been done you should ask?

    And if you wanted to do a more complete job including stations with short runs you can easily interpolate each day’s national temperatures to form a best analysis grid. So even short runs will get a go. Using this technology http://fennerschool.anu.edu.au/research/products/anusplin-vrsn-44

  45. Luke June 28, 2014 at 3:57 pm #

    “people like me are becoming more and more amused and less and less impressed.”

    Perhaps that’s coz you are imbibing more and more b/s and try-ons

  46. Mr Koala June 28, 2014 at 4:14 pm #

    So lets see all those long term temperature records – 50 years or more in the same location – that tell a different story.

    Where are they? They must be hidden away somewhere.

    The only way you can come up with a different story is to apply a “Phil Jones and the boys” trick to make the early records vanish, hide the decline and then you are home and dry. (Or totally wet if you rely on BOM rainfall forecasts).

  47. bazza June 28, 2014 at 5:24 pm #

    You can easily tell an ideoblogger – they ignore the existing peer reviewed research on their silly claim. This topic of NSW temps has been a regular here and all silly claims have been shown to be a joke. None have ever referenced the work by Nicholls (Aust. Met. Mag. 53 (2004) 263-268) which explains it all- not even Mr Koala with his koalafications ( Debs contribution to originality) who disappeared from a previous discussion and Cohenite gave us a laugh when he put up his Newcastle Nobbys example.
    So for annual anomalies, rainfall explains half the variation to 2002 including the mid-century cooling but not much of the increase since. The analysis is from 1910 – nobody would use earlier data without a thorough check to avoid being a joke like Cohenite at Nobbys where the thermometer must have been on the wall of the lighthouse in the early days.

    MaxT(°C) = 1.92 – 0.0035xRain (mm)

  48. Debbie June 28, 2014 at 6:08 pm #

    Now even more amused and even less impressed 🙂

  49. Luke June 28, 2014 at 6:36 pm #

    Bear tries the bluster technique. What a bullduster.

  50. Luke June 28, 2014 at 6:37 pm #

    “the thermometer must have been on the wall of the lighthouse in the early days” hahahahahaha

  51. Mr Koala June 28, 2014 at 7:51 pm #

    Dear Mr Luke,

    I cannot see your list of sites anywhere.

    Put up or shut up.

  52. Debbie June 28, 2014 at 11:07 pm #

    The stuffed toy/bear is still making more sense.
    There seems to be this new phenomenon in the AGW crowd. . .It’s called ‘excusititus’.
    They’ve even resorted to using the same excuses as ‘the other side’. … and still the weather/climate just doesn’t seem to care.

  53. Belinda Alexandrovics June 29, 2014 at 11:22 am #

    Jennifer, Judith Curryls latest post cites your post on NSW adjusted temps
    with reference to problems with US adjusted temp data.also.

  54. sp June 29, 2014 at 11:42 am #

    Looks like the resident Sockmaster has chosen to shut up

  55. jennifer June 29, 2014 at 11:58 am #

    Thanks Belinda. I’ve just been across and see that Judith has linked to an old post. It would be good if we could get the link to the new paper circulating… http://tinyurl.com/lcgk68v

  56. Belinda Alexandrovics June 29, 2014 at 12:12 pm #

    Yes, sp, Blatant infilling data of dead stations is hard ter swallow fer a sock.
    WUWT has the analysis too. (Belinda is Beth the serf’s er sock, sorta …
    Using Belinda’s computer in the bush. infilling fer my missing computer.

  57. Luke June 29, 2014 at 1:54 pm #

    Well sp you could have engaged on any of my points as could have Debbie but the silence is deafening. No debate here just bogus sceptic proclamations and cheer squads. I see Lucia has just shredded that creep Giddard on how to stuff up analysing met data. Over at the Blackboard.

    Luke, I could delete this comment, but I will leave it, as it shows you are reading this thread, while at the same time refusing to provided data to substantiate your original claims that Table 1 contains ‘cherry picked’ sites. Please provide the sites missing from Table 1 (as Mr Koala has suggested “put up or shut up) or apologise to myself (and coauthors Jensen, Stewart and Abbot) for suggesting that our paper was misleading/devious. I repeat, please show which stations are missing from Table 1 or apologise at this thread. Kind regards, Jen

  58. bazza June 29, 2014 at 2:08 pm #

    Seems like I owe Jen an apology. . I once queried the use of Sydney Observatory temperatures in an analysis of forecasting Queensland rainfall. After all, Queensland had separated in 1859. Jen’s bizarre choice was allegedly based on the series being available from the 1850s I recall, and no Queensland station had such a record. But she only analysed from 1900 and she is not reticent in using old records site unseen and not understood. Curiously as you now know Sydney has a different trend to much of Queensland. (remember this was the neural net paper in some journal I cant recall and it was claimed by Jen it would overthrow the CO2 paradigm!). Temperature data for Queensland stations with their upward trend was clearly inconvenient.

    So today the State of the Tropics 2014 report is being launched. The tropics are marching polewards at 10km/year. I read that later this century Sydney’s climate will be more like Brisbanes. It was indeed very prescient of Jen to use Sydney temperatures a s proxy for Queensland.

  59. jennifer June 29, 2014 at 2:16 pm #

    Yes Bazza. When John Abbot and I first used temperature data for our ANN work, we had little familiarity with the temperature record and thought a good long record from Sydney would serve our purpose. At the time the site had the longest record for anywhere on the east coast and was considered a “high quality” site by the BOM. While with hindsight, we would have been better to use somewhere else, the data did get us started.

    We have since progressed in leaps and bounds with our ANN work, and taking on board your advice, no longer use Sydney Observatory.

    We now understand that Sydney Observatory data shows an apparent UHI effect and is best not used, because it is not representative.

  60. bazza June 29, 2014 at 3:28 pm #

    You are talking some sense Jen, but which station is now deemed representative and how would you know unless you did a spatial analysis? Are you ready for that?

  61. Luke June 29, 2014 at 5:11 pm #

    Come on Jen – don’t try to ram raid.You could just answer some simple questions. But you’d rather not obviously and do a Debs by asking more questions.

    (1) No substantiation of site condition changes at Bourke or Amberley. And lots of gilding the lily bluff about how diligent the post office workers were. How pretentious. Cue the violins. Maybe they were maybe they weren’t. Do you know?

    (2) No treatment of major trends in many many sites in minima nicely ducked) – except bluff that’s its UHI – any substantiation of that?

    (3) And if you and bear are UNABLE to read, I have above been listing stations not included. June 28 10:25. But it’s not that hard is it. Just put the lot with more than 30-50 years of data. That’s the lot !

    Then we can discuss the pros and cons of each site’s history.

    Why leave anything out after those stoic bronzed Aussie postal workers dutifully gathered all that data. Don’t we owe it to them. Salutes the flag !

  62. Mr Koala June 29, 2014 at 5:15 pm #

    The bears have experimented with different inputs to predict rainfall using neural networks. The selection of which input is most useful for a forecast is not necessarily obvious.

    Foe example, SOI can be a useful input parameter to predict Queensland rainfall. This is derived from pressure differences in Tahiti and Darwin. Neither location is in Queensland, as far as I am aware. Correspondingly, input of a temperature series from Sydney may, and does, produce useful results. Last time I checked that location is not in Queensland either.

    It is a matter of what gives a good result, not necessarily what may expected from very limited understanding of interactions between variables.

  63. Luke June 29, 2014 at 5:51 pm #

    Remember Bazza that beer consumption in Tokyo was a previous SOI proxy of yours ….

    But back to it – so many stations such little time. But it is fun going through BoM’s database with Google Globes on the other screen checking out ground zero.

    Funny that Tibooburra, White Cliffs and Brewarrina don’t look like Bourke at all. But don’t get me wrong Bazza – all top spots.

    But I don’t like Collarenabri as that supports her argument. I’ll delete that one.
    But Bollon doesn’t so that one’s in.
    oooo – Walgett Council Depot – top holiday spot too suits me…

    It’s bloody good fun this isn’t it.

    Maybe I might need to do a few more to get Jen off my back.

    oooo Cunnamulla doesn’t look like Bourke and just up the road.

    Isisford might be a stretch.

    Or we could bait and switch and throw in that it is remarkable that the land is cooling yet SST’s around Australia are warming. http://www.bom.gov.au/climate/current/annual/aus/2013/ But then again all those species are moving south ….. hmmmm

    Maybe that’s why Cape Moreton Lighthouse is warming up.

  64. Debbie June 29, 2014 at 6:03 pm #

    Even more amused Luke 🙂
    Once again the stuffed toy/ bear is making more sense.
    As far as statistical and projective work is concerned it is the results that matter.
    The rest appears to be more about posturing.
    The weather/climate is simply not interested in the posturing.
    I’m not either.
    As I commented above. . .as each season goes by. . .people like me who work every day in the real environment/weather/climate are more amused and less impressed.

  65. Luke June 29, 2014 at 7:39 pm #

    Debbie – I’m not discussing statistical and projective work. I do worry about your level of comprehension. Indeed if you do work in the real environment you might have some experience in met station sites and measuring basic data which you could deploy here. So Debbie was is your independent evaluation of Bourke and Amberley and how they tally with nearby stations.

  66. jennifer June 29, 2014 at 8:19 pm #

    Luke, You are not providing sites that fit any criteria. You are simply cherry picking.

    Bazza, You keep changing the goal posts.

    Clearly neither of you can concede anything, you are both spoilers and history will remember you both badly.

  67. Debbie June 29, 2014 at 8:47 pm #

    🙂 🙂 :-).
    Luke. . .Quite clearly. . you are not really sure about what you’re discussing.
    Real time data is the judge and it has indeed started to judge very harshly.
    The climate/weather is simply not interested in cooperating.

  68. Luke June 29, 2014 at 9:06 pm #

    Now that a swag of sites appear Jen accuses me of cherry picking. You offer no reason for rejecting those sites. Many stations around Bourke look nothing like Bourke. in terms of temperature trends.

    Anyway a good opportunity to engage on the wider issue of trends and station quality missed. BTW I haven’t necessarily supported BoM’s adjustment to Bourke and Amberley, Moreover returning to what you wished which was first principles.

    They are IMO

    (1) reasonable runs of data
    (2) trends in maxima
    (3) trends in minima
    (4) analysis of site quality, UHI and tree shadow effects

    e.g. so how does one assess the Gayndah site https://www.google.com.au/maps/@-25.62578,151.611202,3a,75y,234.81h,87.08t/data=!3m4!1e1!3m2!1sa82gkXS8dO16vO8tmxMS7w!2e0


    Don’t see anyone else here putting up any numbers Jen.

    Spoilng eh?

  69. Luke June 29, 2014 at 9:08 pm #

    Debbie are you serious. Jen isn’t talking real time data. Nobody is ! What on earth are you rabbiting on about. We’re discussing raw historical data. Try to get minimally oriented.

  70. bazza June 29, 2014 at 9:11 pm #

    This is exciting. I checked Deniliquin for the annual max. temp. data. You can see clearly when global cooling first began in the modern era. The date was June 1908. Prior to then, annual max temps had averaged about 25C at Deni.

    You are not going to believe this but no annual max has exceeded 25C since over the period of record to 2002.
    Who would have thought that such a momentous tipping/urning point in global history would have emerged at Deni!
    The tipping point of June 1908 is obvious when you look at the data and at the graph of annual max temps.
    ( I suppose some spoilsport will come along and knock my moment of fame by proving that the pre 1908 data is clearly a different distribution to post 1908 and that a more boring explanation relating to the site and the instrument is more plausible ). And to think I only looked at one station. I will not waste more time on such careless analysis.)

  71. Ian George June 29, 2014 at 9:50 pm #

    Seems like there was a change to Deniliquin’s min temp about that time as well.

    Note however the temps start off lower, rise, then fall.

    Here’s how NASA GISS originally graphed Deniliquin’s mean temp.

    I know Mildura received a SS around 1908 so I wonder if Deni did too. It does appear that Deni has not warmed at all.
    ‘You are not going to believe this but no annual max has exceeded 25C since over the period of record to 2002.’
    And not since 2002, either.


  72. jennifer June 29, 2014 at 10:06 pm #

    Luke et al.,

    You are not providing the names of locations with long temperature series.

    Brewarrina, for example, does not have a continuous record either from 1900 to 1960 or 1960 to 2013. We need stations for which there is enough data to determine a trend.

    If you want to argue against the criteria we have used to construct Table 1, then provide some reasoning. Otherwise, find stations that fit this criteria as detailed at the top of table 1.

    To pick and choose stations that have short temperature series is disingenuous.

  73. jennifer June 29, 2014 at 10:10 pm #

    Bazza et al

    The mean minimum temperature for Deni is interesting.

    Once we have completed our study of maximum temperatures (for sites across Australia) we will begin a study of minimum temperatures, and perhaps then move on to means.

    At the moment, we are focusing on maximum temperatures.

  74. Luke June 30, 2014 at 12:30 am #

    Jen – nope – that’s just one station and the rest are still enough to tell you the trend parts are different.

    You don’t get out of it that easily.


    looks nothing like


    despite the gaps – AND that is simply one station.



    White Cliffs


    You’re now doing what you accuse BoM of and wishing data away.

  75. Ian George June 30, 2014 at 7:05 am #

    After comparing 1911-1940 to 1981-2011 average max temps for the ACORN stations around Bourke we find that Bourke and Cobar have cooled (0.6C and 0.2C respectively).
    Walgett (0.3C), Tibooburra (0.3C) and Wilcannia (0.1C) have warmed.

    Make of this what you will but it appears there has been little warming over the past 100 years in this region.

    The data is from BoM’s Climate statistics for Australian locations and is based on the CDO record. ACORN data may well be different.

  76. hunter June 30, 2014 at 8:20 am #

    Jennifer, you got a nice reference over Judy Curry’s blog regarding your work on the corruption of Australian climate data.
    It is interesting that while skeptics will admit a skeptic is wrong, and forcefully. Yet the true believers can never admit a problem on anything in their side and blame those who dare point out the problems.

  77. Debbie June 30, 2014 at 10:50 am #

    Terminology aside. . .and you’re right. . .I should have written raw data . . .the basics re this post remain the same.
    Jen’s research is based on observations and the research is observing a different story.
    BoM does not OWN the Aussie weather/climate and apparently, the Australian weather/climate is not interested in being a computerised , reportable commodity for BoM.
    The posturing about the ‘settled science’ is merely human, egotistical posturing. . .and the weather/climate is NOT interested.
    As a farmer. . .I encourage all genuine research in this space.. . I would also reasonably expect entities like BoM to be at the forefront.

  78. bazza June 30, 2014 at 10:55 am #

    Luke, for balance your search for cherries should include bad cherries not picked – bad in the sense of not helping the ignoble cause. The first is that there is sadly no hypothesis here to guide the wild goose chase. The second is ignoring the extraordinary change in variance around 1910 which invalidates regression through significant homoscedasticity. Then there is the story of the min temps as well picked up by I George at Deni.

    Then there are start-end dates that are extreme cherries. Change 1960 to 1950 and you halve the early trend showing of course the dominance of rainfall in the trend because of the more even rainfall pattern in NSW and because of the importance of rainfall in explain trends. Change 1900 to 1910 and you reduce the early trend by one third and you also comply with the BOM warning on the usefulness of pre-1910 data which you have conveniently ignored. The beat goes on.

    Koala added to the phony tone yesterday with his profundity on data mining and his ignorance of artificial skill :“It is a matter of what gives a good result, not necessarily what may expected from very limited understanding of interactions between variables”.. Reminds me of the OEP “If you don’t know where you’re going, any road (variable) will get you there.” (I reckon Jen should have another go at grooming Koala – he could be a useful recruit – who could forget Bob FJ’s contributions.)

    Meanwhile Jen will march on with her case studies ( they have been such a big help in developing the marketing discipline) forgetting the dictum “there can be no science of isolated phenomena” and will begin to learn how to integrate a whole lot of poorly sampled point data until one day she comes up with a trend map just like the one already available from BOM. And the rest is hysteria!
    Talk about losing the plot. This is not even kindy-research.

  79. Luke June 30, 2014 at 11:26 am #

    Debbie and if you had any nous that’s what I am talking about – raw data – so your padding on adjustments and projections isn’t relevant to my points. Your insight and take on these posts is trivial and scatty. The fact that I have pointed out a swag of stations around Bourke that don’t match the trends of Bourke obviously means nothing to you.

  80. Debbie June 30, 2014 at 11:27 am #

    Sometimes cross posts at blogs are deliciously ironic.

  81. Mr Koala June 30, 2014 at 11:41 am #

    Luke – where are your computed trends?

    Do the bears really have to do all the work?

    White Cliffs??

    Not a great example because the record runs only from 1907 to 1956. But it is yet another example of a downward trend, consistent with the majority of sites in NSW between 1890 and 1960.

    White Cliffs has a trend of -1.77 degrees per century. Cooling!!!!!!!

    I think you need to finally try and look at evidence and apply common sense. Maybe you have worked in that government office too long and need to have a break. Take some long service leave and go to the beach.

  82. Luke June 30, 2014 at 11:52 am #

    Bazza I’m doing a climate audit in reverse here. I’m not making any hypothesis or any analysis – simply saying that surrounding stations don’t match and that the detail of site effects over time are an issue. Which of course I understand those dreadful leftie greenie world govt virtually-corrupted commies at BoM do spend some time looking at.

    So I’m not sure Jen had case studies or a few cherries plucked from the case.

    Myself I think there is only quality station in Australia which incidentally was serviced by a salt of the earth postal worker who would struggle through crocodile infested swamps to get the mail through and never drank during his entire service period to ensure that the meniscus was always read properly. This is my cherry ! GEORGETOWN and a very relaxing place for a family holiday. I suggest Bazza that we owe to his memory to keep Georgetown alive in the memory of those people who just think its a set of numbers. It’s a beautiful set of numbers Bazza, And unlike Bourke where the living is easy this guy measured data in tropical cyclones and anchored himself to a tree with a rope to stop being blown away.

    Although putting kerosene in the evaporimeter to quell a mossie plague may have invalidated a few datum points.



    Now clearly Jen had had some bad luck in her random picking of case study stations – this is how trends should look and I written to the Minister demanding to know why all other stations do not look like this. BoM should adjust their data until it looks like this as this is clearly the correct answer.

    Incidentally for those into fine points like homoskedalling – Lucia is going to town on the beloved and objective S Goddard but more importantly revealing some of the real issues in analysis –




    I note that the screechy Koala has obviously overdosed on some high quality eucalyptus leaf ganga and is having a well-deserved big nap. Or perhaps after putting up he’s decided to shut up.

  83. Luke June 30, 2014 at 12:01 pm #

    Is see Mr Baaar has woken up and isn’t getting any better after his nap – even crankier it seems Just for Debs who is a sucker and believes what Mr Baaar says:



    So Baaar plucks a bit of data and turns on the bushfire smokescreen hoping nobody will notice. Ignores the right hand side of the graph. White cliffs max temp data looks nothing like Bourke.

    You couldn’t lay straight in bed mate. Although I guess Koalas sleep stoned in trees at odd angles so maybe that’s understandable.

  84. Luke June 30, 2014 at 12:11 pm #

    And Bazza why are there only 3 stations in common between column 1 and 2 in the Table 1 in the post intro.

  85. Debbie June 30, 2014 at 12:20 pm #

    Sorry to disappoint you. . .but I don’t BELIEVE that anyone is unquestionably right or wrong.
    You may be suffering a little from that straw man thingo. . .or perhaps a bit of projection?
    Mr Koala has offered a good suggestion. . .look up from your screen. Maybe go to the beach?

  86. Dennis Webb June 30, 2014 at 12:26 pm #


    There are only 3 stations in common because there are only 3 stations that fit the criteria for inclusion in both columns.

    Bourke is not even included in column 2, because it doesn’t have a complete data set from 1960 to 2013.

    The criteria are clear and reasonable.

  87. Luke June 30, 2014 at 12:32 pm #

    Debbie I am at the beach on the iPad near Q1.

    Dennis – it’s always good to pick criteria to get the answer you want.

  88. Debbie June 30, 2014 at 1:26 pm #

    Hope it’s nice at the beach Luke.
    Remember to look up 🙂

  89. Luke June 30, 2014 at 1:32 pm #

    Well gusty today or I wouldn’t be online and watching Game of Thrones.

  90. Debbie June 30, 2014 at 4:10 pm #

    Actually. . .I would say it’s just too easy to pick criteria to get the answer you want.
    That’s largely the point.
    There’s obviously a lot of criteria eh?

  91. gavin June 30, 2014 at 6:37 pm #

    Nobody yet mentioned how early max min records dealt with a migration from the old verandah to a cozy screened enviro, i. e. no direct or indirect HF radiation.

  92. gavin June 30, 2014 at 6:54 pm #

    Jen; “Once we have completed our study of maximum temperatures (for sites across Australia) we will begin a study of minimum temperatures, and perhaps then move on to means”

    Neither max, min markers or their mean give a true measure of heat one way or another, daily or year to year without respect to time and duration even if we could clamp the inevitable errors inherited in original BoM station data.

  93. Phill July 1, 2014 at 2:55 am #

    Thank you for this article and congratulations on the work so far. I have been mucking around with the raw data for a few years trying to get sensible reconstructions of temperatures mostly from around SE Australia but also NZ and the rest of Australia. Many of my results show the same dip and the rise as your two examples but from larger samples.

    There are many early record stations were the original site has been moved such as Alice Springs where the PO has a record from 1878 to 1953 and then there is an overlapping record at the airport starting in 1941 to the present. In this case the two records splice almost perfectly together and show basically flat lines.

    In my view there a lot of reasons why parts of the late 19C may have actually been hotter than modern hot periods. The obvious one is agriculture. In those days there were no irrigation systems, no dams and at times even the Murray and the Darling stopped flowing. Once the last of the vegetation was gone there would be nothing to stop it getting really hot.

    There is also a long reconstruction of SE Australia temperatures from 1860 to 2000 by Ashcroft, Karoly and Gergis that seems for maximum temperatures to agree with your conclusions above. The maximum over this period seems to show a fairly flat line or perhaps falling line whist the minimum shows a steady rise but small rise.

  94. Don B July 1, 2014 at 5:05 am #

    In the USA, the record high temperature for each of the states has not yet been homogenized.

    Of the 50 states, 35 high temperature records were set prior to 1940 (despite horrendous global warming since then!). Three states recorded their record temperature in the 1800s.


  95. Luke July 1, 2014 at 12:04 pm #

    And for step 3 of Ashcroft et al – if you haven’t done the station metadata and breakpoint analysis at this level how can you claim Bourke and Amberley as case studies. Due diligence.

    Raw data ain’t the pure substance that sceptics would portray. Caveat emptor.

    So all the carry on about BoM adjustments is simply hysterical without due diligence on site metadata – and even further onto site condition changes like nearby brick walls, run-way earthworks, , “cleaning up of site enclosures (e.g. herbiciding grass, concrete paths, or white quartz gravel), growing and expanding trees and gardens and urban expansion.

  96. Mr Koala July 1, 2014 at 12:22 pm #

    Between 1890 and 1960 atmospheric carbon dioxide levels increased, and the temperatures across NSW, Victoria , Queensland, South Australia and the NT went down. Then, carbon dioxide levels increased between 1960 and the present and the temperatures went up.

    Maybe Al Gore can help?

  97. Luke July 1, 2014 at 12:34 pm #

    Trust a bear to back evidence which anchors one of the graph with 1950s Las Ninas. Probably good for eucalypt ganga production too.

    As Bazza said above “So for annual anomalies, rainfall explains half the variation to 2002 including the mid-century cooling but not much of the increase since. ”

    Telltale mins have gone up at many many sites as you’d expect.

    Funny that the SSTs have gone up too eh? And do try some due diligence when discussing station data.

    Get real bear and stop being such a faith driven denier. When you’ve graduated beyond the trivial x-y plot and learned to be truly sceptical report back. Then talk to Al Gore.

  98. Mr Koala July 1, 2014 at 1:01 pm #

    Luke – I see.

    The explanation of the cooling trend between 1890 and 1960 was an El Nino condition that went on for 70 years.

    Thanks – I will ask the pixies that live at the bottom of my garden if they have an alternative explanation.

  99. Phill July 1, 2014 at 1:07 pm #

    Quite right Luke, the raw data isn’t pure and has lots of problems. Nevertheless one can make an educated guess about the nature of these impurities.

    My first guess would be that most of the things you mention are likely to bias the result in a particular direction for a given time of year/day. Of your listing above I would only classify, “growing and expanding trees and gardens and perhaps white quartz gravel”, as having any possible cooling effect on day time temperatures. The others I would guess are generally creating heating biases. Even for gardens the claim can go the other way. ie That on cooler or windy sites the enclosed garden shelters and allows the sun to heats the surrounds.

    My second guess is that many of these biases are increasing over time, particularly the urban expansion and runway upgrades. The problem here is that such changes are insidious and won’t show up as discontinuities in the data stream.

    My simple expectation would be that at very least the BOM adjustments should be net neutral over time and most probably should be actively countering known heating biases. The issue is they seem to be doing the opposite.

  100. Luke July 1, 2014 at 1:34 pm #

    Mr Bear – who says it’s a trend – when you’ve done some due diligence on site metadata and serious break point analysis report back. This is the style of same issue put to those who would claims a drying trend from a wet 1950s to a dry Millennium 2000s drought sequence.

  101. Luke July 1, 2014 at 1:53 pm #

    Phill thanks for a discussion.

    I agree with your points on site issues. If you haven’t seen my comments above, Bourke has the issue of in my opinion of expanding trees and gardens. And a large comms tower installed when? . Which Google Street view and the NSW Digital Globe can illustrate. http://globe.six.nsw.gov.au/

    Amberley enclosure is blackened (?!) in the ACORN station catalogue picture. And Google Earth and historical photos provide evidence of much earthworks and scraping in the area leading to what albedo and therefore what temperature changes I’m not sure.

    But to Jen’s point of UHI – Google Earth Streetview and the Queensland Globe http://www.dnrm.qld.gov.au/mapping-data/queensland-globe shows that the Gayndah post office station is in a nice garden but a Besser brick wall of the tyre shop is nearby which would soak heat. When was the tyre shop built? How does this all balance out?

    So to select stations for case studies I am suggesting that station metadata in terms of instrument and observer changes and site environment changes over time need serious analysis as well as the trends in surrounding stations – are they consistent?

    A site visit to talk to locals and the historical society photos may reveal even more. As for portraying that Bourke staff were super-diligent how would we know – seems to be mythologising not science. Maybe they were – maybe they weren’t.

  102. Debbie July 1, 2014 at 2:36 pm #

    Identifying trends and identifying causes are not necessarily the same thing.
    The Aussie weather/climate is seemingly not really interested in conforming with most of it.
    BoM is an entity made up of people who are just as vulnerable to human foibles as anywhere else.
    That includes diligence.

  103. Mr Koala July 1, 2014 at 6:10 pm #

    I think I will go and talk with the pixies at the bottom of the garden. They seem to have more idea than some about scientific evaluations. And they are not a financial burden on the long-suffering taxpayers of Queensland and Australia.

  104. Phill July 1, 2014 at 10:23 pm #

    Luke and all,

    A teaser for you. This link is a picture of a minimum thermometer.


    What temperature would you read the minimum here? Both my kids got it wrong. Hint: The bulb is on the left and the liquid column is orange coloured.

    Its quite obvious, when you look at some of the data, that some of the past observers got it wrong as well. This is the kind of error the bureau should be mainly looking to correct. I would also like to see them annotate each change.

    I looked at the images of Bourke and Amberley Post Offices and they aren’t the best but at least for Bourke PO the main building has been there since 1879 but obviously not the TV tower. In the case of Bourke a new automatic weather station has been set up at the airport. During the 21 full months in 1995 and 1996 when the two sites ran in parallel the yearly results were around 0.3C/year warmer at the new site for maximums and 1.0C cooler for the minimums. Even adjusting for these amounts it still looks like Bourke was hotter in the early part of last century than it is now and if the records are to believed even hotter at the end of the 19th century.

  105. Debbie July 2, 2014 at 9:23 am #

    Yes it does appear that warming never completely negated earlier cooling . . .and further. . .the longer time series, despite the possibility of human error, are telling a different story about the Aussie weather/climate.
    My question remains unchanged. If ‘IT’ is not manifesting ‘IT’self regionally or locally what is the SOMETHING we are all supposed to be doing about ‘IT’?
    And BoM does not OWN the weather/climate. but I would reasonably expect an entity like BoM to be at the forefront and proactive about these matters.

Website by 46digital