ABC Rural Reports on New Rainfall Forecast Model

ABOUT one week after Central Queensland University put out a media release announcing the publication of our latest paper on medium-term rainfall forecasting, John Abbot and I received a phone call from the ABC, from a rural reporter.

Part of the interview with Professor Abbot can be heard online here, http://www.abc.net.au/news/2014-02-21/nrn-forecast/5275122

ABC journalists Cassie Hough and Kathleen Calderwood have also written a short article based on the interview in which they effectively capture the key points, http://www.abc.net.au/news/2014-02-21/nrn-forecast/5275122

And following is a succinct summary of the interview by one of the media monitoring services:

“New research from the University of Central Qld shows medium-term rainfall forecasts may be more accurately predicted by a form of artificial intelligence, which aim to predict the weather through pattern detection, and predicts not what the weather will do day-to-day, but in a month or more. John Abbot, researcher, says he was motivated to do the research after the huge flood at the start of 2011.

Abbot talks about how his research works, and the problems with modern-day weather forecasting systems, and says farmers and mine-operators need accurate data as opposed to vagaries. Abbot says he would like the Bureau of Meteorology take more interest in what he is doing.”

John Abbot

The paper can be downloaded from Science Direct, with full open access available until March 26, here,
http://elsarticle.com/1ej97n3

[Open access until March 26, 2014]

Highlights include:

1. Artificial neural networks are developed to forecast monthly rainfall for localities in Queensland.

2. Comparative utility of input variables, including local maximum and minimum temperatures, is tested in the artificial neural network.

3. The Inter-decadal Pacific Oscillation, IPO, is found to have utility for rainfall forecasting.

4. Artificial neural networks forecasts are superior to Australian Bureau of Meteorology official seasonal and monthly forecasts.

67 Responses to ABC Rural Reports on New Rainfall Forecast Model

  1. Luke February 22, 2014 at 4:30 pm #

    So did you forecast the latest rain out west? Just asking …

  2. Debbie February 22, 2014 at 5:24 pm #

    Did BoM forecast the latest rain out west?
    Just asking…

    Anyway,
    Congratulations Jen & Abbot.

  3. jennifer February 22, 2014 at 5:29 pm #

    We only got involved in rainfall forecasting in about June 2011. Our focus has been on improving the prototype model and getting peer-reviewed papers published that demonstrate the potential of the method for localities in Queensland. Without a publication record, we can’t expect to be taken seriously.

    Now we have 2 good journal papers already published, 2 conference papers already published, and 3 more journal papers in the pipeline, we would hope to get some additional support so we can further refine the method, including through signal decomposition, and start producing forecast for particular localities.

    For example, we are keen to work with a mine or agriculture enterprise. If they, for example, provided some funding, in return we could provide tailored forecasts and also forecasts that could be benchmarked against other methods into the future.

    We need additional resources to further refine our method (hopefully resulting in a step increase in skill) and to make it operational.

    More recently we have looked at further improving our forecasts by also developing a method for incorporating a lead Nino/SOI value.

    We have a plan for improvements, but there is already reason to go operational. But so far no government or industry funding.

    We have also indicated a willingness to work directly with the Bureau.

  4. Luke February 22, 2014 at 6:48 pm #

    But BoM are no good Debs as you’ve been saying. I thought it was a very simple question. If you want punters to take you seriously you need a running live system and and archive of what you said and what happened. Wouldn’t you like to know too – or don’t you care? Don’t we all live in the here and now?

  5. jennifer February 22, 2014 at 7:19 pm #

    Luke,

    Usually you want peer-review.

    Anyway, we are keen to have a live system and an archive. But we just lack the Bureau’s billion dollar budget.

  6. Another Ian February 22, 2014 at 7:23 pm #

    Jen,

    To contrast Luke’s empixellated* view.

    A friend had a description of “storms that you could put your hat over”. And that has been what has made the msm in this lot in this part of the west.

    I’ve run into more than average anger at the beat-up.

    And re the BOM seasonals – it does not improve the reception of their prognostications when the next one is announced from a soapbox of total failure of the last one. It produces many comments, wry, angry, ribald etc. Jen would have to censor most so I’ll just repeat the “Believe the opposite” one. The last one on expected below average rainfall invoked responses of “A sporting chance then”.

    I also seem to recall that BOM bought software from the British Met Bureau. Given that outfit’s modelling success with “barbecue summers” etc did this software happen to be the new model that BOM is using for seasonal forecasts?

    *Empixellated – one who spends too much time looking at computer screens and not enough looking out windows

  7. Luke February 22, 2014 at 8:06 pm #

    OK then Jen – how about a low maint lite version where you could just run an ongoing thread that only you post to and post a list of Qld stations and predicted rainfalls periodically. Cut and paste and post.

    You have your system, you have your list of stations – so is it that hard?

    Come on – we’re all curious as cats. You get up BoM for not being forthcoming with any data….

    And the big wet out west is big news – so surely you’d be into that action? Beef markets moving on that rain ….

  8. jennifer February 22, 2014 at 8:23 pm #

    Luke, You curiosity is not high on my list of priorities. Neither is being part of the action. Slow and steady wins this race.

  9. Glen Michel February 23, 2014 at 5:34 am #

    Harrr…empixellated-will remember that!further of interest is the suspect modelling of UK MET being applied here.We need more of the citizenry taking an interest in meteorology nationwide.If only..

  10. Another Ian February 23, 2014 at 6:22 am #

    And then there is this fallback, also from British Met

    http://diggingintheclay.wordpress.com/2014/02/17/place-your-bets/#comments

  11. cohenite February 23, 2014 at 8:28 am #

    Peter Helman gave a good paper on the IPO at a coastal conference some time ago; it was called:

    TWO CENTURIES OF COASTLINE CHANGE ON THE NORTH COAST OF NEW SOUTH WALES

    Helman notes:

    “Abstract
    Extensive research on past records on the New South Wales and south east Queensland coast has revealed changes that have been related to sea level trends and multi-decadal phases of Interdecadal Pacific Oscillation (IPO). Indicators suggest that before European settlement of Australia sea level trend was stable or slightly falling. Sea level commenced a slowly rising trend from the 1820’s. The rise is detected in early tide gauge observations in Tasmania during the early 1840’s, as well as tide gauge observations at Sydney (from 1886) and Auckland (from 1899). The rising trend is attributed to: gradual warming that followed the global sea level fall after the Tambora eruption (1815), a long negative IPO phase (from the 1850’s to 1890’s) and global warming during the 1900’s. Observations at Sydney indicate sea level rise of around 100mm/100 years, well below the global average of 180mm/100 years. The relatively small rise on the east coast has resulted in erosion of barrier dunes, deposition in estuaries, and washover of barriers and spits. The timing of coastal events is generally related to the oscillation of the IPO. During positive phases of IPO the east coast experiences recurring droughts, low storminess, sea level below the long term trend and inshore accretion. During negative IPO phases the east coast experiences wet periods, major floods, sea level above long term trend and coastline erosion. Coastline changes at specific sites illustrate the impact of sea level rise, due to climate change and climatic variability, largely influenced by IPO phase.”

    And of course Stewie Franks is an expert on the IPO:

    http://jennifermarohasy.com/2011/01/is-it-time-to-listen-to-so-called-deniers/#more-7359

    http://www.lavoisier.com.au/articles/greenhouse-science/weather/Franks2007.pdf

    Jennifer’s paper is a good read.

    Luke, how about you put up some of the BOM’s forecasting successes since 2011?

  12. Luke February 23, 2014 at 8:45 am #

    Cohenite – I put you onto Helman – have you forgotten. And Folland invented the IPO. Somehow I think you’ll find more than Franks knows about the IPO not withstanding his good work in the area. Talk about revisionism. Pity IPO isn’t forecastable – whoops !

    Cohers – I said above this is now not about BoM – it’s Jen’s system. And she won’t discuss it – I guess the science is settled – any curious comments and I’ll be deleted.

    However, Cohers as a rootin’ tootin’ sceptic I would have thought you’d be curious yourself how a system based on ENSO and IPO indicators would go in a neutral year drought sequence. But then again you’d have to be a real sceptic and not a tribalist wouldn’t you? mmmmm

  13. cohenite February 23, 2014 at 9:16 am #

    You did indeed put me onto Helman and what a great little thing it is.

    Jennifer is discussing it; she has papers coming out; read the one linked to here and we’ll have a yarn.

    ENSO and IPO are main drivers but don’t forget the IOD; what has it been doing in this “neutral year drought sequence”?

  14. Robert February 23, 2014 at 9:57 am #

    Take plenty of time, Jen. I’m guessing you’ll need it. Any system has to be checked long-term against stuff that actually happens. (I know “stuff that actually happens” is a mere triviality for some – which is what got us into this mess.)

    Has to be worth a shot. And since it’s done on your own time and resources, you don’t have to take orders. (Orders will be given – shrieked, even – but you don’t have to take them.)

  15. Luke February 23, 2014 at 10:06 am #

    IOD doesn’t explain Qld drought.

    “Any system has to be checked long-term against stuff that actually happens.” clueless piffle comment from Robby and incredible in the sheer depth of its hypocrisy. What do you think is done Robby? And if work is done from a university it’s hardly a personal operation – otherwise why list the institution’s name?

  16. Robert February 23, 2014 at 10:18 am #

    There’s that shrieking I was talking about.

  17. Luke February 23, 2014 at 10:23 am #

    Well this is an evidence based blog Robby – so you should defend your comment that forecasts are produced with checking historical data. Defend your assertion.

  18. Luke February 23, 2014 at 10:27 am #

    “without checking”

  19. Robert February 23, 2014 at 10:29 am #

    They check…but they never learn. (You never check, so you have an excuse, Supe.)

  20. Luke February 23, 2014 at 12:53 pm #

    OK so now you’ve backpeddled on your quickie slur but have substituted that they never learn. On what basis do you say that?

    And I check alright – on you continuously. As has been said it’s an evidence based blog.

    So given you say learn – what would be the difference between learning and spurious adaptation after the fact (why one needs to put the for real forecast up here and now). And at what frequency would you suggest learning be done. And if you are advocating continuous learning all you’re doing is tuning you see. Are you following me Robby?

  21. Robert February 23, 2014 at 2:04 pm #

    Hard to backpedal from a position one has not taken, even in a quickie slur. I said what I said, not what you wish I’d said for convenience of rebuttal.

    Why they never learn? They keep issuing those long-term outlooks. I’m fine if they don’t do that, and fine if they do it better. (And, yes, I know that if the 75% doesn’t come up that means the 25% did etc etc.) I’m also fine if someone with the money and concerns gets behind someone like Jen who wants to try a different way. I mean, with the dough that gurgles daily down the sinkhole in the name of “climate science”, a few hundred thou…

    As for the rest, no, not following you, Supe.

  22. Debbie February 23, 2014 at 3:13 pm #

    Luke,
    Seriously?
    Your questions at 6:48 are all over the place.
    Yes… BoM has not been improving in this space (regional seasonal forecasting) but think about the entities that ARE making those ‘here and now’ decisions based on the seasonal forecast info from BoM.
    (big hint: It’s not farmers).
    I don’t think Robert or Jen or anyone here is actually slightly interested in ‘back pedalling’.
    As I have mentioned before. . . most people out here in the 24/7/365 climate/weather/environment would like to just ‘get off it and get on with it’.
    BoM has done some really good work over the years and they should and could build on what works well. . . but. . . they also need to recognise that there is much room for improvement in some of their public services that has nothing to do with improving the ‘window dressing’ and/or changing the deck chairs and/or continuing to defend something that is not delivering on stated goals.

  23. Luke February 23, 2014 at 3:52 pm #

    Don’t understand at all Debs you’ll have to explain. You want them to do better but on the other hand you just want to get on with it. You’re real happy with Jen’s approach even though you can’t experience it. And something that only has been available less than a season from BoM is not delivering. What a mental mish mash and probably one could only chew one’s arm off in trying to respond to such disparate unrelated thoughtoids.

  24. Luke February 23, 2014 at 3:59 pm #

    Let’s try a pop quiz – why did BoM build POAMA and why do they wish to move away from statistical methods of forecasting? Besides it just being a cool thing to do or “we just felt like it”.

  25. cohenie February 23, 2014 at 4:15 pm #

    Jennifer’s paper really sums p the issue:

    “In June 2013, the BOM moved to a new system based exclusively on The Predictive Ocean Atmosphere Model for Australia, POAMA, which is considered by the BOM to be a state-of-the-art seasonal to inter-annual seasonal forecast system based on a coupled ocean/atmosphere model and ocean/atmosphere/land observation assimilation systems (Australian Bureau of Meteorology, 2013). While POAMA generates quantitative seasonal and monthly rainfall predictions, the BOM chooses to continue to present its official forecasts simply as the probability of exceeding the long-term average value. A major limitation of such forecasts is that they provide no information about the magnitude of the expected deviation from the median rainfall value within the defined forecast period. For many practical purposes, such as management of water infrastructure or scheduling mine operations, the distribution of rainfall within the three-month period is more important than an averaged seasonal value (Sharma et al., 2012). Furthermore, these programs failed to adequately forecast the exceptional wet summer of 2010 to 2011, with significant economic and social consequences (van den Honert and McAneney, 2011).”

    Luke says:

    “IOD doesn’t explain Qld drought”

    That’s not what you said here:

    http://jennifermarohasy.com/2009/06/agw-is-just-a-theory/?cp=all#comments

  26. Noosa Girl February 23, 2014 at 8:36 pm #

    Dear Luke,

    I think you should calm down and think a little more, and write a little less.

    Even a girl from Noosa can understand that, from the computer program’s standpoint, the divide between the known rainfall and what you want it to predict can be whatever you select. In order to check the skill of a particular model, this can only be done by comparing forecasts against known recorded observations. As soon as any observation is recorded, today’s rainfall for example, it inevitably become just another part of the historical record. One cannot travel into the future to collect the actual rainfall data in order to make comparisons with a forecast system.

    You seem to have a tremendous number of opinions about everything, rather than deal in facts and logic. Is it possible you work for the government?

    Noosa Girl

  27. Jennifer Marohasy February 23, 2014 at 9:30 pm #

    The BOM created POAMA because its the fashionable thing to do if you are part of the mainstream climate science community. Never mind that POAMA, and general circulation models the world over, can’t forecast rainfall. They prop up AGW theory.

    BOM staff, and other global warming alarmists, claims statistical models won’t work into the future because the climate is on a new trajectory. That our neural network (a type of statistical model) can produce a reasonable forecast working from historical data (and certainly a better forecast than POAMA) would suggest the climate is not on a new trajectory.

    Indeed our work not only as a practical application by producing a better medium-term forecast, it also directly challenges the theoretical basis of AGW.

  28. Luke February 24, 2014 at 12:17 am #

    POAMA’s reason for development are wide and many. Main reason is to give conspiracy theorists lots of talking points.

    In any case on the minor AGW point – one of BoM’s known previous statistical system issues was the warming trend (oh dear) in the Indian Ocean. You could of course have tried the Oodnadatta Index and found the sub-tropical ridge intensity was also trending (double oh dear). I’ve now graphed all of Qld’s temperature stations – trend in minimum is obvious (oh dear again).

    Cohenite – does the IOD have a significant correlation with most of Queensland’s rainfall?

  29. Debbie February 24, 2014 at 8:00 am #

    Luke?

    ” And something that only has been available less than a season from BoM is not delivering. ”
    Are you seriously implying that POAMA did NOT(!) undergo rigorous testing before it was implemented last year?

    It seems that you think Jen and Abbot should just put their work up BEFORE they have funding and BEFORE they go through all the correct channels?

    I am also unclear why you suggest that wanting BoM to do better in this space and that people who live and work in the real environment/climate/weather day in and day out want to ‘get off it and on with it’ . . .are mutually exclusive comments?

    Maybe you missed the suggestion that we need to build on what works and use what is useful but also recognise what isn’t useful and isn’t working?
    I think you also missed the point that some entities such as NRM authorities ARE making ‘here and now’ decisions based on BoM and POAMA seasonal forecasts.

  30. Luke February 24, 2014 at 8:34 am #

    Debbie – I just shake my head at you. No I’m not implying that BoM did not test POAMA. Sheesh. On season is not enough to really make a pronouncement about any seasonal forecast system. You need a few years. I have told you repeatedly that you have no understanding whatsoever with what is being said and this again demonstrates the issue. If anyone was 100% right with any forecast system they wouldn’t even be talking to us. And Debs you want 100%. You do ! 80:20 would kill you. 60:40 would be impossible to work with for you.

    “It seems that you think Jen and Abbot should just put their work up BEFORE they have funding and BEFORE they go through all the correct channels?”

    NOPE ! But Jen’s making the big splash editorials and media. So I’m asking a very simple question – OK so this is a you beaut new technology that blows BoM into the weeds as the media says – so how did it go over the last two years in Queensland where it’s been developed for? It’s not a hard question. It’s not a trick question.

    In fact dear Debs there are a goodly number of sites which I’m sure you check regularly for some compare and contrast. Don’t you

    e.g. http://www.ecmwf.int/products/forecasts/d/charts/seasonal/forecast/seasonal_range_forecast/

    http://iri.columbia.edu/our-expertise/climate/forecasts/seasonal-climate-forecasts/

    http://www.cpc.ncep.noaa.gov/products/international/cfsv2/maritime_prec_stdanom.shtml

  31. Luke February 24, 2014 at 8:38 am #

    and also

    http://cmdp.ncc.cma.gov.cn/pred/en_cs.php

    http://www.apcc21.org/eng/index.jsp

    http://www.jamstec.go.jp/frsgc/research/d1/iod/sintex_f1_forecast.html.en

    All these sites are in the big global conspiracy Debs. Pretty slick eh? Even the Chinese.

  32. Debbie February 24, 2014 at 9:38 am #

    Yes Luke I do check some of those regularly.
    What global conspiracy?
    My comments were related to what is and isn’t living up to expectations and what would be a USEFUL public service.
    On(e) season may indeed not be enough. . .but why are they arguing that even though their probability forecasts were radically off the mark they are still right (or maybe. . .not wrong?) that we SHOULD be using it as a USEFUL management tool?

  33. jennifer February 24, 2014 at 9:52 am #

    Just to clarify.

    We would also be keen to work with the Queensland state government.

    Our method would have application to anywhere along the east coast and therefore we would also be keen to work with the NSW and/or Victorian governments.

    What we have is a prototype model. With additional resources, in particular the capacity to better signal process, we believe we can significantly increase our skill at forecasting not only rainfall but also ENSO.

    We are keen to work with the climate science community. We are keen to eventually have an operational system.

  34. bazza February 24, 2014 at 10:11 am #

    Jen, I would be interested in how your work challenges the theoretical basis of AGW?. Can you expand a little pls.? You stated above ( 23rd Feb 9:30 pm) :“That our neural network (a type of statistical model) can produce a reasonable forecast working from historical data (and certainly a better forecast than POAMA) would suggest the climate is not on a new trajectory. Indeed our work not only as a practical application by producing a better medium-term forecast, it also directly challenges the theoretical basis of AGW”

  35. jennifer February 24, 2014 at 10:48 am #

    Bazza, It is often claimed by climate scientists that statistical models won’t work because the climate is on a new trajectory (because of AGW) making historical data redundant. Our model is based on the use of historical data and it works. Perhaps because the climate is not on any new trajectory, rather patterns in historical data are still relevant and can be used for future rainfall forecasts.

  36. Debbie February 24, 2014 at 11:09 am #

    Luke,
    You’ve managed to get yourself stuck in a pointless semantic rut. . . and we have had this discussion before.
    A lecture on probability is not going to help you out of that rut.
    Put simply yet again.
    In Australia’s case the significantly lower % probability of a seasonal ‘drier than average’. . .or if you like, below the line. . . and in the UK’s case the significantly lower % probability of a seasonal ‘wetter than average’ does not mean that significantly lower % probability of either SEVERE SEASONAL DROUGHT or SEVERE SEASONAL FLOODING.
    There may possibly have been a % probability that could possibly have been in there somewhere about those ‘WAaaaaaayyy below the line’ eventualities (ie severe drought or severe floods) . . . but neither BoM nor MET claimed so.
    If you are indeed arguing more than silly semantics and you do indeed believe that IS what that lower % probability means. . . then perhaps you need to instruct BoM and MET to write up a clear glossary. . . because out here in the real world. . .drier than average and wetter than average are NOT synonymous with seasonal drought and seasonal flooding.

  37. Hasbeen February 24, 2014 at 11:40 am #

    Perhaps we should stop these BOMs from talking to each other. They do appear to be infesting each other with bullshit.

    I was just talking to a couple of friends in the UK. You know that place where their version of the BOM forecast a dry winter.

    One of them is half way through cleaning up the house after the flood, the other still has about a foot of water in his place.

    Both are hoping their Met will be on its usual form, & completely wrong in their latest forecast. It’s for even more rain.

    There they have got so bad, they no longer prepared to issue long range forecasts to the public, oh apart from the one that says we are all going to fry in 100 years.

    It would be funny if it weren’t so expensively ridiculous.

  38. Luke February 24, 2014 at 1:00 pm #

    Debbie you’re being thick as a brick. ! It is not semantics. It is a mathematically precise statement of fact. Why don’t you engage your brain and read what I’ve said again.

    You’ve been told – if you continue to wilfully misrepresent what’s there – your problem mate. You don’t read glossaries Debbie – waste of time.

  39. Luke February 24, 2014 at 1:02 pm #

    Actually it just dawned on me Debs. You have absolutely no idea what you’re looking at. None !

  40. hunter February 24, 2014 at 1:57 pm #

    Clearly the CO2 obsessed models are working no better than CO2 obsessed energy/water management/flood control/temperature predictions/etc., etc. etc.
    The inability of the GCM’s to make meaningful regional or seasonal forecasts, along with their inability to make decade range forecasts is well established.
    The CO2 obsessed know it and seem to hope that if they can stop all alternatives, then somehow their fixation will become real, sort of like Pinocchio.

  41. Debbie February 24, 2014 at 2:48 pm #

    Luke,
    Of course I know what I’m looking at.

    :-) :-)

    A bad case of ‘excusititis’.

  42. Noosa Girl February 24, 2014 at 2:57 pm #

    Luke,

    You should definitely put your manners in. You are very rude.

    Individual GCMs such as POAMA generally produce poor skill for medium-term rainfall in Australia. Compared against climatology, POAMA is generally of poorer skill using RMSE or correlation coefficients as a quantitative measure. This means that schoolgirl in Haiti equipped with only a pad and pencil can beat POAMA. All she needs to do is add up 30 numbers and divide by 30.

    So the POAMA people cry out for more even more funding for supercomputers and average the output from “ensembles” of GCMs – they average the output from a few dozen GCMs and call in an “ensemble” to make it sound sophisticated. This may give you some improvement (at enormous cost) but it becomes very difficult to relate physical processes to an output if you are averaging 36 GCMs, each with different characteristics. So the justification for using a GCM in the first place is looking rather flimsy.

    So the POAMA people still can’t match a neural network, so they get some more government funding and combine the results from the ensemble with a statistical model. Some improvement in skill, but the justification for using the GCM is now even more flimsy.

    Meanwhile, the 8 year old girl in Haiti has sharpened her pencil and does some more long division, dreaming that one day she will be able to afford a calculator.

    Meanwhile, back at the Bureau………………

    Noosa Girl

  43. bazza February 24, 2014 at 4:04 pm #

    Jen, your response to clarify your strong claim of how your work “ also directly challenges the theoretical basis of AGW” has a weak perhaps as follows “Perhaps because the climate is not on any new trajectory, rather patterns in historical data are still relevant and can be used for future rainfall forecasts.”
    I can more logically state with no need for any perhaps, and no challenge “Because the climate is on a new trajectory, patterns in historical data are becoming less relevant but can still be used for future rainfall forecasts. However users will have less confidence as accuracy declines.”
    Anyway your most recent paper uses temperature data for Queensland stations which of course contains an AGW trend. Check since 1970 particularly and for spring when the forecast is likely to be of most value.

    So you will be able to say your forecasts incorporate climate change – and no perhaps about it.

  44. Ken Stewart February 24, 2014 at 4:27 pm #

    Hi Jen

    My interest has been roused once again- especially that you hope to improve skill at forecasting ENSO, which as you know is one of my interests (whether heavy rain precedes a change in SOI by some months). Having perused these comments, I shall now have a good look at your paper. Best wishes but don’t hold your breath waiting for support.
    Ken

  45. Another Ian February 24, 2014 at 6:07 pm #

    Jen,

    Repeated from

    http://diggingintheclay.wordpress.com/2014/02/17/place-your-bets/#comments

    “Another Ian says:
    February 23, 2014 at 8:41 pm
    Verity,

    Last night I was reading “Red sky at night, Michael Fish’s satellite is on fire” in “The Complete World According to Clarkson”. The opening sentence -

    “I rang the Met Office last week and asked something which in the whole 149 years of the service it has never been asked before. “How come” I began, “your weather lorecasts are so accurate these days?”.

    This was asked in 2003 – should the Met Office be petitioning for their old computer back?”

    And check her reply too – and note my typo.

  46. jennifer February 24, 2014 at 6:40 pm #

    Bazza

    We have a paper in the pipeline predicting rainfall for the Bowen Basin under different climate change scenarios.

    It is clear, however, from the data that while there are climate cycles, there is no evidence of dramatic recent warming as suggested by Flannery, BOM, the ABC and you.

  47. Luke February 24, 2014 at 7:06 pm #

    Dear Noosa Girl aka surfie chick – yep am rude – but this is a matter of some pith and moment and so one may become somewhat robust (and I used to be nice but others here aren’t so one adapts. I am quite nice in real life though as is Bazza. Especially if you rub his chin.)

    “So the POAMA people cry out for more even more funding for supercomputers and average the output from “ensembles” of GCMs – they average the output from a few dozen GCMs and call in an “ensemble” to make it sound sophisticated. ”

    errr WRONG they don’t actually – an ensemble is the same run of the same GCM with slightly different starting conditions to sample the chaos in the mathematics and the real world. It’s not a few dozen GCMs…. In this case it is not a multi-model average used in some climate change work.

    As for GCMs matching a neural network – well who knows – I am presented with a paper that only has two tests against POAMA 2.x and only one statistic. The POAMA 1.5 comparisons are against a devised rainfall index which BoM don’t use … so what have we proved I’m not sure.

    The reasons for using a GCM may be to sample some combination of unique events hitherto unseen. And might especially relevant in a neutral year drought not well predicted or managed by ENSO related indices. Don’t you think?

  48. Luke February 24, 2014 at 9:07 pm #

    No warming trends in Queensland stations?

    Burketown Max
    Burketown Min
    Georgetown Max
    Georgetown Min
    Richmond Min
    Cardwell Min
    Innisfail Min
    Burkekin Max
    Burdekin Min
    Mackay Max
    Mackay Min
    Charters Towers Min
    Isisford Min
    Camooweal Max
    Camooweal Min
    Boulia Min
    Bundaberg Min
    Gayndah Min
    Sandy Cape Min
    Cape Moreton Max
    Cape Moreton Min
    Maryborough Min
    Brisbane Min
    Tewantin Min
    Dalby Min
    Goondiwindi Min
    Cunnamulla Min

  49. Another Ian February 24, 2014 at 9:23 pm #

    Luke

    “The reasons for using a GCM may be to sample some combination of unique events hitherto unseen.”

    Like this?

    “REDBACK

    UNDISCOVERED RICHES: The econo-
    mists at the Canberra-based Austral-
    ian Bureau of Agricultural and
    Resource Economics (ABARE) have
    been tucking into some funny mush-
    rooms judging by item 1003 on their
    1994-95 research program. Project
    1003 aims to “develop a method of
    assessing the value of undiscovered
    mineral resources in a broadly
    defined mineral province as a basis
    of quantitative resource assessment”.
    Listen fellas, how can you assess the
    value of something that is “undiscov-
    ered”? Redback is eagerly awaiting
    an explanation from ABARE’s exec-
    utive director, Brian Fisher.

    Australian Farm Journal Nov 1994″

  50. cohenite February 24, 2014 at 10:00 pm #

    Bazza says:

    “Anyway your most recent paper uses temperature data for Queensland stations which of course contains an AGW trend. Check since 1970 particularly and for spring when the forecast is likely to be of most value.”

    What AGW trend?

    Luke I gave you the link to the post here where you heartlessly attacked poor Neville about the IOD, and I have given you the Ummenhofer et al paper which you did comment on at the Jennifer link; but here it is again:

    http://web.maths.unsw.edu.au/~matthew/Ummenhofer.etal_2009_SEA.pdf

  51. Luke February 24, 2014 at 10:43 pm #

    Well why wouldn’t you attack Neville. I have a reminder on my fridge (which is wind powered of course).

    Now let’s see – the clue might be in the last letters of the Ummenhofer file which do not seem to read N or NEA.

    Another Ian – I surrender. Like economists’ drought losses I guess – wool unshorn from lambs unborn.

  52. hunter February 24, 2014 at 11:37 pm #

    The apologists for AGW are trapped in circular faith-based style thinking:
    There is an AGW trend, therefore history is out the window.
    Yet they rely on short term regional records to prove it up. Records of dubious quality and that have been adjusted, etc.
    It it is not, for the record, at all established that there is an AGW trend that is discernible from historical norms at all. What is established is that the models based on the AGW premise do not work. Dr. Pielke, Sr. and others have demonstrated that. Reality demonstrates this worldwide: AGW based predictions are crap.
    But the apologists simply ignore this pesky inconvenience and bash alternatives, instead of dealing with the failure of their own icon.

  53. Debbie February 25, 2014 at 6:51 am #

    Noosa girl,
    Does he work for the govt?

    :-) :-)
    Based on my observations from the evidence (comments) at this blog I would give a 99% probability that the answer is yes.
    However, according to Luke’s view, even if the answer is no. . .I am still not wrong because there was a 1% probability that he doesn’t work for the government.

    :-) :-)
    Of course. . .all of that is immensely useful for you. .

  54. Luke February 25, 2014 at 7:46 am #

    I love it the way our US mate hunter pops up and makes one of those great content free speeches

  55. Johnathan Wilkes February 25, 2014 at 9:00 am #

    If this claim about changing climate patterns due to AGW is right, thus invalidating any attempt of forecast based on known measured records and weather patterns, then we can safely discard any study and conclusion drawn on proxy records of distant past.

    Climate had wild swings from warm to ice and back again for reasons unknown to us, so why do we believe the “hockey stick” or ice core samples for example?

    Just a question as I’m not sure why the claim was made, patterns are patterns, trend is trend or so we are told here incessantly.

  56. bazza February 25, 2014 at 10:18 am #

    The dreaded drift from denial to delusion – what trend, pines Jen?, what AGW trend/ chirps cohenite. Ask your colleague Nev, even he opines for a climate sensitivity of about one, albeit an outlier .
    Moving right along, cohers did want to bet on jens neural net forecasts. So If cohenite is a serious fan of the A&M model and as a Bayesian presumably, he will offer me odds as deduced from the performance to date . As Jen proclaims “We had a lower RMSE for 16 of the 17 sites “. So offering me any odds less than 16:1 would be a vote of no confidence in A&M. What odds Cohers?

  57. Glen Michel February 25, 2014 at 10:53 am #

    What AGW trend??….. None really folks,but bazza and Luke have convinced themselves there is one;others more discerning or have no incentive to believe what has not been PROVED will treat such with disdain.BoM is about useful as a short-sighted projectionist at the pics .Just maybe Luke can unpixellate himself and get a window installed.

  58. cohenite February 25, 2014 at 10:56 am #

    bazza the punter; Jennifer doesn’t what her site to become a betting forum on account of the riff-raff that will be attracted; that is, above and beyond the riff-raff already loitering around the place.

    So, with that in mind, we should be able to look first at the BOM prediction record. Now Jennifer started her work in 2011 so we could first look at BOM’s prediction record for that year; how did BOM go:

    http://www.warwickhughes.com/blog/?p=931

    My bull mastiff could do better than that! Anyway I’m up however many points you reckon are fair.

    Now, your turn bazza, point out a successful prediction by BOM.

  59. bazza February 25, 2014 at 11:59 am #

    BOM Annual Report 2012-13 Seasonal Forecast Success rate 65%.

  60. Luke February 25, 2014 at 12:15 pm #

    Amazing Glen Michel – I’ve just given you 26 warming trends above. PLENTY REALLY !

    Cohenite wisely lays smoke and tracks to port…. and throws out a rabbit – we’re not talking about previous BoM statistical systems probably compromised by an AGW-ish warming Indian Ocean.

    Cohers, POAMA was started in operational seasonal forecast mode May 2013. IMO it hasn’t been on the money this season. Nobody else has either (one of them thar hirtherto unencountered beasties perhaps).

    So curious if Jen has been on the money, but she tells me she’s not operational – how tedious and boring !

    Ya gotta be operational to spruik it.

    Cohers Like Debs you also need to do uncle Bazza’s stats 101 course – who know he may have even taught such classes when he wasn’t running the Darwin knock shop.

  61. Luke February 25, 2014 at 12:16 pm #

    Bazza that’s no good – Debs needs 100%

  62. Glen Michel February 25, 2014 at 12:21 pm #

    BoM spring outlook2013 – bullish at 80% locally SSTs were never going to support their projections.Covered their arses quickly when real time caught up with them.50:50 is the best that the bureau puts up.Heads or tails?

  63. sp February 25, 2014 at 12:41 pm #

    Bazza: “BOM Annual Report 2012-13 Seasonal Forecast Success rate 65%”

    I have a 50% chance of getting it right by flipping a coin – how million dollars did BOM spend to get it to 65%?

  64. bazza February 25, 2014 at 2:35 pm #

    What extraordinary irony – no gambling here says Cohenite at a blog preoccupied with downplaying the risk to the planet from AGW.

    What ignorance from sp shooting off on seasonal climate funding without bothering to research the millions that farmers have invested over 2 decades , and saved many times more from better informed decisions such as whether or not to invest or defer, for example by a grain farmer in a new header with a forecast of 65% chance of a lower than average spring rainfall.

    What blind loyalty from Glen Michel not taking the obvious cue to ask Jen what % accuracy she is offering not knowing that she will respond that is too simple a metric and someone might understand it.

  65. Debbie February 25, 2014 at 4:24 pm #

    Which system was BoM using in that 2012-13 report?
    BTW, the seasonal forecast for winter/spring 2013 from BoM was 80% chance of wetter than average.
    Not sure how you want to work that in to your farmer argument?

  66. Robert February 25, 2014 at 5:38 pm #

    Just as an example of real-life complexity and the dangers of literal thinking:

    Since 1882 my region’s wettest autumn by far occurred in the latter stages of a neutral year, leading into an El Nino (which was wet enough). In fact, our rainfall records won’t neatly fit any theoretical pattern based on ENSO. The two biggie El Ninos – 1982-3 and 1997-8 – gave rainfall around or above the mean…and that rainfall was considerably higher than in the super-double La Nina of 1916-17! There’s value in knowing about SOI etc year to year, but it’s limited value. Pretending that the great dry periods of that pre-1950 era – independent of El Nino etc – never happened will take some stupendous spinning. (I know, I know…there are some stupendous spinners around here.)

    My guess is that there were more things stifling Eastern Australia’s rainfall in the first half of the twentieth century than we know much about. People have started to give names to these other influences – let’s hope they treat them with a sane mix of curiosity and skepticism.

  67. cohenite February 25, 2014 at 8:05 pm #

    Stop griping bazza and provide a link.

Website by 46digital