Betting

NH Ice Bets: July 2013

97 Comments

Better too late than never, it’s time to bet on the Jaxa 7-day minimum! We’ve run bets every year and there is no good reason to miss this year. On the one hand, we might see a recovery relative to last year; on the other hand, we observed dramatic losses around this time 2007. So, who knows?

For bets:

  1. We will use data available for download at JAXA. Fresh plots are constantly available
  2. I will compute the 7 day running minimum. Everyone likes the winners called but oddly, we can’t be certain of a 7 day minimum has been reached until it’s well and finally over. So, at the risk of calling a minimum too early: If we will call the current a minimum on Oct 3, 2013 at latest. (If the minimum occurs after this, we will be running fresh bets because… oh. Wow!). Also, we will call a minimum, if the 7 day running minimum rises for 21 days straight should this occur prior to Oct 3, 2013.
  3. Whoever gets closest to the minimum 7 day average wins the quatloos.
  4. I have not yet decided on my bet! But note: you should be entering numbers like 5.xx, not 5000000. Though the script will permit you to bet there will be -2 millions square kilometers, you should bear in mind that whoever bets 0 will beat you and take the quatloos.

[sockulator(../musings/wp-content/uploads/2011/UAHBets5.php?Metric=NH Ice Extent?Units=x 106km2?cutOffMonth=7?cutOffDay=16?cutOffYear=2013?DateMetric=NH Ice Weekly Ave Minimum?)sockulator]
The cutoff day is entered as 7/15/2013, which really means you need to bet by midnight 7/14 (in some time zone, somewhere.)

97 thoughts on “NH Ice Bets: July 2013”

  1. There is an SI prefix that eliminates the need for the clunky 10^6km^2. It’s Mm².

    The bottom could still fall out this month or next, but I’m betting on a repeat of 2008 vs 2007, although NOAA near real time had a huge drop in extent on 7/1 from 10.14885 on 6/30 to 9.76338 Mm². They also revised the entire month of June downward. But neither MASIE nor JAXA had anything like that decrease for their data. The average ice concentration calculated from Cryosphere Today area and JAXA extent is also higher than it was at this time last year.

  3. I based my bet on the proven method of blindfold and dartboard. The cat and dog hate it when I do that.
    .
    I bet my entire life savings. But if 42 is the answer to everything, 42^2 should be even better.
    .
    Papa needs new shoes!

  10. Nick Stokes (Comment #117522)
    July 3rd, 2013 at 6:03 am
    “Speaking of dramatic losses – Jaxa lost 0.219 Mm^2 overnight.”

    How much overnight is there in the Arctic at this time of year?

  11. michael hart (Comment #117546)
    How much overnight is there in the Arctic at this time of year?

    Apparently, a lot. Another 0.15mM^2 last night. It’s the hot midnight sun.

  12. Climate Avenger (Jul 3 15:55) —

    Thanks for the link to the informative animation, showing changes to both sea ice extent and sea ice depth over the past month.

    I’m even more impressed by the top counter, which indicates that June 12th to July 11, 2013 is the interval on display!

  14. It’s going to get cold soon! Sorry I just had to do that. Its going to get colder soon but then its going to get wamer again later. Repeat as required.

  17. Speaking of water 🙂 …

    Does anyone know of a web site that has global average precipitation beyond the year 2000? If found a couple that go to 2000, but for some mysterious reason, there are none that I can find that extend past 2000.

  20. Both Cryosphere Today area and PIOMAS volume say that this year the ice isn’t melting anywhere near as fast as last year. The AMO index in June was a lot lower than it was in 2012. The UAH NoPol anomaly has been lower this year than last. Export of ice through the Fram Strait also appears to be lower. All this points to no new record Arctic sea ice lows this year.

  21. Ray (Comment #117608)
    July 9th, 2013 at 12:07 pm
    “Off topic but the UAH betting entry is closed.

    UAH published and up from 0.083c to 0.298c.

    Together with change to v5.6 due to the addition of two satellites.

    http://www.drroyspencer.com/”

    That was close then. Only 0.002c adrift with a mark one human eyeball!

    —

    Richard LH (Comment #115617)
    June 12th, 2013 at 7:50 am
    +0.3 based on this predictor.

    http://s1291.photobucket.com/u…..9.png.html

  22. RichardLH,

    “That was close then. Only 0.002c adrift with a mark one human eyeball!”

    Yep, you should scoop most of the pot this month.

  23. I give it away.

    I am more insterested in informed opinion about what could be next. Me, I’ll pick +0.29c because I think it will be about the same, possibly slightly lower again next month.

  24. Ray (Comment #117614)
    July 10th, 2013 at 7:10 am

    Now if you take into account that they also adjusted what I was aiming for slightly as well, after the bet?

  26. Go, I give up. Posted to WUWT thread.

    I tell you my fathers Nick name will stick to me forever. No matter what I do.

    RichardLH says:
    July 10, 2013 at 9:31 am
    Roy: You do realise that I predicted the figure a month ago (shot my arrow), you moved the target slightly (wind) and I still hit the ‘bull’s eye’ (to within 0.002c – I think I can call that less than the width of the ‘arrow’ ), I come from England, And my intials are R(L)H. Now do I really have to give you my last name? You would be right as well! You just cannot make this stuff up!

  27. And if you want the topper. From the tail end of the interchange at Tamino’s. I tell you, Holywood could never write this stuff.

    Please, please, someone do the cartoon! 🙂

    RichardLH | July 10, 2013 at 11:14 am | Reply [edit]

    [Response: Enough already.]

    RichardLH | July 10, 2013 at 4:40 pm | Reply [edit]

    [Response: I’ve been more than patient with you. I even gave you a “parting shot” already, a final comment after I said “It’s time for you to find another outlet.” Three times I’ve told you we’re not interested in hearing more from you. This is the fourth. Goodbye.]

    RichardLH | July 10, 2013 at 5:20 pm | Your comment is awaiting moderation.

    OK.

    My name is now officially ‘Robin Hood’, Aka RichardLH. And, Yes that last name IS correct!

    I shot an arrow a month ago (I made a prediction of the UAH Global temperature figure for this month and laid my ‘cheat sheet’ out on the web).
    I came here asking nicely if my my caclulations could possibly be correct.
    You said you were busy discussing something else much more mathmatical and to ‘run along boy’ and that nothing I could say was of interest to anyone.
    My ‘arrow’ hits the UAH ‘bulls eye’ to within 0.002c (I think we can call that the width of the ‘arrow’.
    I warn you that I hit it and there is an outside possibity that my maths is right all along.

    And you delete all of my comments.

    I claim the prize!

    Where’s a good sherrif when you need one!

  29. I will ask the question here as well. I look forward for all opinions. It is simply this:

    “Can you please give me a logical or mathematical reason why other more complex methods, based on sin, cos and statistics (higher mathematical functions), should be more accurate predictors to short term Climate futures than this simpler method, which is based simply on the display of long and short term averages of the same data and, given that this new presentation of that data still shows that any long term CO2 trend or other such long term factors are not yet confirmed or determined, that it might also better allow their RMS power ranges and time periods to be better estimated or established when using this method instead?”

  30. RichardLH:

    My name is now officially ‘Robin Hood’, Aka RichardLH. And, Yes that last name IS correct!

    By pure luck I was able to confirm this with photographic evidence that Richard and Robin are one and the same.

    I figure England having all of those cameras helped in my search.

  31. Carrick (Comment #117622)
    July 10th, 2013 at 5:16 pm

    RichardLH:

    My name is now officially ‘Robin Hood’, Aka RichardLH. And, Yes that last name IS correct!

    “By pure luck I was able to confirm this with photographic evidence that Richard and Robin are one and the same.

    I figure England having all of those cameras helped in my search.”

    I’ve been framed!

  32. Carrick,
    Thanks for the link to the model tuning paper.
    .
    One may appreciate how sausage is made, but still find the process horrible when actually observed. I appreciate the honesty of the paper’s authors… but I am still more than a bit horrified by the process. The most troubling part is how ad-hoc it all is; lots of important behaviors are simply ‘tuned’ repeatedly, trading off one behavior against another. So much more-or-less arbitrary human input (over generations!), so little apparent concern with basics like air tight energy conservation…. not to mention ‘adjusting’ cloud properties/coverage willy-nilly to match any specified TOA imbalance. There was not a lot of discussion of code quality control either. Ugly may be too kind a description for the process. It is clear that the modeling of Earth’s climate remains in its infancy.
    .
    The only part of the paper where I think the authors ducked an important issue was their claim they did not tune aerosols to match historical temperatures… the same claim Gavin has made more than once. Please, spare us this nonsense. Of course the models are tuned with aerosols… and why not? They are tuned in many other ways.
    .
    If anybody reads that paper with even a moderately critical perspective, it is hard to imagine they would afterwards think the models can produce credible projections of temperatures over 50, 100, or 200 years.

  34. DeWitt Payne (Comment #117625)
    July 11th, 2013 at 5:37 am
    RichardLH,

    “The link in your post http://rankexploits.com/musing…..ent-117613 is broken.”

    Thanks. Sorry. See below for correct links.

    http://i1291.photobucket.com/albums/b550/RichardLH/uahtrendsinflectionfuture_zps7451ccf9.png

    http://i1291.photobucket.com/albums/b550/RichardLH/UAH-Comparisonofcascadedlowpassfilteroutputmonthsrunningaverage_zpsf1883e44.png

    and riddle me this:

    Why is that it is considered that complex methods, based on higher mathematical functions (sin, cos, stats, etc.), are considered to be more accurate than a much simpler method, which only uses lower mathematical functions (long and short term averages), and when this new peresntation is also capable of determining the RMS power in each period/frequncy band and displaying it?

  35. Especially when I got it right to two decimal places from a month ago (even with a slightly moving target – which also means I could have been WAY out – I know that). But definitely pure luck. No skill or observation what so ever. I mean, what ELSE could there possibly be?

    Go away and play little boy..

  36. SteveF,

    “If anybody reads that paper with even a moderately critical perspective, it is hard to imagine they would afterwards think the models can produce credible projections of temperatures over 50, 100, or 200 years.”

    I did read the paper and came away encouraged by the author’s oft-stated recognition that at the present time gaps exist in data/theory that require an empirical tuning process. They were clear on the dangers of compensating errors when tuning several poorly-understood variables at the same time.

    Models are far from perfect – in light of the overwhelming complexity of the climate system, I certainly never expected them to be otherwise. Your statement above has a strawman element to it. Imperfect as they are, they are still one of the best tools available for understanding climate.

  37. Owen not all models are perfect, but not all models are useful either. 😉

    “The best we’ve got” arguments work better when you have a model that you *know* is useful.

    I suspect if anything they underestimate the effects of tuning because of the ad hoc nature in which they varied the parameters. Ideally you’d use a Monte Carlo method to establish the full range of parameter values that give you results consistent with data.

    That’s not possible here, so it’s a real problem with establishing the reliability of the models, which goes back to the question of “how can you know it’s useful?”

    What makes this parameter tuning problematic is they are varying physics approximations as well as parameters. In many case the approximations are really primitive with no physics justification.

    What makes this paper unique is a partial willingness to admit to the fact they are tuning. Clearly aerosols is a tuning (so is solar forcings history), so the claim that it isn’t is odd.

    Changing physics approximations is a sort of model tweak too (just not a continuous one). They need to investigate how varying that affects the results too.

  38. Richard:

    Especially when I got it right to two decimal places from a month ago

    Just by luck I’m afraid. The reasons sinusoidal basis functions work better here is because the underlying physical system has multi-scale oscillatory behavior. The coupling is also nonlinear, so any attempt to forecast using a linear combination of any basis functions is going to fail.

    I’m afraid for once I’m sympathetic to Tamino.

  39. Carrick,

    Is there a better way (than modelling) to understand the complex set of interdependent processes in the climate system?

  40. Carrick (Comment #117630)
    July 11th, 2013 at 9:57 am
    Richard:

    “I’m afraid for once I’m sympathetic to Tamino.”

    Oh I would have too. I stumbled acros this view completely by accident. Hence my attempts to get someone to explain it.

    Take the standard, well know, 3 pole running average low pass filter. Unzip to extend to a whole cascaded filter bank. Observe exact similarility to well know analogue cirecuit long used in most engineering power transfer work and oscilating systems.

    Deduce that this is the ‘perfect’ digital verisions of its analogue base. It has perfect band allocation characteristics. When unzipped to its full extent, the ‘digital’ errors are very low. All energy will end up correctly allocated somewhere in one band or another. Also there are NO phase errors AFAIK. Because it is not a power 2 series, all the harmonics of the main frequency are untouched.

    In nature, power is not distributed (HF end) by single cycles, rather a mixed combination of ‘tone’, harmonics, etc, all in half cycle chunks (other than the main ‘tone’)a nd loosly constrained to fit some overall ‘beat’ pattern that correctly distibutes energy.

    Once you then realise that 1 day = sampling requency and…..

    Really the rest just follows. The final section in a bandpass filter is ‘ground’. The central, long term (millennia?) temperature that this patch of whatever will achieve after all that frantic scurrying around overhead.

    A standard power circuit, used in an amonymuos way on temperature data to reveal the underyling transfer staucture (time wise) and with te ‘prefect’ digital series.

    Still feel sorry?

  41. Oh, and this is what I do for fun!

    Medium/Low Head Hydro/Waves
    http://i1291.photobucket.com/albums/b550/RichardLH/Lowspeedgravitypoweredwaterengine_zpsbbe0d7c9.png

    Solar heat
    http://i1291.photobucket.com/albums/b550/RichardLH/solar_zps30776dd1.png

    Wave
    http://i1291.photobucket.com/albums/b550/RichardLH/waverpower_zps6822f49e.png

    Wind, Tide and River Flow
    http://i1291.photobucket.com/albums/b550/RichardLH/WindPower_zps4538f6f2.png

    Personal
    http://i1291.photobucket.com/albums/b550/RichardLH/HandPowergenerator_zps1edbc1e7.png

  42. Now does everbody understand by tentaive questions about next = round(1.3371 * previous)?

  43. and i know my posting style is not to everyones satisfaction, but at least you can read what I am saying on a single page/paragraph on the screen, whilst also allowing other to also ‘talk’ round me. So here is the last for now, I promise.
    —

    If you were to setup a lab, with a radiant heater on one wall of a vacuum chamber, would you not setup a single RMS power meter, un-synchronised to the rotation speed of whatever globe you placed at the center to monitor the overall behavior? As a naturally smoothed output function. To keep the monitoring costs lower.

    Now think big RMS meter and power distribution curves and sythersising the viewpoints from the Sun and Moon (probably the best place to put the single observatory). This just happens to be a ‘looking out from the inside’ Globally point of view. You could probably do it cheaper the other way.

  44. Owen,
    “Imperfect as they are, they are still one of the best tools available for understanding climate.”
    .
    Humm… I didn’t say anything about if they are useful tools to help understand climate. I said that there is no reason to believe they can make accurate projections of warming over 50, 100, or 200 years. But sadly, that is exactly what they are being used for… and more often than not, to justify demands for specific public actions to force costly reductions in fossil fuel use. The models seem to me at best a kind of messy research tool, with unverified code and long term ad-hoc tuning of lots of parameters to match target data. What they are not is something upon which to rationally base public energy policy.

  45. Owen:

    Is there a better way (than modelling) to understand the complex set of interdependent processes in the climate system?

    Well, you can look at effects that involve short-time-scale mass transport without having to nail ECS, which depends on effects that take thousands of years to fully develop.

    So when you say “a model is/isn’t useful” you have to specify for what.

    SteveF as he points out was thinking of the utility of these models for setting climate policy. That’s a very different application than e.g. studying sudden warming wintertime stratospheric events.

  46. SteveF,

    ” The models seem to me at best a kind of messy research tool, with unverified code and long term ad-hoc tuning of lots of parameters to match target data.”

    Sounds like a pretty shady enterprise.

  47. Owen,
    The only thing shady about it is inappropriate enlistment of very uncertain (and yes, hand tuned) models for political advocacy. Climate science was a research backwater 30 years ago, and it is still conducted as if it were. The advocacy has gottten way ahead of the science.

  48. Put on your science hats.

    Question: what is your best estimate for the temperature in 2100. What tools would you use and why?

    Forget the purpose of the exercise or how people will use or misuse your answer. Approach this as an apolitical, academic, I just care about have the best prediction, science guy.

    Hint: saying you dont know will lose all bets.

    So, as a person who doesnt care about how others will use or misuse your answer, how would you estimate, what tools would you use and why?

    Simple.

    Next, people’s opinions about what policy makers consider useful information isnt very germane. You dont get to decide what policy makers find useful. They do.

  49. Steven Mosher (Comment #117640)
    July 11th, 2013 at 9:51 pm
    Put on your science hats.

    “Question: what is your best estimate for the temperature in 2100. What tools would you use and why?”

    I would create a long term averaging, time distributed, RMS meter of the Earth’s output function. As we know the RMS input function, we should then be able to characterise the Response more correctly.

  50. And, I just can’t resist it. At least my ‘cheat sheet’ (by pure lucky chance?) gotme to within 0.002c from a month out on UAH with what maybe just such a tool (see many, many, Tigger type posts).

    Now I don’t know if I was lucky or right. Please someone tell me.

  52. Well They COULD be checking the ‘cheat sheet’ I left around when I said my arrow was on target, but that would be IMPOSSIBLE! It wasn’t just a lucky guess. Better check JUST in case.

    I really, really don’t know at this point how lucky I was. Anyone can shoot an arrow.

  53. Mosh,
    I would make an estimate of TCS based on a heat balance using the best available data, then look at how much radiative forcing is likely to change. The first part I have already done (about 1.2C per doubling). The second part is more speculative, since it depends on numerous factors that range from politics (will the nutty greens continue to insist on only the least practical energy sources?) to the rate of technological progress and the rate of capital investment in energy infrastructure and efficiency, neither of which is independent of technological progress. As a SWAG, I’d estimate net forcing to increase by 2.5 watts/M^2, and temperatures to increase by about 1-1.2C compared to today, dominated by warming at high latitudes in the northern hemisphere.
    .
    Now I will take off my science hat and ask: what is the point of your question?
    .
    The point of noting advocacy in climate science is that the field doesn’t consist only, ir even mostly, of people wearing their science hats….. seems to me a fair fraction of their time they are trying to be policy makers instead of scientists.

  54. “Question: what is your best estimate for the temperature in 2100. What tools would you use and why?”

    Near the max for the last 400,000 years shown in Hansen 2011 which is about 1 degree warmer than now. It is clear (to me) looking at the long term graph in Hansen 2011 that negative feedback gets steadily stronger around our current temperature such that temperature appears to be capped a little warmer than at present. Hansen, of course, disagrees with my interpretation of his data.

  55. Carrick,
    Should you return:
    I did a quick evaluation of the current return on investment for roof-top solar panels which are grid-tied but without battery backup. In states with favorable net metering rules but fairly reasonable per KW rates (like Florida), the current return in reduced power cost looks like $1500 per year on a capital investment of about $20,000 (installed cost, maybe $12,000 if DIY), or 7.5% per year. Since there is a federal tax credit of 30% of cost, the return might reach 10-11% per year. What sours the deal is the system depreciation and maintenance, which is probably in the range of 5%-6% per year. A net ROI of 4% to 6% won’t get too many people fired up to install systems. If someone faces higher rates and lives in a sunny location (California?), the ROI could improve quite a lot, so long as favorable net metering rates apply… but who knows if they will change?
    .
    Battery based (off grid) systems still are impractically expensive compared to grid-tie systems… Having to replace the batteries every 5 – 7 years is a huge additional expense, and the system has to be bigger to be practical.

  56. SteveF (Comment #117647)
    July 12th, 2013 at 6:31 am

    Was it just a really, really lucky guess on my part then, no skill or tentative reasoning?

  57. Steven Mosher (Comment #117640)
    July 11th, 2013 at 9:51 pm
    ——————
    Obeying the constraints you’ve supplied, I’d use the dang GCM’s.

    I still don’t understand why it’s not ‘science’ for me to acknowledge that I’m probably going to get the wrong answer for 2100, but then again I’m a software engineer and not a scientist. If I expect my best science to give me the wrong answer, what was the point? Maybe you could elaborate on this some time.

  58. Steven Mosher (Comment #117640)

    “Put on your science hats.
    Question: what is your best estimate for the temperature in 2100. What tools would you use and why?
    Forget the purpose of the exercise or how people will use or misuse …
    …
    what policy makers find useful. They do.”

    Interesting comment. Taken as a whole this sounds much like what I would expect politics guy (manager guy) to say to science guy. ‘Give me a number, your best number. ( I don’t give a bleep about your uncertainties, I need a number.)’ Of course as time goes on manager guy’s landscape changes, e.g., uncertainties become evident and stubbornly confound things, and it is down hill from there. As a poorly stated ‘problem’ to begin with it never stood an ice cube’s chance in hell. It’s a tough business with deja vu smeared all over the place.

    If I were science guy I would either walk away to begin with (wiser days), or jump in with an acceptance of the quixotic nature of the whole affair–hoping to nudge the decision process into more rational waters (younger days).

  59. ibid ;o)

    My answer above shows a certain reluctance or inability to be apolitical. That may be because of the the original question: what is the [my] best prediction for the temperature in 2100? [Note here that I assume that you mean some measure of a global temperature.]

    If I were a science guy and interested in the science of global warming [or not], I would have much more pressing interests at this time than giving anyone an 2100 estimate…the science [measurements, mechanisms, etc.]

    As a practical matter in today’s circumstances the question has political facets. But if I could strip away those facets and look at the bare question I would have to I would conclude: not now, it is too early to expend much in resources on that ‘odd’ question; it is a minor detail.

  60. Steven Mosher (Comment #117640)
    July 11th, 2013 at 9:51 pm
    ————————–
    My guess is +0.90C warmer than today. This is simply a +0.10C/decade trend calculated thusly:
    1) Forcing was estimated to increase linearly at 0.02W/m^2 per year. IIRC correctly, I simply calculated the recent trend of anthropological forcings from the GISS data and used that. This is similar to the 1% to 2x experiments, only my doubling time is 3.7/0.02 = 185yrs instead of 70yrs. I think this is a reasonable option. I give equal weight to the probabilities of acceleration and deceleration.

    2) The response was simply calculated by extending Willis’s CMIP5 Blackbox model (ala Lumpy) to the year 2100 and plugging 0.02 into the delta F column. I don’t necessarily think a one-box model is the best approach, but it’s probably ok in the short term (i.e. <100yrs) and I don't trust any models long term.
    I have my own model that modifies the pipeline accumulation function slightly to produce a fast/slow response. By selecting different parameter values, date ranges, and data I typically get a trend in the 0.10C/dec to 0.15C/dec range. So the +0.9C value I gave is on the low end of my range.

  61. >>>Question: what is your best estimate for the temperature in 2100.
    >>>What tools would you use and why?

    Persistence is not a best forecast, but it is the recommended -STARTING- forecast,
    modified only by identifiable factors.

    The MSU era trends ( various SSTs, NCDC, CRU, GISS, UAH, RSS ) vary from 1.4 to 1.6 K /century.

    Therefore, our starting forecast for 2100, based on persistence of trend, is 1.5 K warmer than 2000.

    What other factors are we confident in that would modify that forecast?

    >>> Hint: saying you dont know will lose all bets.

    Perhaps not betting is the only honest approach.

    To bet is to fool ourselves that:

    – we know how albedo varied in the past, what albedo is presently, and how it will vary

    – what insolation will be
    ( whatever it will be, I’m pretty sure it won’t be flat constant as the IPCC models:
    http://igloo.atmos.uiuc.edu/IPCC/FORCING/solar.constant.png )

    – what the Volcanic Explosive Index will be

    – what human population will be
    ( pretty big range:
    http://populationaction.org/wp-content/uploads/2012/01/asset_upload_file274_13152.gif )

    NOAA keeps a GHG index which exhibits a fairly constant trend ( slightly decelerating ):

    http://www.esrl.noaa.gov/gmd/aggi/aggi_2012.fig4_med.png
    http://www.esrl.noaa.gov/gmd/aggi/

    This is pretty remarkable given that this period covers the massive industrialization of China.

    Other factors aside, if GHGs are the major forcing, the constancy of trend adds weight to a persistent trend forecast.

  62. “If I were science guy I would either walk away to begin with (wiser days), or jump in with an acceptance of the quixotic nature of the whole affair–hoping to nudge the decision process into more rational waters (younger days).”

    most skeptics play the walk away game. That is a losing proposition because some science guy will always come forward to make his estimate. Hell I dont even need to ask, I can just say Arrehenius and work with his number if you refuse to play the game.

    By playing the doubt game skeptics have really ceded power to the people who are willing to lay down a number. Contrast this with Nic Lewis who agreees to play the game. Put another way, if you want to influence the answer you have to work inside the system.

    Skeptics make the fundamental mistake of confusing skepticism as a philosophical position and skepticism as a tool.

  63. Steven,

    First off, I’m grateful for your response above to the general question about not laying down a number.

    It’s not clear to me that a desire to ‘have influence’ is sufficient justification to act as if I have confidence in my ability to predict something when I don’t. To me that’s dishonest. I’d be exploiting the trust decision makers place in me to honestly report my technical evaluations, certain, unknown, and the gray in between.

    Again, I’m not a scientist and perhaps mistakenly generalize from my own work ethics and experience. It’s not my primary duty to give guidance to non technical folk based on technical analysis, but it comes up from time to time. On these occasions, I’ve striven to prepare for such briefings by determining as many important pertinent details as possible ahead of time, by trying to estimate in a methodical and reproducible way other important pertinent details / probabilities and possibilities, and by identifying areas where the answers aren’t available. I try to make all of this as clear as I can to the non technical guys. It’s not my job to give them numbers when numbers are meaningless, regardless of what the other engineers might do (and this certainly does come up regarding schedule / resource allocation estimates, particularly very early in projects before scope is bounded or requirements are defined), regardless of how much they want me to give them a number. Why is the position of a climate scientist different?

  64. “most skeptics…”

    Just for the record, my position is not a manifestation of a skeptical reaction. It is a matter of how I have tried to practice science, no more and no less over four decades. You are right there is a price to be paid at times, but that is a part of the way things are. Whether one walks away or charges the windmill is not always an easy decision. I’d likely walk away here for a number of reasons:

    1.) I’m not vested. Like you I don’t give a rat’s put about the ramifications outside of science one way or another. (CE) Hence walking away for me is not a losing proposition.

    2.) Pretending for the moment that technically I was up to date enough to step into in the arena, I would assess the science to be too incomplete to expend the effort at calculating a number for 2100. That is, I would really would have other priorities and I would not be interested in the diversion–at least in a formal context where I might have to defend something that was not ready.

    3.) The decision making machinery does not seem to be viable at any level. Polarization and misinformation or accusation thereof are rampant. The time for actions (if needed) is uncertain but appears short. The science can not be forced, and it seems to me that the real fundamental question is how to proceed in the face of unresolved uncertainty. This for the time being makes the actual ‘science’ less important in the short term and perhaps an unwelcome diversion when over-exercised. It is politics—not science; that is I think we should focus more on the policy for what we can be doing now with what we have and know.

    4.) Just because I walk away I do not forfeit my right to comment, although walking away definitely can affect how or even if those comments are taken. I can live with that.

    “Skeptics make the fundamental mistake of confusing skepticism as a philosophical position and skepticism as a tool.”

    That seems to me to be an on-target comment. Unfortunately philosophies are worn 24/7 whereas one can put a tool down when it is not needed. Non-stop skepticism feeds intransigence and impedes moving to a consensus. (Blind arrogance from other camps does much the same.) So it goes.

  65. Steven Mosher:

    most skeptics play the walk away game. That is a losing proposition because some science guy will always come forward to make his estimate. Hell I dont even need to ask, I can just say Arrehenius and work with his number if you refuse to play the game.

    Saying “Arrehenius” wouldn’t work for a question about temperatures in 2100. Climate sensitivity tells us nothing without forcings.

    I think it’s perfectly appropriate to “play the walk away game” when a question requires us predict emissions over a 90 year period. We have no way to predict that.

  66. Steven Mosher,

    It’s possible that you intended your response merely to be informative and did not intend for it to be construed as an invitation to argument. If so, I apologize for arguing, 🙂 and I’ll reiterate that I appreciate the explanation, regardless of whether or not I find it persuasive.

  67. Steven Mosher–

    One more go. All that said. Were I to attempt a 2100 prediction, I would struggle between using a statistical model or one based in physical mechanisms. After much frothing about, I would then likely fall into a heap, and simply say the temperature in 2100 will be the same as the temperature this year. Then maybe I would fudge around a little on things, ‘id’ a couple of conditioning variables and maybe try updating the original estimate. (That is, I would probably try some cheap ‘bayesian updating’ trick with pretty pictures and call it a day.)

    Really good question.

  68. mw

    ‘Steven Mosher–
    One more go. All that said. Were I to attempt a 2100 prediction, I would struggle between using a statistical model or one based in physical mechanisms. After much frothing about, I would then likely fall into a heap, and simply say the temperature in 2100 will be the same as the temperature this year. Then maybe I would fudge around a little on things, ‘id’ a couple of conditioning variables and maybe try updating the original estimate. (That is, I would probably try some cheap ‘bayesian updating’ trick with pretty pictures and call it a day.)
    Really good question.”

    Well, then a good policy maker would ignore you.

    Here’s a related question: if it was 1890’s and you where Svante’s grad student, and somebody posed the question..

    ” your director thinks that if C02 goes up the temperature will go up. Suppose that c02 goes up between know and 2000. What do you predict will happen to temperature?”

  69. Brandon

    “I think it’s perfectly appropriate to “play the walk away game” when a question requires us predict emissions over a 90 year period. We have no way to predict that.”

    you make no sense.

    there are many ways to predict emissions, in fact an infinite number of ways of predicting emissions.

    Here is a simple way.
    1. establish a lower bound: 0
    2. establish an upper bound: Burn all the FF.

    Then I can predict that the number will fall in between those.

    Here is another way

    I can predict that the future will be like the past. This is called business as usual.

    Here is another way. I can do ground up and estimate population, GDP etc and predict emissions from that.

    perhaps what you meant to say is there there is “accurate” way to make a prediction. But this too is false. If I use BAU to predict and my prediction turns out to be right, then it is accurate.

    Perhaps you meant to say there is no way to estimate a confidence
    bounds, but this too is wrong, I can use ‘0″ and burn it all to estimate.

    Perhaps you meant to say there is no way to predict emissions that you think is reliable?

    I could go on, but clear there are ways to predict, and we know this because A4 is full of predictions for emissions. So, you cant say that the actual is not possible. There are ways to predict. there are predictions. You must of meant something else or you are speaking nonsense

  70. Mark Bofill (Comment #117649)
    July 12th, 2013 at 7:16 am
    Steven Mosher (Comment #117640)
    July 11th, 2013 at 9:51 pm
    ——————
    Obeying the constraints you’ve supplied, I’d use the dang GCM’s.
    I still don’t understand why it’s not ‘science’ for me to acknowledge that I’m probably going to get the wrong answer for 2100, but then again I’m a software engineer and not a scientist. If I expect my best science to give me the wrong answer, what was the point? Maybe you could elaborate on this some time.

    #################################################

    science always gives wrong answers. the issue is how wrong.
    when the degree of wrongness gets really really small we call that thing truth. what that really means is we dont see any point in debating it

  71. Steven Mosher–I agree that a good policy maker is likely to ignore such an estimate. (I wonder whether an decision analyst at soliciting ‘expert judgement’ might have a chance.) But here I’m not doing it for a good or bad policy maker–the conditions under which the prediction are to be made are apolitical and purely academic. That is, I don’t care about the policy maker. If I have to make a prediction for a policy maker–again I likely would walk. [Understand I am at a point in life where that is no big deal.]

    You also asked:

    Here’s a related question: if it was 1890′s and you where Svante’s grad student, and somebody posed the question..

    ” your director thinks that if C02 goes up the temperature will go up. Suppose that c02 goes up between know and 2000. What do you predict will happen to temperature?”

    Well that certainly is a ‘politics free’ situation ;O). The grad student would be both reasonable and smart to say the temperature goes up. But I can not imagine being the grad student without suspending some beliefs and/or biases, and other ‘information’ that I have.

    An inferred follow on the question of course is: OK, coming back to the present (and 2100) given that CO2 does appear to effect some increase in ‘global’ temperature, i.e., there are plausible physical mechanisms and given that CO2 is increasing is very likely to do in the foreseeable future, the temperature may be predict to rise. (I have no major heartburn with either ‘fact’ or such a prediction at this time.)

    However I have no idea of the magnitude of the CO2’s impact on temperature and I have no idea of other factors that may come into play–well not quantitative at this point. Hence I settle on the current temperature as a rough guess. Still, I can not reasonably dismiss the potential of the increasing CO2 to raise the temperature, and I am lead to consider my initial guess as a Bayesian prior–my guess a prior conditioned by CO2 concentration and some measure of ‘sensitivity’ to changes in the CO2 concentration. [Heavy and thorough literature review is key in such assessment.] How things would shake out I don’t know, other than I would expect and only feel comfortable with a posterior estimate (expected value) greater than the prior.
    I do not pretend that this is a globally best approach, but if I was forced to play the game IN A MEANINGFUL DECISION ARENA and predict I would be strongly inclined to start out like I described. This is primarily because I view that time is a real problem. The time required to ‘resolve’ the science is already possibly longer than the window open to significant influences over future events. We should hedge against both the potential extreme negative outcomes associated with no action and the extreme negative outcomes with a precautionary principle-based action.

    I do not think this approach as precluding estimates based on physical models but consider it as complementary to and. I hope, informing the modeling process.

    Sorry for the length of this response.

  72. Steven,

    science always gives wrong answers. the issue is how wrong.
    when the degree of wrongness gets really really small we call that thing truth. what that really means is we dont see any point in debating it

    thanks.

  73. Steven Mosher:

    there are many ways to predict emissions, in fact an infinite number of ways of predicting emissions.

    To predict the future requires accuracy. Otherwise, all you’re predicting is some hypothetical that never comes true. There is not a single way to predict emission scenarios 90 years out, much less an infinite number. Short of blind luck, there’s no way a prediction would wind up being correct.

    Try exercising some charity. We cannot predict what will be. We can only predict what may be.

  74. I predict this week’s tattlotto numbers will all fall in the range 1 to 45 and there will be 6 of them with two supplimentaries.

    Its inclusive, but not very useful.

  75. Steven Mosher (Comment #117663)

    science always gives wrong answers. the issue is how wrong.
    when the degree of wrongness gets really really small we call that thing truth. what that really means is we dont see any point in debating it

    There’s a whole bunch of people who see a point in debating global warming. That would suggest that the answers given by science is really really wrong.

  76. Anyone know where Lucia is? Vacation? Some of us are desperate for the June temp results payouts. Hope all is ok.

  77. I’m here. But I was feeling ill mid-week! I’m busy over the weekened… so probably be more about on Monday!

  78. Ah, yes predictions and good government policy and in this case about AGW and its effects. Any half observant soul knows this is a perilous process.

    For the best opportunity in attempts to make the process work would first require that relevant science and scientists be totally forthcoming about the uncertainty in predicting future climate and its effects on mankind. The most prominent organization attempting this part today is the IPCC. One can decide for oneself whether that organization does a reasonably objective job or if it is driven by an agenda and advocates/scientists. Science is relied upon to make predictions about the effects of increasing GHG levels in the atmosphere, but is not in a position to determine what those GHG levels will be in the future, i.e. the scenario problem.

    Scenarios might come under the domain of economists, where, like with scientists, we need to be aware of the advocacy influence on their work. How many economists saw the recent global economic meltdown coming? How many of those economists who got it right are those influential in making present economic policy? Given the best intentions of those predicting scenarios for climate how easy is it to predict accurately given all the complicated price and other free choice feedback loops that occur in a reasonably free economy. In command economies one might think that these predictions become less complicated as the creativity of mankind is mostly removed from the process, but unfortunately there are many unintended consequences with which to deal under those conditions.

    Finally we come down to the essence of policy making in the US and rationalizations for taking policy into law. We have an agency, namely the Congressional Budget Office, under the influence of Congress that admittedly has to make dumb ass and over simplified economic calculations based on the prerogatives that agency is given by Congress. Obamacare is a prime example of how a very poorly understood process can be given an aura of certainty by politicians quoting the CBO.

    Be realistic about how policy is made and then put into law and once into law how permanent it becomes regardless of how well the law succeeds in doing what is purported to do. Be very realistic.

  80. Steven Mosher (Comment #117640)
    Question: what is your best estimate for the temperature in 2100. What tools would you use and why?
    Answers to Question
    Common-sense The temperature in 2113 will most likely be the same temperature as this year base on reversion to the mean*.
    Why fudge around with modifications or guesses. The study of the temperature of the earth over the last billion years has shown that it has roughly stayed in the same range,i.e. in a range capable of sustaining oxygen breathing animal life and CO2 using plant life all this time which is why we are alive to speculate on this.
    Why should it change?
    Well, then a good policy maker would ignore you.
    Doubtful, A policy maker might ignore me but a good policy maker who took the trouble to ask me for such advice would doubtless pay attention to it.
    Here’s a related question: if it was 1890′s and you where Svante’s grad student, and somebody posed the question..
    ” your director thinks that if C02 goes up the temperature will go up. Suppose that c02 goes up between know and 2000. What do you predict will happen to temperature?”
    Easy the same answer as I first gave above.
    Why would my answer change?
    Now the third question to make the other two responses look foolish?It must be coming?

  81. angech

    [embold is by mwg]

    “Common-sense The temperature in 2113 will most likely be the same temperature as this year base on reversion to the mean*.
    Why fudge around with modifications or guesses.
    –
    The first sentence here is a guess; even the wording is overtly that of a guess–‘common-sense’, ‘will most likely’. To make a prediction one has to have to have some ‘model’. And reversion to the mean [regression to the mean] is the ‘model’ you have stated to select, and it has assumptions tied to that. There are a lot of attendant assumptions, i.e., you are fudging like every one else has to.
    –
    BTW as you stated things [evoking regression to the mean], haven’t you also assumed that the present value is the mean? Do you really want that? I don’t think so. (I suspect you do want the present value as your future estimate, but that seems more directed to predicting a future value in a random walk at some level, and not reversion [regression] to the mean.)
    =======
    =======
    Well, then a good policy maker would ignore you.
    Doubtful, A policy maker might ignore me but a good policy maker who took the trouble to ask me for such advice would doubtless pay attention to it.
    —
    No guarantees on a policy maker; good or bad they have agendas, multiple inputs, etc.
    =======
    =======
    Why would my answer change?
    –
    You want to graduate and you are not a saint? ;O)

  82. well done mwgrant

    “The first sentence here is a guess; even the wording is overtly that of a guess–’common-sense’, ‘will most likely’. To make a prediction one has to have to have some ‘model’. And reversion to the mean [regression to the mean] is the ‘model’ you have stated to select, and it has assumptions tied to that. There are a lot of attendant assumptions, i.e., you are fudging like every one else has to.”

    and at the bottom even people who claim “i dont know” are applying a model, a model of knowledge.

  83. “There’s a whole bunch of people who see a point in debating global warming. That would suggest that the answers given by science is really really wrong.”

    There’s a whole bunch of people who see a point in debating evolution. That would suggest that the answers given by science is really really wrong.

    There’s a whole bunch of people who see a point in debating tax reductions. That would suggest that the answers given by conservatives are really really wrong.

    There’s a whole bunch of people who see a point in debating The Zimmerman verdict. That would suggest that the answers given by jury is really really wrong.

    #############
    Your argument is an odd lemma of the appeal to authority.

  84. “To predict the future requires accuracy. Otherwise, all you’re predicting is some hypothetical that never comes true. ”

    No it doesnt. I predict that tommorrow no place on earth will be warmer than 100C. All it takes to make a prediction about the future is to actually make one. You keep trying to make requirements for predictions. And you forget that a prediction is nothing more than a sentence. A prediction should be grammatical. That’s about it.

    “There is not a single way to predict emission scenarios 90 years out, much less an infinite number. Short of blind luck, there’s no way a prediction would wind up being correct.”

    There are many ways of making the prediction.

    I predict emissions will be 0. There I just did it. Now, is there a way to make a perfect prediction? Sure. Make a prediction
    and see if it comes true. The question you could be asking is how can I prove that there is no way to make a reliable useable prediction. Claiming there is no way, suggests that you have some knowledge that is proof there is no way.

    However, if your claim is that there is reliable way of making a perfect prediction, then nobody would care because we dont need perfect predictions. We just need predictions that work well enough. I predict the sun will come up. I dont need to know the exact time. Rain is forecast tommorrow. That prediction is vague. They fail to specify the time the rain will start or the number of drops. I still take my umbrella.

    ##############

    Try exercising some charity. We cannot predict what will be. We can only predict what may be.

    We can predict what will be all the time. In fact you cannot stop predicting what will be. Whether those sentences turn out true is a different matter.

  85. Steven Mosher – “Try exercising some charity.”

    So that is what it is when a warmer puts a gun to my head and extracts their pound of flesh (taxation). Thanks for clearing that up.

    “is there a way to make a perfect prediction? Sure. Make a prediction and see if it comes true.”

    Also thanks for stating clearly how warmist predictions are made (pulled out of their A**), I had worried for a while that you had gone over to the dark side.

    In the real world not the warmists fantasy land, we make accurate predictions all the time, it is what differentiates Science from Climatology, Scientology, etc. Climatologists claiming that they will start making accurate predictions if just given more time and funding would be funny if it wasn’t so sad.

  86. Steve Mosher,

    I am not sure what the point was of your original ‘put your scientist hats on’ request, or for that matter, the point of all your later comments.
    .
    I fear much of it is semantic rather than substantive. Yes, policy makers will do pretty much whatever they damned well want… limited only by what their constituents will support. The size of what they ‘damned well want’ depends very much on their political inclinations. Those to the left generally have much bigger wants than those to the right.
    .
    Still, I am uncertain what point you are trying to make. What I have been consistently saying (ok, approaching broken-record consistency) is that climate scientists are not playing it straight… not about uncertainty, nor about assumptions, nor about the likelihood of extreme consequences of future warming. The issue for many (certainly for me) is that public policy ought not be hostage to the generally leftist/green political inclinations of those who call themselves climate scientists. I am reminded of the EPA’s friendly ‘settlements’ of lawsuits with various Green organizations… politicians on the left defacto do the same when they ‘accept’ the extreme projections of climate science.
    .
    So what the heck is the point?

  87. Steven Mosher, your entire response to me rests upon a strawman argument. You discuss predictions about the future, but I repeatedly discussed predictions of the future. You simply used semantic obfuscation to misrepresent my comments.

    If you claim to be practicing charity, I must assume you’re incapable of reading simple sentences. I doubt that’s the case.

  88. SteveF (Comment #117683)
    July 14th, 2013 at 7:30 pm

    “So what the heck is the point?”

    Good luck with that.

    Perhaps it was merely allowing you and me an opportunity to make the same points we so frequently do here.

  89. SteveF regarding your solar power analysis, looks similar to what I’ve looked at. For Florida, it would be difficult to DIY because of their permit system, so $20k would be about right.

    I would also say it’s an unsafe assumption to assume that subsidies would continue indefinitely, so one shouldn’t include them in the analysis (you didn’t of course).

    I think we can also assume that prices of components will continue to fall, while energy prices from fossil fuel will continue to rise, so over time a system like this should become more fiscally attractive. Whether it becomes a substantial proportion of national energy generation… well that seems a bit more dubious to me.

  90. SteveF,

    I’ll say it again, excess power metering rates that are above the wholesale generating cost are a wealth transfer from the rest of the ratepayers. Net metering at the retail power rate is therefore effectively a form of regressive tax. The same is true for subsidies for the capital investment.

  91. DeWitt Payne (Comment #117726),
    Of course. Net metering rules are usually nutty and cost the average power user. They would almost certainly have to change if a large number of people began using solar panels, since the cost of distribution and maintaining grid generation capacity are off-loaded from the alternative energy producer to everyone else. Same thing with tax incentives… nutty and unfair. I do not and never would support the kinds of economic distortions the existing solar and wind incentives produce.
    .
    Still, from the POV of an individual making a purely personal economic analysis, and one based on current reality, rooftop solar has began to approach viable, at least in sunny regions. Under what conditions it would become viable absent tax incentives and subsidies is a very different question, and one that deserves careful public consideration.
    .
    A more interesting economic analysis would be comparison of investment in energy efficiency versus solar power. I suspect that improving efficiency (better windows, better insulation, lower energy lighting and appliances, etc.) would be the obvious way to go in new construction. The problem is the cost of retrofit in existing structures, which is often not economically viable.

  92. SteveF (Comment #117683)
    July 14th, 2013 at 7:30 pm
    …
    So what the heck is the point?

    I don’t know. I appreciate it when Steven talks about this because I still don’t understand why he (apparently) is such a strong advocate of models that give projections of questionable value, and I retain the hope that if he talks about it long enough and if I pay attention I’ll eventually understand his reasoning.
    ~shrug~

  93. SteveF,

    They would almost certainly have to change if a large number of people began using solar panels, since the cost of distribution and maintaining grid generation capacity are off-loaded from the alternative energy producer to everyone else

    I’m pretty sure in most places you pay grid maintenance fees, regardless of whether you have net metering or not. Even if you consume zero net energy over a year (or are a net producer), you’ll end up paying fees for being connected.

    In Mississippi, at least, there is as a clause that limits the amount of buy back required so that alteration of the grid to handle excess capacity isn’t required.

    Wikipedia erroneously states that no net metering program is available in Mississippi, but this isn’t true. The net metering rate isn’t state mandated, but it is mandated that the program exist. See Section IV B in particular.

    A more interesting economic analysis would be comparison of investment in energy efficiency versus solar power.

    I think passive solar has a lot to offer in that respect too.

    Here’e a simple example:

    Just going from central cooling in my household to distributed cooling has reduced my summertime cooling bill by nearly 1/2.

    Bedrooms are kept at 70-72°F, other spaces are typically around 75-78°F. Because of the non-symmetric manner that waste energy is produced (more of it is produced in bedrooms, at least in my house, that’s where we keep our computers), so to keep a bedroom at 70-72°F requires cooling the entire home to 68°F for a central system. In a home that had better air circulation than mine has, you might see less of an offset.

    A distributed systems that meet the reduced cooling demands of a distributed system is actually cheaper and less technical to install than a central heating/cooling system too (at least for DIY people like myself).

  94. Carrick,

    In much of South America virtually all home AC is distributed. The motivation is to only spend money on cooling where it is absolutely needed…. bedrooms at night for example. Of course, in lots of places electricity costs more than US$0.25 per KWH (half of it taxes!) and most people’s income is much lower than in the States. You might think this would lead to lots of solar installations, but that is not the case. The governments add ~100% import duty and sales taxes for all electronic equipment, so even with very high electricity costs, solar cell power remains not economically viable. Most people would never make the investment due to lack of personal assets, even if the equipment cost were lower. Lots of lip service paid by those governments to global warming as well, but they are not serious… it’s just political posturing to be ‘good guys’ at UN conferences. They can’t afford to lose the 100% tax on their populations for electricity use.

  97. Here is my analysis of the betting this year.
    I have excluded the bet for 999999 by Dr. Jay Cadbury, phd.
    Not sure if it was intended to be a real bet for 0.999 or just a test.

    The consensus seems to be around 4.9 and the figures seemed to change very little during the course of the betting.
    With the betting around 4.9 being very crowded, it looks like the result is going to be very tight, unless something unusual happens.

    NO. OF BETS 111
    MAX 8.69
    MIN 3.45
    MEAN 4.91
    MEDIAN 4.85
    STD DEV 0.71
    MEAN PLUS 1 SD 5.62
    MEAN MINUS 1 SD 4.19
    WITHIN +/- 1 SD 91
    WITHIN +/- 1 SD (%) 81.98
    MEAN PLUS 2 SD 6.34
    MEAN MINUS 2 SD 3.48
    WITHIN +/- 2 SD 108
    WITHIN +/- 2 SD (%) 97.30
    ABOVE MEAN 51
    BELOW MEAN 60
    MEAN BETS 1-56 4.86
    MEAN BETS 56-111 4.94

Comments are closed.