back to article NASA launches new climate model at SC14

After 75 days running on the 43,000-core, 1.21 petaflop Discover supercomputer cluster at the Goddard Centre for Climate Simulation, producing 4 PB of data, NASA has put together a global model showing how atmospheric carbon dioxide – CO2 – gets distributed in air currents around the world. Youtube Video The model is based on …

  1. imanidiot Silver badge

    Nature run, so?

    So how wel does the calculated data compare to the real word data measured in the same time period? I can't really find much data on that.

    1. Adrian Midgley 1

      from the article ...

      this is the observed data, presented...

      1. Ragarath

        Re: from the article ...

        Yes, Adrian is right, it seems they have just taken data from various sources and parsed it. Though it confuses me why so much computer power was needed to produce the output.

        I tend to think of climate modeling as being a guess at what would happen. If your just putting in real data and seeing the output placed on a grid, why the computing power?

        There must be some guess work in there somewhere.

        1. Grikath Silver badge

          Re: from the article ...

          they parsed the data and refined the grid, call it "image enhancement" . and that takes a *lot* of cycles...

          As the animation goes, as far as I can tell it should be accurate enough to work with. It does concur with thermal barriers, coriolis effects, seasonal growth ( watch how the Big Red disappears as nature wakes up and gobbles up the CO2 on the northern hemisphere..), population centers, and other stuff, like the brush fires in Africa. Everything you expect to be there is in there.

          As is, it shows a couple of things both sides of the Climate $deity have been crusading on, and debunks a couple of arguments on both sides of the fence. I for one would very much like to see a longer run, say, over the past 20 years, extrapolated or not. It would (hopefully) show whether or not the cycle is progressing or stable, or.. can't tell anything from one cycle except that "world is moving on happily as expected".

  2. codejunky Silver badge

    Interesting

    So not a true observational set of data but another model? I hear models are useful to science and I believe that. However the large divergence between model and reality has been a point of contention in the climate debate unless you already somehow have the answer through magic or consensus.

    1. amck

      Re: Interesting

      @codejunky:

      Evidence?

      There isn't a "large divergece between model and reality". The global climate runs match observations within expected experimental error (ensemble error). The main errors have been places where we have a lack of good observational data - the Arctic and Antarctic, especially in winters.

      1. codejunky Silver badge

        Re: Interesting

        @ amck

        Please tell me you are kidding or have been away for a good number of years with little access to the climate news. The main errors are in the lack of understanding in climate which is being continuously corrected through the process of science. The problem being when the models are pushed as fact by the political/religion groups and then (as one would expect from an incomplete model) not being correct.

        As for margin of error, it really depends how wide the margin is as to how relevant. And for every failure it gives ammunition to the other cult of certain believers that nothing is changing and the whole problem is a lie. The certainty we have is that we certainly dont have the answers/understanding yet which is actually a plus in science as it makes the subject interesting. The obsession with co2 is amusing but only highlights the lack of science in the religious/political pushers.

      2. Anonymous Coward
        Anonymous Coward

        Re: Interesting

        "There isn't a "large divergece between model and reality". The global climate runs match observations within expected experimental error (ensemble error). The main errors have been places where we have a lack of good observational data - the Arctic and Antarctic, especially in winters."

        you mean: "run model after model until one matches current observation. Then claim victory again. Now when the model doesn't match observed data next year, we'll rerun and announce a new model that matches data then!"

        Notice we never hear about a new "successful" model until AFTER the "reality" matches. Like those tabloid psychics who always claim they PREDICTED events long after they happened, but don't have any real records of that prediction beforehand. Those that do, have so much ambiguity or so many "predictions" as to cover any eventuality. How many of these models have failed to match reality by now? How many concurrent models with how many different "predicted" outcomes are queued up awaiting revelation as being the one that matches reality?

        When a model we all knew about a decade ago consistently produces results that match objective observed results, THEN we can celebrate a success. Until then, we don't have a model but the data needed to produce such a model.

    2. DragonLord

      Re: Interesting

      I don't believe that this is supposed to be a predictive model, rather it's supposed to tell us how stuff moves from a to b. When we understand that we can then start to discover other things that are useful such as where's the best place to plant forests to absorb CO2, where's most of the CO2 coming from, what paths does it take, how does it correlate to weather conditions.

      Understanding the why of the past is often key to being able to affect the future in the way you want.

      1. codejunky Silver badge

        Re: Interesting

        @ DragonLord

        I did read the bit about it being a simulation of the period May 2005 and June 2007 and if that is it then I have no issue with it. However I do remember a reg article some time ago about NASA and their historical temperature data where they 'adjusted' it into such a steep incline even the IPCC wouldnt back them and was eventually taken down. Since then and other events from the MMCC co2 theory pushers I do get a lot sceptical about their versions of history.

        I do agree with your post, I just hope NASA have done this honestly instead of trying to manipulate public opinion again.

        1. DragonLord

          Re: Interesting

          From my understanding of historical temperature records, the problems have been that when they were made we didn't understand that things like urban heat islands even existed let along the ramifications it would have on temperature records. They also weren't as rigorous about how the readings were taken, so the time of day, and whether it was in shade wasn't necessarily followed as precisely as needed.

          Based on that once we did find out about those things we needed to go back and make adjustments to the old temperature records based on where the temperature was taken from at the time. This being a judgement call by the interpreter means that individual historical models could be very different. Thus the NASA adjustments.

          1. codejunky Silver badge

            Re: Interesting

            @ DragonLord

            That didnt seem to be the interpretation from what I saw. They publicly put out a heavily edited upwards graph after adjusting the data and then took it down when criticized it didnt reflect reality nor the edited upward versions pushed by the other religion pushers. I may have to try and find it again but it gave little impression beyond complete screw up or blatant lie. Either way it didnt inspire confidence in a field which has been taken over by political lies and religious culting.

            I hope NASA is putting out correct data but I will wait and see if any real criticism comes from this first.

            1. NomNomNom

              Re: Interesting

              "They publicly put out a heavily edited upwards graph after adjusting the data and then took it down when criticized it didnt reflect reality"

              your recollection is a pile of steaming shit then. What you describe never happened.

              The temperature records have been independently confirmed again and again.

              1. codejunky Silver badge

                Re: Interesting

                @NomNomNom

                I wondered how long until you appeared to make your claims. Hi again.

                Unfortunately I have yet to find the specific instance of the reg article (you would disregard it anyway as it was a Lewis article) but while searching I found plenty interesting stuff including-

                http://www.theregister.co.uk/Print/2008/06/05/goddard_nasa_thermometer/

                Related to-

                http://www.theregister.co.uk/2008/05/02/a_tale_of_two_thermometers/

                And

                http://www.theregister.co.uk/Print/2008/11/19/nasa_giss_cockup_catalog/

                Not sure if that is referring to the same data I mention in my previous post but should be enough for you to sink your teeth into.

  3. Chris Miller

    Running equations through a computer does not constitute an experiment. - Stanley Miller (1930-2007)

    1. Slartybardfast
      Facepalm

      My quotes better than yours club

      Just because someone once said something it doesn't mean that it's true.

      "I think there is a world market for maybe five computers." - Thomas Watson, president of IBM, 1943

      "Nuclear-powered vacuum cleaners will probably be a reality within ten years." - Alex Lewyt, president of Lewyt vacuum company, 1955

      "There is no reason anyone would want a computer in their home." - Ken Olsen, founder of Digital Equipment Corporation, 1977

      "Two years from now, spam will be solved." - Bill Gates, founder of Microsoft, 2004

      "The coming of the wireless era will make war impossible, because it will make war ridiculous." - Guglielmo Marconi

      1. codejunky Silver badge

        Re: My quotes better than yours club

        @ Slartybardfast

        Well said.

        "We only have 50 days to save the world" - Gordon Brown, class 1 pillock

        "Himalayan glaciers would disappear entirely by 2035, leading to widespread drought, starvation and migration" - 2007 IPCC AR4

        1. Fluffy Bunny

          Re: My quotes better than yours club

          "The map is not the territory" - A.E. Van Vogt

      2. JDX Gold badge

        Re: My quotes better than yours club

        >>"Two years from now, spam will be solved." - Bill Gates, founder of Microsoft, 2004

        When did gmail come out? Since then, it has been 99% 'solved' for me (not only gmail but filters in general).

      3. Chris Miller

        Re: My quotes better than yours club

        I'll see your quotes and raise you:

        Predictions can be very difficult - especially about the future. - Niels Bohr (1885-1962)

        I note all your quotes are attempts to predict the future*, whereas Miller's (no relation, natch) is simply a pithy observation about the nature of scientific enquiry, which anyone with pretensions to being a scientist ought to recognise.

        * Actually, I suspect it took about a decade for Watson to be proved wrong, and Olsen was correct, except he failed to foresee the advent of the microcomputer - (almost) no-one would want a PDP 11/34 in their home.

    2. Paul Kinsler

      Re: Running equations through a computer does not constitute an experiment

      In my experience (as a numerical/theoretical physicist), there are two main usages relevant here:

      (a) "experiment" on its own always refers to a real world experiment

      (b) "numerical experiment" (sometimes "computer experiment") refers to a simulation done in order to test ("experiment on") the numerical model in some way. Perhaps, if the numerical experiment is encouraging, to motivate a real world experiment; perhaps to see if the numerical experiment matches (existing) real-world measurements.

      I suppose if the computational model was actually the system under investigation - i.e it was not only a proxy for some real-world system or specific mathematical model - then one could even experiment on it in the true (a) sense.

      It would be interesting to know when Stanley Miller made that remark - these days, numerical experiments are an invaluable tool, and I think that most would assert the primacy of real-world data over numerical models with a different phrasing.

      1. Fluffy Bunny
        Thumb Down

        Re: Running equations through a computer does not constitute an experiment

        For your '(b) "numerical experiment" (sometimes "computer experiment")', please substitute either "fake experiment", or the slightly more polite "thought experiment" - your choice.

        1. Paul Kinsler

          Re: please substitute either @FB

          I was just providing information of the way "numerical experiment" is typically used in science, by scientists.

          Also note that the phrase "thought experiment" (as used by scientists) is something else again - the most notable being Einstein's "if I was sitting on a photon...".

          Do, though, substitute whatever what you like - I'm not trying to control the vocabulary of anyone. Just, as I said, reporting on usage in one context (ie science).

    3. DNTP

      running equations

      Oh, ok, thank you for reminding me of that. I'll stop running my equations right away.

      Your lab results will take longer though since we'll have to come up with this stuff through trial-and-error instead of being guided by models that are after all only approximations of reality.

      Your medical report will have a great deal more of "who the hell knows"; gene frequency and stability are just approximate models too.

  4. Fading Silver badge

    Very pretty but contradicts observations...

    Like the ones GOSAT has been taking http://data.gosat.nies.go.jp/GosatBrowseImage/browseImage/fts_l2_swir_co2_gallery_en.html

  5. squigbobble

    Enlightening

    I find it interesting that all the CO2 makes a beeline for the North Pole in winter, might be something to do with the greater degree of melting up there vs. the Antarctic.

  6. Martin Budden
    Thumb Up

    airlines will like this

    Knowing where a volcanic ash plume will spread means airlines will only have to cancel flights which actually need canceling.

  7. Faux Science Slayer

    There is ZERO Carbon climate forcing....

    BFD....NASA 'models' a natural, mandatory for life, three atom gas molecule. Free moving gas molecules DO NOT capture, store, redirect or amplify radiant energy photons moving at the speed of light. Some gas molecules will vibrate to some frequencies, filtering out some energy, which on Earth is a COOLING effect.

    See..."Mommie, Can We Play Obombie Truth Origami" at FauxScienceSlayer site.

    1. Rik Myslewski

      Re: There is ZERO Carbon climate forcing....

      Uh, no...

      Quite simply, sir, your "science" (grin, guffaw, bwah-ha-ha) is demonstrably wrong. Your lack of understanding of even the most basic mechanisms of radiative forcing is an embarrassment to your coterie — even the most strident pro-fossilists would be embarrassed by your silliness.

      May I suggest that you re-enroll in University, and take basic physics again? You seem to have either left early, flunked out, or been ideologically (or, as the denialists accuse us scientifically informed folk, "religiously") brainwashed by the Heartlanders and their ilk.

      Sad, really...

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019