Lol. Like I said, it’s you who have no idea what you are talking about.
Let’s summarize your position:
(1) A simulation is “proven” if it lines up well with historical data, regardless of how good or bad its future predictions turn out.
(2) If a simulation’s future predictions don’t match reality then it is “broken.”
(I’m not totally sure what you mean by “proven” and “broken,” and I’m not going to bother asking because I doubt you have any clear idea what you meant by these terms.)
:shrug: The Met Office model failed an extremely simple test. Hansen’s prediction has diverged from “reality”
I’m not sure what this means in reference to the sage rat dialectic.
Telling another poster that he doesn’t know something is an open invitation for him to respond in kind, leading to personal friction.
You have been around long enough to know that brazil84’s standard MO is to post a pretend Socratic dialogue for the purposes of wearing down the posters he opposes rather than presenting any actual points to debate. You are welcome to engage him in that activity or to ignore his games, but please do not allow your frustration to spill over into personal remarks.
No, it didn’t. All the years were predicted within their error bars. Now, you may argue that it wasn’t a very stringent test…and I won’t particularly disagree with you but to claim that they failed the test when in fact they got the correct result every year to within their degree of confidence in the prediction is silly.
Furthermore, as I have noted before, every individual prediction from a climate model will eventually diverge from reality in a statistically significant way given enough time since, among other things, the climate sensitivity is not known to a high enough degree of accuracy for any single prediction to be correct. In particular, Hansen’s model is now believed to have a sensitivity that is on the high side of the most likely value. (Other issues include, for example, how well the forcings assumed in the Scenario B correspond to the actual forcings.)
A more interesting question is how has Hansen has done compared to his detractors. And, I think you will find that on that score, he has done very well.
OK, since I doubt it’s going to be addressed there (I still haven’t heard from brazil84 about how exactly his wife is constrained by PC politics, for example), I’m cross-posting this:
or, in other words, to quote Monty Python “It’s only a model!”.
I think comparing Risk and climate simulations is a stupid analogy, myself. The only connection Risk has with reality is that the outlines of the regions match some real-world political boundaries. That’s it.
Global climate models are not just based on a few simplified parameters, as brazil84 is implying. Yes, some physical effects are consolidated and simplified. That’s what makes it a model and not some kind of real-time simulation.:rolleyes:
So what? That’s what those of us who know how science works expect a model to be.
Now, I’ve never worked with climate science models, but I have worked with very good models of Gondwana, simulating crustal plate evolution. And I can tell you that no-one who works with a science simulation expects them to be RTSes. And we do expect them to differ from each other in their results, in their assumptions, in their parameters, in their outcomes. I never expect two runs of plate simulation on the same model to end in the same exact configuration, never mind two different base models.That’s what models are supposed to be like. The greatest value of modeling *isn’t *in finding some killer simulation that exactly matches reality, past and future (as brazil84 seems to be implying). It’s in comparing what all those different models do, how they differ, why they differ, what assumptions make them differ, and also what they do have in common. What makes models most useful is having many of them (and that’s why the IPCC is such a good effort, because it sums up all these diverse efforts).
What brazil84 is trying to attack as modeling’s greatest weakness, is in fact, its greatest strength.
That’s the danger of having laypeople try and tell scientists what they need to do. They just don’t have a grasp of how we actually do what we do.
That’s not the test I was talking about. The test I was talking about was the simple test of comparing the performance of the model to the “naive” or “no change” model.
The Met Office model would appear to have failed that test miserably.
Unfortunately, I don’t have time at the moment to do that sort of statistical analysis. However, there is actually a simple thought experiment which will show that there is a serious problem with Hansen’s model:
Suppose Hansen was locked in your basement the last 20 years and he had no idea what had happened with CO2 or other GHG emissions in the world. He has no idea if some sort of treaty was adopted, or if some new emissions-free energy source became widespread.
The only thing you show him is this chart and you ask him to guess which of his emissions scenarios has actually taken place.
Well, I imagine he would be rejoicing, having inferred that humanity chose path “C.” Just using one’s eyeballs, it’s pretty clear that “C” is the closest fit to the “reality” of the instrumental temperature lines. And of course Hansen would be dead wrong.
Of course, it’s also possible that Hansen would claim it’s impossible to tell and that scenario “A” or “B” could have easily produced the same temperature record. But in that case, it would be wrong for anyone to claim that his predictions have been corroborated in any meaningful way.
I don’t think that’s an interesting question at all. On the contrary, it’s a red herring.
As a “detractor” of Hansen, I don’t claim to be able to predict future temperatures better than Hansen. I claim that he can’t predict them very well and neither can I.
How is this any different from what we are talking about?
How is this any different than your position in this thread?
I can only answer that with my own opinion.
Seven different, independent groups have spent the last 40 to 50 years taking the best verified knowledge on how various things like plant life, volcanoes, the ocean, El Nino events, the sun, different atmospheric gasses, etc. affect the climate and incorporated it into massively complex simulators. As time progressed, the output of the simulators converged, indicating that they were all taking into account the same things and giving them the same importance. And as time progressed, when various factors like the historical recorded status of all the preceding factors like volcanoes, atmospheric gasses, the ocean, etc. were fed in, all seven jived with what had actually occurred in real life as a result of those factors. This indicates that the initial data on how the various factors effect the real world appears to be correct and the amount of importance the factors have relative to one another is also correct.
Certainly I don’t think that the simulators will perfectly simulate the world, but they have been developed in the correct way and the output agrees with both real life and mutually among all seven independent simulators.
And the only proof there ever could be against a simulator–beyond all seven of them mysteriously being found to have stopped jiving with real life tomorrow–would be for someone to develop a simulator that could model historical periods of the earth given historical readings of the factors, and come up with a different result. No one has been able to do that. Six times, groups of people tried to do just that, and each time they failed. And so we have seven simulators.
I’d say that there’s little to no argument against the viability of the output of the simulators. The input, perhaps, but minus any reason to doubt the modeling capability of the models, there’s no reason to doubt the modeling capability of the models. It’s redundant to even write, let alone argue.
What’s the difference between the alien machine and the global climate? Why would you vote against the best, most researched and comprehensive understanding of X in one case, independently tested and verified for a large sum of decades by thousands of skilled people, and yet vote for the best, most researched and comprehensive understanding of Y–or vice verse? What’s the difference in either example but the names, which can just as easily be swapped out with X and Y?
Hey, what’s that black line? Is that the linear trend? Would that be the one you would use, that fits the instrument record better than models? I must admit I’m not seeing the better fit, myself.
There are a lot of differences, but the most obvious one is that the prediction is for the alien machine to continue doing essentially the same thing as in the past.
A better analogy would be if the machine has been flashing blue and green lights 90% of the time for the past 40 years, and red lights for the remaining 10%. The panel of experts predicts that on January 1, 2012, the red light percentage will rise to 50% of the time.
That analogy is stupid. People are dumping 27 billion tons of carbon into the atmosphere, volcanoes pump around 150 million tons. We are measurably altering the composition of our atmosphere. So by your analogy, people have started messing with the “increase red light” dial.
Please try to think before you post. This kind of ignorant flailing only undercuts your argument. Which is weak and ineffective to begin with.
Fine then, let’s go back to the giant bean machine with the striking hammer. We’ll even take your version where each pin that is struck has some sort of complex algorithm that effects the odds of which side the bean will bounce off of dependent on how it has been struck before. The size of the bean machine needs to be just big enough that the hammer can continue to strike against the side simply through bouncing for X-thousand years, and thus will likely need to be tall enough to be free of most of the gravity and atmosphere of earth as well as having something like a rubber head.
Our scientists have studied, in varying scale models, tests like launching people into outerspace and the moon, simulators that have been shown to model fluid dynamics in 3D environments, Newton’s laws and higher physics disciplines, and other such related things. They split up into ten groups that work independently and competitively with one another to create a new fluid dynamic simulator (since there will be atmosphere inside the lower regions of the machine given the effort of creating a vacuum of sufficient size) that includes large gravity differences and models the interior shape of the bean machine, the algorithmic pins, and randomly imperfect spherical beans within expected ranges.
The scientists work for 40 years, each group perfecting their simulator as anyone comes up with any objection or other factor that needs to be included, and at the end of the 40 years, they’ve added all the major factors they can think of and each of the ten simulators come to within 99.9% of the same result of one another, and give you the plans for a device that they feel will work as necessary.
Now it is up to you, when you start to finally build the 400 mile tall bean machine and hammer, whether you are going to set up bins that are just big enough to store the expected amount of beans which land in them–it costing significant amounts of money to build these, so you want to minimise costs–whether you will make the machine as tall as the scientists say you will need to, whether you will build the hammer out of the recommended materials, and so on and so forth.
In this example, we have no “past.” There is only a prediction based upon the conglomeration of small scale tests of various factors, combined to make a prediction of an expected result, assuming that the specified materials are used, that the specified sizes are met, that the assumed densities of the atmosphere is correct, that the existence of the structure on the face of the planet will effect the global wind patterns in such and such a way and so cause the machine to waver “just so” and so on.
This is a much more difficult challenge to meet than simulating the climate since there is no full scale history to compare the simulator’s output to.
Do you vote to go with what the scientists recommend, or are you going to go with what it seems like it “ought to be” to you?
You are assuming that Hansen displays the same misunderstandings as you. In fact, I would expect Hansen to say, “I can’t tell for sure since as I said in my 1988 testimony, ‘There are major uncertainties in the model, which arise especially from assumptions about (1) global climate sensitivity and (2) heat uptake and transport by the ocean, as discussed in attachment A.’ However, the magnitude of the temperature changes computed with our climate model for various test cases is generally consistent with a body of empirical evidence (reference 2) and with sensitivities of other climate models (reference 1).” In other words, I never meant it to be a precise prediction and meant the various scenarios to primarily illustrate the range of differences that occur with different forcing scenarios but not to be so precise in their exact magnitudes as to constitute a precise prediction of future climate."
If I was persistent I might ask Hansen, “But look at the trend over the past several years where it seems to have flattened out a bit. Doesn’t that tell you that we might be following something like Scenario C where we stabilized our emissions starting around year 2000?” At that point, I might imagine he would say, “Well, this is one possible explanation. However, note that even in my Scenario A which has consistently higher forcings over time, there are periods when the temperature can be fairly flat for several years. In fact, as you can see Scenario A shows no perceptible upward trend between ~2002 and 2008…and even the trend since ~1996 is considerably flatter than the trend before that or the trend that is predicted after that. Hence, you have to be carefully in concluding too much from trends over short time periods, as my model clearly illustrates.”
Well, the meaningful way in which they have been corroborated is that the temperature has continued to increase significantly over the last 20 years (which many others were not predicting)…And, it is not too far off of what Hansen has predicted. In fact, there is no evidence that it is even outside the margin of error.
And, of course, this is just one small piece of evidence in favor of the rise being due to AGW. There is other evidence from the “fingerprint” of the warming. And, from various tests of the various pieces of the model physics, such as the water vapor feedback and so forth, that give us more and more confidence in being able to ascribe the current warming to AGW and to project what the future warming will be.
Well, it would be interesting to see which detractors of Hansen were claiming that the earth would likely be warming at rates of close to 0.2 C per decade since the late 1980s but not be due to AGW (and particularly those who would say this would happen even with a major volcanic eruption and with no significant trend in solar luminosity).
And, if I must choose between people who claim that they can’t predict the climate and have proven it…and those such as Hansen who say that they can and have demonstrated that they can do a reasonable job even if not perfect, then why should I go with the ones who have a proven track record of no competence over one with a track record of pretty decent competence.
Seems to me the problem is that you have a classic alarmist double-standard.
If temperatures fall in the ballpark of scenario B, that’s a “reasonably good prediction” according to you. On the other hand, if the temperatures are even closer to scenario C, well, that’s just variation or something.
Let me put it to you this way: What temperature pattern over the next 10 years would force Hansen to admit that his prediction was dead wrong?
Sure, it would be fascinating to see what people said. Personally, I think it’s entirely possible that the Earth will resume the warming trend from the last century.
The problem is that it’s easily possible to do a “reasonable job” just by blind luck. For example, suppose I have a model which I claim can predict seismic activity. I predict a major earthquake (6.0 or greater) in California to take place some time between 2010 and 2015. Suppose that a 5.5 earthquake hits Los Angeles in June of 2016. Arguably I did a “reasonable job,” but would it really give you confidence in my ability to predict earthquakes?
Then how do you know that it isn’t anthropogenic greenhouse emissions?
Normal sense of the word?
…
Normal sense of the word?
You find me ten things in the universe besides ghosts, quantum mechanics, and the creation of the universe that could plausibly have some sort of “causeless” description and I’ll take that into consideration as an argument that will fly in grade school. Include any particular reason on why climatology would fit into the class of things that could be in this list, get it printed in Nature magazine, and I’ll personally campaign for you to win the Nobel Prize for Physics cause Einstein couldn’t answer that question. Man would balk at contemplating what the hell, “Normal sense of the word” means.
The same way I know it isn’t because of a failure to sacrafice enough virgins. There’s no a priori reason to think so, and the evidence presented is unconvincing.
Actually, that’s not totally true. Because I do agree that greenhouse gases, in particular CO2, can cause warming. It follows that some portion of the warming in the last centure was likely to have been caused by CO2 emissions.
It’s a fact that things can happen in chaotic systems with no clear, tangible cause. For example, if you through the dice in Vegas and get snake-eyes, what caused you to roll a “2” and not a “12”?
What caused Hurricane Katrina to hit New Orleans and not Biloxi?
Anyway, I don’t have time for a semantic debate over what the word “cause” actually means. It’s clear enough what I’m saying.
If you prefer to use a different defintion of the “normal sense” of “cause,” then feel free.
A priori is not an argument. Period. Either testing confirms or disproves a hypothesis. And the only argument you’ve ever made in this or any other thread about the convincingness of the evidence is that it “doesn’t make sense to [you]” or that it “ought to make sense to [you]”. Again, that’s not an argument.
Chaotic systems fall within certain rules. It would be very difficult to predict what value the dice will come up, certainly. But it is not so hard to tell what range the dice will fall between (2 and 12); it is not so hard to tell roughly where the dice will fall upon the tabletop based on the initial direction and force of propulsion; it is not so hard to determine whether the dice will still be the same white and red cubes or if they will crumble into dust on impact, if you know the material from which they are formed; nor any other number of things.
So on what are you founding your knowledge that the average global temperature is something which falls more in the scope of “which side will come up on a die” instead of “what area of the table will a die land if I throw at a certain angle and speed”? You seem to have believed 40 posts or so previous that negative feedback can be trusted to bring the climate back to where humankind wants it, and yet now you are arguing that the climate follows no discernible rules at all and could veer off to being a 300 degree mist of boiling steam tomorrow if the chaotic principles behind it happened to roll that way. It seems rather handy to be able to swing from “It can be trusted to do what I think it ought to do” to “It can’t be trusted enough to be simulated in any way regardless if God himself came down and personally made a simulator” depending on your need at the moment.
And don’t forget to reply to post #91 as promised.
A priori means “an assumption.” An assumption is not an argument. We’re talking about science here, and if you want argue science, you do it on the basis of evidence. Otherwise, you’re better off to say that you really don’t know and leave it at that until you’ve found an answer on one side or the other.
So…what evidence do you have that AGW falls in this range of things that can’t be pinned down? It’s a simple question.
I have no idea what your point is.
