|
Dear Darkwhite,
Climate science automatically attracts people who believe that the climate is greatly affected by human activities. Making a career out of trying to show effects of paramount importance makes much more sense than devoting decades of study to disprove something you find questionable. It is hardly surprising that most climate scientists are positive to idea of human activities affecting the climate. You will similarly find that most astrologers believe that constellations affect people's destinies. Consensus, by itself, means very little. A fine example of scientific consensus gone wrong is given below. It details how a very influential priming experiment, which has been cited thousands of times in psychology papers, has probably been a false positive all along. Note that this is a very simple and straightforward experiment, which can be replicated on a budget of less than ten thousand dollars, and the original paper is from 1996. http://chronicle.com/article/Power-of-Suggestion/136907/Plenty of scientists in all fields - particularly the ones who make headlines in the media - make very bold claims with very dubious evidence. The only vaccine against false prophets is continuous and rigorous experimental verification, which is precisely the thing which is nearly impossible in climate science. The only way to test the accuracy of your models is to wait for the years to tick by. Historically, the short-term accuracy has been terrible, which leaves very little credibility for the long term projections. I won't bother replying to or quoting the last half of your post, which is desperately arguing against claims I have never made. My position is simply to regard climate science with the same sort of moderate skepticism as economics and psychology. All these sciences generally deal with much more complicated systems than physics and cannot test their models with the same relative ease, and also have much less stellar credentials.
you raise a few interesting points. The problem of group think, or that it is easier to think the same thing as the majority of your peers is well known and is discussed in climate science extensively. One of the most famous German climate scientists, Hans von Storch, is a proponent of that argument and repeats it so often, that there is probably no German climate scientist who could say that he has not though about the problem =)
There are also a few people who liken some of climate science to empirical psychologiy, i.e., we ask experts and take the statistical answer as representative of the true state.
this has to do with a few unique aspects of the climate system, as for example in the fact, that there cannot be a single Best model due to the infinite amoutns of subsystems that you can either incorporate or paramterize.
so while I agree with many of your points, I cannot agree with two points:
a) "the only way to test the accuracy of your models is to wait for the years to tick by"
That is untrue. You can check the IPCC chapter on model evaluation of AR4 or AR5 or the hundreds of papers released each year on model evaluation. The line of argument is multifold
- the models should be based on physical principles (i.e., Navier Stokes is essentially Newton)
- the models should be verifiable in parts that have not been tuned( i.e., you tune the mean temperature, if yourmodel gives you the yearly seasonality automatically, you would argue that it has some inherent merits)
- the models should react to external forcings on the short term in a plausible way (i.e., volcanoes. The dissipationfluctuation theorem in physics sais that a model reacts usually in the mode of its biggest variability. So, if we can say that our models react "correctly" in their spatial and temporal fingerprint towards a perturbation as in a volcanoe, we can infer that it is plausible that it reacts similiar to other perturbations"
- connections between seasonality and climate change: the change throughout a year is massive and fast. if the models capture it, it is highly plausible that they capture a much smaller and slower change through increase in co2
b) not all climate scientists are here because they believe in human interference. Actually, probably a majority of us does not care at all. There are so many nerds here as in most natural scientists, and many of us care only about the system, about the laser, the radar, the cloud, the computer and so on. these people just study the climate system. They dont care much for climate change, they are driven by the urge to understand the Earth system better. I just wanted to defend them, even if I fall into your category, I went into climate science because I thought the system was interesting and I wanted to know for myself if all this climate chance talk was true or not =)
Best regards from one fellow physicist to another,
W
|
On August 21 2013 06:57 dabbeljuh wrote:
not all climate scientists are here because they believe in human interference. Actually, probably a majority of us does not care at all. There are so many nerds here as in most natural scientists, and many of us care only about the system, about the laser, the radar, the cloud, the computer...
This is so true. I think the people who claim conspiracy w.r.t scientists and global warming don't realise just how little we tend to care about the real world implications of our research.
I was at a seminar a few months back where this was made more clear to me than ever. The talk was given by some dude from MIT who is studying mutations in the HIV genome as a response to various forcings, model vaccines etc.He had a really cool model for this that he managed to map onto a frigging ising model. Initial clinical tests were looking very promising. The audience was very impressed. As a response to a question/suggestion from a member of the audience he then said: "We welcome all the help we can get, people are dying!" To which he was met with roaring laughter.
Point here being that no one in that room gave a single fuck about the fact that this amazing research was probably going to prevent children from dying and lives from being ruined. All that mattered was the math game. Aids was just a way to get money for playing it.
My guess is that most climate physicists are exactly the same. The mathematics matters, the details of the system matters. Whether future human civilization is going to be able to deal with the earth warming? Who cares?
|
On August 21 2013 11:42 KlaCkoN wrote:Show nested quote +On August 21 2013 06:57 dabbeljuh wrote:
not all climate scientists are here because they believe in human interference. Actually, probably a majority of us does not care at all. There are so many nerds here as in most natural scientists, and many of us care only about the system, about the laser, the radar, the cloud, the computer...
I was at a seminar a few months back where this was made more clear to me than ever. The talk was given by some dude from MIT who is studying mutations in the HIV genome as a response to various forcings, model vaccines etc.He had a really cool model for this that he managed to map onto a frigging ising model. Initial clinical tests were looking very promising. The audience was very impressed. As a response to a question/suggestion from a member of the audience he then said: "We welcome all the help we can get, people are dying!" To which he was met with roaring laughter. Point here being that no one in that room gave a single fuck about the fact that this amazing research was probably going to prevent children from dying and lives from being ruined. All that mattered was the math game. Aids was just a way to get money for playing it. My guess is that most climate physicists are exactly the same. The mathematics matters, the details of the system matters. Whether future human civilization is going to be able to deal with the earth warming? Who cares?
That's interesting that most climate scientists do not care much about the implications of their work. In general I'm sure a fair number of scientists do what they do because they do care about the real world implications their research creates. Doug Melton, one of today's most influential scientists in stem cell and developmental biology, focuses much of his work on the pancreas. One major part of his work focuses on diabetes and ways to cure the disease. Clearly he finds the disease system and cell biology itself very interesting and motivating, but a primary drive behind his decision to pursue a cure for diabetes is the fact that his two kids suffer from it. In other words real world implications following his work very much interest him. I'm sure this attitude is applicable to many scientists across all disciplines.
|
On August 20 2013 19:10 radiatoren wrote:Show nested quote +On August 20 2013 15:05 Darkwhite wrote:On August 20 2013 13:00 radscorpion9 wrote:On August 20 2013 06:51 Darkwhite wrote:On August 20 2013 06:31 GreenGringo wrote:On August 20 2013 05:08 TricksAre4Figs wrote:
Now, having said that, we know that 99% of people are scientifically illiterate and cannot interpret data themselves. We must rely on those professionals who study the field to give their best interpretations. If there is never a debate on the issue and only one side is allowed to speak then I'm afraid I can't believe they are telling the whole truth. A dissenting opinion must be allowed to make it's case in order for the truth to come out. Imagine if astronomers discovered an asteroid heading toward Earth. They checked their calculations, double-checked them, triple-checked them. Every time they found that Earth is in the path of trajectory of the asteroid. In that circumstance would you jump up and make a big song and dance about "dissenting opinion"? If you did, would you be surprised if you tested the patience of most astronomers? The problem is that climate science neither has nor deserves the same level of trust as physics. Climate science is largely in the business of making poorly based claims about the far future while consistently having their models contradicted by the short term developments. Climate science should be lumped with economics and psychology, other fields which similarly apply naive models to much too complicated systems and pretend they have the scientific rigor of a natural science. As Gringo pointed out, most climate scientists are actually trained physicists. If their field really was so complicated, that they couldn't make accurate predictions and that the models all made "embarrassingly" poor predictions as you write, then why is there such a strong, nearly unanimous consensus that human beings are causing global warming? Why is there the same near-unanimous agreement on the degree and range of consequences that will occur as a result? They wouldn't make statements of such clear certainty if they didn't know, or if the models were so poor. Maybe you're right though, maybe they're *all* deceiving themselves or lying to the public for some hidden agenda. Climate science automatically attracts people who believe that the climate is greatly affected by human activities. Making a career out of trying to show effects of paramount importance makes much more sense than devoting decades of study to disprove something you find questionable. It is hardly surprising that most climate scientists are positive to idea of human activities affecting the climate. You will similarly find that most astrologers believe that constellations affect people's destinies. Consensus, by itself, means very little. A fine example of scientific consensus gone wrong is given below. It details how a very influential priming experiment, which has been cited thousands of times in psychology papers, has probably been a false positive all along. Note that this is a very simple and straightforward experiment, which can be replicated on a budget of less than ten thousand dollars, and the original paper is from 1996. http://chronicle.com/article/Power-of-Suggestion/136907/Plenty of scientists in all fields - particularly the ones who make headlines in the media - make very bold claims with very dubious evidence. The only vaccine against false prophets is continuous and rigorous experimental verification, which is precisely the thing which is nearly impossible in climate science. The only way to test the accuracy of your models is to wait for the years to tick by. Historically, the short-term accuracy has been terrible, which leaves very little credibility for the long term projections. I won't bother replying to or quoting the last half of your post, which is desperately arguing against claims I have never made. My position is simply to regard climate science with the same sort of moderate skepticism as economics and psychology. All these sciences generally deal with much more complicated systems than physics and cannot test their models with the same relative ease, and also have much less stellar credentials. Comparing climate science to economics and psychology is wrong. Economics has developed some pretty good mathematical tools already to predict the incredibly complex system, so there is some relevant comparisons there. Psychology relies almost completely on non-objective measures and the systems are very rudimetarily understood. Even psychiatry is moving away from psychology and towards medicine. It is far more relevant to compare climate science to engineering since eventually that will be what the young discipline turns into...
The original post I quoted claimed that climate scientists deserved the same level of trust as astronomer's predicting the trajectory of an asteroid. This is disingenuous. Comparing climate science, as far as subject matter goes, to engineering or physics is fine by me. Pretending that their understanding of Earth's climate is on the same level as engineers who design their own systems, test them thoroughly and carefully study the effects of all sorts of tinkering, is fairly misguided.
On August 20 2013 21:45 GreenGringo wrote:So no, the basic problem of climate change is not especially complicated as far as physics is concerned. There's hundreds of problems in quantum field theory that make this problem look easy.
I disagree.
On August 21 2013 00:17 kwizach wrote:Show nested quote +On August 20 2013 15:05 Darkwhite wrote:On August 20 2013 13:00 radscorpion9 wrote:On August 20 2013 06:51 Darkwhite wrote:On August 20 2013 06:31 GreenGringo wrote:On August 20 2013 05:08 TricksAre4Figs wrote:
Now, having said that, we know that 99% of people are scientifically illiterate and cannot interpret data themselves. We must rely on those professionals who study the field to give their best interpretations. If there is never a debate on the issue and only one side is allowed to speak then I'm afraid I can't believe they are telling the whole truth. A dissenting opinion must be allowed to make it's case in order for the truth to come out. Imagine if astronomers discovered an asteroid heading toward Earth. They checked their calculations, double-checked them, triple-checked them. Every time they found that Earth is in the path of trajectory of the asteroid. In that circumstance would you jump up and make a big song and dance about "dissenting opinion"? If you did, would you be surprised if you tested the patience of most astronomers? The problem is that climate science neither has nor deserves the same level of trust as physics. Climate science is largely in the business of making poorly based claims about the far future while consistently having their models contradicted by the short term developments. Climate science should be lumped with economics and psychology, other fields which similarly apply naive models to much too complicated systems and pretend they have the scientific rigor of a natural science. As Gringo pointed out, most climate scientists are actually trained physicists. If their field really was so complicated, that they couldn't make accurate predictions and that the models all made "embarrassingly" poor predictions as you write, then why is there such a strong, nearly unanimous consensus that human beings are causing global warming? Why is there the same near-unanimous agreement on the degree and range of consequences that will occur as a result? They wouldn't make statements of such clear certainty if they didn't know, or if the models were so poor. Maybe you're right though, maybe they're *all* deceiving themselves or lying to the public for some hidden agenda. Climate science automatically attracts people who believe that the climate is greatly affected by human activities. Making a career out of trying to show effects of paramount importance makes much more sense than devoting decades of study to disprove something you find questionable. It is hardly surprising that most climate scientists are positive to idea of human activities affecting the climate. You will similarly find that most astrologers believe that constellations affect people's destinies. Consensus, by itself, means very little. Holy fallacious analogy, Batman - astrology is not a science, in case you didn't know. The reason climate scientists believe the Earth is getting warmer because of human activity is that the evidence, and the scientific studies on the subject, overwhelmingly point to it being the case. And consensus does mean something (although it is certainly not a guarantee for truth) when it is about interpretation of data and scientific rigor.
I pointed out why consensus does not guarantee scientific validity, both with the example of astrology and an example of psychology, which is indeed a science. The analogy isn't supposed to be an exact match, it is merely supposed to show that the argument from consensus - which is a fallacy - does not hold up.
Of course, climate skeptics will usually always aim for the messenger and the predictions (which, by the way, are made with ceteris paribus assumptions) because they're usually completely unable to dispute the actual data and scientific studies which show that:
I am pointing out why climate scientists are well-educated professionals who have studied this topic for decades, should be viewed with some skepticism. The argument from authority is, fundamentally, about the reliability of the scientists. This has nothing to do with shooting the messenger.
1) the Earth has steadily been getting warmer since the 19th century
2) human activity is almost certainly the driving cause
Instead of making fallacious analogies to social sciences, how about you tell us with which part of the data & the scientific analyses behind these two statements you disagree?
The first is true, and a simple matter of reading data. The second is speculation about the causes of observed effects in a complicated system. There is certainly a component to climate change entirely unrelated to human activity, and it is far from settled which one is the dominant. My specific disagreement, though, is with neither of these two claims, but rather about the reliability of:
- projected climate changes fifty plus years into the future
- the supposedly catastrophic consequences of these changes
By the way, to further counter your initial point about the consensus only existing because of climate scientists' apriori beliefs, here's an example of the exact opposite, namely a physicist (Richard Muller) who had some doubts over the existence global warming itself and who led a research project (which the Koch foundation partially funded) on the matter to determine the validity of the claims that it was real and that mankind was the principal force behind it. He ended up changing his position when he was confronted with the results of his study which established that 1. global warming is real 2. human activity is almost without a doubt the driving cause. He still remains skeptical about other aspects which he has not studied (the magnitude of the consequences on sea levels, for example), but he's now absolutely convinced of the validity of the two propositions I mentioned. Show nested quote +Call me a converted skeptic. Three years ago I identified problems in previous climate studies that, in my mind, threw doubt on the very existence of global warming. Last year, following an intensive research effort involving a dozen scientists, I concluded that global warming was real and that the prior estimates of the rate of warming were correct. I’m now going a step further: Humans are almost entirely the cause.
My total turnaround, in such a short time, is the result of careful and objective analysis by the Berkeley Earth Surface Temperature project, which I founded with my daughter Elizabeth. Our results show that the average temperature of the earth’s land has risen by two and a half degrees Fahrenheit over the past 250 years, including an increase of one and a half degrees over the most recent 50 years. Moreover, it appears likely that essentially all of this increase results from the human emission of greenhouse gases.
These findings are stronger than those of the Intergovernmental Panel on Climate Change, the United Nations group that defines the scientific and diplomatic consensus on global warming. In its 2007 report, the I.P.C.C. concluded only that most of the warming of the prior 50 years could be attributed to humans. It was possible, according to the I.P.C.C. consensus statement, that the warming before 1956 could be because of changes in solar activity, and that even a substantial part of the more recent warming could be natural.
Our Berkeley Earth approach used sophisticated statistical methods developed largely by our lead scientist, Robert Rohde, which allowed us to determine earth land temperature much further back in time. We carefully studied issues raised by skeptics: biases from urban heating (we duplicated our results using rural data alone), from data selection (prior groups selected fewer than 20 percent of the available temperature stations; we used virtually 100 percent), from poor station quality (we separately analyzed good stations and poor ones) and from human intervention and data adjustment (our work is completely automated and hands-off). In our papers we demonstrate that none of these potentially troublesome effects unduly biased our conclusions.
[...]
How definite is the attribution to humans? The carbon dioxide curve gives a better match than anything else we’ve tried. Its magnitude is consistent with the calculated greenhouse effect — extra warming from trapped heat radiation. These facts don’t prove causality and they shouldn’t end skepticism, but they raise the bar: to be considered seriously, an alternative explanation must match the data at least as well as carbon dioxide does. Adding methane, a second greenhouse gas, to our analysis doesn’t change the results. Moreover, our analysis does not depend on large, complex global climate models, the huge computer programs that are notorious for their hidden assumptions and adjustable parameters. Our result is based simply on the close agreement between the shape of the observed temperature rise and the known greenhouse gas increase. ![[image loading]](http://i.livescience.com/images/i/000/029/506/i02/co2-temperature.jpg?1343734352) Source.So, if you're done shooting at the messenger, could we have a discussion over which part of the data and scientific studies which establish the validity of the two propositions I mentioned you disagree with?
I will be done shooting the messenger whenever people stop arguing from authority. See this post, which I initially objected to:
On August 20 2013 06:31 GreenGringo wrote:Show nested quote +On August 20 2013 05:08 TricksAre4Figs wrote:
Now, having said that, we know that 99% of people are scientifically illiterate and cannot interpret data themselves. We must rely on those professionals who study the field to give their best interpretations. If there is never a debate on the issue and only one side is allowed to speak then I'm afraid I can't believe they are telling the whole truth. A dissenting opinion must be allowed to make it's case in order for the truth to come out. Imagine if astronomers discovered an asteroid heading toward Earth. They checked their calculations, double-checked them, triple-checked them. Every time they found that Earth is in the path of trajectory of the asteroid. In that circumstance would you jump up and make a big song and dance about "dissenting opinion"? If you did, would you be surprised if you tested the patience of most astronomers?
I have also claimed, specifically, that there have been worrisome discpreancies between predictions and actual developments. A graph showing disagreement between temperature increase projections and the actual behavior of nature was posted earlier, which is reproduced below. I would like to hear why this keeps happening, and how this is consistent with the idea that people need to shut up and trust the research: Maybe this isn't even one of the accepted models - I honestly don't know.
On August 20 2013 08:43 JonnyBNoHo wrote:Show nested quote +On August 20 2013 08:38 GreenGringo wrote:On August 20 2013 08:16 JonnyBNoHo wrote:On August 20 2013 07:35 GreenGringo wrote:On August 20 2013 06:51 Darkwhite wrote:On August 20 2013 06:31 GreenGringo wrote:On August 20 2013 05:08 TricksAre4Figs wrote:
Now, having said that, we know that 99% of people are scientifically illiterate and cannot interpret data themselves. We must rely on those professionals who study the field to give their best interpretations. If there is never a debate on the issue and only one side is allowed to speak then I'm afraid I can't believe they are telling the whole truth. A dissenting opinion must be allowed to make it's case in order for the truth to come out. Imagine if astronomers discovered an asteroid heading toward Earth. They checked their calculations, double-checked them, triple-checked them. Every time they found that Earth is in the path of trajectory of the asteroid. In that circumstance would you jump up and make a big song and dance about "dissenting opinion"? If you did, would you be surprised if you tested the patience of most astronomers? The problem is that climate science neither has nor deserves the same level of trust as physics. Climate science is largely in the business of making poorly based claims about the far future while consistently having their models contradicted by the short term developments. Climate science should be lumped with economics and psychology, other fields which similarly apply naive models to much too complicated systems and pretend they have the scientific rigor of a natural science. I'm sorry, but you don't understand enough about science to even appreciate the distinctions between different scientific fields. The majority of climate change scientists publishing relevant papers are highly trained in physics and mathematics (like the gentleman running this thread). Many scientists would simply call them "physicists working in climate science". Most physicists would be proud to call them their own. The methods of physics can absolutely be applied to climate science. Thermodynamically, the Earth is not a "complicated system" in the least. It's a system with an atmosphere that's exposed to a radiator. That is why the vast majority of physicists from other disciplines are not in the least skeptical about climate science: because it's so easy to arrive at a first approximation to the problem. Obviously deducing exactly when we're going to die is not easy. But it isn't needed, and that is why anyone who knows the first thing he's talking about on this subject tends to get so testy. I'm pretty sure the Earth is a complicated system. There are quite a few feedback effects that can make increased CO2 output more or less important than the simple greenhouse effect would imply. Ex. increased CO2 can lead to increased water vapor, another greenhouse gas. Or increased CO2 can lead to other things that diminish the effect. I'd agree with you that at first blush it is very intuitive that an increase in CO2 will result in some increase in global temperatures. But that basic intuition isn't very useful when crafting public policy that deals with real costs and real choices. No, you don't name any such "feedback effect". The fact that there isn't any "significant feedback" effect is partly why this is so simple to analyse. Find a reason why pumping more carbon dioxide into the atmosphere means less greenhouse gas density overall and win yourself the Nobel prize for chemistry. And of course intuition isn't sufficient. That is why we have a community of specialists devoted to studying this problem. That is why we don't a bunch of ignoramuses mouthing off and discrediting climate change science with no argument other than a misunderstanding of thermodynamics. I'm quite happy to stop posting and defer to the professionals provided that you do the same. Something tells me you don't like the idea. ![[image loading]](http://cdn.static-economist.com/sites/default/files/imagecache/290-width/images/print-edition/20130330_STC334_1.png)
|
On August 21 2013 06:57 dabbeljuh wrote:Dear Darkwhite, Show nested quote +Climate science automatically attracts people who believe that the climate is greatly affected by human activities. Making a career out of trying to show effects of paramount importance makes much more sense than devoting decades of study to disprove something you find questionable. It is hardly surprising that most climate scientists are positive to idea of human activities affecting the climate. You will similarly find that most astrologers believe that constellations affect people's destinies. Consensus, by itself, means very little. A fine example of scientific consensus gone wrong is given below. It details how a very influential priming experiment, which has been cited thousands of times in psychology papers, has probably been a false positive all along. Note that this is a very simple and straightforward experiment, which can be replicated on a budget of less than ten thousand dollars, and the original paper is from 1996. http://chronicle.com/article/Power-of-Suggestion/136907/Plenty of scientists in all fields - particularly the ones who make headlines in the media - make very bold claims with very dubious evidence. The only vaccine against false prophets is continuous and rigorous experimental verification, which is precisely the thing which is nearly impossible in climate science. The only way to test the accuracy of your models is to wait for the years to tick by. Historically, the short-term accuracy has been terrible, which leaves very little credibility for the long term projections. I won't bother replying to or quoting the last half of your post, which is desperately arguing against claims I have never made. My position is simply to regard climate science with the same sort of moderate skepticism as economics and psychology. All these sciences generally deal with much more complicated systems than physics and cannot test their models with the same relative ease, and also have much less stellar credentials. you raise a few interesting points. The problem of group think, or that it is easier to think the same thing as the majority of your peers is well known and is discussed in climate science extensively. One of the most famous German climate scientists, Hans von Storch, is a proponent of that argument and repeats it so often, that there is probably no German climate scientist who could say that he has not though about the problem =) There are also a few people who liken some of climate science to empirical psychologiy, i.e., we ask experts and take the statistical answer as representative of the true state. this has to do with a few unique aspects of the climate system, as for example in the fact, that there cannot be a single Best model due to the infinite amoutns of subsystems that you can either incorporate or paramterize. so while I agree with many of your points, I cannot agree with two points: a) "the only way to test the accuracy of your models is to wait for the years to tick by" That is untrue. You can check the IPCC chapter on model evaluation of AR4 or AR5 or the hundreds of papers released each year on model evaluation. The line of argument is multifold - the models should be based on physical principles (i.e., Navier Stokes is essentially Newton)
But also, these physical principles are generally removed from the context where we know how to apply them and have verified our understanding. Modelling high-level behavior on the basis of low-level regularities can be anywhere from accurate to useless to misleading.
- the models should be verifiable in parts that have not been tuned( i.e., you tune the mean temperature, if yourmodel gives you the yearly seasonality automatically, you would argue that it has some inherent merits)
This segues right into the machine learning problem of fitting and overfitting. It is somewhat reassuring to have your model correspond with supposedly unfitted data, but it is far from sufficient. With the rampant selection and publication bias in all scientific communities, these sorts of fits can be largely due to chance. Furthermore, the sort of backwards matching the models deal in will be accounting for much less extreme levels of, for instance, CO2, than will supposedly be present in the futures they are intended to predict. Models have a tendency to break down outside of a rather narrow scope of validity.
At this point, climate science has produced countless models with reasonable backwards accuracy which have nonetheless failed the test of time. Whatever else might be true, this clearly shows that backwards accuracy is insufficient and that the predictive power of our models is limited.
- the models should react to external forcings on the short term in a plausible way (i.e., volcanoes. The dissipationfluctuation theorem in physics sais that a model reacts usually in the mode of its biggest variability. So, if we can say that our models react "correctly" in their spatial and temporal fingerprint towards a perturbation as in a volcanoe, we can infer that it is plausible that it reacts similiar to other perturbations"
Again, while these sort of tests should help, they have historically shown themselves insufficient to weed out bad models. Had the models from 2000 matched perfectly, or even reasonably well with the last thirteen years of development, the credibility of climate science would have skyrocketed - though the very long term projections would still have deserved scrutiny.
- connections between seasonality and climate change: the change throughout a year is massive and fast. if the models capture it, it is highly plausible that they capture a much smaller and slower change through increase in co2
Seasonality, however, is a fundamentally different mechanism than longer scale climate change. It can be crudely modeled on the basis of the Earth's axial tilt. For the purposes of proving your model's accuracy in capturing the effects of CO2 levels, whether or not it accounts for entirely different mechanisms is completely orthogonal.
b) not all climate scientists are here because they believe in human interference. Actually, probably a majority of us does not care at all. There are so many nerds here as in most natural scientists, and many of us care only about the system, about the laser, the radar, the cloud, the computer and so on. these people just study the climate system. They dont care much for climate change, they are driven by the urge to understand the Earth system better. I just wanted to defend them, even if I fall into your category, I went into climate science because I thought the system was interesting and I wanted to know for myself if all this climate chance talk was true or not =)
I have never said that all climate scientists believe religiously in human interference. I have said that the consensus is circumstantial and weak evidence. Particularly, scientists agreeing with each other is very suspect if their models don't agree with nature.
|
On August 21 2013 15:07 Darkwhite wrote:
I have also claimed, specifically, that there have been worrisome discpreancies between predictions and actual developments. A graph showing disagreement between temperature increase projections and the actual behavior of nature was posted earlier, which is reproduced below. I would like to hear why this keeps happening, and how this is consistent with the idea that people need to shut up and trust the research: Maybe this isn't even one of the accepted models - I honestly don't know.
No, your graph doesn't show "disagreement". The "actual" line falls within the confidence interval allowed by the computer models at almost every single point on the graph. Furthermore, a general upward trend is predicted within the confidence interval, which is a non-trivial prediction.
If you can't even understand your own cherry-picked data, there can be no point in continuing.
|
Dear Darkwhite,
On August 21 2013 15:45 Darkwhite wrote:
I have never said that all climate scientists believe religiously in human interference. I have said that the consensus is circumstantial and weak evidence. Particularly, scientists agreeing with each other is very suspect if their models don't agree with nature.
just a very quick answer: I agree that it would be suspect, if -as you put it - their models dont agree with nature AND if those models are the only tool or indication for a given scientific hypothesis (manmade CO2 is warming Earths atmosphere in a significant way on a time scale of decades to centuries).
The -AND_ part is completely missing in your argument. Their are multiple lines of evidence, from cores and other paleorecords to simple observations of current temperature increase and TOA (top of the atmosphere) radiations, that show that the increased CO2 concentrations lead to a higher boundary layer and thus an increased surface temperature. (btw the effective radiative temperature of Earth is clearly in the equilibrium case identical with or without CO2, thats simple energy budgets).
I agree with your point of being criticical of the plot you showed. As I have explained several times throughout this plot, one major difference is that the model plots show the ensemble response to a forcing, without singular events like volcanoes or internal variability. It is therefore expected that "reality" as in observations will vary from being on the top to the bottom of the modelled range. The plot is also not chosen very well from a communication point of view, because people dont see that the spread that is shown is the result of many singular model runs that look much "more realistic" than the average, but that cannot be in phase with reality, by construction (climate models for those projections are started WITHOUT initial data from an euqilibrium state with boundary conditions of the 1850s. If they would fit reality exactly, that would most likely be a sign of a bug, they are not constructed or expected to do so).
There is a lot in the upcoming IPCC report on the validity of the projections from the beginning of the 90s (we have not been doing this business for long. The high performance computers of the 90s were less capable than your typical smartphone °) and it appears that so far everything agrees quite well. If the current hiatus continues for a couple of years, we will have to revise the way our model ensembles are structured. This does not change at all the full field of climate science. Just as a guide, probably half of all climate scientist will never use a computer model more complicated than matlab. And less than 10% will ever touch a full GCM. Our science is _NOT_ based exclusively on these complicated models, they are used to hellp us understanding feedback processes, but all projections for policy uses are always a combination of these models, of simple energy balance models and theoretical work based directly on observations.
Best regards, I hope you find some of this interesting, even though maybe technical (but I think that your background will allow for that =)
W
|
On August 21 2013 20:11 GreenGringo wrote:Show nested quote +On August 21 2013 15:07 Darkwhite wrote:
I have also claimed, specifically, that there have been worrisome discpreancies between predictions and actual developments. A graph showing disagreement between temperature increase projections and the actual behavior of nature was posted earlier, which is reproduced below. I would like to hear why this keeps happening, and how this is consistent with the idea that people need to shut up and trust the research: Maybe this isn't even one of the accepted models - I honestly don't know.
No, your graph doesn't show "disagreement". The "actual" line falls within the confidence interval allowed by the computer models at almost every single point on the graph. Furthermore, a general upward trend is predicted within the confidence interval, which is a non-trivial prediction. If you can't even understand your own cherry-picked data, there can be no point in continuing.
You do understand that the predictions of the model were not made in 1950? As far as I have been able to figure out, this particular model was designed in 2008, but I might be misinterpreting something. I hope the Intergovernmental Panel on Climate Change are recognized as serious: http://cmip-pcmdi.llnl.gov/cmip5/
If I understand the data correctly, the relatively good match from 1950-2008 is a trivial matter of backwards fitting, which is merely a question of how many free variables you make room for and says very little about the validity of the model. On the other hand, in the five years since 2008, the only period for which we are actually comparing predictions and results, there is a very significant discrepancy, such that we currently are at the very border of the 5%-significance level. Is any of this incorrect?
On August 21 2013 21:34 dabbeljuh wrote:Dear Darkwhite, Show nested quote +On August 21 2013 15:45 Darkwhite wrote:
I have never said that all climate scientists believe religiously in human interference. I have said that the consensus is circumstantial and weak evidence. Particularly, scientists agreeing with each other is very suspect if their models don't agree with nature. just a very quick answer: I agree that it would be suspect, if -as you put it - their models dont agree with nature AND if those models are the only tool or indication for a given scientific hypothesis (manmade CO2 is warming Earths atmosphere in a significant way on a time scale of decades to centuries). The -AND_ part is completely missing in your argument. Their are multiple lines of evidence, from cores and other paleorecords to simple observations of current temperature increase and TOA (top of the atmosphere) radiations, that show that the increased CO2 concentrations lead to a higher boundary layer and thus an increased surface temperature. (btw the effective radiative temperature of Earth is clearly in the equilibrium case identical with or without CO2, thats simple energy budgets). I agree with your point of being criticical of the plot you showed. As I have explained several times throughout this plot, one major difference is that the model plots show the ensemble response to a forcing, without singular events like volcanoes or internal variability. It is therefore expected that "reality" as in observations will vary from being on the top to the bottom of the modelled range. The plot is also not chosen very well from a communication point of view, because people dont see that the spread that is shown is the result of many singular model runs that look much "more realistic" than the average, but that cannot be in phase with reality, by construction (climate models for those projections are started WITHOUT initial data from an euqilibrium state with boundary conditions of the 1850s. If they would fit reality exactly, that would most likely be a sign of a bug, they are not constructed or expected to do so). There is a lot in the upcoming IPCC report on the validity of the projections from the beginning of the 90s (we have not been doing this business for long. The high performance computers of the 90s were less capable than your typical smartphone °) and it appears that so far everything agrees quite well. If the current hiatus continues for a couple of years, we will have to revise the way our model ensembles are structured. This does not change at all the full field of climate science. Just as a guide, probably half of all climate scientist will never use a computer model more complicated than matlab. And less than 10% will ever touch a full GCM. Our science is _NOT_ based exclusively on these complicated models, they are used to hellp us understanding feedback processes, but all projections for policy uses are always a combination of these models, of simple energy balance models and theoretical work based directly on observations. Best regards, I hope you find some of this interesting, even though maybe technical (but I think that your background will allow for that =) W
I have no problem with the technical parts, though I honestly don't consider them relevant. If the best models climate science can deliver keep making poor predictions, even in the very short term of five to ten years, this raises serious doubts about the validity of their long term warming predictions.
|
On August 21 2013 15:07 Darkwhite wrote:Show nested quote +On August 21 2013 00:17 kwizach wrote:On August 20 2013 15:05 Darkwhite wrote:On August 20 2013 13:00 radscorpion9 wrote:On August 20 2013 06:51 Darkwhite wrote:On August 20 2013 06:31 GreenGringo wrote:On August 20 2013 05:08 TricksAre4Figs wrote:
Now, having said that, we know that 99% of people are scientifically illiterate and cannot interpret data themselves. We must rely on those professionals who study the field to give their best interpretations. If there is never a debate on the issue and only one side is allowed to speak then I'm afraid I can't believe they are telling the whole truth. A dissenting opinion must be allowed to make it's case in order for the truth to come out. Imagine if astronomers discovered an asteroid heading toward Earth. They checked their calculations, double-checked them, triple-checked them. Every time they found that Earth is in the path of trajectory of the asteroid. In that circumstance would you jump up and make a big song and dance about "dissenting opinion"? If you did, would you be surprised if you tested the patience of most astronomers? The problem is that climate science neither has nor deserves the same level of trust as physics. Climate science is largely in the business of making poorly based claims about the far future while consistently having their models contradicted by the short term developments. Climate science should be lumped with economics and psychology, other fields which similarly apply naive models to much too complicated systems and pretend they have the scientific rigor of a natural science. As Gringo pointed out, most climate scientists are actually trained physicists. If their field really was so complicated, that they couldn't make accurate predictions and that the models all made "embarrassingly" poor predictions as you write, then why is there such a strong, nearly unanimous consensus that human beings are causing global warming? Why is there the same near-unanimous agreement on the degree and range of consequences that will occur as a result? They wouldn't make statements of such clear certainty if they didn't know, or if the models were so poor. Maybe you're right though, maybe they're *all* deceiving themselves or lying to the public for some hidden agenda. Climate science automatically attracts people who believe that the climate is greatly affected by human activities. Making a career out of trying to show effects of paramount importance makes much more sense than devoting decades of study to disprove something you find questionable. It is hardly surprising that most climate scientists are positive to idea of human activities affecting the climate. You will similarly find that most astrologers believe that constellations affect people's destinies. Consensus, by itself, means very little. Holy fallacious analogy, Batman - astrology is not a science, in case you didn't know. The reason climate scientists believe the Earth is getting warmer because of human activity is that the evidence, and the scientific studies on the subject, overwhelmingly point to it being the case. And consensus does mean something (although it is certainly not a guarantee for truth) when it is about interpretation of data and scientific rigor. I pointed out why consensus does not guarantee scientific validity, both with the example of astrology and an example of psychology, which is indeed a science. The analogy isn't supposed to be an exact match, it is merely supposed to show that the argument from consensus - which is a fallacy - does not hold up.
I never said that consensus guaranteed scientific validity - in fact, I wrote the opposite: "[consensus] is certainly not a guarantee for truth". What I said was that the analogy you were making with astrology to argue that it was "hardly surprising that most climate scientists are positive to idea of human activities affecting the climate" was completely fallacious. Indeed, the reason "most astrologers believe that constellations affect people's destinies" is a shared belief, arguably pre-existing, unrelated to evidence. The reason climate scientists believe that human activity is affecting average temperatures, meanwhile, is rooted in data and rigorous analysis of the data.
On August 21 2013 15:07 Darkwhite wrote:Show nested quote +Of course, climate skeptics will usually always aim for the messenger and the predictions (which, by the way, are made with ceteris paribus assumptions) because they're usually completely unable to dispute the actual data and scientific studies which show that: I am pointing out why climate scientists are well-educated professionals who have studied this topic for decades, should be viewed with some skepticism. The argument from authority is, fundamentally, about the reliability of the scientists. This has nothing to do with shooting the messenger.
On August 21 2013 15:07 Darkwhite wrote:
I will be done shooting the messenger whenever people stop arguing from authority.
First, an argument from authority is not necessarily a fallacy. Second, like I wrote in my first post, I am not arguing that the two propositions I put forward in my previous post (1. the Earth has steadily been getting warmer since the 19th century and 2. human activity is almost certainly the driving cause) are true because scientists say they're true - I'm defending their validity because the data and rigorous scientific analysis of the data point towards them being true. Third, my reply stemmed from your statement that "consensus, by itself, means very little". The point is that
1) The consensus we're talking about here is not "by itself", it's like I said based on robust data and data analysis
2) The consensus we're talking about does hold more weight than complete disagreement among the scientific community, even though like we mentioned it's not a guarantee for truth.
On August 21 2013 15:07 Darkwhite wrote:Show nested quote +1) the Earth has steadily been getting warmer since the 19th century
2) human activity is almost certainly the driving cause
Instead of making fallacious analogies to social sciences, how about you tell us with which part of the data & the scientific analyses behind these two statements you disagree? The first is true, and a simple matter of reading data. The second is speculation about the causes of observed effects in a complicated system. There is certainly a component to climate change entirely unrelated to human activity, and it is far from settled which one is the dominant. My specific disagreement, though, is with neither of these two claims, but rather about the reliability of: - projected climate changes fifty plus years into the future - the supposedly catastrophic consequences of these changes
Why would you say it is "far from settled"? What are you basing that claim on?
I have also claimed, specifically, that there have been worrisome discpreancies between predictions and actual developments. A graph showing disagreement between temperature increase projections and the actual behavior of nature was posted earlier, which is reproduced below. I would like to hear why this keeps happening, and how this is consistent with the idea that people need to shut up and trust the research: Maybe this isn't even one of the accepted models - I honestly don't know.
As GreenGringo pointed out, the "actual" line falls within the confidence interval on the graph, and as dabbeljuh mentioned, prediction failures have been overstated. Regardless, my reply was not about predictions but about the points I mentioned earlier.
|
On August 22 2013 00:20 Darkwhite wrote:
If I understand the data correctly, the relatively good match from 1950-2008 is a trivial matter of backwards fitting, which is merely a question of how many free variables you make room for and says very little about the validity of the model. On the other hand, in the five years since 2008, the only period for which we are actually comparing predictions and results, there is a very significant discrepancy, such that we currently are at the very border of the 5%-significance level. Is any of this incorrect? Yes. All of it.
First off, you have not cited anything to corroborate your assertion that climate scientists arrive at their models by merely twiddling with free variables. I would be very surprised if such a procedure survived the peer review process.
Second, your idea that a reliable model wouldn't allow for fluctuations at a local level is simply ridiculous. There's fluctuations all over the graph. The ones to occur since 2008 have been pretty standard and in line with the others.
Third, why does it even matter? Nobody thinks climate scientists have a magic crystal ball. They don't need to predict the exact temperature of the Earth, any more than you need to predict the exact moment you'll be attacked if you go into Tehran and burn a copy of the Quran in public. The point is not that they can make predictions which are perfect at a local level...it's that they have a huge body of data which points to global warming having an anthropogenic cause.
|
On August 22 2013 01:08 GreenGringo wrote:Show nested quote +On August 22 2013 00:20 Darkwhite wrote:
If I understand the data correctly, the relatively good match from 1950-2008 is a trivial matter of backwards fitting, which is merely a question of how many free variables you make room for and says very little about the validity of the model. On the other hand, in the five years since 2008, the only period for which we are actually comparing predictions and results, there is a very significant discrepancy, such that we currently are at the very border of the 5%-significance level. Is any of this incorrect? Yes. All of it. First off, you have not cited anything to corroborate your assertion that climate scientists arrive at their models by merely twiddling with free variables. I would be very surprised if such a procedure survived the peer review process. Second, your idea that a reliable model wouldn't allow for fluctuations at a local level is simply ridiculous. There's fluctuations all over the graph. The ones to occur since 2008 have been pretty standard and in line with the others. Third, why does it even matter? Nobody thinks climate scientists have a magic crystal ball. They don't need to predict the exact temperature of the Earth, any more than you need to predict the exact moment you'll be attacked if you go into Tehran and burn a copy of the Quran in public. The point is not that they can make predictions which are perfect at a local level...it's that they have a huge body of data which points to global warming having an anthropogenic cause.
Not to comment on the actual discussion, but about your faith in the peer review process...you would be surprised at the amount of pretty shitty twiddling of free variables that make it to publication. Justifications are given some times, but alternatives are not, aka what would the model look like with a different set of free variables.
Now as a reader/peer what can you do to get to the bottom of it? Go read the paper(s) itself. Otherwise your doubt that "such a procedure" is completely pointless to what Darkwhite wrote, it doesn't matter if YOU believe the methods are biased (pro tip: a lot of science is belief driven), what matters is if Darkwhite's questioning of model validity is based in something substantial.
The concept of the 5% significance itself is a joke when 20 years or so ago in academia, the gold standard was 10%. So actually go look at some of the methodologies being used, and then comment. Darkwhite reads like he has to some extent, others do not.
Edit:
Also people really need to brush up on model validity, along with the various criteria.
Edit2:
Significance levels are completely arbitrary and can be completely meaningless in some contexts. I am not sure why people are buying so much into the significance interval here. Also there's a few statistical concepts people are messing up. Then again, I don't study climate science/meterology or anything related to it, but I would assume the statistical concepts behind it are the same as any other field working with them.
|
On August 22 2013 01:08 GreenGringo wrote:Third, why does it even matter? Nobody thinks climate scientists have a magic crystal ball. They don't need to predict the exact temperature of the Earth, any more than you need to predict the exact moment you'll be attacked if you go into Tehran and burn a copy of the Quran in public. The point is not that they can make predictions which are perfect at a local level...it's that they have a huge body of data which points to global warming having an anthropogenic cause.
It matters a lot, because there is a big difference between an increase of say 2 degrees C by 2100 vs an increase of 4 degrees C by 2100. Policy-makers are trying to use science to make decisions about what laws to pass and how to allocate tax dollars. The mere fact that it's getting warmer -- without being able to predict how much or how fast -- is not that much help in making these decisions.
|
It matters alot because confidence intervals are NOT uniform in their probabilities across the entire interval; in other words, it's not a flat 95% chance for all values within the interval.
|
On August 22 2013 02:12 Judicator wrote:
Not to comment on the actual discussion, but about your faith in the peer review process...you would be surprised at the amount of pretty shitty twiddling of free variables that make it to publication. Justifications are given some times, but alternatives are not, aka what would the model look like with a different set of free variables. Not really. I've had papers rejected for far more nitpicky reasons than a content-free theory whose free variables are tweaked in an arbitrary and ad hoc way to fit to the data. That is the experience of most scientists. The peer review process is stringent. Commit a howler like that and your competitors to scientific glory will seize upon it and make you look like an idiot.
On August 22 2013 02:12 Judicator wrote:Now as a reader/peer what can you do to get to the bottom of it? Go read the paper(s) itself. Otherwise your doubt that "such a procedure" is completely pointless to what Darkwhite wrote, it doesn't matter if YOU believe the methods are biased (pro tip: a lot of science is belief driven), what matters is if Darkwhite's questioning of model validity is based in something substantial. So I'm supposed to read every link posted by every denialist on a forum? No. The burden of proof is on HIM, not on me for subscribing to the overwhelming consensus of climate scientists, borne out by all kinds of disparate and mutually reinforcing evidence they have.
We have a professional climate scientist in this thread and if Darkwhite is curious about the tweaking of free variables he could simply have asked Dabbeljuh.
On August 22 2013 02:12 Judicator wrote:The concept of the 5% significance itself is a joke when 20 years or so ago in academia, the gold standard was 10%. So actually go look at some of the methodologies being used, and then comment. Darkwhite reads like he has to some extent, others do not. And if you read my post, you'll see that that wasn't even a part of the objection. I couldn't care if the projection falls outside the confidence interval altogether as long as it does only a local scale. It's not the job of ANY theory in science to explain every every anomaly and every fluctuation that occurs on a graph.
|
On August 22 2013 02:44 ziggurat wrote:Show nested quote +On August 22 2013 01:08 GreenGringo wrote:Third, why does it even matter? Nobody thinks climate scientists have a magic crystal ball. They don't need to predict the exact temperature of the Earth, any more than you need to predict the exact moment you'll be attacked if you go into Tehran and burn a copy of the Quran in public. The point is not that they can make predictions which are perfect at a local level...it's that they have a huge body of data which points to global warming having an anthropogenic cause. It matters a lot, because there is a big difference between an increase of say 2 degrees C by 2100 vs an increase of 4 degrees C by 2100. Policy-makers are trying to use science to make decisions about what laws to pass and how to allocate tax dollars. The mere fact that it's getting warmer -- without being able to predict how much or how fast -- is not that much help in making these decisions. Sure, it matters if you ALREADY accept global warming and you're just curious about which policy to pursue.
That's not the context of this discussion. We're not talking about the finer points of policy, but we're arguing with denialists.
|
On August 22 2013 02:53 GreenGringo wrote:Show nested quote +On August 22 2013 02:12 Judicator wrote:
Not to comment on the actual discussion, but about your faith in the peer review process...you would be surprised at the amount of pretty shitty twiddling of free variables that make it to publication. Justifications are given some times, but alternatives are not, aka what would the model look like with a different set of free variables. Not really. I've had papers rejected for far more nitpicky reasons than a content-free theory whose free variables are tweaked in an arbitrary and ad hoc way to fit to the data. That is the experience of most scientists. The peer review process is stringent. Commit a howler like that and your competitors to scientific glory will seize upon it and make you look like an idiot. Show nested quote +On August 22 2013 02:12 Judicator wrote:Now as a reader/peer what can you do to get to the bottom of it? Go read the paper(s) itself. Otherwise your doubt that "such a procedure" is completely pointless to what Darkwhite wrote, it doesn't matter if YOU believe the methods are biased (pro tip: a lot of science is belief driven), what matters is if Darkwhite's questioning of model validity is based in something substantial. So I'm supposed to read every link posted by every denialist on a forum? No. The burden of proof is on HIM, not on me for subscribing to the overwhelming consensus of climate scientists, borne out by all kinds of disparate and mutually reinforcing evidence they have. We have a professional climate scientist in this thread and if Darkwhite is curious about the tweaking of free variables he could simply have asked Dabbeljuh. Show nested quote +On August 22 2013 02:12 Judicator wrote:The concept of the 5% significance itself is a joke when 20 years or so ago in academia, the gold standard was 10%. So actually go look at some of the methodologies being used, and then comment. Darkwhite reads like he has to some extent, others do not. And if you read my post, you'll see that that wasn't even a part of the objection. I couldn't care if the projection falls outside the confidence interval altogether as long as it does only a local scale. It's not the job of ANY theory in science to explain every every anomaly and every fluctuation that occurs on a graph.
I'm on your side here, but you aren't understanding how the burden of proof works. You're saying that the burden of proof is on him, and thats true, but if he provides a bunch of sources then that IS his proof. You don't have to read them if you don't want, but you can't go around saying hes got no proof when you just don't want to read the proof he is providing.
|
On August 22 2013 03:17 packrat386 wrote:Show nested quote +On August 22 2013 02:53 GreenGringo wrote:On August 22 2013 02:12 Judicator wrote:
Not to comment on the actual discussion, but about your faith in the peer review process...you would be surprised at the amount of pretty shitty twiddling of free variables that make it to publication. Justifications are given some times, but alternatives are not, aka what would the model look like with a different set of free variables. Not really. I've had papers rejected for far more nitpicky reasons than a content-free theory whose free variables are tweaked in an arbitrary and ad hoc way to fit to the data. That is the experience of most scientists. The peer review process is stringent. Commit a howler like that and your competitors to scientific glory will seize upon it and make you look like an idiot. On August 22 2013 02:12 Judicator wrote:Now as a reader/peer what can you do to get to the bottom of it? Go read the paper(s) itself. Otherwise your doubt that "such a procedure" is completely pointless to what Darkwhite wrote, it doesn't matter if YOU believe the methods are biased (pro tip: a lot of science is belief driven), what matters is if Darkwhite's questioning of model validity is based in something substantial. So I'm supposed to read every link posted by every denialist on a forum? No. The burden of proof is on HIM, not on me for subscribing to the overwhelming consensus of climate scientists, borne out by all kinds of disparate and mutually reinforcing evidence they have. We have a professional climate scientist in this thread and if Darkwhite is curious about the tweaking of free variables he could simply have asked Dabbeljuh. On August 22 2013 02:12 Judicator wrote:The concept of the 5% significance itself is a joke when 20 years or so ago in academia, the gold standard was 10%. So actually go look at some of the methodologies being used, and then comment. Darkwhite reads like he has to some extent, others do not. And if you read my post, you'll see that that wasn't even a part of the objection. I couldn't care if the projection falls outside the confidence interval altogether as long as it does only a local scale. It's not the job of ANY theory in science to explain every every anomaly and every fluctuation that occurs on a graph. I'm on your side here, but you aren't understanding how the burden of proof works. You're saying that the burden of proof is on him, and thats true, but if he provides a bunch of sources then that IS his proof. You don't have to read them if you don't want, but you can't go around saying hes got no proof when you just don't want to read the proof he is providing. No, I'm afraid the tactic of referring to a large study and making vague claims about it is not an honest debating technique even between scholars who CAN be relied on to read the studies.
In a forum like this it is utterly inappropriate. Because frankly, lots of people who do it in forums haven't read the papers they refer to. They're merely pretending to have read stuff. Anyone can do it. A 14-year-old who doesn't know Newton's laws could come in here and post a link to an arcane paper and assure us that he's read it.
If he's going to claim that they tweaked the free variables to fit the data, then I want proof. Which particular study does it? On what page do they make it clear?
|
The models don't have to be right (to make right predictions) to be true. It is something that is very important for social sciences such as economy, where most models are "accepted" as "true" when they refer to mecanisms that can be quantified and observable empirically.
The problem is that models (in economy) are way much simplier than what really happen, and even economists agree with that - for exemple, the principal model used by the FMI doesn't have a banking system, or the model used by most central bank doesn't recognise the existence of money.
Going back to climate change, I suppose (I don't know shit about that) that earth is a complicated system (even if the physic mecanisms that are needed in order to understands it are really simple) where a thousands of variables are needed in order to really "understand" and thus predict how the climate could evolve, but that doesn't really mean that the idea of an anthropogenic induced climate change is wrong, just that we don't really know where we are going.
Being too ambitious with models is, in my opinion, a grave mystake in this regard.
|
On August 22 2013 02:53 GreenGringo wrote:Show nested quote +On August 22 2013 02:12 Judicator wrote:
Not to comment on the actual discussion, but about your faith in the peer review process...you would be surprised at the amount of pretty shitty twiddling of free variables that make it to publication. Justifications are given some times, but alternatives are not, aka what would the model look like with a different set of free variables. Not really. I've had papers rejected for far more nitpicky reasons than a content-free theory whose free variables are tweaked in an arbitrary and ad hoc way to fit to the data. That is the experience of most scientists. The peer review process is stringent. Commit a howler like that and your competitors to scientific glory will seize upon it and make you look like an idiot. Show nested quote +On August 22 2013 02:12 Judicator wrote:Now as a reader/peer what can you do to get to the bottom of it? Go read the paper(s) itself. Otherwise your doubt that "such a procedure" is completely pointless to what Darkwhite wrote, it doesn't matter if YOU believe the methods are biased (pro tip: a lot of science is belief driven), what matters is if Darkwhite's questioning of model validity is based in something substantial. So I'm supposed to read every link posted by every denialist on a forum? No. The burden of proof is on HIM, not on me for subscribing to the overwhelming consensus of climate scientists, borne out by all kinds of disparate and mutually reinforcing evidence they have. We have a professional climate scientist in this thread and if Darkwhite is curious about the tweaking of free variables he could simply have asked Dabbeljuh. Show nested quote +On August 22 2013 02:12 Judicator wrote:The concept of the 5% significance itself is a joke when 20 years or so ago in academia, the gold standard was 10%. So actually go look at some of the methodologies being used, and then comment. Darkwhite reads like he has to some extent, others do not. And if you read my post, you'll see that that wasn't even a part of the objection. I couldn't care if the projection falls outside the confidence interval altogether as long as it does only a local scale. It's not the job of ANY theory in science to explain every every anomaly and every fluctuation that occurs on a graph.
My post wasn't pointed at yours specifically even though I did quote yours.
Some thoughts, if you really think the peer review process is that stringent, then you have never submitted a paper, sat on a review committee, or in general tried to publish with your livelihood on the line. It's not as stringent as you think, and biased results gets by review committees all the time even at the prestigious journals like Nature/Science. I am not saying review committees are chaired by idiots, but I am saying that review committees are very much human. There's no guideline for model variables, and different people have different opinions on model construction, it's totally up to the researcher and their reasons for building their model.
As for mutually reinforcing evidence, what's going on here has happened before in the history of science, just because the academic masses agree with something, does not make it the truth. Since when did reading what the other side has to say become such a no-no? Just because a denialist links something, doesn't mean you shouldn't at least read it and give some consideration. I mean if you care enough to argue/debate the topic on an internet forum, what precludes you from considering both sides of the debate?
Edit:
So why do you choose to believe the methodologies of the most while not believing methodologies or a few? Solely based on numbers? That's pretty dangerous line of thinking in science and one that would still have us disregarding quite a few important scientific theories if that was the case.
Edit2:
Just to be clear, I am being skeptical here of both sides.
|
Dear Darkwhite,
On August 22 2013 00:20 Darkwhite wrote:You do understand that the predictions of the model were not made in 1950? As far as I have been able to figure out, this particular model was designed in 2008, but I might be misinterpreting something. I hope the Intergovernmental Panel on Climate Change are recognized as serious: http://cmip-pcmdi.llnl.gov/cmip5/If I understand the data correctly, the relatively good match from 1950-2008 is a trivial matter of backwards fitting, which is merely a question of how many free variables you make room for and says very little about the validity of the model. On the other hand, in the five years since 2008, the only period for which we are actually comparing predictions and results, there is a very significant discrepancy, such that we currently are at the very border of the 5%-significance level. Is any of this incorrect?
Yes, most of it. As I mentioned in my earlier post (where you dismissed the technical aspects as not so important), those runs are not backwards fitted (btw fitting in a highly nonlinear high dimensional system with models that are incredibly expensive to run is not as trivial as a polynomial fit, it is a challenge, though). They are started as free running runs in 1850 from non initialised equilibrium runs. While some mean quantities are fitted in the model construction process, things like the warming in the 20C are not. This is a model response. the full behaviour in the 20C can be used as a test, not only starting from 2003. The only difference is the forcing (i.e., the model inputs): from 1850 to 2001 it is "observed", from 2001 onwards it is projected. This does not really matter for the short period, though. It is amazing that the model runs are so close to reality (or better, that reality behaves so much like the ensemble mean.), some of that can be explained by the anomaly reference period of 1960-1980 (in this period by construction the anomalies are zero for all runs and reality). It is expected and correct behaviour that reality will over and undershoot the xx% signifinance interval of the multimodel ensemble - for 100-xx% of the time.
I have no problem with the technical parts, though I honestly don't consider them relevant. If the best models climate science can deliver keep making poor predictions, even in the very short term of five to ten years, this raises serious doubts about the validity of their long term warming predictions.
its a pity that you dont consider them relevant, as they are for understanding the peculiarities of the climate system.
For example, contrary to intuition, It is much harder to do shortterm then longterm. I know the metaphor is old, but imagine a hot pot of water. It is easy to predict that it will boil after a certain while of increased radiative forcing, it is however physically impossible to predict when and where the first bubbles will appear. Chaotic properties of the climate system _prevent_ exact predictions, the only thing that people can do with models are probabilistic expectations. Given a million Earths, we expect with our current models that - on average - they would warm as the ensemble mean. It is physically and mathematically impossible to predict the singular trajectory of the system beyond its "memory" ... while it is possible to predict, that a change in boundary conditions will lead to a change in mean behaviour after a while, when the forced change is bigger than natural variability (see again detection and attribution).
Best regards to everyone, (and say hi to the NSA°)
W
|
|
|
|
|
|
EPT
