Suppose that astronomers around the world alerted us that a large asteroid is headed in our direction, and might collide with the earth in the year 2012. The astronomers cannot give us a precise probability of collision because of many imponderables. The laws of physics are, of course, uncontroversial, but there is some disagreement about the precision of the instruments used to measure the location of small objects at great distances. It is also, of course, possible that the asteroid could be deflected by another object before it reaches the earth. And astronomers concede that they do not know everything there is to know about outer space. When pressed, the astronomers will say only that the scientific consensus is that a collision is “very likely,” and that, if it occurs, the consequences will be catastrophic. To build a defense system—say, rockets that would intercept the asteroid and knock it off course—would cost hundreds of billions of dollars.
As is always the case, there are a few dissenters. A highly regarded astronomer at MIT points out that astronomers have made incorrect predictions in the past and argues that a collision is possible but not likely. A few physicists who do not specialize in astronomy argue that scientific instruments are not as precise as some people think, and the astronomers’ risk estimate should therefore be distrusted. A scandal erupts when emails at the West Anglia Space Research Unit are released, and shows that some scientists tried to arrange a boycott of a journal that published a few articles of the skeptics. At the same time, thousands of astronomers not connected with the West Anglia Unit continue to insist that the risk of a collision is very high.
A few questions. In this scenario, would there emerge an industry of non-credentialed “astronomy skeptics” in the press and public comparable to the current batch of “climate skeptics”? My instinct is that the world would quickly get to work building the rocket system, and disregard the views of the skeptics. Is this right or wrong? If it is right, is there some reason to think that climate science and astronomy are different, justifying the skepticism about climate science that does not (yet) exist about astronomy?
[I fear that some will snipe that responsible astronomers would not make such a prediction because of the difficulty of making long-term forecasts given current technological limitations. If so, let’s just move the hypothetical into a future period when the technology is good enough to produce a “scientific consensus” but not complete agreement about all particulars among all people.]
Roach says:
What if these scientists use “computer models” that don’t predict that Venus is hotter than Earth and closer to the sun with a shorter year?
And judging by our foot-dragging on a million and one issues, not least entitlement reform, I’m not so sure we’d get to work building the rocket. Instead, we’d probably have a million blue ribbon commissions working on endsuring that the work force on that rocket “looks like America.”
December 3, 2009, 4:52 pmAnon Y. Mous says:
You forgot to include the part where they rig the “normalizing” of the data to produce the desired result. And, oh by the way, they had to throw out the original data because they were short on space.
December 3, 2009, 4:56 pmMatthew Carberry says:
If they got rid of or refused to provide the original telemetry data (from gov’t owned telescopes no less) that they used to make their calculations and used computer models, also not willingly shared, with degrees of error and non-repeatability as great as the chance of impact, and there ws good evidence that only a few centuries ago human civilization did some flourishing during a similar impact event, then yeah, I might not jump right on board the rocket train.
December 3, 2009, 4:59 pmSoronel Haetir says:
I would say that even now asteroids are better understood than climate. There are only a very few potential types of interaction, gravity, jetted material like that from comets and light pressure.
The first dwarfs the second and third, and the second dwarfs the third just as much.
On the other hand climate is an incredibly complex system with many unknowns.
I would also say that building an asteroid defense system, no matter how expensive, is not asking people to permanently reduce their standard of living. Once the known asteroid danger is past the greatest portion of the sacrifice is over. That is true whether the danger passes because the asteroid missed or the defense system worked. A second asteroid defense would likely be far less expensive due to lessons learned and having a pre-made design.
December 3, 2009, 4:59 pmRichard Aubrey says:
You forgot another issue:
December 3, 2009, 5:02 pmSuch collisions have happened and been catastrophic. Remember the end of the dinosaurs? Hell, Carl Sagan used that to buttress his nuclear winter schtick.
So far, global warming has not happened in the last fifty million years except to the extent that it’s been a great relief to shivering cave men. Other critters don’t count.
In this case, the astronomers can say, it’s going to happen AGAIN and be terrible AGAIN.
To make the analogy, you’d have to be certain that other astronomers don’t have access to current ballistic data or the means to gather it. Don’t even need past data. Just this week’s and next week’s and a big computer.
Poor analogy.
JBD says:
In your hypothetical it seems like how large the asteroid is is settled….making it a much easier call. Plus, we can’t ask for tribute from the asteriod maker(s) for the damage they may cause, or require them to change its course, so we just have to get on with mitigation/adaptation plans, if we do anything at all….again much easier rallying call.
December 3, 2009, 5:02 pmBlue says:
The problem, Professor Posner, is that it is absolutely clear that dropping an asteroid on my head is a Very Bad Thing. It’s a binary Very Bad Thing as well–either it occurs or it does not. Climate change on the other hand will have a mixture of positive and negative effects, will occur over time, and can be ameliorated short of preventing the phenomena.
Oh, and of course the astronomers will also be able to direct us to a specific location of the sky where anyone with a large enough telescope can track the motion of the object precisely.
In short, the hypothetical diverges too far from climate change to be instructive.
December 3, 2009, 5:02 pmAnderson says:
Wow. Reading Posner’s post, I actually *felt* this blog getting appreciably smarter.
December 3, 2009, 5:02 pmDavid Walser says:
Other than “wasting” the money spent on building the rocket system (which may be needed sometime in the future even if it’s not in 2012), there would be little downside to trying to avoid the worst case of the asteroid hitting the earth with cataclysmic results. That’s not true with the steps being proposed to prevent AGW. Cutting CO2 emissions back to 1900′s levels would have cataclysmic results that are only preferable when compared with the worst case predictions for AWG. We’re being asked to accept the lower living standards — complete with the associated shorter lifespan — to avoid a 1 or 2 degree increase in the world’s mean temperature. Since science cannot even tell us for sure whether or not such an increase would be “bad”, it seems foolish to take steps that we KNOW will produce a very bad result in order to prevent something that might not be bad at all.
December 3, 2009, 5:02 pmBlue Neponset says:
If one large industry, say the energy industry, for example, were to benefit from doing nothing then I am sure they would find a way to outrage enough luddites to sow confusion and prime the “astronomy skeptics” pump.
December 3, 2009, 5:04 pmCrunchy Frog says:
It’s not going to hit us until sometime in the late 21st Century (we don’t know when and if, exactly) but we have to DO SOMETHING, and we have to do it RIGHT NOW.
December 3, 2009, 5:04 pmDavid Mader says:
Isn’t a key difference that the physics of an asteroid collision are essentially entirely uncontroversial, and the only real debate would be about likelihood, whereas the physics of anthropogenic climate change remain precisely the subject of disagreement? It’s not just that there’s skepticism over the likelihood of its occurring; it’s that there’s skepticism over the mechanism by which it is thought to occur. If an asteroid of any considerable size hits the earth, bam – bad news bears. But the comparison only works if you take for granted that climate change equals bad news bears – and that’s precisely the question raised by (a) the… quirky data used to date and (b) the admitted inability of even that data to explain recent climate trends.
December 3, 2009, 5:05 pmKenvee says:
Did the same astronomists predicting the asteroid strike in 2012 also predict that there would be a strike in 1995… and then when it didn’t happen, in 1998… and then when it didn’t happen, in 2001… and then when it didn’t happen, in 2005? One reason I’m skeptical of the AGW advocates is that they’ve been predicting dire consequences that would happen right now, but then they keep shifting their timeline back when the deadline approaches and nothing has happened. (And in fact admit in the emails that they’ve been changing their data to hide the fact that their predictions aren’t coming true.)
December 3, 2009, 5:07 pmSteve Goldman says:
I think your intuition is correct. In this situation, I believe the world would get busy on a plan to send Bruce Willis to destroy the asteroid.
I also think the analogy fails, because the effects of climate change are diffuse and not obviously catastrophic. A massive asteroid strike would be. We’d face a choice between a miss and disaster. In the case of predicted climate change, we are dealing with uncertainty as to the chance of the actual change as well as uncertainty of its effects (bad vs. good).
December 3, 2009, 5:07 pmAlyssa says:
If, like climate science, legitimate questions were raised as to
(i) the likelihood of the asteriod hitting the Earth,
(ii) mankind’s ability to stop the asteriod, even if it was going to hit the earth,
(iii) the consequences of the impact, assuming it hits, and
(iv) mankind’s ability to mitigate the impact,
then yes, I think that at least Americans would have varying degrees of support for immediately building a rocket costing in the trillions.
December 3, 2009, 5:07 pmLou Wainwright says:
Where I believe the analogy breaks down is that it equates the two catastrophes. The Asteroid event has these properties: Boolean, discontinuous, unquestionable global harm, short time horizon. None of these are present in Global Warming. Assuming warming is occurring it is not clear what the magnitude, rate, or equilibrium temperature of the warming is. It is a continuous process that can be constantly monitored and evaluated, and whose costs and benefits can be reevaluated as knowledge changes. Depending on the magnitude of the warming the harm may be minor, major, or even slightly negative. Certainly, unlike in the Asteroid case, it would not be a homogeneous impact all over the Earth. Some regions would benefit/pay unequally.
Most critical is the time horizon. Take your Asteroid case and change nothing but the date. Set it at 2052. Do we start building rockets now? Of course not, we study, debate, plan, for roughly 40 years. So let’s extend the hypothetical to include a cost benefit for rapid response. Let’s say that it takes 10 nuclear tipped rockets to make us safe if they are launched in 2010, and 100 if launched in 2050. Let’s further say that each rocket costs $250B. Would we pay today? Almost certainly not. Not if there was even a 10% chance that the Asteroid wouldn’t hit. No government is going to pay that kind of money until they are driven to by a deadline…certainly not to save money in the future. When the economic costs of global warming become directly measurable we will start to see action get taken, not before, no different than any other major catastrophe.
By the way, considering the fact that we don’t actually know what the chance is of a devastating asteroid strike in 2052, and taking into account the discussion above, I would strongly recommend that taking funding from Global Warming research and putting it into asteroid monitoring would be a rational action.
December 3, 2009, 5:08 pmOren says:
Not really. Even if you knew the current position/velocity of the object, calculating its trajectory is an exercise in numerical solutions (small mercy here — at least the gravitational Hamiltonian is symplectic) so it’s not trivial to figure out where it’s going.
December 3, 2009, 5:08 pmCrunchy Frog says:
Smarter in this case defined as “the degree that the OP agrees with Anderson”.
(Period placed outside quote marks, where it clearly belongs. Obviously.)
December 3, 2009, 5:09 pmKen Arromdee says:
Others have said what amounts to the same thing, but basically the answer is that this isn’t a very good analogy, and the changes you’d need to do to make it a good analogy would mean that yes, there should and would be a lot of skeptics around.
Here’s yet a third hypothetical for you. For some reason (perhaps 9/11, but perhaps not), a group of people are convinced that Islamic terrorism is a threat to our society. They come up with the solution of expelling all the Muslims. Of course, if they are wrong, the solution will end up hurting a lot of people who are causing no problems. There are a few dissenters who disagree, some thinking there’s no threat from Islamic terrorism at all, some thinking that there is a threat but not one that justifies extreme measures that hurt a lot of people in the name of preventing a catastrophe that may or may not happen. Would this situation lead to an industry of anti-expulsion skeptics?
I should darn well hope so.
December 3, 2009, 5:10 pmJim says:
Until recent events, I had never taken a look at how climate science was done and I have to say I was shocked. Given how important this is, the way the data is being treated is simply appalling.
Why the asteroid analogy doesn’t work for me is that we can’t (rightly or wrongly) imagine the people hunting down the asteroid being so cavalier.
Could you imagine an observatory saying that they made a number of observations of the asteroid’s location but they weren’t going to show anybody what they were?
And if they used tree rings to forecast the asteroid’s position, wouldn’t the first thing you’d do is go measure as many trees as you could find?
I’d be seriously upset if this is how scientists were behaving in the face of a serious disaster. So what am I suppose to conclude now.
December 3, 2009, 5:11 pmBlue says:
True, over the medium to long term the influence of countless gravitational wells leads to uncertainty. What we DO know precisely, however, is where it is NOW…which was the point I was trying to make.
December 3, 2009, 5:13 pmAnderson says:
Very few of the comments are addressing the point of the post.
However sensible it may be to wonder whether global warming is a good or a bad thing, whether it’s preventable, etc., what we see is “an industry of non-credentialed ‘climate skeptics’ in the press and public” who are attacking both the existence of global warming, and the contribution of human civilization to that warming.
The part about building a missile system is, I take it, Posner’s way of illustrating what general acceptance of the scientific consensus would entail, in the case of his analogy.
… Personally, I think we’re screwed, because China and India just are going to raise their emissions exponentially, for political reasons, whether the rest of the world likes it or not, and they have like two billion people between them.
December 3, 2009, 5:16 pmD.R.M. says:
To make the situation truly parallel:
1) The asteroid impact date should be estimated for some point after 2050, not 2012, and starting construction of the rocket system construction in 2030 instead of today should have no actual impact on readiness.
2) The loudest advocates of building the rocket system should have been pushing for the building of the system back ten years before the asteroid was discovered.
3) The loudest advocates of the rocket system should be unalterably opposed to even studying cheaper, alternate defense mechanisms.
Okay, now that we’ve made your question actually parallel, do you think anybody would actually think we need to start building the defense system today, instead of spending the next twenty years thoroughly reviewing the science and chasing down all the loose ends?
December 3, 2009, 5:17 pmMe says:
No, the skeptics in your hypothetical would not be anywhere as near as influential as climate skeptics.
This is because climate skeptics are actually proponents of maintaining a global status quo. Global warming proponents are essentially saying that everyone all over the world has to change his or her way of life immediately. This is an easy proposition to be skeptical about.
I don’t know much about climate science, but I think the influence of climate skeptics comes more from the incredible inertia behind modern day life, than problems with the science.
There are also powerful arguments on the skeptics side that render the science unimportant.
For example, many skeptics argue that, human-caused or not, warming and cooling cycles have always occurred in the Earth’s climate, and people should adjust to them, rather than try to prevent them from happening.
December 3, 2009, 5:18 pmNaG says:
The main thing that makes the analogy fail is that a rocket system is a creation that could have various other uses (including defense-related), while the proposed solution for AGW is to curb creation. Paying too much money for something of little benefit is preferable to having your standard of living forcibly reduced.
December 3, 2009, 5:20 pmAnderson says:
human-caused or not, warming and cooling cycles have always occurred in the Earth’s climate, and people should adjust to them, rather than try to prevent them from happening
Human-caused or not, meteor collisions have always occurred in the Earth’s history, and people should adjust to them, rather than try to prevent them from happening.
December 3, 2009, 5:21 pmCDU says:
I think that’s because the unsuitability of the analogy is overwhelming the the point Eric was trying to make.
December 3, 2009, 5:24 pmDavid Bernstein says:
My understanding,and correct me if I’m wrong, is that climate scientists acknowledge that the reason they have confidence in their models of how carbon etc affects global weather is because the historical weather data has confirmed what their models would predict. But what if it turns out that the weather data was manipulated, or inexact? Or what if the earth simply happened to be in a natural warming trend over the last century, so the correlation with the theory was just accidental? This doesn’t strike me as having the exactness of physics. (And what if the astronomers had lowered their prediction of the damage from the collision substantially over the last few years, after having expressed great confidence in their previous estimates?) And is it really true that the climate scientists agree that the consequences of global warming will be catastrophic? I don’t follow the controversy as much as I might, but I have never quite understood why warming along the level that occurred naturally in the past, when humanity was poorer and more vulnerable, will be “catastrophic,” as opposed to a “serious nuisance” or some such.
December 3, 2009, 5:27 pmwws says:
You have left out a very crucial part of your analogy. You begin with universal scientific acceptance of the danger – and by doing that you assume your way past the only truly crucial issue. It’s a trick, to steal a phrase from Phil Jones.
It would be much more analagous to the current situation if you said “Suppose a Prophet arose who stated that God had told him that a massive asteroid was coming to destroy the Earth, even though no one could see it yet. Would the world be justified in building a rocket defense system on the basis that the Prophet might just be right about what God had told him to do?”
Would you support building the Rocket Defense System under those circumstances? What if I told you that the Prophet and his supporters were going to become fantastically wealthy from your efforts – would you still do it?
On that topic, Al Gore just canceled his trip to Copenhagen, the conference that was supposed to be the crowning jewel of a decades long effort for him. Looks like the rats are jumping off the sinking ship.
December 3, 2009, 5:30 pmDavid Schwartz says:
I would like to think that it would matter who was right. I think to try to analyze this without considering how much merit the arguments on each side have is madness.
December 3, 2009, 5:30 pmBlue says:
“The part about building a missile system is, I take it, Posner’s way of illustrating what general acceptance of the scientific consensus would entail, in the case of his analogy.”
Actually, there would likely NOT be consensus on the action to take. As it turns out, the composition of the asteriod makes a great deal of difference in how it reacts to, say, explosions. Observations, including a manned mission to scout it out, would need to be undertaken. There are also potential solutions (e.g., solar sails, low impact mass drivers) that could be employed short of a flotilla of nuclear missiles.
December 3, 2009, 5:32 pmNickM says:
The attempt to compare the 2 situations is so loaded that it’s hard to know where to begin.
Let’s start with the presumption that consequences will be catastrophic. 20-inch sea level rise over the next century should not strike any rational person as catastrophic.
“Daddy, the polar bears” makes for a cute commercial, but most Americans would have a hard time finding catastrophe in anything that happens to a predator they only see in zoos and on TV.
Heat-related deaths are likely more than offset by a drop in cold-related deaths.
Precipitation – higher in some areas, lower in others.
All sorts of ecosystem changes talked about (movement of animals and crops, etc.) – not catastrophic, especially if you live in a developed country.
Any mention of Bangladesh – most people already recognize that Bangladesh has been a hellhole since long before it was Bangladesh
As another reason that comparison is absurd, a major meteorite impact either happens or it doesn’t. It’s not like climate, which has naturally fluctuated greatly (it wasn’t AGW that ended the Ice Age).
Also, the existence and size of the asteroid would presumably be unquestioned. Climate science doesn’t even have that agreement as to historical temperatures.
To answer your loaded question, though, astronomical predictions of meteorite impact don’t appear to be a cover for Luddites trying to shut down modern technology and society. Nor do they appear to be a method for the environmentalist movement to lecture others as to moral superiority.
Nick
December 3, 2009, 5:33 pmMike McDougal says:
What if the asteroid researchers stood to gain massive research funding and prestige by exaggerating the risk?
December 3, 2009, 5:33 pmDangerMouse says:
… Personally, I think we’re screwed, because China and India just are going to raise their emissions exponentially, for political reasons, whether the rest of the world likes it or not, and they have like two billion people between them.
Anderson, we’re not screwed. Global warming is a scam, so there’s no need to cower in bed anymore over the Climate Monster.
Given that Al Gore is still a jet-setting fat cat, it’s obvious that even he doesn’t think we’re screwed.
I’m wondering if ANYONE really believe this stuff at all. Actions speak louder than words.
December 3, 2009, 5:34 pmMe says:
In response to the Anderson’s mocking parrallelism, thanks for that;
Warming and cooling cycles are a lot less sudden, thus it would be possible for people to adjust to them in the long term. At least in the popular conception of asteroid armageddons, this is not the case.
Besides, building a missile to blow up asteroids would be a very good example of adjusting to falling asteroids anyway.
Another reason building a missile system would not stir up many skeptics is that is inherently a pro-active solution, and it would be good for economies all over the world. Most global warming policies focus on negative controls like restricting carbon output.
December 3, 2009, 5:34 pmPersonFromPorlock says:
Incidentally, Professor Posner, what do you propose to do about the Mayan end-of-the-world prophecy for 2012? It was, after all, produced by an official government agency using standard methodology, achieved consensus (oh, those obsidian knives!) and has not been disproved.
December 3, 2009, 5:35 pmSunTzu's Nephew says:
And there is no record of the ‘normalization’ that they did…..
December 3, 2009, 5:35 pmRandy says:
” Cutting CO2 emissions back to 1900’s levels would have cataclysmic results that are only preferable when compared with the worst case predictions for AWG. We’re being asked to accept the lower living standards — complete with the associated shorter lifespan — to avoid a 1 or 2 degree increase in the world’s mean temperature. Since science cannot even tell us for sure whether or not such an increase would be “bad””
Wrong on several accounts. First, I haven’t heard a scientific consensus that we need to bring CO2 levels back to 1900 levels. Rather, they are arguing that first we need to stop the *increase* in levels. That alone will be a tough enough job, but it can be done. And it doesn’t require that we go back to horse and buggy days. Rather, it means that we use new technology which already exists to reduce those levels. Buildings, for example, account for a huge amount of CO2 and global warming. We can first require that all new buildings meet or exceed LEED standards. China does, but somehow we can’t find the will to do it.
There are a million small things that we can do that collectively will make a dramatic impact that won’t change your standard of living one iota. In Europe, hotels are routinely fitted with sensors that turn on hall lights only when people are actually walking in the hall. I hardly think that our standards of comfort are dropped when empty halls are dark. And, not surprisingly, that would save you money.
December 3, 2009, 5:36 pmAnderson says:
when humanity was poorer and more vulnerable
Coastal metropolises do not make us “less vulnerable” to rising sea levels.
People genuinely curious about “what we know for sure” might wish to see this article, whose last section has that subtitle.
December 3, 2009, 5:36 pmCGordon says:
I think there are clear differences between the asteroid and climate change. First, no one claims that climate change will happen so soon. Second, and related, the proper method of dealing with climate change is likely to be just as divisive as whether it is in fact a threat. Plus, I don’t know that the reaction would be different if the prescription to the asteroid was to curtail industrial output and impose “big international government” on libertarian Americans.
December 3, 2009, 5:37 pmmatt d says:
Alternate hypothetical:
Let’s say that the CDC discovered a strain of syphilis that was incurable, turned people into mass-murdering lunatics, and was considerably easier to transmit than previously. Then suppose that number of cases were anecdotally increasing, leading to a few high profile polar bear deaths, er, I mean shopping mall shootings.
Then some scientists make a computer model suggesting that increasing numbers of infections will cause us all to be dead within twenty years, unless everyone stops having pre- or extramarital sex. They propose international treaties banning pr0n and sex toys and creating police powers to monitor singles bars and make sure nobody leaves with someone they’re not married to.
Lots of people welcome this. They figure the rampant immorality in the west is finally coming home to roost, and that western countries had better hurry up and quit rolling in the hay quite so much, and, ind fact, most of the cases are in the west. Meanwhile, the developing world’s cases are increasing much faster, but they’ve got little capability to police singles bars so everyone concentrates on the west.
Now, those scientists refuse to show their work. They won’t explain how their models work except in vague generalities. They won’t give up their raw data, or explain how everyone’s either going to die from infection or die because they got murdered by a raving syphilitic.
Skeptics complain that its all overblown, that a few cases here and there don’t make a deadly epidemic, and that the loss of freedom isn’t worth it. They’re denounced as fornicators in the pay of big porn.
Now you’ve got a parallel hypothetical.
December 3, 2009, 5:38 pmSkyler says:
Clearly this is an attempt to compare the hypothetical to the now proven fraud of global warming.
The difference is that an asteroid would be one sudden impact, and if the likelihood of an impact is significantly high, that could be verified mathematically in a straightforward way. There might be disagreements on results but it could be debated without hiding massive amounts of historical data.
The now proven fraud of global warming is different. No one has ever been good at predicting weather and the change expected would happen over time. Even were the debate honest instead of the proven pack of lies that it was, there would still be time to wait and see preliminary results to prove the theory were it not for people lying and using thuggish tactics.
With an asteroid, we would only want to wait until a certain point where we could no longer act to prevent the disaster. If we had to, we could build a lot of rockets really fast.
I think if we put enough of these now proven lying scientists in jail for their attempt to defraud the world’s governments, then we’ll have done a lot to restore our trust in science again.
December 3, 2009, 5:38 pmRandy says:
One thing we know for sure: Certain religionists would say that we shouldn’t do anything because slamming an asteroid into earth is clearly God’s plan for ending the world. Some would no doubt welcome it because it would mean the End Times are arriving, and would prepare for the 1000 year reign of Christ.
I am struck by the fact that many of these religionists blogs anticipate have anticipated and welcomed the coming rapture whenever there is new violence in the middle east.
December 3, 2009, 5:39 pmSunTzu's Nephew says:
At least you can point your scope at the hypothetical corrent portion of the sky, and see it. What if it was there on some nights, but other nights with equal viewing conditions it was simply not there? The AGW people can’t explain the tree-ring discontinuity for the last near 50 years, they can’t explain the medieval ice age, they can’t explain the lack of warming the last ten years (except to say that it’s normal variation), they can’t explain the effects that the sun’s output (or lack thereof) has…
Smoke and mirrors, mirrors and smoke. Pay no attention to the man behind the curtain.
And to call science what they’ve been peddling: When any apostate is attacked, threatened with having their professional credentials revoked, where they engage in criminal conspiracy to avoid FOIA inquiries, where they engage in fraud against the US government…. Where their computer code looks like it was written by a thousand chimps banging away on a thousand keyboards, and runs about as well….
December 3, 2009, 5:43 pmPintler says:
I concur – people would all be happy about battling global warming if it meant using cloth grocery bags, but when you start talking about lowering the thermostat, not flying, turning in the SUV, and so on – well, even Al Gore isn’t willing to do those things.
That is my assessment as well.
FWIW, my usual analogy: you are in a lifeboat at sea. In the mad scramble off the sinking ship, no one got an exact location, and the reasonable expectation is that you are somewhere between 2 weeks and 2 months from shore. You have two weeks rations. Should you go on half rations now? Quarter rations? Or hope for the best and stay on full rations?
December 3, 2009, 5:43 pmRoach says:
If I may make a somewhat silly analogy: astronomers and physicists put a man on the moon. The weather man, even today, is often wrong to the point of comedy. In other words, there is science and there is science. And people who don’t know much about physics or meteorology can see this difference very clearly.
December 3, 2009, 5:44 pmDrGrishka says:
Well, isn’t there already a general agreement that some asteroid may hit Earth at some point in time? In other words, there is no specific asteroid that is likely to do it soon, but given the history of astroids hitting our and other planets there is a real non-zero risk that it would happen again. Yet, no one is rushing to build defenses. In part, I suspect because the risk is not commensurate with the cost.
It is even worse with climate science. At least we KNOW that asteroids have hit Earth before and therefore can again. We can judge the risk based on history. With climate science we don’t even know that much.
December 3, 2009, 5:44 pmAnderson says:
the now proven fraud of global warming
Oh, brother.
December 3, 2009, 5:45 pmSunTzu's Nephew says:
Got a cite, because AFAIK, their models do not correctly back-predict the known conditions….
December 3, 2009, 5:45 pmmariner says:
Another question: what exactly do you consider credentials?
The rhetorical sleight-of-hand by alarmists is to pretend that only those published in “peer-reviewed” journals are credentialed, when the alarmists themselves get to decide what is published.
Skeptics include people with PhD-level education and experience in fields relevant to the debate, but alarmists dishonestly dismiss them as non-credentialed.
In short, to global-warming alarmists “non-credentialed” means “doesn’t agree with us”.
A nice gig for them, and I hope it’s coming to a screeching halt. Too bad we can’t tar and feather the bastards.
December 3, 2009, 5:45 pmDoc Merlin says:
I think mild global warming would be a good thing, not a bad thing.
December 3, 2009, 5:46 pmAn asteroid crashing into the earth would be a bad thing.
wws says:
david bernstein wrote: “My understanding,and correct me if I’m wrong, is that climate scientists acknowledge that the reason they have confidence in their models of how carbon etc affects global weather is because the historical weather data has confirmed what their models would predict.”
You’ve put your finger on one of the crucial problems – great effort has been put into building models that can *back* calculate observed records, but *None* of those models has ever successfully predicted the next years temperatures successfully – they have repeatedly failed miserably. The same problem comes up in software programs that can successfully back calculate stock market moves but are completely useless in predicting future moves. (those programs are a dime a dozen, btw)
In fact, from the leaked e-mails we now know that the CRU models were not even able to successfully back calculate the observed data, which seems to be why the data-tweaking began. There was a steep temperature “blip” in the forties which could not be accounted for, so rather than fix the model they fixed the data by erasing the observed temperature rise.
And of course the models did not predict the recent temperature activity, which is why Phil Jones was so eagerly admonishing his colleagues to “hide the decline.”
The models Do Not Work. They never have, and that’s one of the biggest problems with the entire schema. To restate your opening sentence, Climate scientists claim to have confidence in their models because they will be ostracized, denied access to journals, and blocked from all future funding unless they carefully toe the Party Line.
And that is how all ideological orthodoxies are maintained.
December 3, 2009, 5:49 pmyankee says:
I would expect the reverse. In the case of global warming, there are plenty of well-heeled interests—the coal industry, the oil industry, the auto industry, etc.—who will lose out big if any significant steps to mitigate global warming are taken. They have every incentive to manufacture a controversy and plenty of money to do it with.
In the asteroid case, the people who stand to lose out are future taxpayers who will have to pay off the debt taken off to fund the asteroid system. Since they don’t exist yet, they have no ability to manufacture a controversy. If anything, I’d expect moneyed interests to exaggerate the threat, since defense contractors would stand to make big money off it.
December 3, 2009, 5:49 pmSunTzu's Nephew says:
And to carry the analogy a bit further, lets not forget the UN’s Chairman of the Rocket Defense Committee who’s telling everyone that they have to suffer, while he lives life large on the public dime.
I mean for Christs’ Good Sake, his flying from NYC to India for a DAY, to attend an amateur cricket team PRACTICE? But flying is evil and the rest of us need to be taxed heavily to prevent us from doing it?
December 3, 2009, 5:50 pmgeokstr says:
And even had a guru who was so unintelligent he flunked out of divinity school (how stupid do you have to be to do that?) set to become a billionaire by hyping the whole thing?
December 3, 2009, 5:52 pmCalderon says:
In this scenario, would there emerge an industry of non-credentialed “astronomy skeptics” in the press and public comparable to the current batch of “climate skeptics”? My instinct is that the world would quickly get to work building the rocket system, and disregard the views of the skeptics. Is this right or wrong? If it is right, is there some reason to think that climate science and astronomy are different, justifying the skepticism about climate science that does not (yet) exist about astronomy?
So I think a lot of the posts have answered the last question; I’d like to go back to the earlier one. I’m surprised, especially for someone steeped in L&E, that you’d think the world would quickly get to work building the rocket system. There would be the same sort of collective action, free rider, etc. problems that there are now. Who would pay for the missile system? How would those costs be apportioned? Where would it be built? Public actors who wished to lessen the burden on their countries, or private actors who did not want their resources pulled away to build the rocket system, would argue over the probability of a collision as one way to try to reduce or minimize the costs they would have to pay. Thus, there would be reasons for a similar kind of skeptic to developed as the climate change skeptics there are now.
Also, in reading this post I couldn’t help but think of how people like Gregg Easterbrook have expressed concerns over possible asteroid collisions and argued NASA should be working on such issues, and how Brad DeLong (among others) stated that the probability of a catastrophic asteroid collision was so low that there was no point in worrying about it compared to other issues.
December 3, 2009, 5:53 pmAnderson says:
This Scientific American article is also helpful, for those who pay attention to such things.
December 3, 2009, 5:53 pmtvk says:
I think one important difference is that you are postulating the asteroid striking in 2012, with immediate bad consequences. Change the hypothetical to the strike occurring in 2112, and ask the world to spend billions to build the rockets today, and the calculus changes considerably. I should add that this is not mere myopia. One of the most difficult parts of the climate change calculus is that, although everybody agrees that the danger is still some time away, the alarmists are arguing that we must act now or it will be irreparable. To take the asteroid analogy, it is saying that we must start building the rockets today even though the asteroid cannot be detected yet. That alarm may be warranted, but it does not comport with most of human experience and intuition, so it naturally faces skepticism.
December 3, 2009, 5:56 pmSkyler says:
Anderson says:
“the now proven fraud of global warming
Oh, brother.”
And the fact that there are people who STILL don’t see the fraud is pathetic.
It was obvious to anyone who had ever studied how to measure data. You can’t get accuracy to tenths of a degree from data that is measured, at best, to within 3-5 degrees. It was always a fraud and now we’ve just got their own word on it.
But religions are like that. People will believe in whatever magical mystical things that they want to believe no matter the evidence. With many religions they even claim that it’s important NOT to have evidence because it is belief without proof that is more important. Let’s hope that this doesn’t become a trait of the now proven global warming fraud’s adherents.
December 3, 2009, 5:56 pmShelbyC says:
Prof Posner, would you quit your job and work on the rocket system for, say, minimum wage, under those conditions?
December 3, 2009, 5:56 pmgeokstr says:
This has got to pretty much be the worst analogy I’ve ever seen. Let me guess, Mr. Posner, you believe in Mann-made global warming, right, and that we have to just do something, anything, before the planet melts?
December 3, 2009, 5:56 pmBrett says:
At the very least, there would be a massive rush to get new instruments in place to verify it as soon as possible, with time being such a critical factor. A corresponding situation in the global warming situation would be if the scientists had “very likely” but not “assured” information that the West Antarctic Ice Sheet was going to collapse entirely by 2012, causing sea level to rise rapidly by several meters.
I’d support the major action in case of the asteroid, but I admit that I have ulterior motives – a side-effect of that mobilization would probably be a super-charged space program (manned and unmanned).
December 3, 2009, 6:04 pmwws says:
That Scientific American article is a good example of how the AGW industry has taken once reputable scientific outlets and turned them into pseudo-relgious shills. This is why the intentional corruption of the peer review process as exposed in the CRU e-mails has been so damning. I see no hope for “Scientific American” to ever regain it’s once stellar reputation; it’s science is now no better than that found in the National Enquirer, and may in fact be a bit worse.
Within 5 years, “scientific american” will be bankrupt and another once proud name will be in the dustbin of history. Articles insulting the bulk of its readership such as this one will be the prime culprit.
December 3, 2009, 6:05 pmDavid Chesler says:
Would I have to listen to folks say “See, the asteroid is coming!” every time an acorn drops on their heads?
December 3, 2009, 6:05 pmAnderson says:
That Scientific American article is a good example of how the AGW industry has taken once reputable scientific outlets and turned them into pseudo-relgious shills.
Right after making that statement, you might want to illustrate your point with one or two facts to corroborate.
Otherwise, some of us might go, “hm, Scientific American versus some blog commenter guy … whom to believe?”
December 3, 2009, 6:08 pmguy in the veal calf office says:
As many have pointed out, that’s a terrible analogy and a waste of a thought experiment.
Why change teh subject to the widening interest in how measurements are compiled, what natural systems can cause forcing, how the GCMs are coded, etc? That stuff is juicy and conversation worthy. Instead of talking about those issues and tehir component parts, you laboriously squeez out a clunky analogy?
Last, when a paleochronologists compiles measurements through statistical normalizing, you sound silly in dismissing the statistician who looks it over as an amateur.
December 3, 2009, 6:09 pmroad2serfdom says:
The astronomers would have to be claiming humans are causing the asteroid to hit earth. Their theory might be that asteroids are attracted to human flesh and our population growth is steering the asteroid toward us. Their solution would be free condoms, abortions, forced sterilization, and targeted mass starvation.
Skeptics of the man-made asteroid theory but who think the asteroid heading toward earth would be the ones proposing a missile defense system, but they would be kicked out of their jobs by the consensus scientists for disbelieving the human flesh theory. Skeptics of the asteroid hitting earth would not want to spend money on a defense system.
The asteroid defense solution is analogous engineering solutions some are taking about now to cool the earth by putting chemicals in the upper atmosphere.
December 3, 2009, 6:12 pmDonald Kilmer says:
You forgot to add that Dick Cheney and Halliburton are marketing the asteroid rocket defense system.
December 3, 2009, 6:14 pmTomM says:
A lot of people have suggested why the comparison here is inapt for a variety of very specific reasons. I actually do a bit of work in searching for asteroids that might destroy the Earth, so I thought I’d jump in with a more general comment.
The 19th century view of science was strongly influenced by the success of celestial mechanics in making incredibly accurate predictions on the basis of very simple equations of just the type that Professor Posner uses in his scenario. This led to a view of science as something where we just needed to understand the underlying principles and then we would be able to answer any question about the system.
The 20th century showed that this was a very simplistic view. Not only were the underlying principles far more complex — or at the very least less intuitive — than we might have anticipated. It also became clear that the properties of complex systems are emergent, they are not really deducible from first principles. E.g., you can’t understand simple objects in the real world as a direct application of quantum mechanics. Even when the underlying laws are straightforward, complex non-linear interactions make prediction hard. The time evolution of the three-body problem is still a subject of active research in celestial mechanics and that is practically the simplest system we can think of.
So we’ve learned in the past century is that it may simply not be possible to answer all questions about a complex system, including what its time evolution is, certainly not without experiments on the scale of the system itself.
In understanding the interaction of scientific inquiry with public policy, I think it’s better to understand that celestial mechanics and the view of science that it suggests is entirely at odds with what we have discovered over the past century or more. We need at least a 20th century — or better yet 21st — view of science when we make policy decisions based upon it.
December 3, 2009, 6:28 pmdr says:
Or do you go to double rations, sneer at the people arguing for half-rations, and spend the next two weeks telling everyone not to mess with their “lifeboat way of life”?
December 3, 2009, 6:31 pmMichelle Dulak Thomson says:
Is it worth pointing out that there is, in fact, general agreement that an incredible amount of stuff is floating around the solar system that we haven’t located, and that the chances of our getting clobbered by something large enough to do serious harm, sooner or later, are pretty good? IIRC a rather large rock came within a few Earth diameters in March; a smaller one came damn near (~1 Earth diameter) last month.
Yet not only don’t we see general agreement that we ought to build some sort of missile defense pronto, but the only people who seem seriously interested in doing anything are those who want more money going to space exploration and exploitation in any case.
I don’t know what we’d do with a convenient two-year lead time to a possible asteroidal impact. My pessimistic suspicion is: not much, and (of that) too much panic relative to thought. Larry Niven and Jerry Pournelle gave Earth six months’ lead time in Lucifer’s Hammer, and I’m afraid they got the likely reaction just about pitch-perfect.
December 3, 2009, 6:37 pmDavid Schwartz says:
If you don’t think the SciAm piece is a sham, read their answer to claim six and do just the most minimal critical analysis.
December 3, 2009, 6:39 pmkeypusher64 says:
I felt the same way when reading the comments.
December 3, 2009, 6:42 pmSkyler says:
Or, if you live next to a grocery store, and the stores been open every day for the past twenty years, there’s no reason to think you need to go on starvation rations when the only rationale for doing so is because Chicken Little down the street has been proven to be lying that the store’s delivery truck is broken and can’t be fixed unless you fork over a few trillion dollars.
December 3, 2009, 6:44 pmdr says:
Wait, there’s a grocery store boat?
My head hurts.
December 3, 2009, 6:45 pmMatt s says:
There are no monied business interests that suffer from a decision to build a rocket (or bomb). There are no structures in society (social or economic) that we would expect to be radically changed by building a rocket.
What concentrated and monied interest group loses if the US or the industrialized west or the world pays to build a rocket?
December 3, 2009, 6:47 pmRobert Bloomfield says:
Most of the commenters have missed one important difference. While it might be expensive to counter the asteroid, money is fungible–it can come from any source, and does not cause particular industries to suffer. Nor would the response require changes in daily lifestyles. Responses to global warming harm some industries and create immediate inconveniences. Both factors generate a strong motivation for skepticism.
December 3, 2009, 6:47 pmMatthew Carberry says:
Anderson says:
This Scientific American article is also helpful, for those who pay attention to such things.
It cites to climate data which is now seriously under question and models running that questionable data built from shaky code that have clearly demonstrated weaknesses and non-reproduceability.
It then hand-waves alternative explanations (water vapor, solar, etc.) and recent temperature observations by tweaking the math and then claiming the flawed models “take them into account”.
How is that at all definitive?
It’s a house made of bricks without straw built on the sand.
December 3, 2009, 6:47 pmSkyler says:
It’s bad analogy day at the Volokh Conspiracy!
December 3, 2009, 6:53 pmRobert Ayers says:
Other commentators have refuted most of the claimed parallels. But they have left one out: “thousands of astronomers not connected with the West Anglia Unit continue to insist that the risk of a collision is very high”.
December 3, 2009, 7:06 pmIn fact it is false that thousands of climatologists think that imminent “warming” will be catastrophic.
People oft quote “2500 scientists …” but that is the number of people who provided input to the IPCC. By no means all of them agree with the IPCC paper. And in fact the IPCC does not predict imminent doom and destruction — that is a fringe of Gore-ists.
SG says:
My understanding is that most asteroids & comets are discovered by non-credentialed amateur astronomers.
I don’t think non-credentialed carries as much weight as you seem to think. Arguments ought to be judged without regard to the letters following the arguer’s name.
December 3, 2009, 7:06 pmSkyler says:
Okay, I’m reviewing the Scientific American article.
Claim 1: They use circular logic to defend the CO2 argument by relying on a computer model, which we’ve now proven are based on lies.
Claim 2: Hockey stick was a fraud, but despite the evidence they still believe.
Claim 3: They still claim to be able to measure the “global” temperature within tenths of a degree when the base temperature measuring stations are only accurate within 3-5 degrees in modern days and much worse 100 years ago.
Claim 4: They rely on the proven lies of climatologists again. No credibility.
Claim 5: They claim that since some guy 150 years ago had a theory (based on what science from 150 years ago?) about global warming, then the attempts to prove him correct must be legitimate even though the argument has been based on modern lies. Since everyone relies on the lies of the IPCC, then it’s not a huge conspiracy stretching back 150 years ago. It was only an unproven theory that was only later claimed to be true when a small, modern set of people are now proven to have conspired to defraud us. They also denigrate those that recognize the NOW PROVEN conspiracy by comparing them to UFO chasers. In the range of bad arguments, that’s about the lowest.
Claim 6: The conspiracy need not have been based on money for the researchers to be true. Power is also a good motivator.
Claim 7: This is just absurd. They put in this one to throw us off. If the now proven fraud of global warming doesn’t need to be staved off, there’s no need for any action whatsoever.
Yeah, if SciAm was ever a good organization, it’s certainly not now.
December 3, 2009, 7:07 pmEMB says:
I don’t see no need for something like this to make a valid analogy. In fact, I suspect a large part of the point of this post is that in the current climate debate, people are too caught up in a moral “irresponsible humans caused this crisis, so they must fix their evil ways” point of view when they should simply be asking the questions “with what likelihood and to what extent is humanity in trouble?” and if the answer is that there’s some chance of serious problems, “what should we do to deal with it?” (And yes, the question of whether human activity has caused global warming has led to perhaps too much emphasis on changing that human activity and not enough on other possibly solutions.)
Replacing a possibly man-made disaster with a clearly not man-made disaster shouldn’t matter in figuring out what to do about it, though in practices from a public policy point of view it somehow seems to matter a great deal.
The real problem with this analogy is that the science is in much better shape for predicting the trajectories of asteroids than it is for the climate. There are certainly a few unexplained anomalies in this sort of thing, but my impression is that for the most part with good telescopes and a ton of computer power, this sort of thing can be predicted fairly reliably (and those resources would surely be made immediately available if someone discovered a potential hit).
With climate, there are lots of models out there, and none of them are especially good. The system they are trying to predict is just absurdly complicated (e.g. every living thing on earth contributes), and there are many complicated models, each of which oversimplifies a great many important aspects of the problem.
As a result, the probability estimates for a catastrophe vary a great deal, and are likely to fluctuate a great deal in the future as new models are developed, etc. By contrast, with the asteroid example, I feel like the basic models are generally agreed on, and it’s just a question of getting better raw data and running more careful computer simulations, so I would expect the estimates would quickly improve.
December 3, 2009, 7:12 pmBillB says:
But SciAm’s once considerable reputation is in tatters under its present ownership and recent editors. The amount of material by ‘journalists’ and ‘science writers’ has increased tremendously, pushing out material by actual working scientists.
Under former editor John Rennie, pushed out earlier this year after significant decline in revenues, SciAm became a shill for the Climate ‘Chicken Little’ movement and a spokesman for the environmental left, its hysterical polemics nowhere more obvious than in its crusade to pillory Bjorn Lomberg and his work.
It’s too early to tell whether SciAm can recover; much of the current decline clearly follows a fifteen year dumbing down of the content and a shift to political shouting instead of reporting. Much of what we’ve seen to date was probably in the pipeline before current editor Mariette DiChristina was appointed. Given her background at Popular Science it is probably too much to hope for.
December 3, 2009, 7:12 pmThe Awful Truth says:
If an asteroid were really headed toward Earth, does anyone think the current administration could actually do anything about by 2012?
Or do you think they would just issue a report after impact showing how many millions more lives would have been lost if their healthcare bill hadn’t passed?
December 3, 2009, 7:13 pmwws says:
re: Michelle Dulak Thomson – Lucifer’s Hammer was a great read, and I agree, Pournelle got the reaction just about perfectly.
Now we’re not seriously talking about building an asteroid defense system, but if we were, I would very much oppose it, and here’s why – Every big new technical advance or system created by man throughout our history has always ended up being used in a war by the people who’ve got control of it against the people who don’t. It seems inevitable to me that any system capable of deflecting an asteroid away from Earth is going to be just as capable of deflecting an asteroid *into* the Earth, and that as soon as it is built someone (or several powerful someones) are going to realize this and try to seize control of it. Doesn’t have to be an asteroid big enough to destroy the world, just one big enough to destroy the continent that whoever’s in control doesn’t like. (The fight for that control alone could be enough to set off the 3rd world war.) Sooner or later, that weapon is going to get used and thus our attempt to avoid catastrophe could end up creating the very thing we were trying to avoid.
That seems to be a constant in human history – I’d rather not spend a whole lot of money to create the gun that is going to end up being pointed at my own head. I’ll take my chances with a random nature rather than a too powerful and capricious man, thank you very much.
December 3, 2009, 7:18 pmDavid Ong says:
Global warming will, at worst, cost a small fraction of GDP, but mostly in many decades to come.
Your hypothetical asteroid would, at worst, wipe out most of humanity.
99.9% GDP cost vs. 2-3% GDP cost: 2 orders of magnitude difference.
And that assumes that climatology is on as solid grounding as physics.
December 3, 2009, 7:19 pmMalvolio says:
You mean Scientific American of “Science defends itself…” fame?
The article is entitled (for those of you wise enough not to bother going to the site) “Seven Answers to Climate Contrarian Nonsense”. That’s some even-handed reportage you’ve got there, Lou. It has some laugh lines like:
“anthropogenic CO2 amounts to about 30 billion tons annually—more than 130 times as much as volcanoes produce.” (Because volcanoes aren’t big CO2 producers)
“95 percent of the releases of CO2 to the atmosphere are natural, but natural processes such as plant growth and absorption into the oceans pull the gas back out of the atmosphere and almost precisely offset them, leaving the human additions as a net surplus.” (How stupid do they think their readers are? Plants’ consumption of CO2 precisely balances the net release, not because of some huge lucky coincidence but because the plants are gobbling up all the CO2 they can — and if we release more, the plants will grow faster to consume more.)
“Contrarians holding out for a natural explanation for current global warming” (Uh, the world is currently getting colder, remember?)
Most of the article is just quotes from the usual suspects, IPCC et al.
December 3, 2009, 7:19 pmSealionii says:
Further questions: Does a trillion-dollar industry which forms (for better or worse) the engine, more or less literally, of the world economy benefit from public skepticism about the asteroid threat?
Is one or both sides of the asteroid-collision debate strongly associated with a given social or political philosophy?
If either of those questions could be answered “yes”, then I think there would be swarms of uncredentialed skeptics about the asteroid threat, from a) economic motives, b) sheer bloody-mindedness.
December 3, 2009, 7:21 pmMichelle Dulak Thomson says:
SG,
My understanding is that most asteroids & comets are discovered by non-credentialed amateur astronomers.
Quite right. Astronomers are the last people I can imagine sneering at uncredentialed “outsiders.”
EMB,
[...I]n the current climate debate, people are too caught up in a moral “irresponsible humans caused this crisis, so they must fix their evil ways” point of view when they should simply be asking the questions “with what likelihood and to what extent is humanity in trouble?” and if the answer is that there’s some chance of serious problems, “what should we do to deal with it?”
Very true. Whether, and to what extent, warming is human-caused ought to matter only in the sense that everything we can find out about what’s happening here is potentially useful. It should make no difference to the possible solutions who (if anyone) caused what. A firefighter, analogously, might find it useful to know that a fire was set deliberately, in that it might provide clues to where it started, where accelerant, extra fuel, &c. might have been put, and so forth. But the point is to fight the fire.
December 3, 2009, 7:26 pmRyan Waxx says:
Suppose that the Pope alerted us that our souls are in danger of going to hell. The clergy cannot give us a precise assessment of the chances of our going to hell, because exactly like AGW “science”, religious vision is completely unfalsifiable. The clergy concede…
I think you get the picture. The “we need to do this because there is a possibility of a catastrophe” is the worst kind of fearmongering anti-logic. It seeks to drown out rational discussion of the probabilities and the alternatives because of the terrible, terrible consequences of being wrong.
The real question is… why are you trying to drown out rational discussion? Is there some problem with your fact set or the credibility of your movement? I wonder why that might be…
December 3, 2009, 7:30 pmDavid says:
Astronomy has a 6000 year pedigree.
Climate “science” has a–oh, let’s be generous–60 year pedigree.
…Any questions?
December 3, 2009, 7:31 pmAlanW says:
Calderon touched on it, but I think the criticism of the analogy is misplaced. As I understand it, the most effective way to divert an asteroid would be to land a series of rockets on it and gradually shift its orbit so it no longer intersects the Earth’s (simply blowing it up would, of course, shower the entire planet with asteroid remnants, which is unlikely to be a preferable outcome). This would require a lot of time, money, effort and involve a great deal of uncertainty. The costs would be borne by the richest, most technologically advanced nations, while the benefits would accrue to the whole world (although especially whoever was under ground zero, which might be, for purposes of the analogy, Bangladesh or the Maldives or wherever).
Shout about the science all you like, but I think the analogy itself is pretty decent.
December 3, 2009, 7:32 pmMichelle Dulak Thomson says:
wws,
It seems inevitable to me that any system capable of deflecting an asteroid away from Earth is going to be just as capable of deflecting an asteroid *into* the Earth, and that as soon as it is built someone (or several powerful someones) are going to realize this and try to seize control of it.
Well, but “first, catch your asteroid,” you know. I think it’d be a lot easier to deflect one than to get one close enough to use it as a weapon … oh, you mean, in the moment someone might make it land “accidentally on purpose” on top of someone else? Maybe. Some people really do hate to let a good crisis go to waste.
December 3, 2009, 7:32 pmarch1 says:
The quality of discussion concerning even the factual aspects of these issues among even a well-informed group such as this is IMO way below where it could be if the available information and intelligence were at all effectively marshalled.
As one concrete example, I keep seeing comments from people who appear to think that if there is no AGW threat, then there is no GW threat worth our consideration. I think that I could make a pretty conclusive argument that this narrowly defined assertion is just plain wrong. But I also think that there’s very little value to my doing so, because even in the modest context of that one topic on this one blog, the argument would convince few, be forgotten by most of those, and be ignored by the vast majority.
I guess what I’m acknowledging in a grumpy way is that blog commenting is almost entirely nonprogressive. That’s life, you say, don’t expect a cat to fetch sticks. Well, maybe.
December 3, 2009, 7:33 pmMichelle Dulak Thomson says:
arch1,
As one concrete example, I keep seeing comments from people who appear to think that if there is no AGW threat, then there is no GW threat worth our consideration.
Um . . . there’s EMB just above arguing that it doesn’t matter whether GW is “A” or not, and me just below that agreeing with him/her.
December 3, 2009, 7:43 pmShelbyC says:
Isn’t the point that if it’s not “A” there’s not likely to be much we can do about it?
December 3, 2009, 7:50 pmvic says:
I find the current controversy interesting, not the least because it allows one to evaluate the analytical thinking abilities of many of our public intellectuals.
Sorry Dr Posner, with all due respect, I dont care how accomplished you are …. you kinda-sorta failed that test.
But wait… you are in exalted company… The publication I had hitherto thought as the pinnacle of science… the journal “Nature” just had an editorial. http://www.nature.com/nature/journal/v462/n7273/full/462545a.html
The editorial is sooooo reflective of dispassionate scientific discourse and so not a political hack job…. not!
But seriously, Dr Posner, you shouldn’t be too distraught, we as a specie seem to be hardwired towards religious belief. I use that term somewhat unconventionally to suggest that we are hardwired to uncritically accept certain dogma’s and rituals. Many of todays elites have loosely distanced themselves from traditional religious belief systems. However many of these do not have strong intellectual underpinnings to their atheism/ agnosticism- they were just following the crowd. They feel, how should I say, an emotional longing for a belief system to replace the one that they have abandoned. “Gaiia” and a bunch of other inane new age fads are often convenint replacements, especially if we remember that for these poor souls the rejection of traditional belief systems was not a voyage of intellectual discovery, they were merely fellow travelers…and so they are still.. fellow travellers.
While I have not really closely followed your writings much in the past, I had a reasonable measure of respect for you. Perhaps it was unjustified.
I am not going to comment on the patent absurdity of your analogy.. many before me have.
Perhaps the only way I can excuse your intellectual failing is if you have not exerted the effort of spending a few hours on the web figuring out that the problem is more than just a few loosely worded emails.
I am not capable of commenting on whether the conclusion of the AGW camp will be proven to be correct or not, I have however, enough of a background in the scientific method and enough confidence in MY analytical reasoning to say that data warrants reevaluation and gaps in the generation of plausible hypothesis abound.
December 3, 2009, 7:52 pmwws says:
to arch1, re the quality of the discussion – the biggest problem is that we *don’t* have the facts because they are being intentionally hidden. As long as that goes on, there can be no true substantive discussion about anything else. That is the skeptics entire point in a nutshell.
Here’s an example of what I’m talking about – it’s not just the CRU data that’s being hidden, it’s data from NASA as well.
http://www.washingtontimes.com/news/2009/dec/03/researcher-says-nasa-hiding-climate-data/
“The fight over global warming science is about to cross the Atlantic with a U.S. researcher poised to sue NASA, demanding release of the same kind of climate data that has landed a leading British center in hot water over charges it skewed its data.
Chris Horner, a senior fellow at the Competitive Enterprise Institute, said NASA has refused for two years to provide information under the Freedom of Information Act that would show how the agency has shaped its climate data and would explain why the agency has repeatedly had to correct its data going as far back as the 1930s.
“I assume that what is there is highly damaging,” Mr. Horner said. “These guys are quite clearly bound and determined not to reveal their internal discussions about this.”
The numbers matter. Under pressure in 2007, NASA recalculated its data and found that 1934, not 1998, was the hottest year in its records for the contiguous 48 states. NASA later changed that data again, and now 1998 and 2006 are tied for first, with 1934 slightly cooler.”
(end quote)
Exit question – How come the temperature data for 1934 keeps changing? How come we keep getting new versions of 75 year old data, but we aren’t allowed to see the original data or the algorithms being used to back calculate it? I’ll wager that most people think that once the data is recorded, it stays the same, but that isn’t happening at all. And that – *That right there* – is the heart of the problem. The answer so far to the question “what was the temperature in 1934?” appears to be “whatever NASA wants it to be.”
You don’t have to be a scientist to know that That Ain’t Right.
December 3, 2009, 7:53 pmKen Arromdee says:
I do. Because how much credibility to give either side inherently depends on trusting them. If one side just happens to be ideologically committed to some position that it just so happens is advanced by protecting Earth from the asteroid, you’ll have to work harder to get me to believe they’re correct.
December 3, 2009, 8:00 pmA Law unto Himself says:
Randy, have you ever BEEN to China? Northern China (say Jilin Province)? In winter? And stayed in a brand new hotel in a smallish city? They just barely have double pane glass in the windows. No thermal breaks. No weatherstripping.
I’m sure they SAY that they meet LEED standards, but there is no way that they actually MEET the standard.
December 3, 2009, 8:04 pmGuy says:
Are you talking about the change in global average temperature? Statements like these make me doubt the education of the speakers. You absolutely can use statistical methods to get estimates that are more precise than the individual measurements, even without repeated measurements of the same quantity, as long as you have an assumed or theory-supported ersatz function to describe the expected type of dependence, with parameters. Are you claiming the scientists not only failed to engage in basic error analysis, but also no one else called them this “obvious” mistake? I’m not a climatologist, so I don’t know the models being used, but I can tell you know what you’re talking about even less.
December 3, 2009, 8:15 pmKazinski says:
The two situations are hardly analogous. I think we know with certainty that if an asteroid hits the earth that we are all toast. The impact of climate change is likely to be mostly beneficial.
And of course the other major difference, the postulated asteroid defense would cost “hundreds of billions”, less than what Obama squandered in the stimulous. Then we could get on with our lives.
With climate change the cost will be multi-trillions, and would profoundly impact our lives, negatively, forever, or at least generations.
December 3, 2009, 8:17 pmarch1 says:
Michelle, I think your reply is pretty much orthogonal to the (admittedly grumpy) point I was making. (If OTOH you could get three people to admit that their minds were changed by your and EMB’s well-put arguments, that would not only be on point but might even shame me into a less cynical perspective:-)
wws, if you’re saying that no substantive discussion can occur until we get all of the raw data, I disagree. But I agree that gaining and preserving access to as much raw data as possible is crucial, not only for present purposes but more importantly for future purposes (and FWIW I think actions which permanently impair this access and preservation to be reprehensible).
December 3, 2009, 8:18 pmGuy says:
What’s the basis of your claim that “[t]he impact of climate change is likely to be mostly beneficial”? Since we’ve adapted to live in our current climates, I would expect (in the absence of evidence to the contrary) that any climate change in any direction would be most likely negative.
December 3, 2009, 8:26 pmPaul McKaskle says:
Whatever the merits, or lack thereof, of various degrees of climate change skepticism, I wonder if Prof. Posner might concede by an update to his post that the analogy he posed isn’t really the best one under the circumstances to illustrate the point he is trying to make?
December 3, 2009, 8:29 pmJohn A. Fleming says:
This article summarizes current knowledge on a real-life asteroid situation. And it shows that right now, we can’t say with certainty that Apophis won’t hit us in 2036, but if it does, it’s gonna be bad. By 2013, we’ll know some more, but we still won’t know for sure. If the data looks bad, people are going to start getting anxious, enough to even do something.
But Posner’s analogy is still strained. An asteroid coming at us in 2012 is almost too late to do anything. First, our space-faring capabilities simply aren’t good enough. Moving big rocks in space takes time, lots of time. Second, there’s no consensus on what could be successfully done, in so short a time, so that if anybody tried something, skeptics would have plenty of cause to claim that the attempt being tried could actually make things worse (e.g. turn a just miss into a sure hit). And it would be difficult to judge who’s right. Third, there are observed and unexplained anomalies in the motion of objects of our solar system (satellite flybys of Earth, Voyager trajectories) that mean our predictions are not necessarily accurate enough on such short notice.
The analogy is further strained because the asteroid danger is a specific binary event two years away, with a very small range of uncertainty between hit or miss. That an asteroid will fall exactly in this uncertainty range is very very, unlikely.
Hypothesized secular and rapid climate change caused by human activities is nothing like this analogy, because the uncertainties are very large, and the time-scale is long.
Nice try, I liked it, very creative, excellent effort, thanks for playing, but no cigar.
December 3, 2009, 8:29 pmvic says:
here is a clear elucidation of some of the controversy:
http://bishophill.squarespace.com/blog/2008/8/11/caspar-and-the-jesus-paper.html
December 3, 2009, 8:34 pmHarryEagar says:
If the orbital calculations were as crummy as the climate alarmist temperature series, they would.
And if the head of NOAA gave Congress a demonstration equivalent to the one it got yesterday, comparing the chemistry of the ocean to club soda, there wouldn’t be a scientist in the world who wouldn’t.
December 3, 2009, 8:35 pmJagermeister says:
A more compelling analogy to the “climate warming” crowd would be if the scientists tell us that the asteroid is actually a space ship come to take the enlightened few away to a paradise world, and that we need to start funding social welfare projects to prove that we’re worthy of hitching the ride.
The science of global warming is about as convincing as the Hale-Bopp comet beaming up the entire IPCC to nirvana.
December 3, 2009, 8:48 pmOren says:
No, I can quite easily count all the relevant gravitational sources on my hands and toes (my computer is far better still, despite the lack of appendages). The uncertainty is not because we don’t know what’s out there, but because we cannot solve exactly the equations of motion. That is, we know exactly where everything is right now but there is a lot of uncertainty about where they are going.
We also know exactly what the average world temperature is now without being at all certain about where they will be 2020.
December 3, 2009, 8:48 pmJ1 says:
You’re leaving some things out of your analogy, like that the emails reveal that the astronomers’ data does not, in fact, suggest a colllision will occur, that they privately acknowledge that to each other, intentionally miscode their analysis programs to say it will, and conspire to silence equally knowledgeable scientists who question their theory.
Also, your analogy doesn’t include a demand from these scientists that we completely destroy our economy to stop or deflect the asteroid.
One of the most disconcerting aspects of this whole scandal has been the revelation that academics simply cannot believe or accept that other academics would be this dishonest. It reflects a frightening level of naivete in your profession.
December 3, 2009, 8:52 pmSkyler says:
Why? Because they use algorithms to decide what the “global temperature” is.
To use a value such as a “global temperature” is useful for some purposes, I’m sure. But it is not an absolute value that can be used for every purpose. By the very nature of defining such a value, it is highly dependent on somewhat arbitrary (at least) parameters. There is no “global temperature” except in a very limited theoretical sense, and that sense is highly dependent on how it is created.
By hiding the raw data, they put us all at the mercy of their own – now proven fraudulent – agenda.
December 3, 2009, 9:01 pmSkyler says:
Yes, that is exactly what I’m claiming, and that’s exactly what has been NOW PROVEN. Although there are many who did call them on this obvious “mistake” (note the change in which word is in quotation marks) but since most of the world didn’t expect “scientists” to be brazen liars defrauding world governments, they were not believed. And since the now proven lying “scientists” refused to share their data, there was no way to confront them in a scientific way.
The problem with your gobbledy gook is that you’re assuming that people are using a tried and true theory-supported ersatz function. They aren’t. They’re using lies and propaganda and inimidation and political coercion. No weather prediction model has ever been accurate. No climate prediction model has ever accurately back predicted climate changes. It has been a complete fraud.
And it’s been obvious to anyone that has ever had to use the scientific method. You can’t conclude with an accuracy within tenths of a degree when your raw data is not representative and is not accurate within less than 3-5 degrees. This is basic stuff, but they used gobbledy gook like you did to make their ridiculous conclusions sound otherwise convincing.
December 3, 2009, 9:08 pmloki13 says:
Anderson,
Don’t waste your time. You have to remember, most of these commenters used to be cackling in glee and citing to some Russian chemist’s analysis of sunspots to explain their position. Now that there is something tangible for them to latch on to, they are like terriers with a new chew toy. Of course, when we check back with them in a year, and the science continues to be the same, they will continue to cackle gleefully about these emails, but it is of no import- facts or evidence do not matter to them. You could have Rush, Beck, all the brass from the DoD, Ronald Reagan’s Rotting Corpse, Resurrected Einstein, every single scientist in the world, the CEO of ExxonMobil, and a time traveler with photos from 100 years hence SHOW them the global warming, and they’d still claim it was a conspiracy.
The takeaway, I hope, is that
1) there is increased transparency in the process. The best way to combat loonies is sunlight.
2) We can move to the point of having rational conversations about effective ways to deal with the problem, which includes cost (something hardcore environmentalists wave away) as well as energy sources (including nuclear). In the end, we don’t matter, and while I don’t believe in “Gaia”, I do believe that we are relatively insignificant in the grand scheme of things and the pottery barn will is in effect for humanity.
December 3, 2009, 9:11 pmSkyler says:
Loki (named after the god of mischief, of course) wrote the most laughable comment in his attempt to criticize those who point out the NOW PROVEN FRAUD of global warming:
Loki, if it were even slightly possible to claim with REAL science that there is global warming, don’t you think they would have done it without lying, hiding, intimidating, and coercing?
Personally, I think we’re very significant. You might think yourself a petty and insignificant slug, and maybe you are, but speak for yourself.
December 3, 2009, 9:24 pmGuy says:
That’s not true, If you have a machine that can measure position to within 1 mile, and you measure the position of a moving object 10,000 times, each at a different location, and you know that the object has a well defined, constant velocity, then you can use that to determine the position to a much greater degree of precision than within 1 mile. And if the model is fitting a curve inappropriate to the data, that’s why you calculate the Chi^2 value or some other measure of how appropriate the fit is. To say that the fact that global average temperature is reported more precisely than the measuring stations automatically invalidates the study is simply wrong.
December 3, 2009, 9:24 pmloki13 says:
Skyler,
Just so you know, writing things authoritatively and in all-caps doesn’t make it so, no matter how much you wish it to be.
And I still don’t have a pony.
I don’t think of myself as a slug (a black-footed ferret, oh, heck, really a sloth, especially with the SEC championship coming up), but I do think I’m pretty insignificant. I applaud your tremendous self-esteem, and hope that your Stuart Smalley-esque routine of anonymously posting will enable you to carry your grandiose sense of purpose into the real world, so that somne day, you too can be as pleasant and humble as Bryant Gumbel.
December 3, 2009, 9:39 pmfsfsfsfsfsfsfs says:
But would the rocket system be carbon friendly?
No point in saving ourselves from the asteroid, only to perish from the global warming precipitated by the construction and operation of that same rocket system.
December 3, 2009, 9:40 pmJoshua R. Poulson says:
The analogy is flawed because you could take a reasonably sized telescope and look out in space and SEE the asteroid. Multiple independent observers could all look at it over several days and estimate a trajectory and ephemeris. A bunch of people could grab Lang’s Astrophysical Formulae (Volume II) and dust off their college calculus and see, ahem, where this is going.
Asteroids move much faster and more predictably than climates.
The main issue is that there are serious problems with the source data, and huge error bars (how accurate were thermometers in 1850? In 1945?) and giant holes in the historical observations because someone wasn’t recording them at every station all the time. There’s just huge debates on HOW to normalize the data from thousands of thermometer observations over hundreds of years, let alone what those observations can do to predict the future.
Add to this that the climate has a significant amount of variation, although many apparent cycles have been suggested. There’s even conjecture that efforts to combat erosion of the ozone layer have noticeably altered Antarctic climate!
What does come out of this whole debate are a few things: pollution is probably bad, the way we get our energy now is not necessarily sustainable, asphalt really messes up thermometers nearby, and it’s not unusual to notice that we suddenly get weather data in places we like to vacation via airplane.
December 3, 2009, 9:42 pmLeo says:
Are the people advocating building the anti-asteroid rocket saying that the only way we can build the rocket is (conveniently) by following all the tangential political positions they’ve been pushing for decades? And are they vehemently opposed to any other possible solutions?
December 3, 2009, 9:46 pmGuy says:
Is the proof of fraud that you’re talking about the “trick” to “hide the decline”…. which was discussed openly in Nature before the emails were released?
December 3, 2009, 9:52 pmKen Mitchell says:
In fact, building an asteroid moving system will most assuredly be quite useful in the future, as a means of bringing raw materials from the asteroid belt or water from Saturn’s rings, and delivering this material safely to Earth orbit. Even WITHOUT the near certainty of ANOTHER impact similar to the one that killed off the dinosaurs, or the Burckle impact in the Indian Ocean 5000 years ago that may have caused Noah’s Great Flood, or the Clovis impact 14K years ago that precipitated the Younger Dryas mini-ice age that caused the extinction of the wooly mammoth and saber tooth tigers, and…. And does NOBODY remember that for a couple of hours accidental difference in time, the Tunguska impact in 1908 might have leveled London or Paris?
On the other hand, the AGW folks are trying to DENY the past, by glossing over the Medieval Warm Period and the Maunder (and Dalton) Minima.
December 3, 2009, 9:56 pmKen Mitchell says:
In fact, building an asteroid moving system will most assuredly be quite useful in the future, as a means of bringing raw materials from the asteroid belt or water from Saturn’s rings, and delivering this material safely to Earth orbit. Even WITHOUT the near certainty of ANOTHER impact similar to the one that killed off the dinosaurs, or the Burckle impact in the Indian Ocean 5000 years ago that may have caused Noah’s Great Flood, or the Clovis impact 14K years ago that precipitated the Younger Dryas mini-ice age that caused the extinction of the wooly mammoth and saber tooth tigers, and…. And does NOBODY remember that for a couple of hours accidental difference in time, the Tunguska impact in 1908 might have leveled London or Paris?
On the other hand, the AGW folks are trying to DENY the past, by glossing over the Medieval Warm Period and the Maunder (and Dalton) Minima.
December 3, 2009, 9:57 pmJoeSixpack says:
I still like the idea of building a giant chimney into outer space to let out some of the heat.
http://www.superchimney.org/
This seems like a more fitting analogy to the asteroid cure as well. Why don’t we try that before we ask everyone to stop using energy and cut their standard of living? I think Leo has the answer as to why.
December 3, 2009, 10:03 pmDavid Chesler says:
Anyone remember the cartoon flowchart for something is broken with decision points like “Did you touch it?” and “Did anyone see you touch it?”?
If whatever the climate is doing is not anthrogenic the key questions are “What are the changes going to be?” “What are the practical effects of them?” “Are they bad?” “Can we influence it?” “Can we work around it?” “Can we live with it?” “What’s cheapest?”
The current discussion seems to be stuck at “This is the best of all possible worlds, we’re changing it, so we’d better stop.”
Or back to the asteroid, people are claiming that the falling acorns are evidence that we are attracting asteroids, so we’d better stop doing some things that are the mostly likely asteroid attractors.
December 3, 2009, 10:12 pmKen Mitchell says:
Ummm… Not so much, no. So far, astronomers have on ONE occasion observed an asteroid (admittedly, a fairly small one) and predicted its impact on the Earth. 16 hours before it did so. Asteroids big enough to see that far out are big enough to cause global extinctions. If the asteroid Apophis does hit us on April 13, 2036, (the probability is currently estimated in the thousands-to-one against it) it’ll be all over. Nothing larger than beetles will survive.
Even far smaller objects – still capable of wiping out all life on a continent and making things quite uncomfortable for the survivors – are floating around, and are too small to see long enough to do anything about.
December 3, 2009, 10:32 pmBruce Hayden says:
In other words, are you asking whether or not if this one instance, if the alleged perps (and/or all their defenders) can make a plausible argument about hiding the decline, that they are now off the hook as far as “fraud”? I don’t think so, esp. since in a situation like this, the evidence is cumulative.
Besides, in the present discussion, it really isn’t all that relevant whether the CRU, etc. scientists have the requisite scienter for fraud. Rather, the question is whether or not we can trust their results. Pointing out why hiding the decline is not fraudulent does not address whether or not it helps the debate as to whether there is, and the extent of, man caused global warming (AGW). And, I would suggest that it does not. Rather, I think that it reduces our overall level of comfort with their results. Instead of assuming that their data were wrong (or, somehow no longer representative), they maybe should have looked more closely at their theories.
December 3, 2009, 10:35 pmNew Pseudonym says:
I was going to leave a comment, but then I realized this was not the thread about the best fantasy books of all time. Easily confused, though.
December 3, 2009, 10:45 pmBruce Hayden says:
I would agree to some extent.
As I see it, the identification of any warming as being anthrogenic turns the debate from being about science into one about religion. Questions such as what is the correct or optimal climatic temperature are now off the table. Would we be better off with a warmer climate? There is some evidence that we would be – such as that man has historically thrived in warmer times, and suffered in colder ones. And, look at all the farmland that would be opened up by unfreezing tundra (just look at a globe and note the shape of the continents). Similarly, even if we did come to a conclusion that man would be better off without any more warming, the question of what is the optimal way of reducing or countering GW is also off the table. The only acceptable solution to the problem of AGW is to stop the thing that supposedly causes it, by significantly reducing our CO2 output on a global basis.
And this is another difference with the asteroid heading for earth – we could look to the most expedient way of countering it, and not be constrained by religion to one possible solution, regardless of cost or effect.
December 3, 2009, 10:46 pmTomM says:
A lot depends upon the nature of the errors. You can reduce random errors by repeating measurements, but many sources of error are not random (including I suspect many of the error sources in the climate data).
December 3, 2009, 10:48 pmBruce Hayden says:
And, for a lot of people in Congress, the demonstration was probably fairly effective.
December 3, 2009, 10:49 pmSkyler says:
Guy,
Your comparison of a “global temperature,” which is an artificial construct of some limited usefulness not exceeding the limitations of how someone might define it, to measuring the location and path of a body moving along a constant path is flawed. “Global temperature” is not a constant value. To the extent that it exists, it changes every day and every year and is never quite the same twice in any one spot.
December 3, 2009, 10:49 pmCDU says:
No, it won’t be all over. Not even close, actually. 99942 Apophis is only about 270 meters in diameter. Asteroids 1 kilometer across hit earth about every half million years and life manages to go on just fine. Apophis would be devastating in the region around the impact site, but it wouldn’t even come close to wiping out life on earth.
December 3, 2009, 10:52 pmSebastian Tombs says:
My principal objection to the hypothetical is that, given the conditions stated, the method of dealing with the problem is clear. It might work and it might not, it might be needed or it might be a waste of money. But taking steps to divert the asteroid would not (well, with a high probability would not) cause it to trash the planet. Climate “science” (I’m not sure it deserves the term, but let it go) is not yet at that level. The steps the “warmers” want to take might help, might be needless, or might make things a lot worse. That’s why we need some real science applied to the subject, not political posturing.
I think a better comparison is between “climate science” and the belief that we know enough about economics to fine-tune things, eliminating booms and busts. It’s pretty clear that we don’t, though that hasn’t stopped anyone from trying. Monkeying with the planet’s ecology on the basis of a faith-based theory could do more harm than the economic tinkerers ever did.
December 3, 2009, 10:53 pmHugh says:
I am re-reading LUCIFER’S HAMMER right now! I love that book. And I have had the pleasure of meeting (and breaking bread with) both authors.
December 3, 2009, 10:53 pmSunTzu's Nephew says:
How about their controlling the peer-review process? How about their attacks on supposedly independent scientific journals? How about their refusal to be open about their data, their code, their methodology?
None of that is ‘science’. And if there isn’t any science, what is there? Faith. AGW is a religion to these people, and anyone who argues with them is apostate.
December 3, 2009, 10:55 pmSimon Kenton says:
“It seems inevitable to me that any system capable of deflecting an asteroid away from Earth is going to be just as capable of deflecting an asteroid *into* the Earth, and that as soon as it is built someone (or several powerful someones) are going to realize this and try to seize control of it. Doesn’t have to be an asteroid big enough to destroy the world, just one big enough to destroy the continent that whoever’s in control doesn’t like.”
– wws
An asteroid big enough to destroy a continent (even one merely large enough to knock a big hole in one) triggers mantle and core waves that focus on a point on the opposite side of the earth from the impact, shattering the crust there and inducing vulcanism. If, instead of an equal and opposite reaction that thumps my right hand each time I trigger it, my .45 simultaneously fires equipotent and directly opposed bullets, I will decline to use it, or change my usual grip. Your hypothetical global criminal would need to bring the asteroid down on and allow the transglobal vulcanism to well up on 2 enemy states orthogonally positioned, while his own state was offset 1/4 of an earth circumference from the target states.
December 3, 2009, 11:15 pmvic says:
http://chiefio.wordpress.com/2009/11/09/gistemp-a-human-view/
December 3, 2009, 11:23 pmBruce Hayden says:
Part of the problem here is that historical weather data is really not all that good. It is extremely unreliable over 100 years ago, and fairly unreliable throughout much of the 20th century. So, they have tried to use proxies, but those have their own problems, and are looking more and more unrealiable. Part of the controversy revolves around attempts to extrapolate and interpolate mostly historical data, but also some current data.
It is clear that both are true. The former for good reasons, but how it was done is highly questionable now, with the recent revelations.
Actually, the correlation with the theory seems to only be for about 20 years, ending in maybe 1998.
But one of the more questionable things that CRU, etc. have apparently done is to essentially remove from the temperature data by the massaging of data mentioned above the Little Ice Age that ended in the mid-19th Century. The relevance of that is that if the Little Ice Age actually happened (and there is an awful lot more evidence for it, than against), then much, if not most, of the warming that we have seen is a natural result of coming out of it. This is, BTW, one place where I think that the scientists involved should have double checked and questioned their proxies a bit better.
Agreed
I too find this questionable.
This is, BTW, why the testimony of the head of NOAA was so ludicrous. This is one of the places where the debate goes from the scientific to the purely speculative. The problem is what is the result of the sort of CO2 buildup that we have been seeing? And key to that debate is the issue of feedback. Is the feedback positive or negative? The assumption for most of these models is apparently that the feedback is decently positive. That means that after a certain tipping point, all is lost, and the climatic temperature will spiral out of control, melting all the ice, raising sea levels 20 feet in a short time, causing massive droughts, etc.
There are a lot of reasons to believe that the feedback is not nearly so positive, and maybe even may be even negative. One is the historical record – it has been hotter before, and it has been colder (within recorded human history). CO2 has been higher, and it has been lower. Indeed, there is a lot of evidence that CO2 is a lagging, not leading variable in relation to temperature. Plant life thrives in a warmer environment, and maybe even more so with a higher CO2 concentration, and plant life changes the Earth’s albedo. Also, warmer air has a higher carrying capacity for water. There are also massive interactions involving the oceans and the clouds (and many of the models taking clouds into account fail to distinguish between different types of clouds, which have very different affects on temperature), and the physics involving both is still pretty much computationally intractable.
Remove the assumption of major positive feedback, and the disaster scenarios pretty much disappear.
December 3, 2009, 11:25 pmKazinski says:
Guy:
First of all previous history. During the Holocene, and MWP, temps were likely 1-2c warmer than current temps, and both periods saw very prosperous periods in human history. Despite attempts to wipe the MWP out of the climate record, the prescence of wine grapes in England, higher latitude, and higher elevation tree lines than present, are very good evidence for much higher temps.
Second observations (i.e. scientific evidence) show a much lower sensitivity to increased CO2 than the climate models (i.e. speculation) indicating that even with the doubling of CO2 from levels a century ago isn’t going to raise temps more than about 2C total, leaving only about 1.2c left. That of course is much less warming than the 6c that some of the wildest speculation conjures up.
December 3, 2009, 11:27 pmvic says:
http://pielkeclimatesci.wordpress.com/2009/12/01/global-warming-and-glacier-melt-down-debate-a-tempest-in-a-teapot/
December 3, 2009, 11:33 pmgeokstr says:
Leo brings up the most salient point in this whole AGW issue – what a coincidence that the only solutions allowed by its proponents will lead to exactly the types of political changes the left would like to have happen even if there were no such thing as global warming, but that they have been unable to accomplish so far through the messy ballot box or even via revolutions.
Solutions that enhance the push towards greater central control of the entire world’s economy, supra-national taxation, near complete control over the actions and choices now made by individuals and corporations, spreading the wealth of the developed world to the undeveloped world in something that looks an awful lot like reparations for our having “exploited” them, elimination of capitalism and eventually of national sovereignty.
Looks like international socialism is the natural result of the only solutions we are being allowed to consider. Gosh, whoda thunk that nearly every leftist on the planet would be a True Believer in AGW, eh? I say the hell with that.
If we can actually show through objective, impartial science that the planet is warming and/or that we are causing it, then we have to consider the whole range of questions brought up by others here – is warming better or worse, is devoting resources to adapting to it more cost-effective than fighting it, etc, etc, etc.
December 3, 2009, 11:38 pmwws says:
re: simon kenton and the hypothetical attack on the earth: since we’re already well into the realm of science fiction, why not go all in? Building a planetary deflection weapon implies much more orbital infrastructure than we have now, so why wouldn’t it be someone who has a permanent base in the orbital spacedocks? He could blackmail the entire planet and not need to worry too much about the effects you describe. Or he might even be one of the Mars terraformers, damned spacer scum!
Seems to me that a mass driver would be the most technically feasible method of deflection, but turn this thing towards the earth and you’ve got a heck of a weapon.
December 3, 2009, 11:40 pmJames N. Gibson says:
The two points are:
1) where in the asteroid scenario we have a proven formula for determining the motion of bodies in space there is still argument as to the mathematics used for modeling climate.
2) The accuracy of the data inputs considering that we are comparing temperature data over 150 years were the data prior to 1960 was done with mercury thermometers and after by high accuracy thermacouples. One is accurate to a quarter of a degree, the other to plus or minus 2 degrees.
And just putting the data into some program doesn’t increase the accuracy of the data. And if the model is fitting a curve inappropriate to the data, that’s why you calculate the Chi^2 value or some other measure of how appropriate the fit is. That is forcing a conclusion which is what the Email is saying they were doing.
December 3, 2009, 11:40 pmGuy says:
…no, but this one instance is particularly ridiculous, since the “trick” was apparently a well-known and published practice even before the emails came out, and no one seems to have thought there was anything methodologically unsound before then. I’m not sure what you mean about the evidence being cumulative, unless maybe you feel there is no “smoking gun” but there’s some stuff that maybe casts a little doubt… not something I’m sure is true but in any case it’s very different from AGW having been “proven fraud”.
My comment was in response to Skyler saying the results must necessarily be made up/unscientific solely because they had more precision than the measured data, whether or not systematic error occurred isn’t relevant to that.
Yes, but that’s not relevant to whether the change can be measured to a great deal of precision – you can measure the rate a sand dune moves even though it’s made up of grains of sand that might all be moving different directions at different times. And, technically, average global temperature is a quantity that, in theory, is fairly well defined: Integrate the temperature over the region you care about (like a spheroid a certain distance above the ground), then divide by the area/volume/whatever of that region. That quantity is no more “imaginary” than the temperature at a specific location is imaginary.
December 3, 2009, 11:45 pmJim Owen says:
On the subject of GW, Scientific American has been in the pocket of the IPCC for at least the last ten years. And the basic charter of the IPCC for the FAR was to determine “the extent of the human contribution to global warming.”
Note carefully – There is NO question there about whether GW was real or not. Nor any question about whether there WAS a human contribution.
The IPCC basic charter was to prove AGW, with the human contribution to be as large as possible. Scientific American, like many other organizations, followed the IPCC lead without question – simply because they were a UN organization that claimed to be disinterested, neutral and independent – all of which were lies if viewed in the context of the original IPCC charter.
The actual operation of the IPCC then proceeded to reinforce those lies – by refusing to publish or even acknowledge any data or opinion that did not fit the agenda. As just one example – look
here. And he’s not even a “skeptic.”
That’s just one of a number of recent revelations by climate scientists who were excluded from the process because they failed to support the “cause.” there are others – MANY others.
For many of us who are/were “skeptics” the emails held few surprises – only confirmation of what we already knew to be true.
December 3, 2009, 11:49 pmGuest12345 says:
Not really. Any asteroid that calls for a deflection system will already be on a trajectory to hit the planet. Depending on the lead time it would take a relatively tiny amount of delta-v to change the asteroid’s trajectory such that an Earth impact is impossible. On the other hand the vast majority of asteroids in the Solar System are not aimed at the Earth. Compared to a deflection course adjustment, the adjustment necessary to move from a non-Earth impacting orbit into a impacting orbit would require millions of times more energy.
Additionally, the most likely successful technique for deflacting an asteroid would be to fly up to it, attach an ion thruster pretty much anywhere and turn it on. To actually hit the Earth would need a very specific aim.
December 3, 2009, 11:50 pmGuy says:
Using the Chi^2 value to test how well your model comports with reality is forcing a conclusion? I’ll need help understanding that. If you’re saying they selectively weighted their data in a way that was based on something other than previously known confidence values then I get you, but would like to see evidence of that, if you’re saying something else I’ll need some clarification.
December 3, 2009, 11:54 pmSoronel Haetir says:
We know what a few groups claim global average temperature is. Small changes in the thermometer sets, in the methods of calculation can have effects as large as the claimed warming.
December 3, 2009, 11:57 pmbbbeard says:
Oh, for heaven’s sake, don’t go there, Eric. The sight of a scientifically illiterate public intellectual trying to formulate a gedanken experiment about how science works is just pitiful.
The analogy you are trying to make is rife with fallacies and counterfactuals. At this point we know a lot about celestial mechanics, enough to drop robots on particular craters on Mars for a look-see, enough to slingshot spacecraft from planet to planet and out beyond the solar system. The significant uncertainties are observational, not computational. But even the climate modelers bemoan their inability to retrodict the last ten years of global “warming”. And FWIW the asteroid-watching community, unlike the IPCC’s cabal, is extremely open about codes, data, ephemerides, etc. As far as I know it’s been awhile since a group of astronomers secretly torpedoed the editor of a journal for indulging skepticism. Phil Jones’ CRU should be fired en masse.
This is a weak exercise, made weaker by your evident confusion about how science is conducted.
BBB
December 4, 2009, 12:19 amElliot says:
Is the asteroid on a collision course with earth because of human activity? Too many radio or TV broadcasts? Too much space junk in orbit? Is man the cause of the impending collision?
If man is not the cause, then we have a case of a natural event, and we take action to deal with that impendng event. We adapt. In this case, adaption might mean building missiles, spaceships, or stocking seeds and provisions in caves.
However, we would not demand that all TV and radio broadcasts be halted, nor would we demand an international crash program to clean up orbiting space junk because of Anthropogenic Celestial Collision (ACC). And we probably wouldn’t pay much attention to people who advocate for such measures. Why should we? We can see asteroids zooming all around our neighborhood.
We would demand those supporting ACC present their data. We would not follow them if they insisted on keeping it secret. We would not follow them if they adjusted data to ensure the asteroid trajectory hit the earth. We would also ask how many of the scientists insisting on ACC are actually non-credentialed in astronomy or astrophysics. We would ask why evidence for the demise of humanity is a secret.
In answer to your general question, we would take action to deal with the event, but we would not waste our time and resources pretending man is responsible for it.
December 4, 2009, 12:23 amloki13 says:
Isn’t that always the case? Must be comforting.
December 4, 2009, 12:36 amCareless says:
And after reading Anderson’s third comment, I actually *felt* this blog getting dumber.
Yes, Anderson, the end of ice ages is just like cataclysmic meteor impacts that destroy most life on Earth
December 4, 2009, 12:37 amCareless says:
Random personal note: I was halfway through Lucifer’s Hammer when my plane landed in Medan, North Sumatra a few hours after the Indian Ocean tsunami of 2006 hit. I haven’t been able to pick that book up since.
December 4, 2009, 12:48 amOren says:
If you care to share that formula with me, I can assure you a front-page papar in Nature or Science (your choice) and probably a Nobel. As it happens, we physicists are not in possession of a formula that determines the motion of many-body systems in space.
What is done, in actual practice, is that the equations of motion (which are analytically unsolvable) are solved numerically to form a trajectory of motion. There are two manifest problems here — the first is that error builds up, since the process is iterative. That much can be estimated and solved by throwing computational resources at it. The larger problem is that very similar initial conditions can lead to widely divergent trajectories. As it happens, the error in where you project a body will be in 2 years is exponential in the present-day uncertainty.
In short, I would advise you to either pick up a book on classical mechanics or leave the asteroid calculations to those that have.
December 4, 2009, 12:50 amArtifex says:
To answer Eric’s question directly, I would have to say query contains so many unknowns as to be relatively unanswerable. If we are looking at this from a scientific viewpoint, my take on the problem is going to vary based on the arguments put forth by the scientists.
I view this as the equivalent of me asking Eric: “If you were arguing a case based on the first amendment, would you win ? It sort of depends on the arguments in question doesn’t it ?
Science is not some priesthood accessible only to the only to the initiates. We aren’t given some form of decoder ring so that once we are part of the club the mysteries of the universe are revealed. On any problem, I am going to ask, “how does this fit with what I already know ?”. I will never start out knowing all of the answers. The best I can do is pull in relevant information, have the intellectual honesty to shove none of it under the rug and make my best guess.
How does this relate to Eric’s question ? I may not be an expert in astrodynamics, but using basic scientific principles and watching the arguments of the actors, I can make a pretty good guess as to how seriously I should treat the astroid problem without having to resort to blind faith. Here’s my three step method.
1. Watch the arguments. As a veteran of many design reviews and thesis defenses, I know one thing with near absolute certainty. The strong side of any scientific argument that has valid data and theory will attack based on those resources. The weaker side will seek to obfuscate and confuse. It may take a while but the stronger side to make their point, but they can simplify and explain very point in finer and finer detail if required. The weaker side does not have this luxery. Refusals to explain, appeals to authority and failure to engage are generally signs of extreme weakness of argument. A physicist with a strong argument will explain all the way to the basics while constantly emphasizing how blockheaded you are for not seeing these things. A physicist with a weak argument will simply tell you they won’t argue with you because they are so much smarter than you are.
So, in step one, I check to see if the scientists predicting the astroid strike are actually engaging the arguments of the skeptics. It’s a very good sign if they are taking the skeptics main arguments, doing the work and producing quantitive counter arguments. It is a very bad sign if they are avoiding questions and simply claiming that the skeptics “don’t understand their work”
2. I then use what I do know to spot check their results. I may not know astrodynamics, but if they make claims about basic mechanics or statistics that I can check directly, I will. It is a good sign if my spot checks show no errors. It is even a better sign if I or others turn up small errors (and almost always, at least one turns up) that are rapidly admitted and corrected. This shows a lot of integrity in the process. It is a very, very bad sign if they fail to correct small errors or promptly engage in spin control.
As an example, suppose our astro-doom-prediciters make available their calculations (another very good sign) and I find a mistake in the entry of some gravity parameter where the sign is reversed. If their response is “Ooops” followed by a quick correction and they update their results to show that the effects were minor and they have a quantitative correction, I am very likely to believe them. On the other hand, if they start by refusing to explain clearly how they got their results, then claim that the sign is not inverted, then claim that the sign doesn’t matter and finally claim that the changes to the results are minimal, I am probably going to think they are frauds.
3. Watch the general integrity of both sides of the debate. This is not as important as my first two criteria, but it can cast the winning vote. This is more general, but it depends on factors such as:
A. Is one side or the other censoring the debate ? Can legitimate questions be ask ? If I ask both the astro-doom-predictors and the astro-doom-deniers questions, does a specific side try to censor or control the questions I am allowed to ask ?
B. Is one side’s argument revolve mostly around motive and/or funding or are they really arguing the science.
Even considering all these points, I might still be incorrect on the final outcome, but I would be able to choose a side to support.
December 4, 2009, 12:52 amCareless says:
Somehow I suspect that the next few hundred eclipses predicted will happen on time and that Neptune’s orbit is predictable enough we could plan a satellite launch for 2020 that would fly by it. On the other hand, our climate models made for long term predictions are useless for predicting the next three months, next year, next 5 years, next decade, etc.
It’s just not comparable. Again,the climate scientists are probably right, but they clearly can’t prove it by scientific standards.
December 4, 2009, 12:59 amCareless says:
Funny I haven’t seen anyone mention it before: the climate predictions of these guys are ultra vires. It doesn’t mean they’re wrong.
December 4, 2009, 1:01 amJim Owen says:
Comforting? No.
Satisfaction, yes.
Now allow me to introduce a more valid analogy than the one being discussed here. This isn’t my creation – but it’s certainly worth passing on.
A section of politically active scientists, policy makers, politicians and NGOs in effect put on white coats and told us that our planet was gravely ill, and that we needed to follow their prescriptive advice to save ourselves from a deadly disease. That’s really how they framed the discussion, and they classified everyone who disagreed as a denier, like a smoker dismissing his cough and waving away the X-Rays.
That’s not a crime. But it’s pretty close to it to change the readouts on a patient’s condition to convince him to undergo expensive treatment, label other doctors as quacks if they disagree with the changed diagnosis, and to refuse to show the patient the data underlying the charts.
They may protest that the diagnosis is too technical for the patient to understand and that their actions are for the patient’s good. They may even believe it. But I call it quackery.
And the crime is malpractice. Deliberate and conscious malpractice. And since they arrogated the power unto themselves to diagnose the disease and prescribe a cure, they might also be charged with practicing medicine without a licence.
That’s a good summary of the last 20 years of “climate science.” Read it all here
December 4, 2009, 1:18 amSkyler says:
Guy asked:
The evidence? How about the emails and the lack of sharing data and the programmer that couldn’t make sense of the slip shod baseless adjustments made to their absurd computer models, and their threats against colleagues and . . . .
Or are you just going to ignore all that evidence?
Just how smokey does a smoking gun have to be for some people to agree that it was fired?
December 4, 2009, 1:25 amJim Owen says:
The evidence is in the code – not the emails. The emails just give us reason to dig into the code. And to do our best to prosecute those who’ve stolen the last ten years of productive science from the world.
To paraphrase a line from a very old, very bad movie – Trouble’s not here yet, but it’s coming. When the results of the code analysis are released.
If you know what you’re looking at, it’s not at all hard to see where and how they diddled the data.
But just for something to think about – if the tree ring data doesn’t match temps for the last 50 years, why would anyone with more than two brain cells to rub together believe that that same data would match the temps 1000 years ago.
December 4, 2009, 1:42 amA closer analogy says:
There has to be some way of including all the reporters and politicians whose argument, repeated over and over, never gets beyond the idea that if you disagree with them you’re not modern or smart. That’s a huge part to leave out: all the mindless name-calling substituted for facts. The people who don’t know the first thing about the last million years. People who couldn’t tell you about the last eleven years. This absence of any discussion of how our tiny fraction matters most, more than all the CO2 that we aren’t putting out, and water vapour, and high clouds, and the sun.
There has to be an equivalent to what I hear everyday: not scientific explanations for why the models don’t work. For why predictions and reality never seem to match up, and why emissions and temperatures keep failing to track. In real life, the newspapers are full of this stuff, about “the children” and “consensus” and what a “progressive” person thinks, but not justifying the hate that our betters in the ruling class feels for us proles. They hate us, and we have to shut up. No explanations are necessary. We just have to do what we’re told.
December 4, 2009, 3:44 amGaryC says:
One of the oldest stories about astronomy refers to Hsi and Ho, two Chinese astronomers who were executed for failing to predict an eclipse.
Astronomy takes its predictions very seriously!
http://www.nasa.gov/missions/solarsystem/f-eclipse.html
According to NASA, they were actually supposed to prevent the eclipse, so the emperor killed them for that failure, but they wouldn’t have still been in town (and drunk) if they had managed to predict it. That was in the 22nd century B.C.
December 4, 2009, 5:27 amGaryC says:
It is also a good sign if at least some of the errors don’t work in their favor. If you compare the astrometric data that they used with the raw measurements, and you keep finding “adjustments”, and all of those make the asteroid more likely to hit the earth, then it is suspicious. It might have an innocent explanation, but it becomes much more important to have a full explanation of every adjustment, and proof that the same logic was used throughout the data set
December 4, 2009, 5:55 amGaryC says:
I was stunned by the 1989 book review in the LA Times by David Graber, a National Park Service ecologist, who responded to the claim that man was part of nature by writing,
“I know social scientists who remind me that people are part of nature, but it isn’t true. Somewhere along the line – at about a billion years ago or maybe half that – we quit the contract and became a cancer. We have become a plague upon ourselves and upon the Earth.
It is cosmically unlikely that the developed world will choose to end its orgy of fossil-energy consumption, and the Third World its suicidal consumption of landscape. Until such a time as Homo sapiens should decide to rejoin nature, some of us can only hope for the right virus to come along”
This is incredibly ignorant. Half a billion years ago is in the Cambrian Explosion, when the first vertebrates can be found in the fossil records. If our ecological original sin can be dated back that far, then that virus would have to also target polar bears, whales, condors, northern spotted owls, the desert tortoise, pandas, puppy dogs, and kittens. And I’m not even certain that you can grant an indulgence to the cockroach, tarantula, lobster, salamander, rattlesnake, or octopus. We’re all guilty, according to Mr. Graber, because we share a common ancestor that stopped being part of nature.
December 4, 2009, 6:16 amGordon says:
“What’s the basis of your claim that “[t]he impact of climate change is likely to be mostly beneficial”? Since we’ve adapted to live in our current climates, I would expect (in the absence of evidence to the contrary) that any climate change in any direction would be most likely negative.”
So all those New Yorkers move to Florida when they retire to give them some inadaptation stress, in case they get too complacent in their old age?
December 4, 2009, 6:55 amPerhaps there is a future for sheltered accommodation on Baffin Island.
A. Zarkov says:
How about doing rational arithmetic? I’ve run calculations in Mathematica that produced huge round off errors, but when when I put the rational wrapper around all the arguments, everything worked.
There is also something called “symbolic Newton.” Here you can turn what is a numerical procedure into an analytic formula. I’m sure we could do this for iterative perturbation analysis used in calculating orbital trajectories.
December 4, 2009, 7:07 amHavoc Jack says:
How much does that sensor cost?
How much does it cost to install? How many hours will the lights be off because of it? How much electricity does the sensor run on? Given the lifetime of the sensor, will I make up those costs?
If I answer those questions with the sensor actually being cheaper, then I might install sensors in my hypothetical hotel. Whether or not it reduces my carbon output is irrelevant.
Now maybe the sensor costs more. But maybe I actually do need to reduce my carbon footprint. So I buy the sensor, either to do my part to save the planet or because new government regulations require me to. In this case my standard of living has been reduced, even if I value the outcome (I certainly prefer to live on habitable planets.) The darkness of the halls doesn’t even enter into it. (Assuming nobody stubs a toe waiting for the sensor to trigger.)
Now suppose that reducing my carbon output doesn’t actually save the planet. In this outcome, I’ve lowered my standard of living. I might have surrendered some liberties or privacy to the state (“Hello, I’m the inspector. I need to go into your hallways to make sure the light isn’t on when I’m not in them.”) I’d be pretty unhappy about that.
In the long run, that sort of sensor probably does save me money; I’ve got one in my garage. But not every environment saving device saves money, or makes my life more convenient. Given that, I’ve got an incentive not to use them just because someone with a Ph.D after their name told me to.
December 4, 2009, 7:24 amA. Zarkov says:
I think you need to study the what has been going on a little more carefully. Why would Michael Mann refuse to release his data and codes for his the “Hockey Stick” calculation? Congress had to force him to do that so the Wegman team could go over it. Of course we know why: the analysis was flawed as the guys at ClimateAudit showed. You can drive their algorithm with random noise and it will give you a “Hockey Stick.” Gavin Schmidt has tried to spin this by saying it doesn’t matter, but it does. The Hockey Stick misses the Medieval Warming Period. So they have to say that it was either local or did not exist. If it’s local than that busts Mann’s assumption that the temperature where the trees were taken provides a surrogate for the whole world. So they have to say there never was a Medieval Warming Period.
Before you run off any more go read the Wegman report. Wegman is a real expert on statistical calculations. Go to his web site and check his resume. Bringing up Rush or Beck is an attempt to misdirect people. You might also read the Willie Soon1, 2,*, Sallie Baliunas paper, “Proxy climatic and environmental changes of the past 1000 years” in Climate Research in 2003.
The fact is there are lots of problems with AGW.
December 4, 2009, 7:28 amA. Zarkov says:
We can’t do that here because of crime problems. Do you really think American hotel patrons are going to want to enter a dark hall, and get mugged? Many hotels are in high crime areas and the hotel hallways are monitored by closed circuit TV cameras. How can you do that in the dark? If they don’t do that, then the patrons who gets robbed, or raped will sue the hotel. Remember Connie Francis got raped in her motel and successfully sued for a lot of money. That changed the whole hospitality industry’s attitude towards guest safety. Paying a little for electricity is cheaper than big legal fees and damage awards.
December 4, 2009, 7:36 amHavoc Jack says:
“Every big new technical advance or system created by man throughout our history has always ended up being used in a war by the people who’ve got control of it against the people who don’t.”
Huh, that’s a very interesting thought. Interesting enough to get me wildly off topic. My first attempt at a counterexample was scientific medicine, but then I thought of Smallpox blankets and biological weapons. Does the ratio of lives saved to lives destroyed enter into it? If we avert a world destroying asteroid but encourage a continent destroying asteroid, is that a net positive?
Is it necessarily bad to exploit new technologies? One of the first computers was used by British Intelligence to decrypt Nazi communications. Should we be glad that it was used for the benefit of the good guys, or saddened that it was used to take lives at all? Do we blame the technology for the destruction caused, or is it inherent to people. To mangle an aphorism, “Nuclear weapons don’t turn cities into burning heaps of rubble, People turn cities into burning heaps of rubble.”, as Oliver Cromwell could no doubt tell you.
Is this only relevant to technology? If the north had more industry than the south, should the north be blamed for using that advantage when they went to war? Is it in human nature to grasp at anything that could provide an edge and twist it to their advantage?
To bring this back around, If Iran doesn’t believe in global warming, isn’t it in their best interest to say that they do in order to weaken the U.S. economy and therefore America’s ability to mess with Iran? Maybe not in this particular example because of the oil.
December 4, 2009, 7:48 amSk says:
The analogy fails because of the status of the two sciences. Asteroid science is hard science: it is probably the most accurately predictable future event that we have.
Climate science is more akin to political science or sociology; not a hard science, but rather a soft science: it is probably the most unpredictable science we have.
So, a better analogy would be between a sociologically predicted event (say, ‘sociologists predict that if we don’t increase the tax rate to 50% in the next 2 years, civil war will break out throughout the West’). Such a scientific prediction wouldn’t be taken seriously even without evidence of data manipulation and suppression of dissent: with data manipulation and suppression of dissent, they would be justifiably perceived as fringe element nonsense.
The harsh reality is that climate (and natural science) models just aren’t that good (yet, or possibly ever). It is truly bizarre that anyone in environmental sciences (who use air modelling, hydrology modelling, weather modelling, contaminant transport modelling) takes climate science modelling seriously: they all know better. Their hearts are getting the better of their heads, unfortunately.
Sk
December 4, 2009, 8:02 amStephen Goldstein says:
Because SOMETHING is ALWAYS better than NOTHING, right? Sorry, wrong.
Here’s my own poor analogy . . . .
You’re driving along on a barren stretch of highway. You run out of gas. You don’t have a spare gas can but you do have a gallon of windshield washer fluid. Something is better than nothing, right?
You add the washer fluid to the gas tank. Your car still does not run, but worse, add gas and it still won’t run!
But nobody would every do something so stupid you say. No? Ever heard of adding MBTE to gasoline and the subsequent contamination of water supplies or the millions who have died from Malaria as a result of the DDT ban?
December 4, 2009, 8:12 amDiversityHire says:
Of course we have to let the asteroid hit the earth—it’s the only way to prevent Anthropogenic Global Warming. Boogedy-boo!
December 4, 2009, 8:16 amFantasiaWHT says:
The differences are astoundingly huge:
1) The science behind moving objects is undeniably “solved” and universally agreed on. It’s provable, reproducible, all the things that actually make science “science”. Climate “science” is a bunch of theories that have not been validated, wherein many scientists disagree with many of the basic premises.
2) The possible result – asteroid colliding with earth, is largely understood as something devastating that would kill millions, if not billions, and destroy civilizations. Warmer temperatures nobody really knows how widespread any damage might be, and it would be just as likely to HELP people as harm.
3) The “solution” is just money, not changing the lifestyles of billions of people and telling us that the air we breathe (out) is a pollutant. The solution doesn’t require changing the fabric of society.
December 4, 2009, 8:24 amAnderson says:
Don’t waste your time. You have to remember, most of these commenters used to be cackling in glee and citing to some Russian chemist’s analysis of sunspots to explain their position.
True enough. The resemblance to creationist arguments is also striking.
The ice is melting off Greenland, etc. *faster* than IPCC predicted. But of course global warming can’t be true, because … LOOK AT THE E-MAILS.
December 4, 2009, 8:35 amwws says:
Funny that Anderson is comparing skeptics to “creationists”, when HE is the one who has taken a true Fundamentalist Religious position on this matter.
No criticism allowed! No skepticism allowed! Ye Must Believe, or be cast out of the Congregation!!! Heresy must be punished!
This new religion has got a Priesthood, (Jones, Mann, Hansen) dietary restrictions, (Gaia cries when you eat meat!) indulgences (er, carbon credits), and it’s even got the promise of hellfire and eternal damnation for all unless we mend our wicked ways!
REPENT!!! ye will all all perish in flames unless ye REPENT! Hear the word of the Prophet and REPENT!!!
and yet Anderson claims his opponents are the religious ones. HA! At least you could try to get some new tunes – the hymns you’re singing are already getting pretty stale.
December 4, 2009, 9:18 amDavid Schwartz says:
Say, Anderson, do you happen to know what year the ice sheet *started* melting?
December 4, 2009, 9:20 amDavid Schwartz says:
Anderson: If, by some remote chance, you are an honest intellectual, ask yourself this question — How does the fact that the IPCC mispredicted the rate of ice melt *support* the claim that they correctly understand the cause of it? Isn’t it more likely that they misunderstood it, since the climate didn’t warm as much as expected over the same time period at which the ice melted more than expected. Shouldn’t they have overestimated the rate of ice melt?
(And, the ice has melted in precisely the same way many times in the past, 8,200 years ago and 12,000 years ago are the best matches, if I remember correctly. Can anyone find graphs online?)
December 4, 2009, 9:43 amrarango says:
David Schwartz: can’t speak for Anderson, but my guess is that the general melting started maybe 15,000 to 18000 years ago as the last ice age came to a close. Of course, I am not a climate scientist, but I did stay in a holiday inn last nite.
December 4, 2009, 10:06 amStephen Goldstein says:
Nope. Sorry. One, absolutely, can not improve the accuracy of data over the accuracy of the original measurement.
Oh. To be sure, one can, compute an “estimate” with more significant figures but that is not the same as “accuracy.”
Take a look at http://en.wikipedia.org/wiki/False_precision
December 4, 2009, 10:20 amDavid Schwartz says:
Stephen: You have a record of temperature measurements for a hotel room that go back a year, one each second. They are all taken with a thermometer with an accuracy of plus or minus one degree except the reading for January 1 as 12:00 which was taken with a thermometer with an accuracy of +/- 10 degrees. They look like this:
30.2 +/- 1
30.1 +/- 1
30.8 +/- 1
36.4 +/- 10
30.6 +/- 1
30.2 +/- 1
For all the rest of the data, there is never a variation of more than plus or minus 2 between successive readings.
Will you seriously argue that none of that other data allows you to improve the “36.4 +/- 10″ you have for that second?
I think what you are missing is that everything is an estimate. Every temperature reading is an estimate of the actual temperature of the room. Don’t think about significant figures, think about tolerances.
Suppose we make the tolerance rigorous. By N +/- X I mean that for 95% of the samples, the measured value will be between N+X and N-X. Are you seriously arguing you can’t reduce the X part in the measured “36.4 +/- 10″ based on the fact that:
1) We know the temperature a second ago has a 95% chance of falling between 29.8 and 30.8 degrees.
2) We know the temperature a second later has a 95% chance of falling between 29.6 and 31.6 degrees.
3) In a year of observation, we have never seen the temperature change more than 2 degrees in a second (which is within error for the actuall change being much less).
Will you maintain that the best 95% interval we can give is 36.4 +/- 10?
One can realistically give data even where one has no measurement. If I have a thermometer in my kitchen, I can give you some data about the temperature in my living room. The accuracy will be lower, but your argument would suggest that I can say nothing at all — it might be 800 degrees.
December 4, 2009, 11:01 amOren says:
Ask Mathematica to solve the 3-body gravitational problem rationally, tell me what you get. There is no rational (analytic) solution, period.
As for symbolic iteration, the problem is that you don’t know when the perturbative assumption holds — how close do two object need to come before the resultant trajectory is not longer a small deviation from the two-body (analytic) solution. This is why nobody does numerics that way, instead doing first-principles solutions of the underlying equations of motions.
December 4, 2009, 11:03 amSoronel Haetir says:
David Schwartz ,
Except that is only the error bintroducedy the measurement instrument, as has been shown by the surfacestations.org survey there are lots of problems with the stations themselves. They are supposed to be as similar as possible, but they are not.
An example of why this matters, you have a room with a wood stove that is going and two thermometers. One thermometer is on the floor, the other on the ceiling. There is going to be a significant difference between the two.
What is the temperature of the room?
December 4, 2009, 11:40 amJPG says:
Well, A. Zarkov, there’s an easy answer to your question: American hotel patrons wouldn’t have to wander in dark hallways because – precisely what Anderson was talking about! – lights would go on as soon as they reach the door. Most hotels in Europe, Asia and South America have either a timer or a movement detection system to turn the lights on and off, even in highly criminal zones (such was my personal experience, at least, in Honduras). I don’t see the safety issue to be particularly relevant. It has more to do with our North American conception of electric power as a cheap resource vs their culture of cost-efficiency.
December 4, 2009, 11:42 amMike Keenan says:
We don’t how much warming there will be. We are only weakly certain about the cause. We don’t really know a practical way of stopping it. And we don’t know when it will really start affecting us in a net detrimental way — if ever. Also, our models are garbage with almost no way to test the garbage they put out.
Can you really compare the response to that to the response to an asteroid hitting the earth in two years?
December 4, 2009, 11:46 amShelbyC says:
Who’s arguing that it “can’t” be true? Of course it can be true. Global cooling can also be true. And which side is it who, like the creationists, is arguing that they know something is true that can’t be proven?
December 4, 2009, 12:01 pmDaveW says:
Good old Scientific American. I sadly canceled my subscription about 3 years ago. Way too much political advocacy for me, science content way down and salted with politics.
If I want to read leftist politics I’ll subscribe to The Nation or Mother Jones – not Scientific American. Every once in a while I pick up a copy in the drug store to see if anything has changed, but so far, no.
December 4, 2009, 12:25 pmwws says:
david schwartz wrote: “One can realistically give data even where one has no measurement. If I have a thermometer in my kitchen, I can give you some data about the temperature in my living room. The accuracy will be lower, but your argument would suggest that I can say nothing at all — it might be 800 degrees.”
Actually No, one cannot give any “data” under those circumstances. You *can* give your feeling as to what the data probably was, and given your experience that may be a fairly good estimate – but how could you dispute someone who had a feeling that you were stricken with a fever that day so your personal estimates were off? Or someone who claimed that you were purposely slanting your estimates for some reason? The entire exercise degenerates into a he-said, she-said mess that the use of real “Data” is meant to avoid.
If you’re not taking real measurements from real instruments, you’re just guessing. That doesn’t mean all guesses are wrong, but it does mean that anybody else’s guesses are probably just as good as yours.
And in your example about the 36.4 +/- 10 degree measurement – no, as long as you include that in your measurements you will *never* be able to increase the accuracy of the series – it would be much better to throw that out and simply say that an outlier had to be tossed, with an explanation. Here’s why: you could never eliminate the possibility that the actual temperature measurement that day was 46.4 degrees just as you can not eliminate the possiblity that it was actually 26.4 degrees. That is what your instrumentation told you, and that is ALL it told you. The other measurements have no causal ability to eliminate the range for that outlying measurement, and you have no right to assume that just because you haven’t seen big variability in other measurements you didn’t see it that day. (Read Nick Taleb’s Black Swan for a good lesson on the misuse of statistical methodology)
December 4, 2009, 12:35 pmJPG says:
But the same questions can be asked in regards to the perspective of a comet hitting the planet: we wouldn’t know for sure if and how we could stop it from hitting Earth; we would also be uncertain about the effects a collision would take place; all our models to evaluate the how likely a collision is to occur and its predictable effects on our conditions of living would be imperfect projections, at best.
Many commentors here have stressed the differences between the analogy used by Posner and the true issue at stake. But lets not forget Posner’s message over here, if I may rephrase his questioning: lets pretend scientists agree there is a slim chance a comet will hit our planet in a relatively short time. What are the set conditions for us to take deperate measures to avoid a catastrophic scenario to happen? Lets say they evaluate the odds to 1/5000 chances we will be struck, are they high enough for us to drill into the nation’s funds to send Bruce Willis on a costly mission? Lets pretend some astronomers would opine the comet may not be massive enough to have a significant impact, while most others would not be sure about the outcome of a probable collision to happen. Should we act and, if so, at what cost?
December 4, 2009, 12:54 pmLaura(southernxyl) says:
Want to cosign this, and also point out another difference between the current state of astronomy v. climate science.
It is not an infrequent occurance that a comet is discovered by an amateur astronomer who points his backyard telescope at the sky every night and makes observations. And when the amateur does this, and reports his find, the world-class astronomers don’t close ranks and refuse to admit that he might have seen something they missed.
December 4, 2009, 12:58 pmA. Criminal says:
My instinct is that the world would quickly get to work building the rocket system, and disregard the views of the skeptics.
No, because the rocket’s Carbon Footprint would be so large that its contribution to AGW would be worse than getting hit by the asteroid.
December 4, 2009, 1:01 pmJKB says:
Of course, if the asteroid impact prediction was based on telemetry data that “was lost” after corrections for undiscovered gravity sources that just have to exist but no one has been able to prove and the “experts” work tirelessly to prevent non-believers from learning the current location of the asteroid.
Not to mention, rockets sound cool but would have little impact on a body whose trajectory is being governed by large masses like the sun, planets and distant celestial bodies.
December 4, 2009, 1:14 pmStephen Goldstein says:
To quote the character Mona Lisa Vito in the film My Cousin Vinny, “That’s a BS question.”
I will seriously argue with your scenario. We’re talking about instruments and data series, right?
You have an instrument that measures temperatures to tenths with a error of ±1.0 degree, right? Swell. From what instrument did you obtain that measurement of 36.4±10?
If it’s the same instrument used for the other readings, what happened to it? Did you check the calibration sometime after recording the value and found it out of tolerance? Then your reading is suspect and the value, IMO, should be thrown out. If you need to account for every second you get to interpolate between two “good” readings for the missing second but understand that you’ve got data problems?
If it is a different instrument then, IMO, you have no business including this reading in your series.
What do you think is meant by “error?”
If your instrument reads 30.2 ±1.0 degree it means that the temperature is somewhere between 29.2 and 31.2.
Let’s say the reading an hour ago was 30.1 ±1.0.
You can SAY that it is 0.1 degree warmer but you don’t Know that it is warmer, at all ’cause the calibration error is larger than the difference.
Similarly, you can SAY that the average over the two hours is 30.15 because that’s how the arithmetic worked out. But, as I wrote above, if your instrument measures to tenths you can not claim a temperature record that is accurate to hundredths!
And one more small point though I know I’m being fussy . . . .
You write “30.1 +/- 1″ It makes no sense, IMO, to claim that your instrument reads to tenths but is accurate only to whole degrees.
This is why I wrote 30.1 ±1.0 to show that the instrument has been calibrated and it has been determined to be accurate to ±1.0 (its accuracy has been determined to the closest tenth so the error is not ±0.9 or ±1.1 but it is ±1.0).
December 4, 2009, 1:23 pmHarry Eagar says:
Bruce Hayden sez:’And, for a lot of people in Congress, the demonstration was probably fairly effective.’
For all of them, apparently, as well as Seth Borenstein of AP and Harris of NPR.
I wouldn’t mind if Lubchenco had put chalk in a class of real ocean water and asked Congress to watch until it dissolved. That would solve LOTS of difficulties.
December 4, 2009, 1:31 pmSteve Hayward says:
Eric:
Actually this imaginary scenario has already played out on a slightly different time scale. A few years back a handful of astronomers announced the possibility of an identified asteroid that might strike the earth about the year 2030. Subsequent calculations reduced the probability to near zero. But the initial reaction was telling. I clearly recall the New York Times editorializing that we shouldn’t get ahead of ourselves and panic about this remote possibility, which is quite different than the response about climate change from the the Times and others. If we’re going to apply the precautionary principal to climate, shouldn’t we apply it to asteroid risks, too? I’m going from memory here; I don’t have any links to the stories and editorials. Another thing I recall is that the first time I ever heard about the potential risks of asteroid strikes was in the early 1990s–from a prominent climate scientist!
December 4, 2009, 1:38 pmSkyler says:
Schwartz, your example is good in that you’re claiming all the thermometers but one are measuring within plus or minus one degree.
It “might” be reasonable to toss out the outlyer if you have a rational reason to do so. The remaining data can lead to a conclusion plus or minus one degree.
But the now proven fraud of global warming has been concluding an accuracy of a tenth of a degree. That doesn’t wash.
December 4, 2009, 1:40 pmSimon Spero says:
Oddly enough I’ve been trying to find the current best estimates for how long a warning would be available before a large asteroid impact (and how long before the bands of the earth’s surface which might be impacted could be determined).
Even odder, the reason for looking for these order of magnitude figures was to figure performance constraints for a global datanet for preserving digital scientific data.
The best place to go look is the Near Earth Object program at NASA JPL.
December 4, 2009, 1:44 pmLaura(southernxyl) says:
Stephen Goldstein, I am appalled at the number of people I’ve worked with who thought the concept of significant figures was something you learned for a test and never had to think about again.
December 4, 2009, 1:49 pmHugh says:
Asteroids do not concern me Admiral; I want that ship!
December 4, 2009, 1:57 pmDavid Chesler says:
Even more confusing to me than the question of why liberals support banning guns is why people who are concerned about AGW and carbon footprints aren’t screaming for more nuclear plants.
December 4, 2009, 2:03 pmDavid Schwartz says:
I guess you don’t understand the hypothetical. There is no reason to suspect that reading is incorrect, and in the hypothetical, everything is fine and it is correct.
If a stick is 10.1 inches long, 10.08 +/- .4 is correct. 8 +/- 3 is also correct. Neither is an “outlier”. In fact, their 95% probability ranges overlap, just like the readings in my example. They are all fully consistent.
If you measure the length of a stick with a ruler and a micrometer, both will give you completely valid measurements so long as you keep the tolerance with the measurement. There is no reason to throw out either, provided your subsequent steps keep the tolerances.
The math is complicated, but if you understand Fourier transforms and how things like MRIs work computationally, it would be obvious that throwing out the “outliers” would be needless. If you throw out the outliers, all you do is lose the information they added. (Which was very little, but not none.) The subsequent steps will understand that these readings carry very little information.
It is not difficult to design a system such that any true information adds to the knowledge (though how much will depend on its quality) and true information can never reduce it. That you got that reading from that instrument is true. Discarding it is only needed if you do not have subsequent steps that can keep the accuracy data with the primary data.
Have you ever looked at a CT scan? The computer does not have one single direct measurement of *ANY* particular volume of tissue. Yet it can produce remarkable information about the bone density of *EVERY* bit of tissue. Why is that? Because it has a *lot* of low-quality data that’s sort of about all the tissue (every bit of raw data it collects is about large areas of tissue), and it’s smart enough to infer what the tissue had to have been to produce that particular lot of low-quality data.
In simple terms, if all the large areas of tissue that include bit of tissue X are, on average, blocking less X-rays than all the large areas of tissues that don’t include bit of tissue X, we can infer that bit of tissue X is less bone dense than the average tissue. That is, we synthesize direct data about X purely from indirect data that is about something other than X.
Skyler: I am not saying toss out the outlier. I am saying keep the outlier because it is 100% as correct as the others. It may even add some value to the others (though not much). The point is to keep the outlier *with* its known accuracy level so that it can add any information it is capable of adding and cannot harm anything. It is well understood how to do this.
December 4, 2009, 2:12 pmSkyler says:
Okay, keep the oddball data if you need to.
I don’t know anything about ct scans except that my wife had one two days ago.
But to rely on your representation of it I would say that your ct scan to measure the temperature of the room is not quite the same. The temperature of the room changes continually and somewhat randomly, at least to a degree that we’ve never been able to reliably predict. The Bones and organs of a body are being held rigidly I place and don’t change appreciably during the time the measurements are taken.
That’s a big difference. But I admit that I’m not a ct scan engineer.
December 4, 2009, 2:45 pmXanthippas says:
It depends. If reasonable liberals supported doing something about this forthcoming problem, then conservatives and libertarians would impugn it with scientifically unsound arguments that either cherry-pick available evidence, mis-state facts, or both.
December 4, 2009, 3:00 pmStephen Goldstein says:
First, I agree, I do not understand your hypothetical.
Second, I did not use the terms “incorrect” or “outlier.”
Third, may I say, there you go again, having an instrument that is purported to measure more accurately than its calibrated accuracy (measuring the stick at 10.08 ± 0.4).
Fourth, in scientific measurement, one does not KNOW the objectively true length of the stick — one can only record and report what one’s instruments say to the accuracy of those instruments.
So, I’m running an experiment in my lab and I ask two assistants to measure your stick. One reports that the length is 10.08 ± 0.4 and the other reports 8 ± 3.
You are correct, there is no reason to doubt either of these but what should I record in my book? Multiple choice: A: 10.08 ± 0.4 B: 8 ± 3 or C: 9.04 ± 1.7
My choice? A. Oh, and I also tell my assitant to stop using his shoe to measure sticks. :-)
Fifth, we’re are way off Mr. Posner’s expectations for this thread so I’m done.
December 4, 2009, 3:12 pmBruce Hayden says:
All this uncertainty has got me thinking a bit about the climatic temperature readings. And I have been bothered a bit as a result thinking about how the databases, etc. have been constructed.
The uncertainty of much of the actual data, at least up until recently, has had a fair bit of uncertainty. But maybe just as bad, if not worse, there has been a lot of interpolation and extrapolation for missing data points. Actually, still apparently is. I find it questionable that these climate scientists could have done all this as well as the MRI designers, possibly because the MRI data is likely a lot more regular and well behaved. My gut feeling is that it is likely impossible to do with worldwide temperature data over time what has been done with MRIs, due to the fact that the temperature data is so irregular, but gut feelings don’t have scientific merit. So, I would be appreciative if anyone would take a whack at this.
December 4, 2009, 3:35 pmRob in CT says:
“Very true. Whether, and to what extent, warming is human-caused ought to matter only in the sense that everything we can find out about what’s happening here is potentially useful.”
I agree. The focus on whether or not it’s OUR FAULT is misguided, IMO. Climate is not static, with or without humans. If there is a warming trend ongoing (and, while I tend to believe those who say it is, I also figure this hasn’t been seriously studied for long enough to know much for sure) then we should be pondering how to deal with it. Can we slow/stop it? What will such efforts cost, and what will the likely impact be? Can we adapt to the warmer world and what does that entail? What does that cost?
I’m not a scientist, and I don’t have the ability to sort this out myself. I simply don’t. This is true of most people. We end up having to trust what others are saying, based on how reasonable they sound to us. Best we can do, really.
It seems to me that there is a pretty strong incentive to disbelieve GW, particularly if the prescription is a significant lifestyle change. Who wants to cut back on all manner of things they like? There’s a lot of people with vested interests in the status quo, both individuals and corporations. I certainly have a vested interest in the status quo. Also, change is just generally scary. Given all of that, I tend to view some of the more strident skeptics with, well, skepticism of my own.
My general position on this is that we should be looking to get “greener” for multiple reasons, not just the possibility that we’re causing GW. Many of the things we do that release lots of carbon also release a bunch of pollutants into our environment. They may not mess with the climate, but they can mess with plants, animals and people. That harms our standard of living and costs a ton of money (yanking contaminants out of an aquifer isn’t cheap). I’d rather levy taxes on dirty processes than do complicated “cap & trade” schemes that can (and will) be gamed, though.
Re: nuclear… it has upside, but the waste is a concern for me. That stuff takes forever to break down. If we built a lot more nuke plants, we’d have a lot more waste (even if we were more efficient in using the nuclear fuel). That’s a downside. Well, that and a tiny chance of catastrophe. For all that, it’s probably a better option than “clean” coal, and if you did a full-on analysis of the process from start to finish maybe it stacks up well with solar, wind, etc.
December 4, 2009, 3:56 pmGuy says:
The Wikipedia article does not support the proposition you claim it does, which is not surprising, because that proposition is false. Suppose you have measurable quantity x(t), which is a function of time, and suppose your equipment can measure x to a precision of .1 units. Now suppose you take 1,000,000 measurements of x, once each second, and you see the results appear to be roughly quadratic ( x = a*t^2 + b*t + c, where a, b, and c are constants – these are your parameters). So you find the best quadratic fit to the data, and find the uncertainties in the a, b, and c, values you calculated. Since you took so many measurements (one million of them!) you can actually use the calculated a, b, and c to go back and calculate an estimated value of x with more precision than your initial measurements. Of course, you need to be careful in your error analysis to make sure that the quadratic dependency is the only major factor – but there are ways to analyze that, like calculating Chi^2.
December 4, 2009, 4:03 pmGuy says:
Another example is when you’re receiving a sinusoidal signal with enormous amounts of white noise over it. the white noise reduces the quality of your measurement, but if you perform a Fourier transform on the signal then take an exponential average of the result over time, you can filter out the noise, getting far better precision than any individual measurement had. Why? because all the measurements you made had some degree of interdependency: each measurement gives you some information about the surrounding measurements. Surely you would agree that measuring the same quantity, which you know is constant, 1,000 times gives you much better precision than just measuring it once, even though the precision of the measuring equipment didn’t change. It’s the same basic principle.
December 4, 2009, 4:10 pmSkyler says:
The obvious point you’re ignoring is that temperature throughout the globe is not constant nor continuous and changes randomly to an extent that has eluded our ability to predict or even understand.
December 4, 2009, 4:41 pmGuest12345 says:
Remeasuring will only give greater precision if the instrument has a known error distribution. If your measuring stick is constantly off by a fixed, but unknown, amount remeasuring won’t improve your accuracy.
Your FT comment however raises an interesting question about how an daily/monthly/annual temperature is determined. Wouldn’t it make sense for a climate station to take many samples through a given day, do a FT against the data and take the DC component as the daily temp?
December 4, 2009, 4:44 pmGuy says:
If you still doubt me, and have access to Matlab or similar software, perform this exercise and it should convince you I’m right: First take a quadratic function like x=.0028t^2-3.2t+7400. Then create an array of 1000 values for x from t=1 to t=1000. Now add errors to each value by creating an array of 1000 randomly generated numbers – create them according to a normal distribution with a mean of zero and a standard deviation of 10 – add this array to the one you calculated according to the formula. Now find the quadratic equation of best fit to these values, and calculate a new array of “expected” values of x from t=1 to t=1000, If you average the squares of the disagreement between this array and the original array, then take the square root of that average, you should generally find a degree of fidelity to within a greater precision than 10 units.
December 4, 2009, 4:53 pmSkyler says:
And in fact if you measure the same identical, unchanging quantity 1000 times with the same flawed instrument or instruments all identically flawed then you still do not improve your measurement.
There would seem to be a difference between error introduced by a property’s variability and error introduced by the measuring instrument itself.
That is, if your ct scan can accurately measure x-Ray signal strength to a thousand decimals but the x-rays bounce around with some variability your method would seem to be valid. But if your thermometers are only accurate within 3-5 degrees in a room that stays at constant temperature then it is not valid. Add in the globe as your object to measure that is constantly changing and haphazardly sampled and you’re into the realm of make believe to think that you can conclude to a certainty within tenths of a degree.
And now that the fraud of global warming has been proven there’s no need for anyone to suspend disbelief based on the say so of proven liars.
December 4, 2009, 4:54 pmSkyler says:
That’s one heck of an assumption to apply to weather stations of varying and sometimes downright poor quality.
December 4, 2009, 4:58 pmGuy says:
Skyler said you can’t ever get greater precision than any one individual measurement has, the possible existence of systematic error doesn’t affect whether it’s possible, I was just pointing out that he was wrong – random error can be filtered out, and random error is usually the bulk of the error in most direct measurements. If someone wants to allege that a systematic error is biasing the results in favor of warming, they would have to provide evidence of that (If the systematic error is only a fixed amount added to the measured values, it would cancel out when you calculate the change in temp, so you would have to show the systematic error is more positive today than it was in the past).
FT’s are more relevant in the signal reception example than here because there you’re trying to measure the frequency of a signal. Taking the DC component of measurements throughout the day is equivalent to averaging those temperatures, and I would assume that they at least average the annual temps, if not the daily ones. I don’t know how often they take measurements, as I said, I’m not a climatologist.
December 4, 2009, 5:01 pmGuy says:
You can set the standard deviation to whatever you want, my basic point is that you can get better precision by taking multiple measurements. As long as you have a lot of data points the error should be reduced by some factor.
December 4, 2009, 5:03 pmGuy says:
I agree an error introduced by a mean of something other than zero wouldn’t be reduced by taking multiple measurements – that would be systematic error, not random error, but presumably any systematic error in the measuring stations is mostly cancelled out when you calculate the change in temperatures.
December 4, 2009, 5:08 pmBruce Hayden says:
My problem here is that the temperature data doesn’t have a constant error rate, but rather, some of it has a lot of error, some has little, and it is all being thrown in together. Some of the data points are real, and some are (often, apparently, manually) interpolated or extrapolated. I just don’t see being able to average everything and assume that the errors got averaged out.
December 4, 2009, 5:17 pmA. Zarkov says:
My experience was differnet in Sweden and other places. I had to walk in a dark hallway towards a lit push button switch. Apartment buildings had a similar system. Now that was a while ago before cheap motion detectors. Nevertheless in the US hotels want surveillance cameras operating full time. You can’t do that in the dark.
December 4, 2009, 5:18 pmHarry Eagar says:
Bruce Hayden sez: ‘So, I would be appreciative if anyone would take a whack at this.’
Look here for an example of how you could goose the temps by choosing one or another method of in-filling data. I don’t know whether they did it just this way, because it’s a deep, dark secret that we are not supposed to know.
December 4, 2009, 5:36 pmAbdul Abulbul Amir says:
But the energy industry is all in favor of the government mandating the use of expensive forms of energy. That industry is more than happy to fund those “scientists” that say the only way to save the planet is to buy expensive energy. Ken Lay was after all a father of cap and trade.
Link
December 4, 2009, 5:37 pmA. Zarkov says:
I did not say there was an analytic solution to the three body problem. I said we can improve the numeric of an iterative solution by rational arithmetic, so round off error don’t pile up. As to tracking asteroids, they must use something like a Kalman Filter. The mechanics are put into the state transition matrix and you predict the future path of the asteroid from the past positions. The process noise would account for the effects of other masses and the observation noise would account instrument error. So the prediction of the asteroid orbit would be constantly updated. I don’t know if this is exactly the approach, but I’ll bet it’s close.
December 4, 2009, 5:40 pmMatthew Carberry says:
Not supporting the “go to sensors for efficiency” argument per se.
If there’s an operating motion sensor, assuming the bad guy isn’t a stealthy slow-moving ninja, the light will come on if anyone is in the hallway and the camera will work just fine. alternately, low-light tech isn’t that expensive anymore. I can buy an infrared motion sensor, low-light camera in the Cabela’s catalog to hang on a tree for goodness sake.
December 4, 2009, 5:46 pmRobinGoodfellow says:
Some of the earliest computers were designed with the goal of handling the vast numbers of computations necessary to predict weather–that’s why they are called “computers” as opposed to “internet machines” or “word processors” (even though most people don’t do much computing with them). Weather is a highly complex, chaotic system that is not well understood.
On the other hand, we sent people to the moon with less computational power than can be found today on a moderately-priced hand-held calculator. The calculations necessary for predictions of the motion of large bodies in space have been understood since Isaac Newton’s time.
We know that, given enough gime there will be another earth-asteroid collision that will destroy most life on earth. We are told we must worry about a 1.7 degree increase in world temperatures that isn’t even higher than the highest temperature earth has survived.
No comparison in my book.
December 4, 2009, 6:14 pmShelbyC says:
But in your actual data, how do you know that your quatratic elements are data and the random variations are error? Maybe the data you are measuring fluctuates randomly and the error follows a quadratic pattern?
December 4, 2009, 6:20 pmGuy says:
It’s a well-known phenomenon that whenever error results from a large number of small independent factors that aren’t collectively biased in a particular direction (as is generally the case in real-world measurements – systematic error in your equipment can be corrected by proper calibration, and large sources of error can be controlled for). Then the total error will roughly obey a normal distribution. At any rate, if you have no idea what the expected instrumental error should be, then you should really take the time to familiarize yourself with the equipment and how it works. If you don’t even have a good model for explaining how your instruments work (and thermometers are pretty solidly based in well-known physics), then why are you using them?
December 4, 2009, 6:37 pmStephen Goldstein says:
We really are talking past one another so I’ll try one more time.
You are G-d and your quadratic equation does, actually, determine the value for x.
Me? I am a mortal. I have a measuring instrument so while the true values for X are determined, by the physical laws, to eight significant digits (the four from t=1 > 1000 and four from your “0.0028,”
December 4, 2009, 6:43 pmmariner says:
An Awful Truth:
First they would blame George W. Bush for leaving them the problem.
December 4, 2009, 6:51 pmmariner says:
arch1:
That example is more like helium than concrete.
A more accurate statement would be that since we can’t even tell that there is any GW threat it is nonsense to be fearmongering about AGW.
December 4, 2009, 6:57 pmjfb2252 says:
is NASA’s table “current impact risks”. Largest probability in next half-century is a 130m rock with probability 3.4e-4. 150 megatons, comparable to the largest H bomb ever tested.
December 4, 2009, 7:08 pmjfb2252 says:
http://neo.jpl.nasa.gov/risk/
is NASA’s table “current impact risks”. Largest probability in next half-century is a 130m rock with probability 3.4e-4. 150 megatons, comparable to the largest H bomb ever tested.
December 4, 2009, 7:09 pmBruce Hayden says:
What has been interesting to me is that the distinction here is routinely ignored or exaggerated. This last decade, it looks like we may have had global cooling, while possibly still having AGW. How could that work? The climate appears to be cooling just a little, and there is reason to believe that that may be because we are towards the bottom of the sunspot cycle, and, therefore, near a minimum of solar radiation (which, in the end, has to be the primary driver of our temperatures). So, possibly, AGW has kept the Earth from cooling quite as much as it would have otherwise.
But getting back to your question, there is a difference between AGW and GW, and its consequences. The reason is that the Earth’s temperature has gone up, and it has gone down, over time. We have huge swings, such as in and out of ice ages, pretty large swings, such as from the Medieval Warming Period into the Little Ice Age, which we are apparently now recovering from, and reasonably short cycles such as the sunspot cycles and El Nino/La Nina. That sort of stuff, and probably some in the middle that real climate scientists worry about, but the rest of us can ignore.
So, should the global average temperature be warmer or colder? I am of a couple minds here. I ski a lot, so I do like cold weather. But, historically, man has done better when the global temperature is warmer, as opposed to colder. When it is warmer, food supply is up, while disease and famine are down.
But when we are talking about AGW, it becomes a moral/religious issue. We have to stop the AGW, not because Manhattan, Florida, and New Orleans will be under water in a decade or so, because they won’t. Not even close. I would suggest that even a foot in the century is fairly unlikely. No, the reason is because it is man caused. Man is interfering with nature, and so that harm must be ended (even, in this case, whether or not there really is any harm). It is essentially some sort of Gaea worship.
Now having constructed such a straw man, imagine the scientists telling us that we shouldn’t worry about the asteroid hitting the Earth because it is in tune with nature, but admitting that if we had caused it, then we would be obligated to redirect or destroy it.
December 4, 2009, 7:35 pmVolokh Groupie says:
Too close to home/my vocation to put down a long response to this but at the very least I think as long as such an asteroid has actually been observed for a suitable period of time, the orbital simulations which determine likelihood of an impact are much more accurate and well understood than the GCM’s you’re trying to create an analogy for.
December 4, 2009, 7:47 pmGuy says:
I think that’s an oversimplification, If humanity is causing the CO2 concentration in the atmosphere to rise to geologically unprecedented levels, then the effect could be catastrophic (in the very long-term). But presumably we can “trust” natural changes in climate to be more minor and slowly-occurring because of past experience. In other words, if AGW should be a source of greater concern because it could potentially dwarf the size of “natural” variations to the extent that AGW is a new process that hasn’t occurred before in the planet’s history, so we have less reason to expect it not to be a problem.
December 4, 2009, 7:57 pmA. Zarkov says:
Hotels like to scan their hallways to see if anyone is hanging around, and that includes stationary figures who could sit there in the dark. I suppose they could use IR sensitive cameras to pick up body heat. But let’s say we could do it all with cheap equipment. There’s still a public relations problem. Most American hotel guests would just not be comfortable with a place where lights that went out. After all they don’t know how good the technology is.
December 4, 2009, 8:14 pmBruce Hayden says:
I think then, though, that you would have to prove that AGW is more potent than natural GW (or a stronger driving force), and I just don’t think that the data is there to support that. And, one indication of this is the climate scientists involved in ClimateGate being surprised when temperatures leveled off and maybe even declined over the last decade. That, to me, is indication that there are much larger forces involved (in particular, in this case, the amount of solar radiation being emitted by the sun).
The only way really to get to where you are suggesting is if there is fairly strong positive feedback in terms of CO2 and temperature. But I think that is looking more and more questionable.
December 4, 2009, 8:16 pmGuy says:
What I’m saying is that there is a natural “cap” on global warming as a result of, say, solar radiation, in that it is extremely unlikely the sun will suddenly begin to regularly emit much more energy than it ever has previously. But if we never stop pumping CO2 into the atmosphere, then temperatures will never stop rising unless some kind of powerful buffering effect kicks in. A small increase in temperature might possibly be a net good, but a large temperature change in either direction certainly isn’t.
December 4, 2009, 8:29 pmGuy says:
If the rapid changes in human civilization that happen within individual generations prove anything, it is that humanity has not reached a stable equilibrium, therefore we should be more careful about major changes resulting from human activity than from other causes.
December 4, 2009, 8:33 pmMike G in Corvallis says:
A few questions. In this scenario, would there emerge an industry of non-credentialed “astronomy skeptics” in the press and public comparable to the current batch of “climate skeptics”?
There already is an enormous industry of non-credentialed “astronomy skeptics” … if you use the “I’m from Missouri — show me!” definition of skeptic. Literally hundreds of thousands of non-credentialed amateur astronomers would be trying to observe the object. Tens of thousands of them would have the software — well-tested commercial software! — and the computing power to duplicate the analysis, at least to the point where they could assess for themselves the plausibility of an impact.
There’s a “dog that didn’t bark in the night” in your analogy: If any group of astronomers made such a prediction, other astronomers would be clamoring for the raw data, and casting aspersions on the credibility and professionalism of any observers who refused to make it available. And the observers themselves would be publishing every bit of data they could get their hands on, and distributing their analysis code along with it, hoping that someone would refute the prediction!
When the doctor diagnoses you with cancer, you’ll want to get a second opinion. If the doctor tells you, “I’m sorry, I can’t show you those x-rays, they’re are copyrighted information,” you’ll really want to get a second opinion, right?
By the way, every astronomer I know is delighted when his or her colleagues are sufficiently interested in the research as to ask for a reprint of a publication. Some of them would expire from ecstasy if a colleague actually cared enough to ask for a copy of the data. And we suffered fools gladly — astronomers put up with more than their fair share of cranks and crackpots than other scientists do. When I was in grad school we’d get two or three of them per month wandering the halls, looking for someone to whom they could explain their Grand Theory of the Universe That Refutes the Impostors Newton and Einstein. We treated them with more far sympathy and cooperation than Jones did McIntyre.
December 4, 2009, 8:39 pmA. Zarkov says:
Goldstein and Guy
Consider this.
X1= u + eps1
x2= u + eps2
X1= reading from instrument 1
X2= reading from instrument 2
u= true value of the item being measured
eps1 = random variable describing error of instrument 1
eps2 = random variable describing error of instrument 2
Suppose eps1 and eps2 are perfectly anti-correlated
Then (1/2) (X1 + X2)= u gives me a perfect measurement. This is an extreme case. The correlation does not have to be -1.
What is not well known is that as long as var(eps1) and var(eps2) are unequal, eps1 and eps2 can be positively correlated and you can still get a measurement that’s more accurate than either X1 or X2. So I can combine non-independent measurements and get a better measurement than either one alone by proper weighting.
December 4, 2009, 9:20 pmDavid Schwartz says:
Why would you deliberately make an error, I don’t get it? Neither of those three answers are correct. A and B both discard valid information and C uses the wrong algorithm to combine the two readings.
I think you’re missing the point — the algorithms for combining multiple readings that include error bars are well known. In this case, one sensible (and very easy to explain/understand) way to do it is to compute the weighted average of the two readings, weighing each one in inverse proportion to its range.
So we have 10.8/.4 and 8/3, since we weight them inverse proportion to their error bars, the first reading gets weight 1/.4 or 2.5 and the second gets weight 1/3 or .333, the total weight is 2.833, so
( (10.8*2.5) + (8 * .333) ) / 2.833 = 10.47
That is the number you should report.
The algorithm for calculating the error bar on this is not worth going into detail here, but it will be very slightly less than the .4.
The fact that the second reading is 8/3 rather than 11/3 makes it very slightly more likely that the true value is lower than 10.8 than higher. And more importantly, it makes it *much* less likely that the true value is, say, 11.2 (which was already highly unlikely, but now it is *much* more unlikely).
Prior to the second reading, we had to consider >10.8 as equally likely to <10.8. With the second reading, we know something new — it is very slightly more likely that it is lower. To simply throw this information away would be stupid, needless, arbitrary, and indefensible.
This assumes both input values do in fact have the correct error bars. Garbage in, garbage out — of course. But the assumption is that both readings were correct, just one known to be less accurate than the other.
And the definition of the tolerance is rigorous — the number after the +/- means some defined fraction (95%, 98%, and 99.5% are commonly used) of the readings will fall inside the specified range.
December 4, 2009, 9:49 pmVolokh Groupie says:
Amateur Astronomers already exist by the thousands and have contributed in everything from observation to simulations.
December 4, 2009, 10:02 pmGuy says:
Why do you name me? Goldstein’s on the other side.
December 4, 2009, 10:02 pmA. Zarkov says:
Not taking sides. Just wanted to bring it to the attention of both of you.
December 4, 2009, 10:12 pmSkyler says:
Sure, but the thermometers in the weather stations are only accurate to within 3-5 degrees, at best.
I think you’re trying to purposefully mask the common real problem by injecting unnecessary and inappropriate statistical methods that don’t apply to the system we’re discussing.
December 4, 2009, 10:14 pmGuy says:
What I was trying to do was explain to you how it’s possible to get a result that has better precision than one individual measurement when you have taken many measurements. You can’t just look at the fact that the change in temperature is less than the accuracy of the measuring stations and immediately conclude that there’s a problem.
December 4, 2009, 10:26 pmSkyler says:
Guy, then you’re talking about completely irrelevent things.
December 4, 2009, 10:38 pmvic says:
I am fairly skeptical of the AGW crowd/ religion, but the argument is really about statistics and statistical methods
December 4, 2009, 10:56 pmA. Zarkov says:
If the errors are normally distributed and the “±” corresponds to 1 standard deviation then the weights are proportional to 1/variance, not 1/std. So the weights are 1/.4^2 = 6.2500 and 1/3^2= 0.1111 divided by the sum of the weights,. Thus the weighted combination is
[(6.25) 10.08 + (.1111) 8]/(6.25+ 0.1111) = 10.0437
Notice the weights are 0.9825 and 0.0175 so the variance is
.9825^2 .4^2 + .0028^2 3^2 = .1572, taking the square root we get
10.0437 ± .3964, the maximum likelihood estimate of the measurement. It pretty much ignores the second measurement as one would expect.
December 4, 2009, 11:15 pmDavid Schwartz says:
Oh, the irony! I made a quick program to perform this ‘average’ in a scientifically rigorous way, and the second reading was so low, it actually *INCREASED* the variance! The correct result is 9.83 +/- .462!
That is, even though the second reading was less accurate, it was far enough off to cast doubt on the first reading. (This makes sense. An accurate thermometer says it’s 82 degrees +/- 2 but you are cold. This less-accurate reading does cast at least some doubt on the accurate one.)
So throwing away the “worse” reading, as it turns out, would be fudging data and claiming more accuracy than you actually had. Serves me right for not doing rigorous calculations.
For anyone interested, I wrote a quick web interface to a statistically correct average w/ variance program. You can find it Here. I can’t promise how long I’ll keep it running, but it’s fun to play with.
December 4, 2009, 11:21 pmSkyler says:
It’s all still irrelevent, Dave. We’re not talking about accurate-but-large-margin-of-error data, or even fake-but-accurate data, but fake-and-purposefully-misleading data.
We’re talking about a system that hasn’t been accurately modeled, even to back predict events that have occurred, and they’re using proven invalid models to claim things that never happened.
We can’t trust anything about climate that they’ve been claiming since Bill Clinton and Al Gore were elected, at a minimum.
There was once a time when people would talk about the weather to avoid political disagreements. Congratulations, international communist movement, you’ve always claimed that everything is political and now you’ve made the weather political too.
As for asteroids, that hasn’t been made too political yet.
December 4, 2009, 11:35 pmDavid Schwartz says:
Skyler: No, it’s completely relevant for at least two reasons:
1) You can’t even throw away “obviously wrong” data, because you don’t know what’s obviously wrong until you’re done. Plus, how much “obviously wrong” data you have suggests how much “unobviously wrong” data you have. And you sure as hell don’t get to throw away that suggestion. (We know this was done to produce the infamous hockey stick.)
2) Our intuitive sense about how to do things, even if we know what we’re doing, is sometimes very wrong. I was wrong even in my “back of the envelope” assessment of the variance. Others can make those mistakes too. That’s why it’s vital that procedures be documented and reviewed by others. And it’s why the raw data *MUST* be kept — you have to retry it with these assumptions changed or procedures fixed. If the “outliers” had been thrown away, we would never know how inaccurate our results were.
Climate scientists are not statisticians. Computer programmers may or may not know much about statistics and or science. Nobody is an expert on everything. And experts can fall victim to incorrect assumptions or bad logic on the back of a napkin.
December 4, 2009, 11:49 pmA. Zarkov says:
Are we using the same numbers? Your error variance for the combined measurement is greater than the error variance of the first measurement– it should be less– (0.462^2) > (0.4)^2. The 9.83 is way too low. The second measurement stinks compared to the first, the error variance is 3^2= 9 compared to .4^2 =.16. It can only pull it down a very little bit. See my calculation above.
December 4, 2009, 11:49 pmGuy says:
Climatologists who believe in AGW are communists? Are you even trying to look credible anymore?
You try to say the science backs you up, but when it’s pointed out you don’t understand what you’re talking about, you declare it doesn’t matter because it’s all a big conspiracy anyway. I’m glad you’re not arguing for anything I agree with and damaging my credibility by association.
And not to play the blame game, (well, I guess I am), but I think the politicization of science is primarily the result of the rise of anti-intellectualism on the right. The religious right in particular had a big hand in this.
December 5, 2009, 12:06 amSkyler says:
Um, but this isn’t the case of not knowing whether to throw away data that might be suspect. It’s the case of purposefully inserting or removing data because it gives results that don’t fit reality.
And it’s why the raw data must be made available for review.
I didn’t say that. But there is a part of the international communist movement that joined in the “green” band wagon.
December 5, 2009, 12:18 amDavid Schwartz says:
A.Zarkov: No, it is precisely correct. This is the danger of seat-of-the-pants math.
Here’s a simple way to get your brain wrapped around it.
How does 9.83 compare to 10.08 +/- 0.4? Answer: (9.83-10.08)/.4 = -.625
That is, 9.83 says the 10.08 is low by .625 of its accuracy.
Now how does 9.83 compare to 8 +/- 3? Answer: (9.83-8)/3 = .61
That is, 9.83 says the 8 is high by .61 of its accuracy.
So it asserts that each reading is about as likely, which makes sense.
An even simpler way:
A thermometer says the temperature in a room is 83 +/- 1. This means the temperature could be anything, but 95% of the time (or whatever percentage we’re using), the actual parameter is between 82 and 84 when the thermometer says this.
So with just the thermometer, we’re pretty sure we know the temperature pretty well.
Now, I walk in and say, I’m cold. We know from studies that when I say I’m cool, 95% of the time the temperature is between 65 and 75. Now, we’re not so sure. It’s unlikely the thermometer is that far off, but it’s unlikely I’m that far off either. So we assume each is (approximately, the exact math is complex) equally far off in units of our own precision and find the actual temperature that makes both readings equally likely (or unlikely as the case may be).
Of course, it very rarely happens that you get two such far off readings, comparable to the odds that when you pick two people from a room, one is in the top 1% and one is in the bottom 1%.
This shows why error bars are so important. They change more than just the final error bars! They control how much different parameters are trusted and thus impact the result.
Sometimes it takes a while to see how obvious and intuitive something is. You can play with the web page I linked to see how the numbers come out in any what-if scenario you want.
December 5, 2009, 12:31 amA. Zarkov says:
Ok let’s go back to basics. We have two measurements, x1 and x2. What is the joint distribution f(x1,x2)? If the measurements are independent f(x1,x2)= f1(x1) f2(x2). Assuming normally distributed errors, f1(x1) is the normal density with mean u and standard dev sig1, and the same for f2(x2) with “1″ replaced by “2″ everywhere.
Now write the log likelihood,
log(f(x1,x2))= log(f1(x1))* log(f2(x2)). This will be a function of the unknown mean u.
We want the value of u that maximizes, log(f(x1,x2)), so take the partial derivative with respect to u, set that equal to zero and solve for u.
The answer is:
u= (v2 x1 + v1 x2)/(v1+ v2)
v1= sig1^2
v2= sig2^2
Plug in
x1= 10.08
x2= 8
sig1= .4
sig2= 3
then you get
u= 10.0437, the maximum likelihood estimate of the true value of the measurement.
This is the correct answer for the maximum likelihood estimate of u. It is the standard formula for combining measurements of differing error variances.
If you believe in maximum likelihood then you must use my formula.
December 5, 2009, 1:10 amA. Zarkov says:
No. The reading with the smaller error variance is the most likely reading. Suppose the reading has zero variance, then it is exact and the has probability= 1 of being correct. Suppose the error variance is 10^1000. This means the reading is almost certainly wrong.
December 5, 2009, 1:16 amDavid Schwartz says:
A.Zarkov: In the degenerate case, the simplification is useless. In the general case, the simplification is valuable.
In any event, your math is incorrect. The definition we were using for the measurement was that half the time it would be above it and half the time below. By picking the peak of the distribution that is asymmetric, you pick the wrong point.
The maximum likelihood reading is not what you want conceptually. Consider a degenerate with a sharp near vertical line and a long rightward tail (like a sawtooth or wave leading edge with back). The peak of the tooth severely underweighs the readings in the big tail.
See my link for more details, the method I’m using is statistically rigorous. (But it’s very easy to see how non-statisticians, even with lots of math knowledge, can get it wrong. I got it wrong at first too, way wrong.)
December 5, 2009, 1:32 amA. Zarkov says:
What is your joint distribution for the two measurements? Give me the function. What is the parameter you are trying to estimate?
December 5, 2009, 2:59 amDavid Schwartz says:
Well, we’re not being precise about what the ‘tolerance’ or ‘variance’ means. I think twice the standard deviation is reasonable. But you can use some other definition. (Thus 10 +/- 1 means 95.4% of readings between 9 and 11.)
I treat both distributions as normal, centered on the reading. I then sum the two distributions and find the point such that half the distribution is above and half is below. That is the output reading. (Our estimate of the parameter all the measurements measure.)
To find the tolerance, I calculate the window such that the cumulative distribution obeys the same initial rules (tolerance is two standard deviations), with appropriate corrections for the fact that the distribution is not normal.
That’s the really tricky part, and I’m not really qualified to rigorously discuss how you estimate the output tolerance. Finding the standard deviation and doubling it is not the same as find the 95.4% point. I just used the published algorithms.
December 5, 2009, 3:07 amA. Zarkov says:
So if x and y are the measurements, what is f(x,y), the density function of the measurements? Or are you taking a Bayesian approach where you condition on the observed data? I don’t understand your sample space. Remember in the frequentist approach the sample space goes on the data, in the Bayesian approach it goes on the parameters. It’s not clear to me what you are doing. Is this some kind of fuzzy logic approach?
December 5, 2009, 3:51 amA. Zarkov says:
David Schwartz:
Are you doing statistical tolerance intervals?
December 5, 2009, 4:00 amSShepherd says:
I propose that your “model,” while amusing, is drastically oversimplified. Let us adjust.
A more realistic model:
Suppose that an object of unknown size were spotted, headed toward Earth on what might be a collision course. The astronomers cannot give us a precise probability of collision because of many imponderables, nor can they narrow down the collision date. Estimates range from “next year” to “several centuries.” However, after looking at the data and mulling it over, prominent astronomers reach the following conclusions (some of them contradictory) and form groups supporting one or more of these predictions:
A. We don’t know how large the asteroid is, but it could be quite large. So we should form an international coalition of countries to develop a space program to nuke it/place a rocket on it, with the purpose of diverting it from its course.
B. We don’t know how large the asteroid is, but it could be quite large. So we should force the richest countries to develop a space program to nuke it/place a rocket on it, with the purpose of diverting it from its course.
C. We don’t know how large the asteroid is. We can’t get reliable data because we don’t know the asteroid’s composition (and therefore its albedo), but if we use *this* model, it nudges the data into a better and more readable curve, and we can use that to make predictions. Using this new technique, we conclude that the best response is A.
D. Same as C except that the preferred response is B.
E. We don’t know how large the asteroid is, and we have published our results, but we’ll be damned if we’re going to show you our methods. Also, we support the response in A.
F. Same as E except that the preferred response is B.
G. Since we do not have a good way to tell how large the asteroid is, we should wait for more data and not do anything drastic.
At the same time, several other events occur:
H. A portion of the general populace takes the position that “the asteroid is a load of bunk and we should ignore it entirely.”
G. A situation similar to that in E causes several amateur astronomers to look into the results, and they conclude that whatever is going on, it ain’t proper astronomy. Perhaps astrology? In any case, the red-faced and angry Mann (uh, persons involved) tells them their critique is worthless because they’re only amateurs, after all, and so their opinion shouldn’t count.
H. A prominent politician starts making statements to the effect of A, publishes a short book with no bibliography or works cited called “An Inconvenient Truth: How We’re Going to be Wiped Out.” MIT students reading the book and watching the related movie conclude that there is no way to tell where he got the figures or data from, except one study, which helpfully provides a common last name. Because obviously searching “Johnson asteroid study” will get you to the right research paper…
I. A team of astronomers get together and pool information. They produce a 2000-page treatise on the asteroid data. Large numbers of politicians insists on a brief summary.
J. As in A, except the politicans confuse the book from H for the brief summary of I.
K. Someone in the back pipes up that we can probably deal with this emergency in a wiser fashion if we wait a year or two for more, and better-quality data. He is ignored.
L. As in K, except the speaker is an MIT professor. He is ignored.
M. The media begins to use the words “asteroid skeptic” to describe anyone who still argues that the asteroid may be much smaller than a car, and essentially harmless.
N. A group of concerned citizens examines the lenses of the local telescopes and concludes that a significant minority of them are so dusty that the data should not be trusted.
O. Investigations into data quality and statistical methods reveal that at least some of the arguments used to prove that the asteroid is larger than a small dog are, to put it nicely, bunk. (Note that the majority of studies are either sound or have not been explored yet.)
P. The asteroid controversy / consensus / doohickey becomes politicized.
This leads to one of several outcomes:
December 5, 2009, 4:42 am1. The asteroid is large and we do something drastic and quickly. Good.
2. The asteroid is large and we do nothing, nor do we wait for data; we ignore it. Bad.
3. The asteroid is medium sized and we do something drastic and quickly. We prevent local damage only. We later conclude that most people would have adjusted to the impact results, some of them easily, some of them with great difficulty.
4. The asteroid is medium sized and we do nothing, nor to we wait for data; we ignore it. We suffer local damage only, and many people eventually adjust to post-impact conditions.
5. The asteroid is smallish, and we do something drastic and quickly. We prevent minor local damage. Also, some people feel self-righteously good about themselves, because they *might* have been preventing something major.
6. The asteroid is smallish, and and we do nothing, nor to we wait for data; we ignore it. We suffer minor local damage only.
7. The asteroid is a small rock. Whether we act or not, not much happens. A plague starts in Guatemala and is blamed on the asteroid.
8. The asteroid is any size. Instead of ignoring the problem or panicking, we devote resources to narrowing the impact window, while simultaneously narrowing down the size of the asteroid. Life continues as usual, and the final decision is nearly guaranteed to be more appropriate to the *actual* rather than theoretical situation.
David Schwartz says:
A.Z: f(x,y) is f(x)+f(y), where f() is a normal distribution with mean equal to the measured value and standard deviation equal to 1/2 the tolerance.
December 5, 2009, 5:11 amLaura(southernxyl) says:
Don’t know about asteroids and all, but in the lab, “precision” and “accuracy” are two different things.
Accuracy is a measure of how close your result is to the true value.
Precision is a measure of how close replicates match each other.
An analogy I have used is shooting at a bullseye target. If your bullets hit all over your target, but the center of your pattern is at the center of the bullseye, you have good accuracy but not good precision. If you end up with a tight little pattern at one corner of the bullseye, you have good precision but not good accuracy.
Obviously, precision is the more important thing to have, as long as you can measure and account for your bias (adjust the sight on your gun, for instance).
If you remove numbers that appear to be outliers, you will improve precision, but you may cause your result to be less accurate.
December 5, 2009, 11:37 amElliot says:
But what do we do about all those non-credentialed people? If non-credentialed people are a problem as skeptics, are they also a problem as boosters?
December 5, 2009, 1:23 pmA. Zarkov says:
Your f(x,y)= f1(x)+ f2(y) is not a proper pdf because it integrates to 2. Do you mean 1/2 (f1(x)+ f2(y))? Do you assume x and y which are the two measurements to be independent? If so then, f(x,y)= f1(x)* f2(y). Why would they not be independent? By f(x,y) I mean the joint pdf of the two measurements. Surely f(x|y) is not a function y? Otherwise the two measurements must somehow communicate.
December 5, 2009, 1:26 pmKen Mitchell says:
I recall a science fiction story many years ago. Three intrepid space explorers boarded a brand-new, state-of-the-art spacecraft and set out to a nearby star that observation had revealed to have habitable planets. Because the destination was several light years away and the trip would take 500 years, they entered their suspended animation capsules.
Upon arrival 500 years later, they were greeted as heroes who had been the first to set out for the stars – but not the first to arrive. In the interim, science had progressed – and when does it not? 150 years after the launch, mankind had developed much FASTER spacecraft, and the destination planet had been colonized and settled 300 years previously.
So too with “global warming”. Impoverishing our civilization NOW to pay for something that won’t be a problem for a while and may NEVER be a problem is to abandon the possibility of future discoveries; discoveries that pay determine that there never was a problem, or discoveries that make the solution to the problem trivial in nature.
December 5, 2009, 2:41 pmDavid Schwartz says:
A.Zarkov: I was implicitly normalizing over the area under the curve, so it didn’t matter to me whether the density function summed to 1 or 2. But you’re right, 1/2 would put it in the proper form.
I’m not a statistician, but I’m pretty sure that you can’t multiply the probability density functions to get a useful probability density function. The value of that function at any point would be proportional to the probability of getting those readings if the parameter was that value. But we need to go the other way around, to deduce the likelihood that the parameter is a given value given the measurements.
The question is not “for what temperature are these the most likely measurements we would get” but it is “for these measurements, what temperature has a 50% probability of being less than or equal to the actual temperature”. (You definitely don’t want the peak of the distribution function because that will underweigh the tail.)
However, how you calculate the spread of the result (or accuracy, or tolerance, or standard deviation, or whatever parameter you want to indicate confidence level) is complex and not easy to explain in common sense. Even just trying to wrap your brain around whether two 10/1 readings results in a 10/1 output or more confidence in the 10 makes my head hurt. (It makes it less likely that the 10/1 is an ‘outlier’, so it actually does contract the confidence interval, assuming the readings are independent!)
December 5, 2009, 3:44 pmepignosis says:
Cannot help you here. I cannot tell which commenter is the better statistician.
One thing is certain. Climate changes. Been getting slightly colder for a few years.
Are glaciers or ice caps disappearing and the mean sea level rising? Maybe, but they will be replenished or restored during the next glacial period. Probably has little to do with human activity.
We need to know how much (money or jobs) it will cost to give the environmental activists a clear conscience. How many trillions of $ do we have to spend to satiate their crusader arrogance? Their need to save the world.
December 5, 2009, 4:06 pmBill Johnson says:
Poor example – the consequences are not on the same level of magnitude – or does ‘unknown’ just cover both examples. The costs are not spoken of – how about we all kill ourselves on the chance something else will come by and attempt it.
And don’t ever use science and consensus in the same sentence. Remember what Einstein said about the 100 editors dissing him: ‘If I were wrong, it would only take one.’
And are you also presupposing perfect would governmental cooperation? Yeah, right.
Quit these damn salvation campaigns. Perfect your own behavior – invent a moral code for those without a moral authority other than ‘it feels right’.
December 6, 2009, 9:43 amruralcounsel says:
Great thread, and despite reading from a few folks in complete denial of the weaknesses of AGW theory, lots of excellant points, analogies, etc.
Here’s just one more that rarely gets brought up… the point that increases in CO2 may not even cause additional warming, because it causes about as much warming as it theoretically can already.
“Now, here’s an important point that global warming scientists don’t mention. Though carbon dioxide definitely absorbs IR, it only absorbs IR in two very narrow ranges of wavelengths, one between 2.5 and 3 microns, and another between 4 and 5 microns. This is a small percentage of the total IR emitted by the surface. I don’t know exactly how small (because I can’t find any source for the wavelength distribution of IR emitted from the surface), but it’s probably less than 10%, and perhaps as low as 4%. And even in those ranges, CO2 has to compete with water vapor, which also absorbs 2.5-3-micron IR. So, even if carbon dioxide in the atmosphere increased a thousand-fold, and even if there was no water vapor, there is a limit to how much IR CO2 can absorb, and that limit is 10% (or less) of all the IR emitted from the surface. And of that 10%, over half of it still ends up escaping into space.
This limit of absorptivity is important because some skeptics argue that, between water vapor and CO2, every available ray of IR within the absorption ranges of CO2 is already being absorbed. Additional molecules of CO2, therefore, will have zero effect on the total absorption of IR. So future warming due to CO2 is simply not possible. The only way CO2 could absorb any more IR than it is already absorbing is if 1) the surface started re-emitting more IR, which could only happen if more sunlight reached the surface, or 2) atmospheric water vapor levels dropped, freeing up more IR to be absorbed by CO2, in which case, warming would not occur, because that radiation was already being absorbed by the water vapor that disappeared. In fact, if the second option occurred, temperatures would in fact drop, because water vapor absorbs IR over a much wider range than CO2, and therefore, CO2 cannot completely offset the loss of IR absorption by water vapor. However, the existence of CO2, replacing the IR absorption of some of the lost water vapor, would mitigate this temperature drop. Therefore, at current levels, CO2 could be said to be somewhat of a stabilizer of the greenhouse effect, taking up part of the slack when water vapor levels drop too low. In this respect, it is good to have an excess of carbon dioxide in the atmosphere.
It is also possible that, even at pre-industrial concentrations of atmospheric CO2, we were already above the “saturation point” of IR absorption by CO2, and therefore, even the warming that has occurred in the last 150 years could not have been caused by carbon dioxide.”
December 6, 2009, 11:15 amSource: http://wiki.answers.com/Q/How_is_Carbon_dioxide_a_greenhouse_gas
sitzpinkler says:
If I have some supplies and a fisherman with me, I go with the full rations.
December 7, 2009, 10:37 amGaryC says:
If you go beyond 14 microns, you will find that CO2 is again a significant absorber.
With respect to the amount of energy in the CO2 absorption bands, check the Wikipedia page for “Planck’s law” to get equations for Intensity as a function of wavelength and temperature, I(lambda, T). The earth is not a true blackbody over this wavelength range, but the emissivity should exceed 85% , so it would be close.
December 9, 2009, 9:13 pmSlow says:
Here’s an analogy. The war in Iraq.
Substitute “the intelligence” for “the science.”
It’s 2002. There’s disagreement about what Saddam’s done with the WMD he’s used in the past. No one knows what he has now.
But surely the consequences will be catastrophic if we do not act, and soon.
December 21, 2009, 1:25 am