Friday, August 01, 2014

Climate Clarification



Update 1:  James talks about modeling this morning over at I Don't Know But...

Update 2:  I am told that Max Planck's theory of science change has been investigated and proven false.


It keeps coming up enough that I find a need to set down a statement of what my default prejudices are on climate controversies.  Apparently it is very easy to be misunderstood.  I’m mostly writing this for my own benefit.  I will include some visuals just to break up the appearance.

There has been slight warming since 1800.  As the Little Ice Age ended in 1870 or so, that is hardly surprising.  Exactly how much depends on where one starts measuring, with advocates trying to manipulate both the time and temp axes to show either dramatic or non-dramatic effect.  (You know how I hate that.) 
 But there is some, even with all the questions about measurement differences, urbanisation heat islands, and the like.

In particular, there was a brief, steep increase from 1985-1997. The temps have held steady since then, so it’s “no further warming, but still at a high level.”  The 1930’s may have been warmer, particularly in the arctic.  There was a very slight cooling trend from 1940-75 (there were others in the past two centuries as well), but even with that, the 200-year trend is up.


So the folks who claim there has been no warming whatsoever have a significant burden to show that there is something in that period’s measurements which causes a consistent overestimate.  So far, they have offered possibilities but not convincing proof of that.


What is causing this is extremely complicated, and many disciplines weigh in on solar radiation, volcanism and tectonics, water vapor, feedback mechanisms and the like.  Most of the pixels are devoted to the effect of human activity, as you well know.  My reading of the Scientific Consensus Ô is that almost everyone agrees that there is some human influence on the climate, from agriculture, deforestation, and use of fossil fuels. There are a few holdouts – and these are people with real credentials and peer-review - who say there is none, or almost none.

There is not, however, such a huge consensus that the anthropogenic part is large. It may be north of 50% who would make that larger claim, but nothing near the 97% often quoted.   (That study measured those who said “some.”) As far as I can tell, no one has measured the percentage of scientists who say that humankind is driving “most,” or “virtually all,” or “a significant amount” of the warming. Perhaps it is indeed up north of 90%, as is commonly claimed.  Or less than 50%.  We don’t know.  It pays to remember that on such questions, a great many researchers are of a professionally cautious bent who would be noncommittal even if they strongly suspected an effect.  Others might err in the opposite direction, dramatising or overstating their claims in an effort to spark action they think necessary.

There is no study of the cautious vs incautious personality styles of climate scientists compared to other disciplines. Even though I just made that up, I’m pretty sure it’s true.

I don’t think that the field is riddled with dishonest scientists.  I don’t think there is a consistent conspiracy to suppress data that is suspected to be true but is politically inconvenient.  Yet all disciplines are partially politicised.

Let me expand on that last just a bit.  The DSM-V is a politicised document in more than one way, as were the previous DSM’s.  However, it contains a lot of useful information in framing psychiatric diagnoses, it helps everyone in the field know what everyone else means by certain terms, it eliminates a good deal of vagueness of thought. It is not invalidated by the fact that much of its final wording came down to bullying, horse-trading, and resigned acquiescence.  You can go online and read people who are convinced that the whole thing is nonsense: it goes too far or doesn’t go far enough.

Supreme Court decisions often have political considerations underlying them.  Journalists, and people who get their ideas from journalists, would have you believe that there is little but politics in those decisions.  That’s because that is what they understand, so they believe that must be the only real framing. I don’t see it that way.  I figure in a 6-3 or 5-4 decision both sides must have some valid, arguable points or we wouldn’t have gotten there.  I think SCOTUS justices can be wrong, even badly wrong, and they certainly have their very general political approaches which show in their reasoning.  They are products of their times and training, as we all are. Just because their decisions have some politics in them does not mean they are entirely invalid, however.

Plate tectonics did not finally win out until the 1960’s.  The importance of vitamins still had important opponents until the 1930’s.  People still believe in Freud, and Kinsey, or that stress causes ulcers. The food pyramid was only scrapped last Tuesday. Economists don’t agree about anything, it seems.  There have been scandals in psychology and sociology of researchers fudging data. Anthropology and Education in general make me crazy because they are so riddled with presuppositions.  Noam Chomsky exercised a general suppression effect on certain lines of cognition research over decades – even according to the people who agree with him.  Steven Pinker, Nicholas Wade,  Luigi Luca Cavelli-Sforza  are all aware that there are things you just can’t say without inviting wrath.  People hang on to ideas they like, and change comes in science – well, I’ll let Max Planck say it: A scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die and a new generation grows up that is familiar with it.

Thus, when I make the rather dull observation that there is much that is political in climate science, the counter-accusation that I must therefore believe in conspiracy theories and “denying” the evidence on a par with creationists or holocaust-deniers is just annoying.  Also, it makes my little ears prick up.  People confident in their knowledge are comfortable with qualifiers, uncertainties, plus-minus, allowing some validity to an opponent’s points and the like.  Those who reflexively go nuclear make me suspicious that they are oversensitive because they are indeed more political than average.  That is what I do in fact believe about climate science at the moment.  I don’t believe it is invalid or worthless.  I do believe that it is more politicised than most sciences – when you’ve got the UN-derived groups involved as some of your biggest players you’ve rather invited that – and I believe that its claims exceed its data.  On this last, the actual researchers may be little at fault.  It may be that they are merely not protesting in favor of cautious interpretation as much as they should be for good science.  And I admit, that’s probably hard to do when everyone is telling you how important your research is.  It would go to my head pretty quickly.  Note: In at least one major instance, the claims have demonstrably exceeded the data. The most recent IPCC report had to reluctantly admit that even the lowest of its previous predicted warming scenarios turned out to be too high.

There’s some evidence that they are worse and do neglect or cover up evidence more than other disciplines.  How much that evidence is overvalued by opponents for political reasons of their own I don’t know.  Whether their failings are actually worse than researchers in other fields, or simply more publicised, I don’t know.  Whether even real shoddiness is ultimately not very crucial to evaluation, I don’t know.


Next, we have the whole question of how much does it matter?


There is a genuine consensus that if there were a whole lot of warmin’ goin’ on, the net effect would be bad for humans. There are big Howevers behind that.

There isn’t evidence that there is a lot of warming.  There are projections, with varying degrees of uncertainty, that there could be a lot in the future.  Those projections are based on computer modeling, which in turn is based on the assumptions the modelers build in. This is simply a very weak link in the chain, and there’s no getting around it.  This is an area where very small changes in assumptions can result in large changes in predicition.  People don’t have to be deeply biased to get things wrong here.  There doesn’t have to be a nefarious conspiracy.

This is a place where some skeptics get things badly wrong.  They conclude that large differences between projected results and actual results means that there is something significantly rotten in the state of Denmark.  The whole batch of climate scientists, from all those disciplines that hardly ever speak to each other, must nonetheless be keeping the Real True Information out but mutual agreement, winks, and nods.  As we have noted above that such soft conspiracies, refusing to question the conventional wisdom, have occurred in other fields, it is of course possible that we are in the midst of some consistent, stubborn, anosognosia among the climate-research power brokers which bends the research of the entire field.  Possible, but not necessary.  A very little shared expectation, natural favoritism, and confirmation bias can do the trick.  I think most climate researchers are honest individuals who genuinely want to find correct answers.  I believe that in the face of enormous contrary evidence, they would change their views (though perhaps gradually).  But absent that, the normal human trend is to stick with one’s original POV, highlighting supportive evidence, downplaying disconfirming evidence.  I’m taking on general faith in people’s desire to be competent in their work that the researchers are more likely to modify their beliefs than the skeptics.

Perhaps I am naïve there, because maintaining ones livelihood, aside from any considerations of saving face, can be a powerful deterent to rocking the boat, but I still hold to it.

All that to say that the reliance on projections is unconvincing. They might be steadily improving and trustworthy now.  But the previous projections have turned out to be badly wrong, and in precisely the same way that one would expect if the initial assumptions were wrong.  Modeling builds in fudge factors all the time.  Ithas to.  

This is not Newtonian mechanics, and we should not be treating the evidence as if it is unequivocal. There are multiple variables, there is uncertainty, there is ambiguity. Climate science is more like economics.  We know some things, but others are uncertain.  Sometimes things happen and we aren’t sure why.

The second However: In all the talk of scientific consensus, there is one consensus that doesn’t get mentioned often.  Results will not be uniform across the globe, nor will they be all negative. As above, a lot of warming would cause net harm, even catastrophe.  But a little warming?  YMMV.  Respectable minds have concluded that a little might actually be a net good.  Therefore, to treat warming as a possible catastrophe, one has to show that there’s going to be a lot, and more quickly than mankind can manage. My understanding of IPCC Fourth Assessment summary (2007) is that only the highest projected trendline was potentially catastrophic.  We didn’t even reach the lowest, and the early releases of the Fifth Assessment backpedal on that.

Some of the observed data is indeed worrisome, such as the part that agriculture has played in warming.  That doesn’t get as much attention, but it could be forcing change. (Why it doesn’t get attention is an interesting discussion.)  But some of the report is puffery, such as the observation that the last fifty years have been the warmest in the last 700.  Well yeah – most of the last 700 was the Little Ice Age. Most of the worrisome warming was that 12-year leap, which though we have not fallen back from, also hasn’t increased, despite the continued increase in greenhouse gases.

Warmer will mean higher agricultural yields.  The largest stores of available freshwater in North America are in the cooler regions.  People die of cold more than heat. The food disruption of where the fish are and how many there are, may be the largest issue, and not insoluble.

Which is not to say that we’re safe, and we should disregard the warnings and be dismissive.  Through it all, there likely is something to this.  It just has not been evidenced strongly enough to convince doubters that drastic action is needed.  There is a little warming, quite recent, and that has paused but not gone away.  What is the evidence for catastrophe?  I don’t mean “What would a catastrophe look like if there were one?”  That’s a very different qauestion. That some of the skeptics are intransigient and will not be convinced by any data is irrelevant.  That is true of every field.  Treating all skepticism as invalid is evidence that the most popular explainers simply don’t know what they’re talking about. That isn’t done, in any discipline.

There are no scientific organisations which commit to a statement that human influence is neglible – there are individual scientists, * but not groups – but there are some which are neutral.  Organisations of geologists are most prominent among them – I have no idea why.  More cautious?  Longer perspective? Still, through all the hoopla there are respectable bodies which are at least not counseling urgency.  It doesn’t mean they’re right yet it’s worth noting.

The precautionary principle always sounds simple and obvious to those who already believe. Gee, why wouldn’t we invest heavily in alternative energy sources and tax petroleum like crazy to reduce consumption?  Okay, then, why don’t we go to war with a new small country every year, just because one of them might develop nukes?  Think how bad it would be if Haiti got nukes.  Why not say the Sinner’s Prayer word-for-word in one breath, tearfully, just in case it’s the only sure way into heaven?  Why not ban bicycles because of injuries?  It doesn’t hold.  First, you have to establish that there is some real likelihood. If tipping point arguments are supposed to be so persuasive, then why are they dismissed in discussions of taxation and regulation?

The third However may be the largest of all.  Even if true, what do we plan to do about it?  As PJ O’Rourke noted “There are 1.3 billion people in China, and they all want a Buick.”

* Warning:  Biased source.  But the info seems accurate.

11 comments:

  1. The main problem is ignorance. Ignorance of chaos theory.

    "The theory was summarized by Edward Lorenz as follows:
    Chaos: When the present determines the future, but the approximate present does not approximately determine the future."

    It is impossible to determine the initial conditions for the system. The instruments would necessarily weigh many times the mass of the atmosphere.

    IOW, it is impossible to model the climate.

    ReplyDelete
  2. The infamous hockey-stick aside (an unstable fit to the data), the biggest problem I see is that of defining what a "global" temperature is, and then trying to accommodate that definition using existing sensors (What is the temperature profile of the atmosphere, not just the ground level temperature?) which don't by any means span the globe, and then trying to weight older measurements with fewer sensors and different environments (when did they pave the parking lot near the sensor?). What is the effect of the oceans and clouds? I've never seen any estimate of the systematic uncertainty of the models, but it can't be small.

    ReplyDelete
  3. Nobody will specify what the median standard climate is/should look like in the various places around the earth. And give us believable reasons to justify it.

    ReplyDelete
  4. Most sources are biased. The honest ones tell us what their biases are.

    ReplyDelete
  5. Reprinted with permission from a friend who does not comment online but emailed me.

    David,

    I'm enjoying reading about your perspective on Climate Change, and especially the struggle to understand what's true. My experience, both lifetime and recent, has produced a perspective on the influence of government on truth.

    There doesn't need to be a conspiracy on the part of politicians to corrupt the research process, or to control its results. It is simply a process of a selection bias. For instance, my professional career has been 40 years of providing research and products to the government, primarily the defense department. My consistent observation is that good ideas don't get funded; rather, products that match what the government thinks it wants, and research that supports what it wants to believe, get funded. Proving your benefactor wrong is a career and business limiting result.

    At least in defense there is a correction process. If you build the wrong stuff, your adversary will gain an advantage, and when that becomes apparent, what the government wants will change. In the case of climate change, there is no such countervailing force. There is no organization with deep pockets and a vested interest in "proving" that climate change is not a crisis to fund balanced research. The closest we have is major countries, like China, Russia, and India who just won't play along. They are perfectly willing for the US to stifle its economy chasing environmental purity.

    I have another, more recent comparison that informs my assessment of how desire for political power influences research and the definition of crisis. I've recently taken a liking to free, on-line courses, offered from major universities. I've taken a course in climate change from the University of Chicago, and several courses in macroeconomics from UCal and Yale and a business school in Spain.

    While the climate change shills have the vaunted "hockey stick" graph, showing temperature going out of control, the Congressional Budget Office has a similar hockey stick graph, showing the US debt going out of control ... and it's sharper, and happens sooner, than even the most extreme climatologists project.

    It's interesting to compare the science behind the graphs. In the climate change case, there is, as you observed, a considerable amount of complexity, requiring models with inherent assumptions that are hard to justify, but radically affect the results. In the national debt projection, which is virtually entirely driven by the coming expansion of retired boomers collecting Social Security and Medicare (and the expansion of Medicaid and Obamacare), it also takes models to project where we'll be in 20 years. However, in this case, the assumptions are rather simple and defendable ... it's pretty easy to count the number of people who will be collecting, (and how many will be paying) and life expectancies are fairly well understood and slowly changing, as are medical costs, without some major breakthrough.

    So, it's interesting to compare the political response to these models. In the climate change case, where politicians stand to gain power by collecting more taxes to pass back to their benefactors, there's lots of hype. In the national debt case, which might require politicians to restrict the government largesse, no comment.

    ReplyDelete
  6. Oh NOOOOOOOoooeeeeeees! Gummint has a vested interest in the outcome? Say it ain't SO, AVI!

    ReplyDelete
  7. Sam, your comment is prompting another post, about the default position, even among fairly cynical and skeptical liberals who will make jokes about The Feds, that government is essentially benign and wants to help.

    ReplyDelete
  8. "...that government is essentially benign and wants to help."

    Essentially benign--with armed forces and groups, and those "fairly cynical and skeptical liberals" don't believe that regular citizens should have guns because "they don't need them".

    "...wants to help." That we don't want; or if we do want help, it's not the kind they are willing to give, or want to give, because they know what's best for us.

    I eagerly await that next post. Don't hurry on my account, take the time to say it the way you want.

    ReplyDelete
  9. Anonymous5:07 PM

    whether climate's changing (pun:) - i'm less concerned about mean temperature than i am about variability - is variability increasing? going by the last few years, maybe - BUT, compared to the past say 1000 or 100,000 years (tree rings, etc.) it seems variability is less these days than in the ol' ol' days. as david cross once said, "it's got me all bugaboo!" :) i'd worry more about extreme variability (harder to adjust to) than a slowly increasing mean (slow changes in same direction: easier to adjust to). thank you for an excellent post - & your friend's email/comment is superb - especially their last paragraph!

    ReplyDelete
  10. Your observation about Geologists being more cautious is spot on. Their education covers scales of billions of years where temperatures varied widely along with the concentrations of oxygen and co2 (far more change than is postulated by climate models). Over geologic time cold had been far worse for life than heat as covering the earth with mile high glaciers is not conducive to most land animals. They cannot adequately explain many of the most dramatic changes in earths climate so how can the short term (100's of years) climate models do so with any precision.

    ReplyDelete
  11. There is a genuine consensus that if there were a whole lot of warmin’ goin’ on, the net effect would be bad for humans.

    Mrph. A warmer world is a more hospitable world for life in all ways. Just because our species has spent all its lifetime in an ice age - a cold, dry, hostile climate, with glacial periods always threatening to destroy us - doesn't make that the optimum.

    We do tend to take "recent history" as "the way things oughta be" without question. (The Sahara is supposed to be a desert, and having it turn to savannah as it does during warm periods ain't natural. Northern Canada and Russia are supposed to be treeless and near-barren. CO2 levels are supposed to be so low as to keep plant life relatively stunted. Etc. If the subterranean rivers in the Sahara came up to the surface and the permafrost melted and plants had an easier time of it then it'd be chaos! Pure chaos!)

    Just one more perspective.

    ReplyDelete