The truly enlightened pagen worships jove. Full emacs needs to go on a diet, and vi, well, let's just say that they spell it e-VI-l for a reason...

Don't be silly. Any climatologist will tell you that's the first thing you have to take into account for the climate. We've had good measurements of insolation since the 1950's and very good satellite measurements since satellites started measuring it in the 1970's (or early 1980's). The simple fact is that the Sun's output hasn't varied enough to account for all of the temperature change. At best it can account for less than 10% of it. On top of that if the warming was caused by the Sun the whole atmosphere would be warming. But in fact the stratosphere has been cooling, a signature of greenhouse gas caused warming.

This is what I meant, actually -- that the insolation variation alone is too small to account for the warming. But that's also very likely true over the last 300 years post LIA, and it warmed nevertheless. And yes, I look at the thermal proxy and otherwise temperature records and reconstructions.

And I agree that CO_2 almost certainly causes some warming. Where I disagree is that I do not think that one can determine how much, and I vehemently disagree that the climate sensitivity is as high as has been asserted. I also think that our probable degree of knowledge and confidence is being overstated, although in the working group reports a lot more doubt is expressed than ever makes it out to the public.

Finally, I absolutely agree that the right thing to do is stick to the science. The GHE is certainly real -- clearly visible in TOA vs BOA spectroscopy. However, the CO_2-linked component of this is just one factor in a very complicated nonlinear chaotic system that is never quite in a state of dynamical equilibrium, with a huge degree of natural variability visible in the past that we do understand and cannot (retroactively) predict or convincingly explain. The net CO_2 linked GHE-driven AGW could range from basically zero (nearly completely cancelled in the long run by negative feedbacks or other bad habits like air pollution aerosols) to substantial, but I think its effect is probably being overestimated and certainly being overstated, and is likely to be less than disastrous, or less disastrous than some of the measures being urged to ameliorate or prevent it. I'm not completely alone in this, even among physicists interested in the climate.

My specific personal interest in the science is more in the statistics of it. There is an amazing series of papers by a guy named Koutsoyiannis who has been studying hydrology "forever" out of Athens (Greece, not Georgia). He long ago observed that water levels and periods of drought and flood follow patterns of "punctuated equilibrium" that are describable by what he calls Hurst-Kolmogorov statistics. Those same patterns are clearly visible in the thermal record, and indeed the 33 years of the satellite era are precisely that -- stable temperatures for 15 years, a sudden jump over 2-3 years associated with a single event (an unusually strong El Nino), followed by stable temperatures for 15 years (where by "stable" I mean specficially that there is a great deal of noise and oscillation, but that the linear trend is not resolvable from zero).

Two intervals does not a theory make, but it is very suggestive, especially when those temperatures are very likely to be (in some sense) a projection of a complicated poincare cycle around an occult multidimensional attractor. The interesting question from this point of view is what moves the attractors (and what keeps them locally stable!)

This is a macrodynamics question, not a microdynamics question. I agree that we know a lot of the components of the microdynamics, but there are macrodynamical features that I don't think can be captured in a detailed model because they represent large scale self-organization of the underlying heat flow. Even only within the Holocene, phenomena like the Younger Dryas suggest that there is some serious variability that can be triggered strictly out of internal non-linearity completely independent of CO_2. CO_2 could even cause the buffering of that variability and stabilize the climate in only, or mostly, good ways -- as Lindzen (and I, in other contexts) have pointed out, a uniformly warmer world is likely to be less stormy because storms are caused by thermal gradients, not absolute temperature.

Overall, I think there is a lot of science left to be done before literally betting the ranch on a single hypothesis, especially when there is remarkably little evidence supporting that hypothesis. Seriously. SLR will be extreme! But it isn't yet. Temperatures will go unrelentingly up! But they haven't, not compared to what they were doing anyway post LIA seeking some new "equilibrium" we cannot predict, describe, or even heuristically understand any more than we can understand why the LIA itself occurred. More droughts! Compared to the 30s? Not happening, at least not yet. Worse storms! We are in the longest interval ever recorded without a major hurricane making landfall in the US, an interval almost certain to extend by at least seven more months before there is any reasonable chance of one, and this year was subnormal for tornadoes (and really, pretty unremarkable except for one warm spell in the US with a specific and understood cause). With the ENSO meter currently precisely on neutral (down from a very short period of very mild ENSO conditions and falling), we could have a third La Nina in a row without an intervening ENSO. What does this even mean? What is its cause? Why does the world stubbornly refuse to warm (and maybe even cool a bit) in association with La Nina and warm in association with El Nino in proportion to its strength, and why does it ignore CO_2 as far as one can tell in between, at least as far as the LTT or SSTs are concerned?

Eventually some of the not yets may come to pass. Or not! In the meantime, there is little empirical reason to panic, and a lot of pure unadulterated hype that transforms Sandy into evidence of "climate change". One needs to be a skeptic with this sort of egregious manipulation of public opinion occurring unopposed by actual scientists who clearly know better.

rgb

By the thirty year standard, there has barely been enough time to resolve any climate variation in the modern post-CO_2 world, let alone ascribe the fractions to AGW and natural variation. Indeed, the entire satellite era comprises just one of these intervals, and IMO that is the only stretch where we have truly reliable global data mostly free from the possibility of a variety of biases that have to be estimated, often badly. In that stretch, nearly 100% of the warming can arguably be ascribed to a single El Nino event, as the UAH LTT data is basically flat from 1979 to 1997, goes steeply up in 1998 to overshoot a 0.3C total rise that more or less is flat thereafter up to the present. One could easily be tempted to conclude that "a strong ENSO causes global warming", based strictly on this data, were it not post hoc ergo propter hoc, like so much of the discussion on both sides (which also begs the question of "what causes ENSO and what modulates its strength", which is AFAIK rather unknown).

As to whether 16 years is too short, I'm sure that I don't have to quote the 2008 report to you, which excluded a 15 year stretch without statistically significant warming at the 95% or better confidence level. It is the stubborn perpetuation of this stretch that is requiring reconsideration of earlier, often egregious, estimates of climate sensitivity. They're coming down, and they'll come down more every year without warming. I know that it is physics-based dogma that solar variation cannot possibly affect climate, but historically there has been a fair bit of correlation between solar state and climate (and there are a few proposed plausible causal mechanisms, which I will not presume to judge) and we are very likely to be at the peak of the lowest solar cycle in 100 years, with the prospect of the next cycle being even lower or the sun entering a Maunder minimum. At the end of this we may know a lot more than we do today -- either warming will resume with a vengeance (as it has a lot of catching up to do at this point) or it will remain flat or even cool. Either of the latter two will force substantial revision of everything -- flat by gradually reducing sensitivity still more but perhaps leaving the GCMs alive, actual cooling might cause the GCMs to be thrown under the bus and rebuilt from scratch.

Remember that there are substantial, poorly understood nonlinearities in the climate system, and that even very small non-CO_2 influences can be amplified. There has been a substantial and (as far as I know) unexplained reduction in stratospheric water vapor content in the current solar cycle, for example. This in turn can actually lower the troposphere (permitting escape from greater depth) and reducing ALR warming. Is this a chance fluctuation (quite possible) or evidence of a process we hadn't anticipated? There is a lot more science undone than done in climate science, because we have so little high quality instrumental data over such a very short time frame -- basically a single "minimum interval" by the very 30 year standard you cite.

Climate models are tested by hindcasting, sure. Can they hindcast the MWP and LIA? Can they explain why there was a warming trend from the beginning of the thermometric era until the present in the absence of CO_2? In other words, can they explain the baseline climate variation over geological timescales? I don't think so. They are attempting to fit/predict local anomalies without anything like certain knowledge of local baseline behavior and where they are literally incapable AFAIK of reproducing it outside of a very narrow time window. As I said, I love models and modelling. It's one of the things I do. But it helps to have a model that first works in the big strokes arena, getting the gross behavior right and THEN worrying about the details. I rather think that current models have this backwards, and are thus confusing signal and noise.

It is easy to fit nonlinear functions with an overcomplete basis in a completely "convincing" way, when one doesn't have anything like a property of uniform convergence to help you out. Is this what is happening? I don't know. It might be.

Regarding c) -- what made summer Arctic sea ice decline dramatically in the 1930s? It certainly wasn't CO_2. Would it be another "I don't know" in a long list of unknown things about past climate? What caused the dust bowl, also in the 1930s? In a lot of ways, the present resembles the 1930s (starting at a different baseline temperature, of course, given the ongoing post-LIA warming that we do not understand). Why?

Of course one can always turn the 1930s into anecdotal evidence and ignore it, because we didn't have satellites (we barely had airplanes). Antarctica was a great big question mark. Siberia was mostly wilderness. Radar wasn't even on the horizon (so to speak).

But that just emphasizes my point -- we have 33 or so years of moderately reliable satellite-based data. We might eke that out with soundings on a less global scale back another 20 or even 30 years, but it is also less reliable. Beyond that we have little that is reliable enough to do more than inform us of very general trends with huge chunks of the globe essentially unsampled. We have roughly 15 years of satellite sea level measurements (although I'd argue that tidal gauge data is likely to be adequate over a much longer time span because the ocean is basically isostatic so that -- unlike temperature -- going up any reasonable number of locations is a good measure of its mean height, and because tidal gauges are likely to be fairly mechanically accurate in their simplicity). We have even less than that of anything like global ocean temperature measurements at depth.

In other words, we have barely started to accumulate the kind of detailed information required to inform a global climate model, and to be frank, until we have modern instrumentation sampling the planet on a global basis through at least one full cycle of the major decadal oscillations, it is very, very doubtful that GCMs built without that data will be globally accurate on a longer timescale. This makes even the thirty year timeframe too short. ENSO is "easy", and yet it is exhibiting behavior that is difficult to understand with what looks like a triple El Nina without anything like a significant El Nino in between starting (at least the current El Nino is remarkably weak and keeps bouncing out of the El Nino zone). The PDO "just" changed phase for the first time in the satellite era. We still await phase change in the atlantic, where the oscillation is erratic at best. How can we possibly predict -- accurately -- what will happen to the climate when it does? Then there are SSTs and the THC.

Uncertainty in each of the major drivers of the climate is cumulative. A small uncertainty in the effect of the PDO or NAO or ENSO or THC each reduce the certainty of any given climate model that makes assumptions about what the entire, globally unobserved consequences of in some cases major variations in circulation might be. How does solar and/or orbital state couple to all of this? I am a "skeptic" simply because I think that the science here not only isn't "settled" -- it may not even be approximately correct. How can we know? We simply lack the data needed to be able to know, and aren't likely to get it in less than decades more of observation with modern instrumentation. At the moment, we're making inferences with egregious claims of certainty informed by data intervals that are absurdly short compared to the known timescales of natural variation (which we cannot predict or explain on a geological timescale within a factor far greater than the entire warming ascribed to CO_2 so far).

BTW, I appreciate the tone of the discussion. Usually when I post anything skeptical on /. all that happens is that people call me names or accuse me of being a shill of the oil industry, which gets tedious and encourages me to crank up logical fallacy bingo, with popcorn...:-)

rgb

Personally, as a medical researcher, I like to try it and see what happens, but the patients usually prefer we have some theoretical justification and do most of our experimenting in animals (imperfect models) first. In terms of altering global climate I can see how the experimental approach might also have a few issues.

Excellent point. So, for a disease that has never in human-recorded history occurred due to the causes ascribed to its future occurrence, for which there is no empirical evidence in recorded history or prehistory that it is occurring other than an unproven theoretical argument, you would, I'm sure, endorse the medical adage "do no harm" by not prescribing what may be quack remedies for a disease that might not actually exist, it only "theoretically" exists, when those remedies have many severe side effects.

I agree.

It's really rather like not prescribing daily antibiotics for the entire human population because we are pretty sure -- theoretically -- that some disease is likely to evolve that the antibiotics might -- or might not -- prevent. Or taking any other sort of extreme measure in response to Pascal's Wager, however it is formulated. By not placing the predictions of future damage far, far above any visible sign of damage now one justifies the human cost and side effects of any measure taken to combat it, and of course transfer a rather lot of political power into the hands of those we elect to take those measures without any possible mechanism for them being held accountable, either politically or economically.

A recipe for disaster.

rgb

I don't argue with most of this. A few points:

a) Global temperatures have largely levelled off over the last 16 years. Yes, this is also looking at the end of a time series, but over that time global CO_2 levels have risen dramatically. The lower troposphere temperature, as one of the few truly global indicators, is simply not showing the sort of growth it did over the first 17 years, and most of that growth is associated with a single discrete event -- the 1998 El Nino. Sea surface temperatures follow this trend even more directly, being nearly flat on both sides of the El Nino. It is also very, very difficult to separate out the CO_2 derived "warming signal" from the natural rise in temperature the planet has experienced after the little ice age, almost all of which had nothing to do with CO_2. The evidence that CO_2 forced warming is associated with a high climate sensitivity is weak already and weakening further. I do not know what sort of confidence one should place in high sensitivity predictions -- there isn't even good agreement among climate scientists or climate models, and the uncertainty is well-represented in the AR working group reports, just not in their summary for policy makers.

b) The only way to test climate models is to wait for decades and compare them to actual data. This is a test they have not done particularly well with over the last 16 years. In the meantime, one has to assign a lot less confidence to their predictions than is commonly done, given that they are trying to solve what is literally the most difficult problem in computational physics in the world, out to truly absurd future times. Their ability to hindcast and e.g. explain the last 1000 years of climate data is essentially nil. I personally just think that we don't yet know the right physics, or perhaps we do know the basic physics itself but that the complexity of the model is not yet computable. Tiny errors in a highly multivariate nonlinear system can have profound effects the further away you go. I also don't have a lot of confidence in various input assumptions -- not when they are applied to the geological data over long time spans. I think it will take as long as the rest of the century just to get the physics right, and if we were LUCKY we might get it mostly right in 20-30 more years of satellite data (the only data I have a lot of confidence in -- too many thumbs on too many scales in the thermometric record, as evidenced by the increasing divergence between reported land surface temperatures and LT and SS temperatures.

c) We know ice is forming as well as melting. Total sea ice isn't even changing a whole lot, and again it went through a very similar cycle back in the 30's, without CO_2. We simply don't have enough observational data to tell whether what is happening is mostly normal. And nearly all of the observed SLR is from the normal thermal expansion of seawater, and is not happening at an alarming rate.

But the main point is that I completely agree that we need to look carefully at cost vs benefit, based on the actual evidence and not unproven models. The actual evidence does not support drastic and expensive action, it supports research into alternative energy resources that might -- when mature in a decade or three -- be able to reduce the consumption of carbon based energy without causing a worldwide energy depression worse than the problem it seeks to "cure". There is a substantial human cost to most of the steps being taken now to "ameliorate" the problem. In fact, they form a real-time "catastrophe" of their own, one definitely affecting the world right now, not in 80 years, maybe. Every day of energy poverty in the third world is another day of misery, and like it or not, carbon-based energy is cheap and plentiful compared to the currently available alternatives.

Then there is the politics of it all. Nuclear power, for example, could substantially reduce our reliance on coal using well-understood technology -- it is a truly feasible short-term solution. Does that mean that supporters of the catastrophic anthropogenic global warming theory support the emergency adoption of nuclear energy, even as a bridge measure until e.g. PV solar matures or energy storage matures? No, curiously they oppose it almost as vehemently as they do coal based power. Then there is the absurd notion of sending money to countries as some sort of compensation for damages suffered from human-caused warming (as if we can tell what fraction of the warming is caused by humans, which we can't), computed by an equally absurd and arbitrary formula that never seems to look at possible advantages/benefits of the hypothetical warming, only the costs.

It may well be the best possible thing to do nothing expensive until the evidence is clear. In fact, I think it is almost certainly the case that it is. If I'm wrong, we're all screwed anyway no matter what, because if the carbon cycle (Bern) model is correct -- which I doubt -- then our choices really are stop using all energy produced by burning anything now, today, and accept the unbelievably catastrophic cost known as "the collapse of civilization and the pursuit of those who suggested it by peasants armed with pitchforks and torches" right now, or else accept the fact that we've probably already pushed the Earth past the hypothetical "tipping point" to a semi-permanent warm phase, pulling the Earth out of the current ice age, etc.

I really do doubt this -- the Earth has managed ice ages and warm spells with much higher -- and lower -- CO_2 concentrations than we have now repeatedly over geological time -- and we are still far short of the Holocene optimum (which lasted for thousands of years at least a full degree to two degrees warmer than it is now) -- and that is just in the CURRENT interglacial. But we will see.

rgb

One should indeed! One should also be very leary of fitting any kind of fit, linear or nonlinear, to data over only 10-15 years that has varied rather consistently over 140 years: http://en.wikipedia.org/wiki/Current_sea_level_rise. Personally, I think the figure says all one could possibly need to say. Note well that anthropogenic CO_2 was completely irrelevant over almost all of this time series even according to the IPCC. Note also the overall range of the entire chart, roughly 1/4 of which supposedly incorporates "substantial" anthropogenic global warming -- just under 9 inches in 140 years. Note also that there are several periods with rise rates comparable to the present, for even longer periods, e.g. 1938-1950, where CO_2 was again not a factor even according to the IPCC reports.

It may be absolutely true that global climate models predict large amounts of sea level rise, but there is no actual evidence in the form of large rises in sea level to support this! . Perhaps there will be in the future. Perhaps not. I'm a theoretical physicist, and I just love theories. I'm a computational physicist, and love large scale model computations. But at the end of the day, I'm just a physicist, and the theories and computations have to agree with observation. So far, these do not, not even over the last 140 years, and they don't even give us insight into the large climate fluctuations observed over the last (fill in the blank with any number greater than 1000) number of years.

In the meantime, I work every summer literally living at the edge of the ocean facing straight out through the Beaufort inlet in North Carolina. Although I know that tidal gauge data indicates that there has been a sea level rise there over the time I've been working there or otherwise visiting, I certainly can't see it in my own (literal) back yard, where if there were any substantial and consistent rise -- I'm talking a rise of a few inches -- my back yard would be underwater at high tide. My neighbors have lived in their houses for over 40 years, and (yes, I've asked) haven't observed any rise at all, let alone an alarming one, of the highwater marks on their docks or seawalls (where the tops of their docks would be underwater at high tide if there were any consistent rise). This is absolutely anecdotal evidence, although the ocean being isostatic it is difficult to imagine it going up one place and not everyplace else, but see the curve above for the best global tidal gauge and satellite data in summary.

If you looked at this data and didn't know it was "sea level rise" and destined to rise because of Evil Human Activity -- if you were told it was the sales price of widgets, or the mean length of romance novels, over time -- and were asked "is there a statistically meaningful acceleration in trend" visible anywhere in the record -- you wouldn't even bother to do an actual statistical analysis because the answer is fairly evidently "no". If you were asked to estimate how likely it is that any aspect of this trend would justify a final sales figure for widgets of 48 (9 plus 39 more) in only 88 more years -- that would be over four times the entire growth over 142 years in only 90 years -- you wouldn't hesitate to give odds of 99 to 1 against. Bayesian analysis might alter the 99 to 1, sure (depending on how sure you are about your priors) but not even Bayes is going to comfortably make this 99 to 1 for.

I'm just sayin'...

rgb

(I will now wait for the usual "refutation" of this, the assertion I'm being paid off by the oil industry or the like. I wish. Instead, take due note of my Russell quote, below.)

Go for it. Bear in mind that the actual data is that SLR is around 3 mm/year -- depending on how short a segment of the data you are willing to cherrypick to prove a point. Since the assertion is made above that SLR is supposed to be a meter more than previously claimed -- hence around 2 meters or even more -- and since here we are in 2012 with SLR having gone up a whole inch (to the nearest inch) in the last decade -- we have to take something like 78 inches and split it up among 88 years. Hmm, if SLR went up by an order of magnitude next year we might just make it.

Otherwise, bear in mind that people who currently have beachfront property could die of old age before SLR becomes an issue for them. You (dear reader) could die of old age before SLR gets high enough to realize a profit on land you bought inland anticipating that it would become oceanfront. Or not.

rgb

Does AC have parents? How could we tell? They too would have to be AC. AC is indeed his own Grandpaw. Or hers. Its.

Why is this sad? Was the collapse of the Ottoman Empire sad? The breakup of the British Empire? Relax, pop yourself a bag of popcorn, and enjoy the show -- it's been a long time coming and worth every minute.

Wait, wait! AC is talking to itself again! It's so confusing. Maybe we need "first AC" and "second AC" but no, then I suppose it wouldn't be anonymous. In the meantime, it sounds like AC is very confused...

Oh, I'm not arguing -- a lot of our linux-centric sysadmin folks here like apple laptops, and as you say, they've long since gotten to where they've got a full or nearly full complement of unixoid tools and features and most of the important OSS offerings. Of course, the students I'm referring to are not in your (or my:-) geek class. They just like them because they are thin, cool as in socially acceptable, and work pretty well. I'd say the "work pretty well" is one of the most important things in the list, even.

My own gripes against apple are strictly premium price, the fact that I need cone-head amounts of compute power even on my laptops a lot of the time (e.g. i7 processors and a mountain of ram), the fact that all of my source is written under linux, and the fact that its "coolness" actually annoys me. Oh, and the fact that my wife's iPhone got a SINGLE DROP OF WATER on the charging port, popped the little red tab, and even though it was fully functional (until it ran out of charge) it refused to charge and had to be completely replaced for almost 1/3 of its cost in spite of us having a service plan on it. Otherwise known as "Rip-Off Hardware Design", brought to you by a company eager to take your money in exchange for coolness.

Personally, I like my Casio android phone with its waterproof covers on all ports. You can (reportedly) drop it into a meter of water, pull it out, wipe it off, and it not only works, it works literally untouched. Not that I care to tempt the gods, but all that AND it cost a fraction of the iphone and has all of the android apps available, most for free.

Apple does sell its own version of kool-ade. That's not to assert that sometimes they don't deliver value for money -- it is to assert that they display the same shocking arrogance that any market leader seems to gain when they get on top in the PC universe. All that engineering skill -- so much that they actually bother to engineer in the little red switch! -- but somehow they can't manage to build a charging interface or user interface that a drop of water can't kill.

Or worse, they could easily do so, but prefer to make all of that money when people have to buy second and third iphones etc.

rgb

I was in our local supermall yesterday. They had an interior kiosk set up to sell Surfaces, manned by an easy half dozen earnest young salespeople hired for the season. They didn't have a single customer in view -- not one in all the times I walked by it. Everybody standing around looking bored.

The Apple store about fifty meters away, on the other had was absolutely packed, as it always is, with customers waiting in line. It wasn't even a busy night at the mall -- parking was actually pretty easy for the season.

The really interesting question is -- can Microsoft compete ANYWHERE on a level playing field? If they didn't have the world's computer retailers in a ball-lock with their pricing formula, would they even exist? The answer is not so clear. I've watched student PC and laptop ownership transition from nearly all WinXX PCs to nearly all Apple products in only five years. iPhone, iPad, iPod, thinline apple laptop -- standard operating equipment for current college students. A smattering of Droid tabs and phones in there -- it is the nerd product and also pretty cool. Even linux-based systems -- the choice of the ubergeek -- are starting to compete with Windows systems for a whole generation of kids.

If Valve/Steam works out and games move over the Linux big time, Windows could actually experience the start of its long awaited death spiral.

rgb

Except that you can practice. State learning matters. If somebody DOES drive all the time high, they very likely learn to compensate, but it is those first few times... and too many people would have lots of state learning on "video games". It would need to be a test nobody is likely to be able to practice on ahead of time.

Perhaps a road simulator video game. That would actually make sense. If they can't drive a video simulator well enough to avoid simple road circumstances that might lead to a crash, then they shouldn't be driving no matter what the reason...

rgb

amendment prohibiting redistribution of money to the states

uneven redistribution of money, conditional use of money, surely, but the provision is already there, in the reservation of the powers to the states. The federal government should not be able to coerce any laws out of the states. If they wanted to pass a federal law raising the drinking age, they should have done so. It would have then been challenged, and would have lost the challenge. Using highway funds as an end run around the inability of the Federal Government to pass a law of this sort itself is de facto proof that it violates the constitution.

This isn't about taxation, BTW. One could argue about their right to differentially tax things to accomplish some end as a completely separate issue. It is strictly about bribery and economic coercion to extend federal powers to issues explicitly reserved to the states.

What in the world were they (the SCOTUS) thinking?

I did not know that. Wow, what were they thinking. That is just so wrong.

rgb