Temperature is correlated with and generally lags CO2

Comments (3)

• 

  Florian

  13 June 2012 at 06:27 | #

  Well done Willis. When I read in their paper about potential phyicsal explanations for the correlations between temperature, CO2 concentration and AMOC variability in three transient simulations of the last deglaciation I started wondering about the purpose of all this verbiage. Climate simulations as far as I go have been losers and I certainly can't check any of this stuff myself. After more unnecessary verbiage about Uncertainty analysis and Robustnes of results I realized it was meant to ease us into a belief that they have discovered something big: carbon dioxide did not follow but preceded end-Pleistocene warming. I never would have guessed it from their graphs. It is clear that this paper, as all others emanating from the climate establishment, takes it as a matter of faith that any observed warming is caused by the enhanced greenhouse effect of carbon dioxide and then attempts to prove it. There is just this one problem with their assumption: the chief greenhouse gas on earth is not carbon dioxide but water vapor. They both absorb outgoing infrared (long-wave) radiation and it is their combined absorption of radiant energy that causes the atmosphere to get warm. But now consider this: when we don't change the amount of carbon dioxide in the air we have a stable climate. There are local temperature and humidity variations, to be sure, but long-term drift is absent. What guarantees this? To prevent a long term temperature drift the IR absorption by greenhouse gas concentration that determines IR transmittance of the atmosphere must respond to any such temperature drift. And water vapor is the only greenhouse gas that can easily do that. Starting from this qualitative picture Ferenc Miskolczi brought in radiation theory and showed that for a stable climate to exist the optical thickness of the atmosphere in the infrared had to have a value of 1.86 (15% transmittance). This transmittance is determined by the combined absorption of infrared radiation by all the greenhouse gases present, but the adjustment is maintained by water vapor, the only adjustable greenhouse gas in the lot. The blogosphere was hostile to the idea because it wiped out the sacrosanct Arrhenius law. But Miskolczi went on to test it using NOAA database of weather balloon observations that goes back to 1948. He found that the IR transmittance of the atmosphere had been constant for the previous 61 years as his theory predicted (E&E 21(4):243-262, 2010). During that same period of time the amount of carbon dioxide in air increased by 21.6 percent. This means that the addition of all this carbon dioxide to air had no effect whatsoever upon the absorption of IR by the atmosphere. And no absorption means no greenhouse effect, case closed. This is an empirical observation, not derived from any theory, and it overrides any theoretical calculations that do not agree with it. Specifically, it overrides any calculations based on climate models that use the greenhouse effect to predict warming. In accord with this, a close examination of the temperature history of the last 100 years reveals that there has been no greenhouse warming at all during this entire period. Starting with the twentieth century, the first part of the twentieth century warming started in 1910 and stopped in 1940. There was no corresponding increase of carbon dioxide at the beginning of this warming which means that according to the laws of physics it cannot be greenhouse warming. Bjf8rn Lomborg attributes this warming to solar influence and I agree with him. There was no warming in the fifties, sixties, and seventies while carbon dioxide relentlessly increased. There is no satisfactory explanation for this lack of warming, only various contorted excuses to explain it away. The true reason for this lack of warming is clear from Miskolczi's work. There was no warming in the eighties and nineties either according to the satellite temperature measurements. There was only a short spurt of warming between 1998 and 2002 caused by the warm water that the super El Nino of 1998 had carried across the ocean. And there was no warming from that point on to the present while carbon dioxide just kept on going up on its merry way. And if you still think Arctic warming proves the existence of greenhouse warming think again: Arctic warming is not greenhouse warming either and is caused by Atlantic Ocean currents carrying warm Gulf Stream water into the Arctic (E&E 22(8):1067-1083, 2011). Taking all this history and Miskolczi's theory into account the attempt of this Nature article to explain the end-Pleistocene warming as greenhouse warming is nothing more than hopelessly misguided global warming doctrine.REPLY: Try learning about the revolutionary new feature: PARAGRAPHS- Anthony

  reply
• 

  Joanna

  12 June 2012 at 17:07 | #

  Downdraft says:April 6, 2012 at 10:42 amObviously, all you need to do to arrive at an exact rcerod of global temperatures is average all the proxy results.Your point is not wrong, even if you sarced it!Statistics says, that if you have enough sloppy measurers, the mean (not average) with the st.dev. is a more reliable result than the single number measured by the best measurer. Hard to conceptualize that, I know! Yet it is true science.Well, not exactly. In this particular case it has to be a weighted average. The sites are not determined by a random statistical process (like generating a random point from a uniform surface distribution) and then finding a proxy at that point. Rather, there are some places on the Earth's surface where, for better or worse, some sort of proxy of earlier temperatures is supposedly preserved. Since that is generally not true at arbitrary locations, one has to take what one can get.Those locations are drawn from an environment that we know was not homogeneous (and this is of course immediately apparent in the data). In particular, there is a very clear latitudinal structure, as we might expect, but with surprises like Greenland that confound any simple rule we might use to relate a local measurement to a smooth function of expected temperature variation by latitude. However, beyond that we know perfectly well that temperatures on a seashore are often terribly correlated with interior/mainland temperatures, sometimes interior temperatures drawn from just a few miles away. San Francisco is a marvelous example you can have multidegree differences in average temperature drawn from sites separated by a hill inside the city, let alone the temperatures drawn from coastal sites compared to sites ten or twenty miles inland. In other words, all of the problems we have now with GISS and HadCRUT, except that we have only a tiny handful of samples, we have no knowledge whatsoever about the moral equivalent of UHI effects (local environmental deviations from even a local thermal average because a site happens to be on what was a southwest facing hill 12,000 years ago until a major earthquake rearranged the landscape so now it faces north). Finally, because we are averaging on a sphere there is the eternal problem of the Jacobean.Now a glance at the coverage map Willis placed up at the top reveals major problems. First of all, it looks like around 90% of the sites are coastal. Whoa, huge problem. Surface to perimeter of the continents being what it is, this means that nearly all of the land surface area remains unsampled. Indeed, areas omitted include things like all of Asia (except for enormous oversampling of coastal regions and islands, e.g. Japan) what's up with that? Coastal sites are the worst possible sites to sample, as they are buffered by the ocean and fail to show the full variability of the climate elsewhere, even elsewhere a very short distance away.Second, although it is difficult for me to tell for sure on the 2D map provided, it looks like the coverage oversamples the poles compared to the tropics. I know, it is difficult to be sure, because one cannot see it very clearly, but there is a lot more area in a given latitudinal band near the equator than there is near the poles, and did you notice the lack of samples from this thing called the pacific ocean , which covers a mere 1/2 of the planet, except for oversampled clusters at a few specific locations on its perimeter?I do agree that one learns something by just averaging the renormalized data together, especially if one smooths the average laterally in time as well as vertically across the samples, but what one ends up with is not a valid estimate of any sort of global temperature or global anomaly. Rather it is a curve that one hopes is somehow a monotonic function of the global temperature and/or global thermal variation across the time frame in question.If I wanted to turn this into an actual estimate of global temperature presuming that one had the slightest hope of normalizing proxy data to an actual temperature scale at even 2-3K precision I would absolutely not do a flat average. I might do a flat average of Japan otherwise absurdly overrepresented and then weight Japan alone with a factor (ideally an empirically determined parameter) that I might hope is its correlation function with local areal temperatures in other words to the extent that temperatures in Japan are predictors of temperatures in mid-China, one might create a reverse projection (with a very large error bar, obviously) to assign a very weakly weighted estimate of the probable temperatures in the unsampled place from the sampled place.Ditto the other sites. For example, there appears to be a sample from or near the North Carolina Coast. The Coast is a lousy proxy for the interior. For one thing, it is 2-4K warmer, on average, in the winter, and 2-4K cooler, on average, in the summer. It is often sweltering in Durham but delightful a full 10F cooler at Beaufort when I teach there in the summer. In Durham we can have frost and snow in the winter, but palms and beaugenvillias and oleanders thrive in Beaufort. Durham isn't a good proxy for Chapel Hill the latter is on a hill where spring comes sooner but the summer stays a bit cooler. Neither is great for the Smoky Mountains or Appalachians or for the coastal plains. The sandhills trap and release water differently from the piedmont differently from the mountains differently from the oceanic moisture on the coast so the patterns of precipitation are different across the state. And this is just one state go over the mountains into Ohio, go west to Kansas just how good is that NC coastal proxy going to be at predicting the temperature or even just the temperature variation in Kansas? A change in patterns of prevailing wind over the 15,000 years of the proxies is enough to completely erase any modern correlations and replace them with something entirely different. A change in ocean currents would be far worse, and just such a change is supposed to have occurred and been involved in the Younger Dryas.Speaking of which, just where did the Younger Dryas go? There it is, big as day in the antarctic ice cores, unmistakable and huge. Yet it completely disappears in the dot-o-gram of the collective data. Was the YD an antarctic polar event? I didn't think so, but this data seems to more or less erase it elsewhere. This is perhaps not surprising, given that so much of the data is coastal and thermally buffered, but given the clear trace of the YD on the land where it is correlated with things like severe drought and dust storms on the NA continent, clearly visible as sedimentation layers the gully walls of streams not far from my house this means that it is a terrible proxy for climate elsewhere.The point is that, with care, one might be able to do something with the data, but the data raises far more question than it answers. I remain unconvinced by the assertions that axial precession plus orbital resonance plus a touch of magic are responsible for the bistability of the Earth's climate. Perhaps all of these contribute, but they were all present four or five million years ago more or less as they are today, but the Earth was stably warm in spite of it. Until we really know what went on to alter this and create the bistability, assigning a cause (set) to the warming that started 15000 years ago, trying to resolve CO_2 from the multiplicity of causes or contributing agencies is absurd. And in the end, the data needs to make sense. This dataset is not senseless, but it is filled with puzzles far more than it provides answers.rgb

  reply
• 

  Cobbalenka

  09 April 2012 at 05:31 | #

  WE NEED TO DO SOMETHING ABOUT IT NOW,NOT ONLY FOR US NOW,FOR THE NEW GENERATIONS TO COME!!

  reply