Foreword by Anthony Watts
An essay by Monckton of Brenchley follows, but I wanted to bring this graphic from Dr. Mann’s recent Scientific American article to attention first. In the infamous “hide the decline” episode revealed by Climategate surrounding the modern day ending portion of the “hockey stick”, Mann has been accused of using “Mike’s Nature Trick” to hide the decline in modern (proxy) temperatures by adding on the surface record. In this case, the little white line from his SciAm graphic shows how “the pause” is labeled a “faux pause”, (a little play on words) and how the pause is elevated above past surface temperatures.
Source: http://www.scientificamerican.com/sciam/assets/Image/articles/earth-will-cross-the-climate-danger-threshold-by-2036_large.jpg

Looking at the SciAm graphic (see zoom at right), something didn’t seem right, especially since there doesn’t seem to be any citation given for what the temperature dataset used was. And oddly, the graphic shows Mann’s little white line peaking significantly warmer that the 1998 super El Niño, and showing the current temperature equal to 1998, which doesn’t make any sense.
So, over the weekend I asked Willis Eschenbach to use his “graph digitizer” tool (which he has used before) to turn Mann’s little white line into numerical data, and he happily obliged.
Here is the result when Mann’s little white line is compared and matched to two well known surface temperature anomaly datasets:
What is most interesting is that Mann’s “white line” shows a notable difference during the “pause” from HadCRUT4 and GISS LOTI. Why would our modern era of “the pause” be the only place where a significant divergence exists? It’s like “hide the decline” deja vu.
The digitized Mann’s white line data is available here: Manns_white_line_digitized.(.xlsx)
As of this writing, we don’t know what dataset was used to create Mann’s white line of surface temperature anomaly, or the base period used. On the SciAm graphic it simply says “Source: Michael E. Mann” on the lower right.
It isn’t GISS land ocean temperature index (LOTI), that starts in 1880. And it doesn’t appear to be HadCRUT4 either. Maybe it is BEST but not using the data going back to 1750? But that isn’t likely either, since BEST pretty much matches the other datasets, and in Mann’s graphic above, which peaks out at above 1°C, none of those hit higher than 0.7°C. What’s up with that?
Now compare that plot above to this portion Dr. Mann’s SciAm plot, noting the recent period of surface temperature and the 1°C reference line which I extended from the Y axis:
I’m reminded of Dr. Mann’s claims about climate skeptics in this video: http://www.linktv.org/video/9382/inside-the-climate-wars-a-conversation-with-michael-mann
At 4:20 in the video, Dr. Mann claims that US climate skeptics are part of the “greatest disinformation campaign ever run”. If his position is so strong and pure, why then do we see silly things like this graph given with an elevated ending of global surface temperature (in contrast to 5 other datasets) and not a single data source citation given?
UPDATE: Mark B writes in comments:
Looking at the SciAm graphic (see zoom at right), something didn’t seem right, especially since there doesn’t seem to be any citation given for what the temperature dataset used was. And oddly, the graphic shows Mann’s little white line peaking significantly warmer that the 1998 super El Niño, and showing the current temperature equal to 1998, which doesn’t make any sense.
Explanation of graph including links to source code and data were given here: http://www.scientificamerican.com/article/mann-why-global-warming-will-cross-a-dangerous-threshold-in-2036/
REPLY: Yes, I’ve seen that, but there is a discrepancy, the label on the image is “Historical Mean Annual Temperature” (white)
In http://www.scientificamerican.com/article/mann-why-global-warming-will-cross-a-dangerous-threshold-in-2036/ it is written:
Historical Simulations. The model was driven with estimated annual natural and anthropogenic forcing over the years A.D. 850 to 2012. Greenhouse radiative forcing was calculated using the approximation (ref. 8) FGHG = 5.35log(CO2e/280), where 280 parts per million (ppm) is the preindustrial CO2 level and CO2e is the “equivalent” anthropogenic CO2. We used the CO2 data from ref. 9, scaled to give CO2e values 20 percent larger than CO2 alone (for example, in 2009 CO2 was 380 ppm whereas CO2e was estimated at 455 ppm). Northern Hemisphere anthropogenic tropospheric aerosol forcing was not available for ref. 9 so was taken instead from ref. 2, with an increase in amplitude by 5 percent to accommodate a slightly larger indirect effect than in ref. 2, and a linear extrapolation of the original series (which ends in 1999) to extend though 2012.
“Historical Mean Annual Temperature” is NOT the same as “Historical Simulations” It looks to me like a bait and switch.
UPDATE2: Note the lead in text says “Global temperature rise…”
But in comments, Willis and Bill Illis have worked out that the white line represents only half the planet, the Northern Hemisphere. The white line is HadCRUT NH value, not global.
Obviously we can’t take such statements as the lead in text saying “global” at face value. Imagine if a climate skeptic made a graph like this. We’d be excoriated.
What needs to be done is to create a graph that shows what this would have looked like had Mann not cherry picked the NH and presented it on a graph with the text “Global temperature rise…”.
==============================================================
Mann’s ‘False Hope’ is false hype
By Christopher Monckton of Brenchley
The legendary Dr Walter Lewin, Professor of Physics at MIT, used to introduce his undergraduate courses by saying that every result in physics depended ultimately on measurement; that mass, distance, and time are its three fundamental physical units that every observation in these and all of their derivative units is subject to measurement uncertainty; and that every result in physics, if only for this reason, is to some degree uncertain.
Contrast this instinctual humility of the true physicist with the unbecoming and, on the evidence to date, unjustifiable self-assurance of the surprisingly small band of enthusiasts who have sought to tell us there is a “climate crisis”’. Not the least among these is Michael Mann, perpetrator of the Hokey-Stick graph that wrought the faux abolition of the medieval warm period.
In logic, every declarative statement is assigned a truth-value: 1 (or, in computer programs, –1) for true, 0 for false. Let us determine the truth-values of various assertions made by Mann, in a recent article entitled False Hope, published in the propaganda-sheet Scientific American.
Mann’s maunderings and meanderings will be in bold face, followed by what science actually says in Roman face, and the verdict: Truth-value 1, or truth-value 0?
Mann: “Global warming continues unabated.”
Science: Starting in Orwell’s Year (1984), and taking the mean of the five standard global temperature datasets since then, the rate of warming has changed as follows:
1979-1990 Aug 140 months +0.080 Cº/decade.
1979-2002 Apr 280 months +0.153 Cº/decade.
1979-2013 Dec 420 months +0.145 Cº/decade.
The slowdown in the global warming rate has arisen from the long pause, now 13 years 2 months in length on the mean of all five datasets (assuming that HadCRUT4, which is yet to report, shows a result similar to the drop in global temperatures reported by the other four datasets).
Verdict: Truth-value 0. Mann’s statement that global warming “continues unabated is false”, since the warming rate is declining.
Mann: “… during the past decade there was a slowing in the rate at which the earth’s average surface temperature had been increasing. The event is commonly referred to as “the pause,” but that is a misnomer: temperatures still rose, just not as fast as during the prior decade.”
Science: During the decade February 2005 to January 2014, on the mean of all five datasets, there was a warming of 0.01 Cº, statistically indistinguishable from zero.
Truth-value 0: Temperatures did not rise in any statistically significant sense, and the increase was within the measurement uncertainty in the datasets, so that we do not know there was any global warming at all over the decade. Here, Walter Lewin’s insistence on the importance of measurement uncertainty is well demonstrated.
Mann: “In response to the data, the IPCC in its September 2013 report lowered one aspect of its prediction for future warming.”
Science: In 2013 the IPCC reduced the lower bound of its 2007 equilibrium climate-sensitivity interval from 2 Cº to 1.5 Cº warming per CO2 doubling, the value that had prevailed in all previous Assessment Reports. It also reduced the entire interval of near-term projected warming from [0.4, 1.0] Cº to [0.3, 0.7] Cº. Furthermore, it abandoned its previous attempts at providing a central estimate of climate sensitivity.
Verdict: Truth value 0. The IPCC did not lower only “one aspect of its prediction for future warming” but several key aspects, abandoning the central prediction altogether.
Mann: If the world keeps burning fossil fuels at the current rate, it will cross a threshold into environmental ruin by 2036. The “faux pause” could buy the planet a few extra years beyond that date to reduce greenhouse gas emissions and avoid the crossover–but only a few.
Science: Mann is asserting that on the basis of some “calculations” he says he has done, the world will face “environmental ruin” by 2036 or not long thereafter. However, Mann has failed to admit any uncertainty in his “calculations” and consequently in his predictions.
Verdict: Truth-value 0. Given the ever-growing discrepancy between prediction and observation in the models, and Mann’s own disastrous record in erroneously abolishing the medieval warm period by questionable statistical prestidigitation, the uncertainty in his predictions is very large, and a true scientist would have said so.
Mann: “The dramatic nature of global warming captured world attention in 2001, when the IPCC published a graph that my co-authors and I devised, which became known as the ‘hockey stick’. The shaft of the stick, horizontal and sloping gently downward from left to right, indicated only modest changes in Northern Hemisphere temperature for almost 1,000 years–as far back as our data went.”
Science: The Hokey-Stick graph falsely eradicated both the medieval warm period and the little ice age. At co2science.org, Dr. Craig Idso maintains a database of more than 1000 papers demonstrating by measurement (rather than modeling) that the medieval warm period was real, was near-global, and was at least as warm as the present just about everywhere. McIntyre & McKitrick showed the graph to be erroneous, based on multiple failures of good statistical practice. The medieval warm period and the little ice age are well attested in archaeology, history, architecture, and art. It was the blatant nonsense of the Hokey Stick that awoke many to the fact that a small academic clique was peddling unsound politics, not sound science.
Verdict: Truth value 0. Once again, Mann fails to refer to the uncertainties in his reconstructions, and to the many independent studies that have found his methods false and his conclusions erroneous. Here, he takes a self-congratulatory, nakedly partisan stance that is as far from representing true science as it is possible to go.
Mann: “The upturned blade of the stick, at the right, indicated an abrupt and unprecedented rise since the mid-1800s.”
Science: The graph, by confining the analysis to the northern hemisphere, overstated 20th-century global warming by half. Mann says the rise in global temperatures, shown on the graph as 1.1 Cº over the 20th century, is “unprecedented”. However, the Central England Temperature Record, the world’s oldest, showed a rise of 0.9 Cº in the century from 1663 to 1762, almost entirely preceding the industrial revolution, compared with an observed rate of just 0.7 Cº over the 20th century. The CETR is a good proxy for global temperature change. In the 120 years to December 2013 it showed a warming rate within 0.01 Cº of the warming rate taken as the mean of the three global terrestrial datasets.
Verdict: Truth value 0. The warming of the 20th century was less than the warming for the late 17th to the late 18th centuries.
Mann: “The graph became a lightning rod in the climate change debate, and I, as a result, reluctantly became a public figure.”
Science: For “lightning-rod” read “laughing-stock”. For “reluctantly” read “enthusiastically”. For “public figure” read “vain and pompous charlatan who put the ‘Ass’ in ‘Assessment Report’”.
Verdict: Pass the sick-bucket, Alice.
Mann: “In its September 2013 report, the IPCC extended the stick back in time, concluding that the recent warming was likely unprecedented for at least 1400 years.”
Science: The IPCC is here at odds with the published scientific literature. In my expert review of the pre-final draft of the Fifth Assessment Report, I sent the IPCC a list of 450 papers in the reviewed literature that demonstrated the reality of the warm period. The IPCC studiously ignored it. Almost all of the 450 papers are unreferenced in the IPCC’s allegedly comprehensive review of the literature. I conducted a separate test using the IPCC’s own methods, by taking a reconstruction of sea-level change over the past 1000 years, from Grinsted et al. (2009), and comparing it with the schematic in the IPCC’s 1990 First Assessment Report showing the existence and prominence of both the medieval warm period and the little ice age. The two graphs are remarkably similar, indicating the possibility that the sea-level rise in the Middle Ages was caused by the warmer weather then, and that the fall in the Little Ice Age was caused by cooler weather. The sea-level reconstruction conspicuously does not follow a Hokey-Stick shape.
Verdict: Truth value 0. The IPCC has misrepresented the literature on this as on other aspects of climate science. There are of course uncertainties in any 1000-year reconstruction, but if Grinsted et al. have it right then perhaps Mann would care to explain how it was that sea level rose and fell by as much as 8 inches either side of today’s rather average value if there was no global warming or cooling to cause the change?
Mann: “Equilibrium climate sensitivity is shorthand for the amount of warming expected, given a particular fossil-fuel emissions scenario.”
Science: Equilibrium climate sensitivity is a measure of the global warming to be expected in 1000-3000 years’ time in response to a doubling of CO2 concentration, regardless of how that doubling came about. It has nothing to do with fossil-fuel emissions scenarios.
Truth value: 0. Mann may well be genuinely ignorant here (as elsewhere).
Mann: “Because the nature of these feedback factors is uncertain, the IPCC provides a range for ECS, rather than a single number. In the September report … the IPCC had lowered the bottom end of the range. … The IPCC based the lowered bound on one narrow line of evidence: the slowing of surface warming during the past decade – yes, the faux pause.”
Science: For well over a decade there has been no global warming at all. The pause is not faux, it is real, as Railroad Engineer Pachauri, the IPCC’s joke choice for climate-science chairman, has publicly admitted. And the absence of any global warming for up to a quarter of a century is not “one narrow line of evidence”: it is the heart of the entire debate. The warming that was predicted has not happened.
Verdict: Truth value 0. Mann is here at odds with the IPCC, which – for once – paid heed to the wisdom of its expert reviewers and explicitly abandoned the models, such as that of Mann, which have been consistent only in their relentless exaggeration of the global warming rate.
Mann: “Many climate scientists – myself included – think that a single decade is too brief to accurately measure global warming and that the IPCC was unduly influenced by this one, short-term number.”
Science: Overlooking the split infinitive, the IPCC was not “unduly influenced”: it was, at last, taking more account of evidence from the real world than of fictitious predictions from the vast but inept computer models that were the foundation of the climate scare. Nor was the IPCC depending upon “one short-term number”.
James Hansen of NASA projected 0.5 C°/decade global warming as his “business-as-usual” case in testimony before Congress in 1988. The IPCC’s 1990 First Assessment Report took Hansen’s 0.5 C°/decade as its upper bound. It projected 0.35 C°/decade as its mid-range estimate, and 0.3 C°/decade as its best estimate.
The pre-final draft of the 2013 Fifth Assessment Report projected 0.23 C°/decade as its mid-range estimate, but the published version reduced this value to just 0.13 C°/decade – little more than a quarter of Hansen’s original estimate of a quarter of a century previously.
Observed outturn has been 0.08 Cº/decade since 1901, 0.12 C°/decade since 1950, 0.14 C°/decade since 1990, and zero since the late 1990s.
Three-quarters of the “climate crisis” predicted just 24 years ago has not come to pass. The Fifth Assessment Report bases its near-term projections on a start-date of 2005. The visible divergence of the predicted and observed trends since then is remarkable.
It is still more remarkable how seldom in the scientific journals the growing discrepancy between prediction and observation is presented or discussed.
Verdict: Truth value 0. Step by inexorable step, the IPCC is being driven to abandon one extremist prediction after another, as real-world observation continues to fall a very long way short of what it had been predicting.
Mann: “The accumulated effect of volcanic eruptions during the past decade, including the Icelandic volcano with the impossible name, Eyjafjallajökull, may have had a greater cooling effect on the earth’s surface than has been accounted for in most climate model simulations. There was also a slight but measurable decrease in the sun’s output that was not taken into account in the IPCC’s simulations.”
Science: So the models failed to make proper allowance for, still less to predict, what actually happened in the real world.
Verdict: Truth value 0. Eyjafjallajökull caused much disruption, delaying me in the United States for a week (it’s an ill wind …), but it was a comparatively minor volcanic eruption whose signature in the temperature record cannot be readily distinguished from the la Niña cooling following the el Niño at the beginning of 2010. The discrepancy between models’ predictions and observed reality can no longer be as plausibly dismissed as this, and the IPCC knows it.
Mann: “In the latter half of the decade, La Niña conditions persisted in the eastern and central tropical Pacific, keeping global surface temperatures about 0.1 degree C colder than average …”
Science: There were La Niña (cooling) events in 1979, 1983, 1985, 1989, 1993, 1999, 2004, and 2008 – the only la Niña in the second half of the noughties. There were, however, two el Niño (warming) events: in 2007 and 2010.
Verdict: Truth value 0. There is very little basis in the observed record for what Mann says. He is looking for a pretext – any pretext – rather than facing the fact that the models have been programmed to exaggerate future global warming.
Mann: “Finally, one recent study suggests that incomplete sampling of Arctic temperatures led to underestimation of how much the globe actually warmed.”
Science: And that “study” has been debunked. The numerous attempts by meteorological agencies around the world to depress temperatures in the early 20th century to make the centennial warming rate seem larger than it is have far outweighed any failure to measure temperature change in one tiny region of the planet.
Verdict: Truth value 0. Increasingly, as the science collapses, the likes of Mann will resort in desperation to single studies, usually written by one or another of the remarkably small clique of bad scientists who have been driving this silly scare. Meanwhile, the vrai pause continues. As CO2 concentrations increase, the Pause will not be likely to continue indefinitely. But it is now clear that the rate at which the world will warm will be considerably less than the usual suspects have predicted.
Mann: “When all the forms of evidence are combined, they point to a most likely value for ECS that is close to three degrees C.”
Science: The IPCC has now become explicit about not being explicit about a central estimate of climate sensitivity. Given that two-thirds of Mann’s suggested 3 Cº value depends upon the operation over millennial timescales of temperature feedbacks that Mann himself admits are subject to enormous uncertainties; given that not one of the feedbacks can be directly measured or distinguished by any empirical method either from other feedbacks or from the forcings that triggered it; and given that non-radiative transports are woefully represented in the models, there is no legitimate scientific basis whatsoever for Mann’s conclusion that a 3 Cº climate sensitivity is correct.
Truth value: 0. What Mann is careful not to point out is that the IPCC imagines that only half of the warming from a doubling of CO2 concentration will arise in the next 200 years. The rest will only come through over 1000-3000 years. Now, at current emission rates a doubling of the pre-industrial 280 ppmv CO2 will not occur for 80 years. However, 0.9 Cº warming has already occurred since 1750, leaving only another 0.6 Cº warming to occur by 2280, on the assumption that all of the 0.9 Cº was manmade. And that is if Mann and the models are right.
Mann: “And as it turns out, the climate models the IPCC actually used in its Fifth Assessment Report imply an even higher value of 3.2 degrees C.”
Science: The 2007 Fourth Assessment Report said there would be 3.26 Cº warming at equilibrium after a CO2 doubling. But the 2013 Fifth Report said no such thing. It has fallen commendably silent.
Verdict: Truth value 0. Mann is, yet again, at odds with the IPCC, which has now begun to learn that caution is appropriate in the physical sciences.
Mann: “The IPCC’s lower bound for ECS, in other words, probably does not have much significance for future world climate–and neither does the faux pause.”
Science: This is pure wishful thinking on Mann’s part. In all Assessment Reports except the Fourth, the IPCC chose 1.5 Cº as its lower bound for equilibrium climate sensitivity to doubled CO2 concentration. In the Fourth it flirted briefly with 2 Cº, but abandoned that value when faced with the real-world evidence that Mann sneeringly dismisses as “the faux pause”.
Verdict: Truth value 0. Calling the vrai pause “the faux pause” is a faux pas.
Mann: “What would it mean if the actual equilibrium climate sensitivity were half a degree lower than previously thought? Would it change the risks presented by business-as-usual fossil-fuel burning? How quickly would the earth cross the critical threshold?”
Science: But what is the “critical threshold”? Mann fails to define it. Is there some value for global mean surface temperature that is the best of all temperatures in the best of all possible worlds? If so, Mann’s hypothesis can only be tested if he enlightens us on what that ideal temperature is. He does not do so.
Verdict: Truth value 0. In the absence of a clear and scientifically justified statement of an ideal temperature, plus a further justified statement that a given departure from that ideal temperature would be dangerous, there is no case for a “critical threshold”. Furthermore, there is at present little empirical basis for a global warming of more than 1 Cº over the coming century.
Mann: “Most scientists concur that two degrees C of warming above the temperature during preindustrial time would harm all sectors of civilization–food, water, health, land, national security, energy and economic prosperity.”
Science: No survey of scientists to determine whether they “concur” as to the 2 Cº above pre-industrial temperature that Mann considers on no evidence to be the “critical threshold” has been conducted. Even if such a survey had been conducted – and preferably conducted by someone less accident-prone than the absurd Cook and Nutticelli – that would tell us nothing about the scientific desirability or undesirability of such a “threshold”: for science is not done by consensus, though totalitarian politics is. And it was totalitarian politicians, not scientists, who determined the 2 Cº threshold, on no evidence, at one of the interminable paid holidays in exotic locations known as UN annual climate conferences.
Verdict: Truth value 0. There is no scientific basis for the 2 Cº threshold, and Mann does not really attempt to offer one.
Mann: “Although climate models have critics, they reflect our best ability to describe how the climate system works, based on physics, chemistry and biology.”
Science: Mann’s own model that contrived the Hokey-Stick graph shows what happens when a model is constructed with insufficient attention to considerations that might point against the modeler’s personal preconceptions. The model used a highly selective subset of the source data; it excluded hundreds of papers demonstrating the inconvenient truth that the medieval warm period existed; it gave almost 400 times as much weighting to datasets showing the medieval warm period as it did to datasets that did not show it; and the algorithm that drew the graph would draw Hokey Sticks even if random red noise rather than the real data were used. The problem with any model of a sufficiently complex object is that there are too many tunable parameters, so that the modeler can – perhaps unconsciously – predetermine the output. To make matters worse, intercomparison tends to institutionalize errors throughout all the models. Besides, since the climate behaves as a chaotic object, modeling its evolution beyond around ten days ahead is not possible. We can say (and without using a model) that if we add plant-food to the air it will be warmer than if we had not done so; but (with or without a model) we cannot say with any reliability how much warming is to be expected.
Verdict: Truth value 0. Models have their uses, but as predictors of long-term temperature trends they are, for well-understood reasons, valueless.
Mann: “And they [the models] have a proved track record: for example, the actual warming in recent years was accurately predicted by the models decades ago.”
Science: Here is Hansen’s 1988 prediction of how much global warming should have occurred since then, according to his “Giss Model E”.
The trend shown by Hansen is +0.5 Cº per decade. The outturn since 1988, however, was just 0.15 Cº per decade, less than one-third of what Hansen described as his “business-as-usual” case. Models’ projections have been consistently exaggerated:
Verdict: Truth value 0. The models have consistently and considerably exaggerated the warming of recent decades. The next graph shows a series of central projections, compared with the observed outturn to date, extrapolated to 2050. This is not a picture of successful climate prediction. It is on the basis of these failed predictions that almost the entire case for alarm about the climate is unsoundly founded.
Mann: “I ran the model again and again, for ECS values ranging from the IPCC’s lower bound (1.5 Cº) to its upper bound (4.5 Cº). The curves for an equilibrium climate sensitivity of 2.5 Cº and 3 Cº fit the instrument readings most closely. The curves for a substantially lower (1.5 Cº) and higher (4.5 Cº) sensitivity did not fit the recent instrumental record at all, reinforcing the notion that they are not realistic.”
Science: Legates et al. (2013) established that only 0.3% of abstracts of 11,944 climate science papers published in the 21 years 1991-2011 explicitly stated that we are responsible for more than half of the 0.69 Cº global warming of recent decades. Suppose that 0.33 Cº was our contribution to global warming since 1950, that CO2 concentration in that year was 305 ppmv and is now 398 ppmv. Then the radiative forcing from CO2 that contributed to that warming was 5.35 ln(398/305) = 1.42 Watts per square meter. Assuming that the IPCC’s central estimate of 713 ppmv CO2 by 2100 is accurate, the CO2 forcing from now to 2100 will be 5.35 ln(713/398), or 3.12 W m–2. On the assumption that the ratio of CO2 forcing to that from other greenhouse gases will remain broadly constant, and that temperature feedbacks will have exercised 44/31 of the multiplying effect seen to date, the manmade warming to be expected by 2100 on the basis of the 0.33 Cº warming since 1950 will be 3.12/1.42 x 0.33 x 44/31 = 1 Cº. Broadly speaking, the IPCC expects this century’s warming to be equivalent to that from a doubling of CO2 concentration. In that event, 1 Cº is the warming we should expect from a CO2 doubling, and the only sense in which the 1.5 Cº lower bound of the IPCC’s interval of climate-sensitivity estimates is “unrealistic” is that it is probably somewhat too high.
Verdict: Truth value 0. Here, as elsewhere, Mann appears unaware of the actual evolution of global temperatures during the post-1950 era when we might in theory have exercised some warming influence. There has been less warming than They thought, and – on the basis of the scientific consensus established by Legates et al. – less of the observe warming is anthropogenic than They thought they thought.
Mann: “To my wonder, I found that for an ECS of 3 Cº, our planet would cross the dangerous warming threshold of two degrees C in 2036, only 22 years from now. When I considered the lower ECS value of 2.5 Cº, the world would cross the threshold in 2046, just 10 years later.”
Science: Mann here perpetrates one of the fundamental errors of the climate-extremists. He assumes that the prediction of a climate model is subject to so little uncertainty that it constitutes a fact. This statement is one of a series by true-believers saying we have only x years to Save The Planet by shutting down the West. Ex-Prince Chazza has done it. Al Gore has done it. The UN did it big-time by saying in 2005 that there would be 50 million climate refugees by 2010. There weren’t.
Verdict: Truth value 0. Extreme warming that has been predicted does not become a fact unless it comes to pass. If you want my prediction, it won’t. And that’s a fact.
Mann: “So even if we accept a lower equilibrium climate sensitivity value, it hardly signals the end of global warming or even a pause. Instead it simply buys us a little bit of time – potentially valuable time – to prevent our planet from crossing the threshold.”
Science: No one is suggesting that the Pause will continue indefinitely. Theory as well as observation suggests otherwise. However, a Pause that has not occurred cannot “buy us a little bit of time”. Mann’s mention of “buying us a little bit of time” is, therefore, an admission that the Pause is real, as all of the temperature datasets show.
Verdict: Truth value 0. A low enough climate sensitivity will allow temperatures to remain stable for decades at a time, during periods when natural factors tending towards global cooling temporarily overwhelm the warming that would otherwise occur.
Mann: “These findings have implications for what we all must do to prevent disaster.”
Science: Warming of 3 Cº would not be a “disaster”. Even the bed-wetting Stern Review of 2006 concluded that warming of 3 Cº over the 21st century would cost as little as 0-3% of global GDP. But at present we are heading for more like 1 Cº. And even the IPCC has concluded that less than 2 Cº warming compared with 1750, which works out at 1.1 Cº compared with today, will be net-beneficial.
Verdict: Truth value 0. There is no rational basis for any suggestion that our adding CO2 to the atmosphere at the predicted rate, reaching 713 ppmv by 2100, will be anything other than beneficial.
Mann: “If we are to limit global warming to below two degrees C forever, we need to keep CO2 concentrations far below twice preindustrial levels, closer to 450 ppm. Ironically, if the world burns significantly less coal, that would lessen CO2 emissions but also reduce aerosols in the atmosphere that block the sun (such as sulfate particulates), so we would have to limit CO2 to below roughly 405 ppm. We are well on our way to surpassing these limits.”
Science: What we are concerned with is not CO2 simpliciter, but CO2-equivalent. CO2 itself contributes only 70% of the anthropogenic enhancement of the greenhouse effect. The (admittedly arbitrary) target of 450 ppmv CO2-equivalent is thus a target of only 315 ppmv CO2 – the concentration that prevailed in 1958. Mann’s suggested target of 405 ppmv CO2e would represent just 284 ppmv CO2. And that would fling us back to the pre-industrial CO2 concentration.
Verdict: Truth value 0. We are not “well on our way to surpassing these limits”: we passed them as soon as the industrial revolution began. The current CO2-equivalent concentration of 398 ppmv already exceeds the pre-industrial 284 ppmv by 40%, yet the world has warmed by only 0.9Cº since then, our contribution to that warming may well be 0.33 Cº or less.
Mann: “Some climate scientists, including James E. Hansen, former head of the NASA Goddard Institute for Space Studies, say we must also consider slower feedbacks such as changes in the continental ice sheets.”
Science: The IPCC already takes changes in ice-sheets into account. It says that in the absence of “dynamical ice flow” that cannot happen, the Greenland ice sheet would not disappear “for millennia”. And there is no prospect of losing ice from the vast ice sheet of East Antarctica, which is at too high an altitude or latitude to melt. Even the West Antarctic Ice Sheet, which has lost some ice, is proving more robust than the usual suspects had thought. Sea level, according to the GRACE gravitational anomaly satellites, has been falling (Peltier et al., 2009). During the eight years of ENVISAT’s operation, from 2004-2012, sea level rose at a scary 1.3 inches per century.
Verdict: Truth value 0. There is no reason to suppose the major ice sheets will disintegrate on timescales of less than millennia.
Mann: “Hansen and others maintain we need to get back down to the lower level of CO2 that existed during the mid-20th century–about 350 ppm.”
Science: 350 ppmv is, again, CO2-equivalent. That implies 245 ppmv, a value well below the pre-industrial 280 ppmv. At 180 ppmv, plants and trees become dangerously starved of CO2. Flinging CO2 concentration back to that value would reduce CO2 fertilization and hence crop yields drastically, and would do major damage to the rain-forests.
Mann: “In the Arctic, loss of sea ice and thawing permafrost are wreaking havoc on indigenous peoples and ecosystems.”
Science: The Arctic has not lost as much sea ice as had been thought. In the 1920s and 1930s there was probably less sea ice in the Arctic than there is today. The decline in sea ice is small in proportion to the seasonal variability, as the graph from the University of Illinois shows. And the part of the satellite record that is usually cited began in 1979. An earlier record, starting in 1973, showed a rapid growth in sea ice until it reached its peak extent in 1970. Indigenous peoples, like the polar bears, prefer warmer to colder weather. And almost all ecosystems also prefer warmer to colder weather.
Verdict: Truth value 0. The decline in sea ice in the Arctic is far more of a benefit than a loss.
Mann: “In low-lying island nations, land and freshwater are disappearing because of rising sea levels and erosion.”
Science: On the contrary, detailed studies show not only that low-lying island nations are not sinking beneath the waves, but that their territory is in many cases expanding. The reason is that corals grow to meet the light. As sea level rises, the corals grow and there is no net loss of territory. Also, sea level rises less in mid-ocean, where the islands are, than near the continental coasts. And sea level has scarcely been rising anyway. According to Grinsted et al., it was 8 inches higher in the medieval warm period than it is today.
Verdict: Truth value 0. If the world were once again to become as warm as it was in the Middle Ages, perhaps sea level would rise by about 8 inches. And that is all.
Mann: “Let us hope that a lower climate sensitivity of 2.5 degrees C turns out to be correct. If so, it offers cautious optimism. It provides encouragement that we can avert irreparable harm to our planet. That is, if–and only if–we accept the urgency of making a transition away from our reliance on fossil fuels for energy.”
Science: Mann is here suggesting that a climate sensitivity of 3 Cº would be disastrous, but that 2.5 Cº would not. The notion that as little as 0.5 Cº would make all the difference is almost as preposterous as the notion that climate sensitivity will prove to be as high as 2.5 Cº. As we have seen, on the assumption that less than half of the warming since 1950 was manmade, climate sensitivity could be as low as 1 Cº – a value that is increasingly finding support in the peer-reviewed literature.
Verdict: Truth value 0. The central error made by Mann and his ilk lies in their assumption that models’ predictions are as much a fact as observed reality. However, observed climate change has proven far less exciting in reality than the previous predictions of Mann and others had led us to expect. The multiple falsehoods and absurdities in his Scientific American article were made possible only by the sullen suppression by the Press of just how little of what has been predicted is happening in the real climate. In how many legacy news media have you seen the Pause reported at all? But it will not be possible for the mainstream organs of propaganda to conceal from their audiences forever the inconvenient truth that even the most recent, and much reduced, projections of the silly climate models are proving to be egregious exaggerations.
![earth-will-cross-the-climate-danger-threshold-by-2036_large[1]](http://wattsupwiththat.files.wordpress.com/2014/03/earth-will-cross-the-climate-danger-threshold-by-2036_large1.jpg?resize=640%2C423&quality=83)

![land-and-ocean-other-results-1950-large[1]](http://wattsupwiththat.files.wordpress.com/2014/03/land-and-ocean-other-results-1950-large1.png?resize=640%2C492&quality=75)

Cramer says: I said it averages to zero.
Henry says
if that were the case, it (=a linear regression) should be a straight line (between 1960-1990)
So, clearly, WFT is using original data. You cannot “correct” data. The data is the data. If you “correct” the data you have something else.
Otherwise, I am glad that we both agree that we have started to cool globally.
That helps.
Henry said
if that were the case, it (=a linear regression) should be a straight line (between 1960-1990)
Henry says: this should read
if that were the case, it (=a linear regression) should be a straight line with zero slope i.e. no incline or decline (between 1960-1990)
HenryP says: “if that were the case, it (=a linear regression) should be a straight line with zero slope i.e. no incline or decline (between 1960-1990)”
That is not true. What is the average of this data series: -2,-1,0,1,2. This is a straight line with a slope=1. It averages to zero. Add the five points and divide by five. Why not just take an average of the 360 data points (Jan 1961 to Dec 1990) in HadCRUT4. Add the 360 points and divide by 360. It is equal to -0.00051. Since it’s not exactly zero (due to rounding error), this is why I wanted you to calculate the “trailing moving average,” because you can then see this cross zero at Dec 1990. Surely, you know how to use a spreadsheet, if you are calculating sine waves. Do you? Do you know what a moving average is? Maybe I assumed you did, when you might not.
HenryP says: “So, clearly, WFT is using original data.”
I never claimed WFT was NOT using the original data. It’s the same data from the Hadley Center. The original data is calculated with a baseline of 1961-1990. WFT is using this same data.
HenryP says: “You cannot “correct” data.”
I never said anything about “correcting” data. “Correcting” implies the original data is wrong. Changing the baseline of temperature anomalies is not “correcting” it. You seem to believe these temperature anomalies are ABSOLUTE temperatures. They are NOT. They have to be calculated relative to a baseline. In the graph used above to compare different data, they used a baseline of Jan 1951 to Dec 1980 (although that looks to be a typo — but they did specify it because it is important to do so). I could adjust the data to a baseline of 1945 to 1965. It does not make the data wrong if that is specified. I weigh 170 lbs. If I start saying that I weigh 77.11 kg, does that mean I am not giving my correct weight?
“Otherwise, I am glad that we both agree that we have started to cool globally.”
There you go making assumptions again. I no more believe in your Gleissberg sine wave theory, than I believe in Mann’s 2036 prediction. I’m a skeptic. Those that believe they know the truth (warming or cooling) are not skeptics. When the heat capacity of the oceans are 1000 times the heat capacity of the atmosphere, that’s where the answer lies — and it has not been figured-out yet. No warming from the mid-1940s to the mid-1970s (when we were emitting a lot of CO2), warming from the mid-1970s to late-1990s, then no warming again. These periods do correspond with ENSO activity (el nino vs la nina predominance). See heat capacity here:
http://wattsupwiththat.com/2011/04/06/energy-content-the-heat-is-on-atmosphere-vs-ocean/
Cramer says
I no more believe in your Gleissberg sine wave theory, than I believe in Mann’s 2036 prediction.
\
Henry says
it is not theory
it is fact
my own results merely confirm it
HenryP says: “it is not theory it is fact”
It is a fact that global temperatures will decrease by 0.6 deg C by 2038?
Cramer says
(re-stating Henry’s claim)
It is a fact that global temperatures will decrease by 0.6 deg C by 2038?
Henry says
According to my own data set, we are already down by ca. -0.2 since 2000.
The other 4 data major global sets make this about -0.1 since 2002, on average
Now, as I said before, the major cooling actually is happening from the top latitudes downward,
so between [60-70] it is cooling more
For example, my own results show that it has been cooling significantly in Alaska, at a rate of -0.055 per annum since 1998 (average of 10 weather stations in Alaska)
http://oi40.tinypic.com/2ql5zq8.jpg
That is almost 1 degree C since 1998.
So,
all I am saying that by 2040 temperatures we will be back, more or less, to where we were in 1950
It seems to me that around 1950, it was -0.1?
http://www.woodfortrees.org/plot/hadcrut4gl/from:1927/to:2015/plot/hadcrut4gl/from:2002/to:2015/trend/plot/hadcrut3gl/from:1987/to:2015/plot/hadcrut3gl/from:2002/to:2015/trend/plot/rss/from:1987/to:2015/plot/rss/from:2002/to:2015/trend/plot/hadsst2gl/from:1987/to:2015/plot/hadsst2gl/from:2002/to:2015/trend/plot/hadcrut4gl/from:1960/to:1990/trend/plot/hadcrut3gl/from:1987/to:2002/trend/plot/hadsst2gl/from:1987/to:2002/trend/plot/rss/from:1987/to:2002/trend/plot/hadcrut4gl/from:1930/to:1960/trend
In 2000 we were +0.5 or +0.6
So, what is not to understand of my claim for a drop of 0.6 by 2038?
HenryP says: “So, what is not to understand of my claim for a drop of 0.6 by 2038?”
Now your calling “it” a claim. What were you calling a fact? [And don’t tell me “Gleissberg cycles” because I explicitly said “your Gleissberg sine wave theory.”]
What’s not to understand? Why you cherry picked Nome and Anchorage Intl Airport as a proxy for upper latitudes? Did you look at the “Alaska” data from 1978 to 1999 (be sure to cherry pick those dates)? The 1978 to 1999 temperatures have a downward trend that about as much as your 1998-2013 data. Should the 1978 to 1999 trend be upward or downward according to your “sine wave theory?”
I also understand other upper latitudes locations such as Greenland and Iceland (not just Nome and Anchorage Intl Airport). Have you ever heard of gridded data. If you are going to cherry pick Alaska, at least use gridded data:
http://berkeleyearth.lbl.gov/auto/Regional/TAVG/Text/alaska-TAVG-Trend.txt
http://www.ncdc.noaa.gov/temp-and-precip/alaska/tmp/mon/0
What’s not to understand? Why you cherry picked the Gleissberg Cycle. Do you even understand that the Gleissberg cycle is simply a cycle of the amplitude modulation of the Schwabe Cycle? It has little effect on solar forcing. Can you provide an estimate in the variation of solar forcing due to the Gleissberg Cycle?
Cramer says
What’s not to understand? Why you cherry picked Nome and Anchorage Intl Airport as a proxy for upper latitudes? Did you look at the “Alaska” data from 1978 to 1999 (be sure to cherry pick those dates)? The 1978 to 1999 temperatures have a downward trend that about as much as your 1998-2013 data.
Henry says
I did no such thing. I only chose to show the linear regression lines of those two stations as otherwise the picture becomes too cluttered. I chose 10 weather stations in Alaska precisely from a source other than Berkeley, namely tutiempo.net.
Of the 10 stations observed, only one station showed an upward trend, namely Barrow, and there I put a question mark at one particular result that does not fit in with the other 9.
Nevertheless, to determine the average trend, I took the 10 slopes (including Barrow) and averaged these, to give me an average slope of -0055 degree C per annum or -0.55 degreeC/decade since 1998.
I chose 1998 as my starting point as most data sets (including my own), seem to agree that earth was at its warmest point in 1998, i.e. when its output was at its maximum.
As far as Greenland and Iceland is concerned: most stations are on the coast, which gives a false impression, due to warmer streams that can carry warmth for decades after cooling has started, (just like Barrow, perhaps).
If you chose to believe Berkeley, and not me, that is your choice.
Note my results for the Anchorage Army base station going back to 1942 (for maxima only)
2nd graph, below the global graph, here
http://blogs.24.com/henryp/2012/10/02/best-sine-wave-fit-for-the-drop-in-global-maximum-temperatures/
hence I know that the cooling in Alaska will continue until 2038.
There is a small variation within the TSI when measured over time, especially in the UV sector.
A small difference in the distribution of light coming from the sun affects the production of the ozone, peroxides and nitrogenous oxides lying at the TOA, and that deflects more radiation to space , when there is more of it. The higher the latitude the more pronounced this effect TOA, apparently, e.g. note the difference between the Anchorage- and the global graph for the speed in warming.
Every place on earth is on its own particular Gleissberg wave (of 88 years)
“hence I know that the cooling in Alaska will continue until 2038.”
Alaska started cooling in 1978. That’s 60 years to 2038 (120 year cycle). You used Anchorage Intl Airport and Nome Airport for your linear trend in your graph. 1978 to 1999 had a steeper negative trend than 1998-2013 (about double for Anchorage):
Anchorage Intl Airport (1978-1999): -0.047 K/yr
Anchorage Intl Airport (1998-2013): -0.024 K/yr
Anchorage Intl Airport (1978-2013): -0.007 K/yr
Nome Airport (1978-1999): -0.081 K/yr
Nome Airport (1998-2013): -0.080 K/yr
Nome Airport (1978-2013): -0.027 K/yr
“Every place on earth is on its own particular Gleissberg wave (of 88 years)”
The actual Gleissberg solar cycle peaked about 1960 (max sunspots during Schwabe Cycle 19).
So I guess every place on earth can have it’s own cycle by having different peaks and different durations. Alaska temperatures peaked in 1978 with 120 year cycle. Some cycles are 88 years, others are 30 years, 40 years, …120 years, etc. Some peaks occur in 1960 with the peak of Gleissburg. Others peak in 1978. Others in 1998.
Cramer says
Anchorage Intl Airport (1978-1999): -0.047 K/yr
Anchorage Intl Airport (1998-2013): -0.024 K/yr
Anchorage Intl Airport (1978-2013): -0.007 K/yr
Henry says
those results do not compare good with my results (from tutiempo.net) for Anchorage airport
0.0114 K/yr 1973-2012
0.0104 K/yr 1980-2012
0.0060 K/yr 1990-2012
-0.1234 K/yr 2000-2012
(see 2nd part, 2nd table – means, marked Anchorage a/p)
http://blogs.24.com/henryp/2013/02/21/henrys-pool-tables-on-global-warmingcooling/
Clearly you can see a declining warming trend over time which turned negative before the new millennium and can be put into a binomial with high correlation,
but hopefully my sine wave is the (most) correct best fit, as otherwise we might fall into an ice age.
Perhaps, if you understand the seriousness of the problem here, you could try and see how I obtained these particular results? I explained this at the beginning of my tables and gave an example, e.g. New York, Kennedy a/p
Note that it is better to look at maxima (first table) as a proxy for energy coming through as here you will find a lot less noise due to error and weather.
Do you think Alaska is a good example? You have reduced daylight (on the edge of the Arctic circle and land of the midnight sun) You can not provide much dairy produce because of lack of pastures, although moose is eaten. You are naturally cold anyway. Fishing great. And I’d love to go on one of those cruises to see the glaciers etc. You are more adapted to cold and alpine conditions than most of America. And close to those pesky volcanoes.
@bushbunny
I took samples of the temperature from everywhere. Alaska is a good example of what is happening globally. At the lower latitudes you get more rain and condensation, which compensates for energy loss: when water condenses it releases energy. However, at the +[40] latitudes you will get droughts as a result of global cooling, and ironically, it might get warmer….
NASA also admits now that antarctic ice is increasing significantly.
http://wattsupwiththat.com/2013/10/22/nasa-announces-new-record-growth-of-antarctic-sea-ice-extent/#more-96133
I am merely re-stating that the global cooling is happening from the top [90] latitudes downward, and you might be confused into believing that earth is still warming, by those whose business depends on the carbon scare.
Cramer says:
April 1, 2014 at 9:57 pm
I have no idea what you mean by “Alaska started cooling in 1978”. Using what metric? 10 year trailing trends? 20 year centered trends? And using which dataset?
Also, was 1978 the point when one of those (or something like it) went below zero? Or was 1978 the point when one of those (or something like it) peaked and started decreasing?
You can see the reason for my lack of comprehension.
w.
HenryP says: “(see 2nd part, 2nd table – means, marked Anchorage a/p)”
Your 2nd table is for 2000-2011 (Last 12 yrs). Your graph is for 1998-2013. I used 1998-2013 (YOU CHOSE THE DATES!)
Your data from tutiempo.net gives -0.01 K/yr for 1978 to 2013 (more of a decline than my -0.007 K/yr).
My data came from:
http://berkeleyearth.lbl.gov/auto/Stations/TAVG/Text/43378-TAVG-Data.txt
http://www.arh.noaa.gov/cliMap/akClimate.php
Your tutiempo.net is missing 2002 and 2005. 2013 is 3.7 deg C vs BEST 4.144 deg C. The other years are close. What data source did you used to fill in the missing data? Or did you fudge it?
HenryP says: “Perhaps, if you understand the seriousness of the problem here, you could try and see how I obtained these particular results? I explained this”
What seriousness??? This is very easy stuff. This is not rocket science HenryP. You have been sounding as you have just learned this recently and do not have any formal education in science, engineering or math. Or you have a Dunning–Kruger bias.
Do some work and update your tables. Do some work and calculate the 1998-2013 trend. You chose those dates, not I. You were lazy when going through the HadCRUT data; and now you are being lazy again.
This is the data from NOAA that I used:
http://www.ncdc.noaa.gov/cdo-web/datasets/ANNUAL/stations/COOP:500280/detail
Cramer says
Your tutiempo.net is missing 2002 and 2005. 2013 is 3.7 deg C vs BEST 4.144 deg C. The other years are close. What data source did you used to fill in the missing data? Or did you fudge it?
Henry says
I don’t fix or fudge results. I explained the sampling technique including on how to fill in missing data before the main tables begin.
My main tables merely showed me that there was serious cooling happening in Alaska evident from the results at both the Anchorage Airport and Anchorage Airforce base since the beginning of the millennium. At both stations it showed that it was warming until around 1998 or so.
I subsequently looked at 8 more stations from Alaska where most data only go back to 1995.
All results together show that it has been cooling significantly in Alaska, at a rate of -0.055 per annum since 1998 (that is the average of 10 weather stations in Alaska)
http://oi40.tinypic.com/2ql5zq8.jpg
That is almost -1.0 degree C since 1998.
The work involved is first year statistics
but
I have not seen anyone doing a trend on the speed of warming against time
Apparently you have to be a genius like me to see that it drops like the curve of thrown object.
Hence the reason I am 100% confident that we will be cooling globally until 2038
HenryP says: “I don’t fix or fudge results.”
HenryP wrote on his blog: “take the average of that particular month’s preceding year and year after”
This is fudging it. When you continue to say things like this, it shows your background. What do you think fudging means?
Your fudging method is not even robust. You should have at least attempted to include month-over-month comparison than only year-over-year comparisons. I just read on this blog that Chicago had the coldest winter on record (Dec-Mar). Comparing Feb-2014 to Jan-2014 can give you more info than simply comparing Feb-2014 to Feb-2013 and Feb-2015 (if Feb-2015 existed). But it still is best to find another temperature record for Chicago if you are missing Feb-2014. That can be adjusted for bias (both level and variation) between data sets.
Your TuTiempo data is missing 2002 and 2005 for all ten of your Alaska data sets. This should have been a red flag for you.
HenryP says: “I subsequently looked at 8 more stations from Alaska where most data only go back to 1995.”
THIS IS NOT TRUE. TuTiempo.net gives more complete data from 1973 onwards. 7 out of ten of your Alaska stations are not missing a single year from 1973-1999. Dawson and Tanana are only missing one month from 1978-1999. Nenana is the only station missing a significant amount of data.
DO THE WORK!
Here are the results using TuTiempo.net data:
Anchorage Intl Airport (1978-1999): -0.045 K/yr
Anchorage Intl Airport (1998-2013): -0.037 K/yr
Anchorage Intl Airport (1978-2013): -0.010 K/yr
Nome Airport (1978-1999): -0.104 K/yr
Nome Airport (1998-2013): -0.065 K/yr
Nome Airport (1978-2013): -0.039 K/yr
Ten AK Stations Avg (1978-1999): -0.025 K/yr
Ten AK Stations Avg (1998-2013): -0.052 K/yr
Ten AK Stations Avg (1978-2013): +0.001 K/yr
Again, both Anchorage and Nome have steeper rates of temperature decline for 1978-1999 than for 1998-2013. This is a simple lesson in cherry picking.
Here is the data:
Anchorage Intl Airport (1978-2013) =
4.6,4.1,2.3,3.9,1.3,3.0,3.8,2.3,3.5,3.6,3.2,2.3,1.8,2.8,2.1,4.2,2.6,3.1,1.7,3.6,3.2,1.6,3.5,3.0,4.0,4.1,3.8,4.2,2.3,2.8,1.6,2.6,3.1,2.8,1.7,3.7
Nome Airport (1978-2013) =
0.1,-1.0,-1.5,-0.5,-1.6,-0.5,-3.6,-2.6,-2.0,-2.2,-2.1,-2.6,-3.1,-1.7,-4.3,-0.9,-3.3,-1.8,-2.7,-1.8,-1.6,-5.2,-1.2,-3.2,-0.8,-1.3,-0.6,-1.2,-3.2,-1.2,-4.0,-3.3,-2.7,-2.6,-4.6,-2.0
Average of Ten AK Stations (1978-2013) =
-0.75,-1.54,-2.28,-0.31,-3.08,-1.93,-2.72,-2.48,-1.96,-1.35,-1.61,-2.22,-2.84,-1.82,-3.02,-0.31,-2.34,-1.58,-3.39,-1.34,-0.87,-3.38,-1.42,-1.75,-0.39,-0.93,-1.09,-0.65,-2.24,-1.41,-2.87,-2.00,-1.33,-1.86,-3.45,-1.54
[Note: be careful of line breaks if you copy this data.]
Here’s the missing data points that I used:
Anchorage Intl Airport (2002,2005) = 4.0, 4.2
Nome Airport (2002,2005) = -0.8, -1.2
HenryP says: “Apparently you have to be a genius like me to see that it drops like the curve of thrown object. Hence the reason I am 100% confident that we will be cooling globally until 2038.”
Yes, this confirms your arrogance (Dunning–Kruger bias). No, you don’t have to be a genius — even a ten year old child can understand and recognize the curve of a thrown object.
Let us deal with the various issues separately, in several posts
First of all re. to my applied correction for missing data
Cramer says
your fudging method is not even robust. You should have at least attempted to include month-over-month comparison than only year-over-year comparisons. I just read on this blog that Chicago had the coldest winter on record (Dec-Mar). Comparing Feb-2014 to Jan-2014 can give you more info than simply comparing Feb-2014 to Feb-2013 and Feb-2015 (if Feb-2015 existed). But it still is best to find another temperature record for Chicago if you are missing Feb-2014. That can be adjusted for bias (both level and variation) between data sets.
Your TuTiempo data is missing 2002 and 2005 for all ten of your Alaska data sets. This should have been a red flag for you.
Henry says
Starting with your last remark, remember that the whole year’s data is given missing but usually only a part of one particular’s month daily data was found missing.
Of the given years the other 11 months data were available, so, I collected those 11 months separately. In the month where I found that there were missing daily data, I looked at how many daily data I had. If there were more than 15 (days of daily data), I would take the average for the month. If there were less than 15 days of daily data and let us say that was for November 2002, I would look at November 2001 and November 2003, and take the average of those two months as the figure for November 2002. I would then proceed to calculate the average yearly temperature for 2002 from the 12 months of 2002.
Now, if I understand you correctly, what I hear you say is that, in the above example, I should rather have taken October 2002 and December 2002 and calculate the average of that, to fill in as the results for November 2002.
I can honestly say that I did not think of doing that as the variations within months can be quite dramatic, especially in the arctic. This course of action is debatable. I am not saying your method is better than mine or that mine is better. I honestly don’t think it will make such a big difference as the true amount of “fudging” error is in any case diluted by 12, seeing that we had the true exact daily data of the remaining 11 months of the year.
I hope you agree with me on that?
Cramer says
DO THE WORK!
henry says
Clearly, I don’t have to do any more work once I know what is happening.
But I invite you to repeat my results.
Here are my results for Anchorage airport (again)
0.0114 K/yr 1973-2012 (that means 1973 to 2012 – not including 2012)
0.0104 K/yr 1980-2012
0.0060 K/yr 1990-2012
-0.1234 K/yr 2000-2012
Here are my results for the Elmendorff Airforce base in Anchorage:
0.0245 K/yr 1973-2012
0.0193 K/yr 1980-2012
0.0220 K/yr 1990-2012
-0.1785 K/yr 2000-2012
The last figures 2000-2012 in Alaska seemed a bit steep to me, i.e. the cooling I mean, so I looked at 10 stations in Alaska, getting an average result of
-.0.055 K/yr 1998-2013
Now, going off from the means,
here are my results for maxima (average for 47 weather stations, spread equally Nh and Sh + 70/30 at sea / in-land)
First table, bottom
0.036 K/yr 1974 -2012
0.028 K/yr 1980-2012
0.015 K/yr 1990-2012
-0.013 K/yr 2000-2012
Now,my dear Cramer, anyone who knows a little bit of stats,would see a clear pattern emerging here,
I did a linear fit, setting the speed of warming out against time, on those 4 results for the drop in the speed of global maximum temps,
ended up with y=0.0018x -0.0314, with r2=0.96
I was at least 95% sure (max) temperatures were falling.
On same maxima data, a polynomial fit, of 2nd order, i.e. parabolic, gave me
y= -0.000049×2 + 0.004267x – 0.056745
r2=0.995
That is very high, showing a natural relationship, like the trajectory of somebody throwing a ball…
projection on the above parabolic fit backward, ( 5 years) showed a curve:
happening around 40 years ago. You always have to be careful with forward and backward projection, but you can do so with such high correlation (0.995)
ergo: the final curve must be a sine wave fit, with another curve happening, somewhere on the bottom…
http://blogs.24.com/henryp/2012/10/02/best-sine-wave-fit-for-the-drop-in-global-maximum-temperatures/
Now tell me,
do you honestly still think I am stupid?
Dunning–Kruger bias, is what?
Perhaps I should tell you why I started to avoid anglo saxon stations (BEST etc)
I wrote a report on that as well
but you can query me on that if you are interested..
(clearly you work for one of those institutes that I don’t trust)
HenryP says: “Now, if I understand you correctly, what I hear you say is that, in the above example, I should rather have taken October 2002 and December 2002.
No, you did not understand me correctly. I said, “you should have at least attempted to INCLUDE month-over-month comparison THAN ONLY year-over-year comparisons.” It was not an either-or; it was both. You should use ALL data available. First, you should be using other data sets to check for inconsistencies and to fill in missing data. There are USHCN, GHCN, GISS, and BEST to name a few. Tutiempo is missing 2002 and 2005 on all 10 of your AK stations. This is a problem with Tutiempo, not the weather station. Did you every compare tutiempo to BEST raw station data? The monthly data is close to exact. BEST even has a QC column. Notice that failed=1 for Nov 2013 for Anchorage data (last data point).
HenryP says: “Now,my dear Cramer, anyone who knows a little bit of stats,would see a clear pattern emerging here,”
You have introduced severe autocorrelation into your time series. “Anyone who knows a little bit of stats” would notice this pattern:
12/38 = 32%
12/32 = 38%
12/22 = 55%
12/12 = 100%
Your last 12 data points from 2000-2011 (“not including 2012”?) represent a larger and larger proportion of your data. Try your regression analysis with this data:
1, 0.05
2, 0.10
3, 0.15
4, 0.20
5, 0.18
This data is clearly linear from 1 to 4 with a slope of 0.05. The last point then drops to 0.18 with a slope of -0.02.
slope(1 to 5) = 0.036
slope(2 to 5) = 0.029
slope(3 to 5) = 0.015
slope(4 to 5) = -0.02
This is very close to your results (maxima for 47 stations).
Your analysis is erroneous. R^2 is meaningless for your regression. Why not take your analysis to someone with expertise in statistics? They will tell you the same thing that I have. Know how do do a regression in Excel does not mean you have much knowledge in statistics.
HenryP says: “Perhaps I should tell you why I started to avoid anglo saxon stations (BEST etc)”
That doesn’t make sense. Do you actually believe tuteimpo and BEST have their own stations at Anchorage airport?
HenryP says: “clearly you work for one of those institutes that I don’t trust.”
It’s an irrational bias on your part. Someone could form an irrational Spanish bias because tutiempo.net is missing a lot of data from 2002 and 2005. That’s why you need to look at the data from multiple sources, not just one.
Here’s the BEST vs tutiempo average temperatures for Anchorage Airport in 2012:
Month, BEST, TuTiempo
1, -15.932, -15.9
2, -3.670, -3.6
3, -6.131, -6.1
4, 3.685, 3.8
5, 7.629, 7.6
6, 12.37, 12.5
7, 13.324, 13.3
8, 13.708, 13.8
9, 9.357, 9.4
10, 1.238, 1.3
11, -7.069, -7.1
12, -8.823, -8.8
So if tutiempo is missing April 2012 (3.8 deg C) you would rather fudge your own number than use that evil BEST data?
tutiempo Apr 2011 = 3.1
tutiempo Apr 2013 = -1.3
average = 0.9
BEST Apr 2012 = 3.685
You need to leave your emotions of hate out of your analysis.
Cramer says
So if tutiempo is missing April 2012 (3.8 deg C) you would rather fudge your own number than use that evil BEST data?
tutiempo Apr 2011 = 3.1
tutiempo Apr 2013 = -1.3
average = 0.9
BEST Apr 2012 = 3.685
You need to leave your emotions of hate out of your analysis.
Henry says
I needed a source of data, as unbiased as possible, as globally as possible,
with minima, to see if there was any man made warming,
and with maxima to see what the natural pattern of warming was.
I chose tutiempo and a dealt with the missing data problem as stated, rightly (by me) or wrongly (according to you).
Quite honestly, I don’t think it is going to make any much of a difference to the outcome,
as the other 11 months of data in 2002 and 2005 were available.
Even in the example that you give, which I assume was cherry picked,
it did not affect the final year-average by more 0.3 degrees C which is not much if you look at the variation in the average yearly temperature in Anchorage.
Cramer says
Your analysis is erroneous. R^2 is meaningless for your regression. Why not take your analysis to someone with expertise in statistics? They will tell you the same thing that I have. Know how do do a regression in Excel does not mean you have much knowledge in statistics.
Henry says
(remember we are looking at maxima here)
if the average speed of warming (on a randomly selected, balanced global sample) is (found by me )
to be as follows
0.036 K/yr 1974 -2012
0.028 K/yr 1980-2012
0.015 K/yr 1990-2012
-0.013 K/yr 2000-2012
during the periods indicated\,
you honestly don’t see that there must be a strong relationship between time and the speed of warming?
whether you take it linear or binomial or as a sine wave (we really must hope the latter to be true),
surely anyone who knows a little bit about statistics would be able to do a test and see that the correlation R^2 is significant on the 95% confidence level.
The problems is that at the universities there are too many scientists who refuse to accept these results, because of the “public opinion” , hence nobody. including you, is doing the (little) work required to repeat my results.
HenryP says: “you honestly don’t see that there must be a strong relationship between time and the speed of warming?”
It appears you did not read what I wrote in my previous comment. Your Y-values are not independent of each other. If you want to see a relationship between time and the speed of warming you should fit the original data to a polynomial or do a segmented linear regression:
1974-1979
1980-1989
1990-1999
2000-2012
Segment it how you want (equal or non-equal), but your segments can not overlap.
HenryP says: “surely anyone who knows a little bit about statistics would be able to do a test and see that the correlation R^2 is significant on the 95% confidence level.”
Your R^2 = 0.96 (or R^2=0.995 for polynomial fit) is meaningless because your regression is meaningless. Learn about the assumptions required for regression. Learn how you can still have an R^2 with a high value (>0.95), but you still have a poor fit or a bad model.
HenryP says: “The problems is that at the universities there are too many scientists who refuse to accept these results, because of the “public opinion” , hence nobody. including you, is doing the (little) work required to repeat my results.”
Yes, you are correct. You did “(little) work.” It’s not your results. It’s your methods. I did the work. You did not do the work (including educating yourself). You refused to investigate the criticisms. You simply kept repeating (cutting-n-pasting) the little work that you did over a year ago. I understand your results better than you do. You did not even understand how to correctly average your ten Alaska weather stations. You drew a question mark on the Barrow weather station, not understanding that it is a North Slope location and the results are consistent with other North Slope locations such as Deadhorse/Prudhoe, Wainwright, and Barter Island (which all continue to warm)(oi40.tinypic.com/2ql5zq8.jpg). But you did not like that it was continuing to warm because it went completely against your thesis: “the major cooling actually is happening from the top latitudes downward.” You didn’t want to believe that Anchorage might be affected by the Pacific Decadal Oscillation, but then you eliminate Greenland and Iceland from your analysis because of effects from the gulf stream and weather stations being on the coast. I guess it okay to include Anchorage when the PDO is cooling it, but not okay when the PDO is warming it. You also compared Greenland to Barrow because both are warming.
HenryP says: “I don’t think it is going to make any much of a difference to the outcome,”
I agree. When your analysis methods are scientifically flawed, the quality of your data does not make much of a difference.
HenryP says: “Even in the example that you give, which I assume was cherry picked,…”
Yes, you are correct. It was cherry picked but in the opposite direction of your implied assumption. I chose the April 2012 because it was the worse month (greatest difference between BEST and tutiempo data), not the best.
Cramer says
you want to see a relationship between time and the speed of warming you should fit the original data to a polynomial or do a segmented linear regression:
1974-1979
1980-1989
1990-1999
2000-2012
Segment it how you want (equal or non-equal), but your segments can not overlap.
HenryP says: “surely anyone who knows a little bit about statistics would be able to do a test and see that the correlation R^2 is significant on the 95% confidence level.”
Your R^2 = 0.96 (or R^2=0.995 for polynomial fit) is meaningless because your regression is meaningless. Learn about the assumptions required for regression. Learn how you can still have an R^2 with a high value (>0.95), but you still have a poor fit or a bad model.
HenryP says: “The problems is that at the universities there are too many scientists who refuse to accept these results, because of the “public opinion” , hence nobody. including you, is doing the (little) work required to repeat my results.”
Yes, you are correct. You did “(little) work.” It’s not your results. It’s your methods. I did the work. You did not do the work (including educating yourself). You refused to investigate the criticisms. You simply kept repeating (cutting-n-pasting) the little work that you did over a year ago. I understand your results better than you do. You did not even understand how to correctly average your ten Alaska weather stations. You drew a question mark on the Barrow weather station, not understanding that it is a North Slope location and the results are consistent with other North Slope locations such as Deadhorse/Prudhoe, Wainwright, and Barter Island (which all continue to warm)(oi40.tinypic.com/2ql5zq8.jpg). But you did not like that it was continuing to warm because it went completely against your thesis: “the major cooling actually is happening from the top latitudes downward.” You didn’t want to believe that Anchorage might be affected by the Pacific Decadal Oscillation, but then you eliminate Greenland and Iceland from your analysis because of effects from the gulf stream and weather stations being on the coast. I guess it okay to include Anchorage when the PDO is cooling it, but not okay when the PDO is warming it. You also compared Greenland to Barrow because both are warming.
Henry says
Ok, I will eat the humble pie here and say that you are probably right in cutting it up that way, in fact I think I will cut it up in the actual Schwabe solar cycles that have been apparent…. I still have all the files and I could now also include the results for 2012 and 2013.
The only problem is that at the moment I donot have much time available for this. This is just my hobby.
Having said that, however, I still think that this excercise will not change my analysis that much.
All major data sets say that we have started cooling from just before the new Millennium just as my results predicted.
http://www.woodfortrees.org/plot/hadcrut4gl/from:1987/to:2015/plot/hadcrut4gl/from:2002/to:2015/trend/plot/hadcrut3gl/from:1987/to:2015/plot/hadcrut3gl/from:2002/to:2015/trend/plot/rss/from:1987/to:2015/plot/rss/from:2002/to:2015/trend/plot/hadsst2gl/from:1987/to:2015/plot/hadsst2gl/from:2002/to:2015/trend/plot/hadcrut4gl/from:1987/to:2002/trend/plot/hadcrut3gl/from:1987/to:2002/trend/plot/hadsst2gl/from:1987/to:2002/trend/plot/rss/from:1987/to:2002/trend
In the latter half of your comment you accuse me of bias.
I think I have gone out of my way to explain my sampling procedure to obtain a globally representative sample.
Note that I was more interested inthe pattern of warming being allowed through the atmosphere, than the PDO and AMO etc.
I know that earth has an intricate system of storing energy for years, hence the survival of the planet for such a long time.
However, if you say that I should also have looked at those places on the warmer gulf stream, then I say that you should remember that there is a lag from energy-in and energy-out. Counting back 88 years i.e. 2013-88= we are in 1925.
Now look at some eye witness reports of the ice back then?
http://wattsupwiththat.com/2008/03/16/you-ask-i-provide-november-2nd-1922-arctic-ocean-getting-warm-seals-vanish-and-icebergs-melt/
Sounds familiar? Back then, in 1922, they had seen that the arctic ice melt was due to the warmer Gulf Stream waters. However, by 1950 all that same ‘lost” ice had frozen back. I therefore predict that all lost arctic ice will also come back, from 2020-2035 as also happened from 1935-1950. Antarctic ice is already increasing.