Tuesday, May 19, 2015
Book Review: The Great Global Warming Blunder: How Mother Nature Fooled the World’s Top Climate Scientists – by Roy W. Spencer (Encounter Books 2010)
I read this one a few years ago. Now I know quite a bit more about climate science so this revisiting is interesting. Spencer is a climatologist and meteorologist who worked for NASA but now is with the University of Alabama. He works much with satellite data. His previous book, Climate Confusion, is often praised among climate change skeptics. Spencer makes charges of scientific elitism and over-politicization of the climate debate. While there may well be some of this (as Dr. Judith Currie notes) Spencer does not prove his point persuasively. It should, however, be considered. Surely there may be some research bias towards the climate change consensus community. He does seem to make a good scientific case that the effect of greenhouse gases on global average temperatures has more uncertainty than most climate modelers assume. Overall, this book is not convincing that climate change is far less dangerous than predicted by the main consensus. It may be, but the doubts raised here, though plausible, are not clear cut. There are definitely uncertainties. The book is, however, worth reading, and important for getting a scientifically-based view that differs from most scientists. Spencer helped to develop satellite measurements of temperatures and these can sometimes differ in rate of change from ground based monitoring. The usual uncertainties about climate variables are displayed: effects of clouds, effects of aerosols (particulate matter), solar variation, climate sensitivity, evaluation of past climates, and measurement methods.
Spencer thinks that climate sensitivity is much less than the IPCC consensus depicts, so that human influence on climate is minimal. He also thinks that much of the warming in the last 100+ years is from natural climate variation rather than human influence. Of course, the real question is how much of the observed warming is due to natural processes and how much is due to the greenhouse gases of humans. Spencer thinks that green energy policies like renewable mandates will cause problems like brownouts and that a war on global warming will end up being a war on the poor. He is probably wrong on both accounts. I agree with him that the EU’s over-reliance on Russian gas is problematic but disagree that more coal-burning power plants are the answer. Even if man-made global warming was not happening the pollutants from those plants shorten lives. He seems to have a disdain for climate modelers and thinks they are biased against meteorology. Of course, there is much more to climate than weather. He erroneously states that researchers have assumed “that natural climate change does not exist.”
He introduces the concepts of forcing and feedback as what induces changes in climate. Regarding forcing, he agrees that we are adding CO2 and it is causing heat to accumulate (how can he not agree – it is basic fact) but regarding feedback he thinks that positive feedback has been over-estimated by overestimating climate sensitivity. He could be right. He could be wrong. That is the big question. Climate modelers must make some assumptions based on proxy data from the past and a limited time period of actual measurements so models are conditioned on these bases. Spencer’s main point is that he thinks cloud behavior has been mistaken as a feedback from increased atmospheric heat when in reality the cloud behavior is a forcing that causes warming. Rather oddly, he also suggests that other scientists surely must have figured this out as well and are hiding it. He and colleagues have written some peer-reviewed papers on this subject but he claims that the overly politicized ‘climate community’ has ignored them. While I do believe there is some political pressure on scientists from the IPCC-led climate community (having heard Dr. Judith Currie speak about it) I do not think it is as bad as he suggests or that scientists are hiding things or being stifled, though ignored perhaps.
Cyclical climate variability is noted, including the El Nino/La Nina cycles (Southern Oscillation), and the Pacific Decadal Oscillation. Spencer claims that the IPCC ignores climate variability on a scale of thirty years or more. I don’t think that is the case although there may be yet undiscovered cyclicity on such scales. Spencer does note the value of atmospheric satellite measurements and that they have some advantages over near-surface measurements. Atmospheric satellite avg. temps (~24 deg F) are much lower than near-surface temps (~ 57 deg F) but can be measured more accurately. Satellite measurements show a global warming of about 0.13 deg C since 1979 while the IPCC models predict a warming of 0.3 deg C going forward – which Spencer sees as an overly bold prediction. The global warming of the last 30 years has also been affected by two major volcanic eruptions (cooling effect) in 1982 and 1991 and by the cycles of El Nino (warming) and La Nina (cooling). Spencer does point out potential problems in the temperature proxy records of the past where temps of individual years cannot really be determined, just hinted at.
Regarding Michael Mann’s “hockey stick” graph of annual global temperatures, Spencer notes that earlier (and more widespread) versions of it neglected the Medieval Warm Period and the Little Ice Age. This was at first due to the use of tree-ring data from limited areas and has since been corrected. Spencer seems to think the IPCC deliberately ignored these two well-known climate fluctuations to enhance their case. However, it has also been shown that these events may not have been global in scope and mostly confined to Northern Europe, particularly the warming. It has also been shown that tree-ring data may miss warming events so tree-ring data is probably not a very accurate proxy for temperature. Spencer goes on to note that surface temperature measurements are affected by where they are taken and where they have been taken in the past which usually are places near human habitation that tend to be warmer. Spencer measures temperatures via satellites and notes that even though they are the best indication of global temperatures, there are some issues with them as well that require estimating.
Spencer invokes Edward Lorenz’s observation that weather prediction is chaotic and that it exhibits “sensitive dependence on initial conditions.” The chaotic behavior of weather systems is why it is nearly impossible to predict weather beyond about seven days. He also notes that El Nino and La Nina events are fairly well understood but still quite difficult to predict. Spencer claims that such chaotic behavior also occurs in multi-year climate cycles and gives the Pacific Decadal Oscillation (PDO) as an example. As an oceanic circulation mechanism that transfers heat, the PDO affects weather similarly to the El Ninos and La Ninas (collectively known as the Southern Oscillation) but on about a 30-year time scale. He thinks the PDO has been ignored by many climate scientists and since it has been on its warming phase for the last thirty years he thinks that it may have contributed to warming more than greenhouse gases. Its cooling phase corresponds to the cooling experienced from the 1940’s to the 1970’s and the beginning of its current warming phase also happens to coincide with the beginning of satellite measurements. The Great Climate shift that began in 1977 and coincides with the warming of the Arctic and temperature rises in Alaska in the 1980s also corresponds to the beginning of the PDO warming phase. Although the correlation is good, some scientists see other causes for the 1940s to 1970s cooling such as Milankovitch effects. Spencer contends that events such as the PDO cause changes in average global cloudiness and increased cloud cover from the most widespread greenhouse gas, water vapor, causes warming. Some scientists think the Southern Oscillation causes the PDO. Although some scientists have suggested that anthropogenic greenhouse gases have caused some of the El Ninos, Spencer thinks the events like the El Ninos and the PDO have caused much of the recorded warming. Spencer agrees with Jim Hansen that the key to global warming is climate sensitivity. While Hansen argues that the climate is very sensitive to man-made greenhouse gases, Spencer thinks it is not so sensitive. Spencer is very critical of Hansen and Al Gore, calling them alarmists who place a theoretical threat of climate catastrophe ahead of the very real threat of energy poverty in parts of the world. Spencer gives Hansen and many others’ argument that CO2 affects temperature in the greenhouse effect but in the ice core records temperature increase precedes CO2 increase. Then temperature increases more due to the CO2. This is due to feedbacks caused initially by the temperature increases. It is a vicious cycle of increasing solar radiation hitting various parts of the earth triggered by planetary orbit Milankovitch cycles (a weak climate forcing) just as decreasing solar radiation from the same orbit cycles triggers ice ages. Spencer seems to think this idea is possible but speculative although there is a considerable amount of evidence for it. Spencer seems to refute Milankovitch effects and CO2 forcing, and considers the ice core records irrelevant to what is happening today. His comment on the ice core records being irrelevant surprised me. Spencer goes to some length to show Al Gore’s propaganda machine, although most seeking to win over others to a view of which they are strongly convinced will involve similar tactics.
Spencer introduces the concept of energy balance and notes that an energy imbalance causes temperature change. He also talks about ‘radiative energy balance’ as the mechanism where incoming solar radiation and outgoing infrared radiation to space are in balance. Spencer notes that water vapor and clouds account for about 90% of the Earth’s greenhouse effect, while CO2 accounts for 3.5% and methane for even less. He notes that overall radiative forcing by CO2, estimated at 1.6 watts per square meter, is below the margin of error for the estimated in-flows and out-flows to and from the Earth. Jim Hansen calculates that 0.8 watts per square meter is the global warming capacity that has yet to heat the atmosphere as it is currently in the ocean. Spencer notes that satellite measurements can only measure energy imbalance to an accuracy of about 2 or 3 watts per square meter so Hansen’s calculations can’t be substantiated by measurements.
Next he introduces the concepts of climate feedbacks. If the sum of feedbacks in the climate system are positive (toward temperature rise) then catastrophic global warming is possible. Spencer, however, argues that the sum of feedbacks is negative due mainly to cloud variations. He notes that feedback due to clouds is the most uncertain, also reminding that water vapor and clouds are by far the largest greenhouse gases. He notes that atmospheric feedbacks are faster acting and more global than surface feedbacks such as ice-sheet feedbacks which are local and slow-acting.
The weak Milankovitch effects forcings that may trigger ice ages and interglacials necessarily assume a sensitive climate. If a weak forcing can trigger large climatic change, that implies a net positive feedback. Spencer argues for an insensitive climate so discounts the magnitude of such effects. As to the past, he says we can estimate past temperatures by proxy but it is more difficult to estimate past climate forcings. Uncertainties abound also in measuring climate feedbacks which lead to the wide ranges in temperatures predicted by the IPCC in response to CO2 forcing. It is this inherent uncertainty in measuring feedbacks, Spencer says, that keeps the global warming alarmism alive and well.
The role of clouds involves forcing (cause) and feedback (effect). Climate modelers assume that warmer temperatures cause decreasing cloud cover but Spencer thinks it is the reverse, that decreasing cloud cover causes warmer temperatures. Spencer and a colleague ran a simple climate model that showed positive feedback response to clouds in the model even when negative feedback was specified. Their paper was published in the American Meteorological Society’s Journal of Climate in 2008, but was largely ignored even though the reviewers noted that climate modelers needed to be aware of its potential implications: that climate models may be biased toward positive feedback based on assumptions of climate sensitivity.
Atmospheric feedbacks need to be measured with satellites, says Spencer. He gives the various satellites uses by NASA and NOAA that orbit several times a day and gives various graphs of radiative energy imbalance vs. temperature change to show how these graphs are constructed, calculating their slopes and interpreting them in terms of “total feedback parameter.” This part of the book of the hands-on science I found to be very interesting. He notes that plotting satellite data has some limitations and can be interpreted in various ways based on how it is plotted (ie. monthly avgs, 3-month avgs, etc). His main contention is that it is unknown if these plots are simply a measure of feedbacks as they could be measuring forcings as well, especially cloud forcings as he sees them. We do not know how much is feedback and how much is forcing, how much is effect and how much is cause, which ultimately means we don’t know “how much” CO2 increases actually affect global avg. temperatures. Spencer notes that specified feedbacks in climate models may differ considerably from diagnosed feedbacks. He shows a simple climate model of cloud variability where a strong negative feedback was specified but after plotting the slope indicated strong positive feedback. His contention here is that the forcing obscures the feedback. In order to accurately estimate the feedback, the forcing would need to be removed, he says. IPCC modelers remove known forcings like anthropogenic CO2 but natural forcings like cloud variation (natural quasi-random, chaotic cloud variation) are not known. Other known forcings such as the 1991 Pinatubo volcano can be removed as well. Spencer claims to have found a signature in the connected time data plots that he calls “radiative forcing spirals” that show up regardless of pre-specified parameters.
Spencer next considers non-radiative forcing events, where heat reservoirs: ocean, land, and atmosphere, transfer heat. He give the most prominent example of a non-radiative forcing as the Madden-Julian Oscillation (MJO) that occurs in the tropical Pacific where variation in wind speed, humidity, or temperature in different atmospheric levels can cause changes in evaporation rates. In the MJO this causes the ocean surface to cool for a month or two and the evaporation causes increased rainfall. This causes latent heat in the ocean to be transferred to the atmosphere – considerably more heat than is lost at the ocean surface. Such an event can cause average global temps to increase by 1 degree F for a month or more, equivalent to 50 years of global warming. This, he says, complicates identification and measurement of feedbacks. Hurricanes are another more localized example of non-radiative forcing. Spencer now runs a non-radiative forcing model similar to the previous radiative forcing model of radiative energy imbalance vs. temperature change and shows here that specified feedback and diagnosed feedback are strongly in alignment – which implies that it is a measure of feedback only, without any obscuration from forcing (what he calls a feedback stripe).There are no spirals, only a straight line. He thinks that only non-radiative forcing can be accurately estimated from satellite measurements. In the real world there is a mixture of radiative and non-radiative forcings continually. While I am not competent enough with climate models to know if Spencer’s logic is foolproof, it does seem to make sense and it does seem to convey the complexity and reliance on assumptions that may be inherent in climate modeling. I would, however, like to see a critique of his ideas by other climate scientists. If there is a mixture of radiative and non-radiative forcings in the model then:
“… natural cloud fluctuations in the climate system will cause a bias in the diagnosed feedback in the direction of positive feedback, thus giving the illusion of an overly sensitive climate system.”
He thinks that the reason feedback estimates from satellite data of other researchers have been quite variable is that each model contained different variables of radiative and non-radiative forcing. Next he goes through some IPCC climate models and shows that they also display radiative forcing spirals – assumed to be caused by cloud variability. Feedback stripes are harder to find but he does show some models with them which would seem to indicate that long-term climate change as well as short-term climate change (year-to-year change as in the models with radiative forcing spirals). Spencer suggests that taking the radiative forcing out of some climate models would render a doubling of atmospheric CO2 to producing a 0.5 deg C temp increase which is far lower than the lowest IPCC estimate of 1.5 deg C. His charge is that the IPCC has ignored radiative forcing from cloud variability even though it is well aware of the effects of radiative forcing as it was when it removed the forcing from Pinatubo from the models. Thus, he says their models are overly sensitive. He claims they mixed up cause and effect regarding cloud decrease and temperature and that we now have peer-reviewed papers that show the decreases in clouds can force temp increases, instead of the other way around.
According to Spencer the IPCC considers longer-term events such as SO and PDO to only influence regional weather patterns. He suggests they change global average cloud cover which is the main cause of the global warming we have seen over the last 100 years (75%). He thinks that most climatologists are not well versed enough in meteorology and most meteorologists believe that natural climate variation is greater than depicted by climatologists. He says it would take changes in global average cloud cover of a little more than 1 percent to explain most of the climate change of the last 2000 years but we have only been able to measure such small changes in cloud cover for about 10 years. This makes me wonder – with new instrumentation and satellite data we should be able over the next 20 years or so to get much better ideas about climate change, and the effects of clouds.
Interestingly, he points out that satellite measurements of Arctic sea ice have only been available since the mid-1970’s which happens to be when the PDO went into its warming phase. He notes that loss of such sea ice and possible opening of the Northwest Passage is documented as well from the late 1930’s. One of his contentions is that the IPCC and other climate scientists consider that the Earth’s climate is in a state of overall energy balance unless disturbed by external forcings: solar, greenhouse gases, volcanoes, etc. He thinks that changes in cloud cover generate energy imbalances both on a year-to-year basis and on longer multi-year time frames in response to cycles like the PDO and SO. He and many meteorologists consider energy imbalances to be inherent and the drivers of weather.
He describes the PDO as a driver of preferred types of weather patterns with two phases, each lasting 30 years or so. Graphs of the PDO index since 1900 compared to global average surface temperatures are a very good match, especially since the advent of satellite measurements in the late 1970’s. The PDO entered its warming phase in the late 1970’s and so is close to shifting back to its cooling phase by now. Spencer thinks that recently published changes in the Arctic ocean circulation system may be a signature of the PDO. He notes that his paper on the PDO influence on climate submitted to Geophysical Research Letters was quickly rejected by a single reviewer. Spencer plots his presumed PDO-influenced temperature changes derived from a simple climate model along with Jim Hansen’s presumed CO2-influenced temperature changes. There is a pretty damn good match of PDO and temp – which he says was not at all known before running the model. I assume this graph is what informs his idea that 75% of warming is due to the PDO and presumably 25% due to CO2. Only about nine years of satellite data were available at the time to try and determine if the PDO actually caused changes in cloud cover. Spencer’s methodology was a reverse of climate methodology to remove a forcing (such as a volcano) to measure feedback, in that he used temperature changes to remove feedback in order to measure forcing. There was a fairly good correlation between his predicted model and observed satellite data. This shows, he says, that the PDO does indeed affect cloud cover, and thus the global energy balance. Interestingly, he says that the evidence is good enough that natural influences of climate variation like the PDO can influence warming that it should be investigated further with more research. He also reiterates that climate models are based on assumptions and one can get quite different results based on what assumptions with which they begin.
If Spencer is correct that greenhouse gases affect climate significantly less than thought, then one should consider other consequences of CO2 which are not detrimental but potentially beneficial. He contests the assumption that in-flows and out-flows of CO2 have been in some sort of equilibrium until humans came along and upset the balance. Fluctuations in CO2 levels occur annually with fairly large changes attributable to El Ninos (death of plankton and less CO2 uptake) and La Ninas (greater CO2 uptake by plankton). The CO2 increase rate is roughly half of what humans produce which shows that the other half is being taken in by carbon sinks. Spencer suggests that perhaps more is taken in as some of the CO2 increase can be attributable to natural climate variation. NASA’s 2009 Orbiting Carbon Observatory failed to launch but a stripped down Canadian version is orbiting. This and other satellites, says Spencer, should begin to give us a better idea of carbon recycling, such as how much is taken up by the oceans, the forests, and other land masses, and rock weathering. He predicts it will be revealed that nature is more adept at removing excess CO2 than thought. There is some recent evidence that he is correct but the estimated increased removal capacity is not thought to be massive. He points out that it has been recently discovered that a gradually decreasing amount of the carbon isotope C13 is not a unique fingerprint of anthropogenic CO2 increase, but can be from biological sources as well. He points out that the reason he is emphasizing that some CO2 increase can be naturally occurring (probably a small amount) is that the climate models assume that all of the increase is anthropogenic and if some is natural that would mean that more CO2 is being absorbed by nature than thought and the models would then need to be adjusted. He then mentions ocean acidification. He predicts that more CO2 in the ocean might actually be good rather than undesirable. He gives one reference here but his argument is very weak as much of the evidence points to increased ocean acidification as harmful, especially to carbonate-shelled sea creatures such as corals.
Spencer then goes off on a bit of a tangent, asking why we should reduce greenhouse gas emissions at all, since nature has periods of high CO2. He seems to think “greenism” is dangerous and when adopted by Big Business, helps to solidify global warming alarmist positions. Sure, there is uncertainty but enough is known about the problem – we are increasing CO2 and it is having affects – that even if part of its increase is natural and more importantly even if part of the heat is from natural processes– things are still heating up and all we can control is our own emissions. I do agree with him that renewable energy has been overhyped thus far, though it may become more economic in the near future and will definitely become more economic as fossil fuels deplete.
In the next chapter he gives some climate predictions, with the playful caveat that on some he is wildly speculating. Interestingly, he notes that the PDO is expected to shift to its negative phase or may have as some researchers such as geologist Don Easterbrook from Western Washington University contend. This would mean that a natural cooling phase, possibly global in extent, has begun. This in combination with anthropogenic global warming would make a period, possibly on the scale of 30 years, where on change balances the other so that there is little global temperature rise. Satellite measurements have indeed shown a leveling off of global temperature rise since the early 2000’s. He speculates with funny scenarios depicting what would happen if anthropogenic emissions turned out not be harmful. He thinks debates about climate change by scientists would be useful. One such occurred in 2007 with 3 scientists on each side. Spencer notes that before the debate more people were in the alarmist camp but after the debate it was close to even among alarmists and skeptics. He also rails against carbon legislation of any sort, saying it offers unfair advantages and disadvantage to both technologies and people. He also does not like market-based cap-and-trade since it has in the past led to cheating and corruption and unfair distribution of credits and allowances. Of course, the new EPA clean power plan fast-tracked by Obama is not all about CO2. It is about particulate pollution and improving human health. He charges those in the alarmist camps with “religious environmentalism.”
Finally, he gives a final chapter of Summary and Conclusions. Time will tell whether he is right about the PDO and other non-radiative forcings affecting global cloud cover variations, the climate feedback system, and climate sensitivity, since we now have better satellite measurements and trends will be revealed in time. Still, his charge that the IPCC and most climate scientists are biased, whether through influence, or compliance with corruption, is a bit hard to swallow. Indeed, climate modelers have noted that the models are only as good as the data inputted into them by the assumptions made to condition the models. Of course, there is natural climate variability. He keeps suggesting that the IPCC and their ilk have rejected natural climate variability. I don’t think that is the case at all. His argument that the climate research establishment has lost objectivity through “groupthink” is more of a possibility but more details of such a charge would need to be displayed. He may have a point also about editors unfairly rejecting articles that go against the groupthink. He notes that he has even had editors tell him to change his conclusions to be in agreement with IPCC conclusions.
Overall, I would recommend this book as one of the best books advocating climate change skepticism. Spencer is a scientist, one of the presumably 3% who disagree with anthropogenic global warming as it is commonly depicted though I suspect the real number is about 30% or more. Even so, he would really need to prove his case over the next decade or two as he may be flat-out wrong about clouds acting as forcings rather feedbacks. All I can say is time will tell.