My Thanks and Comments for Dr. Walt Meier

Guest Post by Willis Eschenbach

First, I would like to thank Dr. Meier of the National Snow and Ice Data Center (NSIDC) for answering the questions I had posed (and had given my own personal answers) in “Trust and Mistrust”. I found his replies to be both temperate and well-reasoned. Also, I appreciate the positive and considerate tone of most of those who commented on his reply. It is only through such a peaceful and temperate discussion that we can come to understand what the other side of the debate thinks.

Onwards to the questions, Dr. Meier’s answers, and my comments:

Question 1. Does the earth have a preferred temperature, which is actively maintained by the climate system?

Willis says that he “believes the answer is yes”. In science “belief” doesn’t have much standing beyond initial hypotheses. Scientists need to look for evidence to support or refute any such initial beliefs. So, does the earth have a preferred temperature? Well, there are certainly some self-regulating mechanisms that can keep temperatures reasonably stable at least over a certain range of climate forcings. However, this question doesn’t seem particularly relevant to the issue of climate change and anthropogenic global warming. The relevant question is: can the earth’s temperature change over a range that could significantly impact modern human society?

My comment: Since unfortunately so little attention has been given to this important question, my idea of how it works is indeed a hypothesis. Therefore, “belief” is appropriate. However, I have provided several kinds of evidence in support of the hypothesis at the post I cited in my original answer to this question, “The Thermostat Hypothesis”.

Next, Dr. Meier says that there are “some self-regulating mechanisms that can keep temperature reasonably stable at least over a certain range of climate forcings.” Unfortunately, he does not say what the mechanisms might be, at what timescale they operate, or what range of forcings they can handle.

However, he says that they can safely be ignored in favor of seeing what the small changes are, which doesn’t make sense to me. Before we start looking at what causes the small fluctuations in temperature that we are discussing (0.6°C/century), we should investigate the existence and mechanism of large-scale processes that regulate the temperature. If we are trying to understand a change in the temperature of a house, surely one of the first questions we would want answered is “does the house have a thermostat?” The same is true of the climate.

Question 2: Regarding human effects on climate, what is the null hypothesis?

I will agree with Willis here – at one level, the null hypothesis is that any climate changes are natural and without human influence. This isn’t controversial in the climate science community; I think every scientist would agree with this. However, this null hypothesis is fairly narrow in scope. I think there is actually a more fundamental null hypothesis, which I’ll call null hypothesis 2 (NH2): are the factors that controlled earth’s climate in the past the same factors that control it today and will continue to do so into the future? In other words are the processes that have affected climate (i.e., the forcings – the sun, volcanic eruptions, greenhouse gases, etc.) in the past affecting climate today and will they continue to do so in the future? A basic premise of any science with an historical aspect (e.g., geology, evolution, etc.) is that the past is the key to the future.

My Comment: I assume that Dr. Meier has temporarily overlooked the fact that a null hypothesis is a statement rather than a question. Thus, his Null Hypothesis 2 (NH2) should be:

NH2: The factors that controlled earth’s climate in the past are the same factors that control it today and will continue to do so into the future

However, this formulation has some serious problems. First, a null hypothesis must be capable of being falsified. My null hypothesis (NH1) could be falsified easily, by a showing that measurements of the modern climate are outside the historical values.

Dr. Meier’s NH2, on the other hand,  extends into the future … how can we possibly falsify that?

Second, to determine if the factors that controlled the climate in the past are the same factors that control it now, we must know the factors that controlled climate in the past, and we must know the factors that control climate now. But that is exactly the subject being debated – what controls the climate? We don’t know the answer to that for the present, and we know even less about it for the past. So again, his NH2 is not falsifiable.

Finally, there is a more fundamental problem with NH2. The null hypothesis has to be the logical opposite of the alternate hypothesis, so that if one is true, the other must be false. My null hypothesis NH1 is that the currently observed climate variations are the result of natural variation. The opposite of my null hypothesis is the alternate hypothesis, that currently observed climate variations are the result of human-caused GHG increases.

However, what is the opposite of NH2, which states that the factors that controlled climate in the past are those that control climate today? The opposite of that is the alternate hypothesis that the factors that controlled climate in the past are not those that control climate today.

But I have never once, in this entire decades-long debate, heard anyone make the claim that some factors that affected climate in the past have stopped affecting the climate. As a result, NH2 is a straw man. It is the null hypothesis for an alternate hypothesis that no one is propounding.

Since it is not falsifiable, and since it is a straw man null hypothesis, Dr. Meier has not proposed a valid null hypothesis. As a result, his arguments that follow from that null hypothesis are not relevant.

Question 3: What observations tend to support or reject the null hypothesis?

Let me first address NH2. We have evidence that in the past the sun affected climate. And as expected we see the current climate respond to changes in solar energy. In the past we have evidence that volcanoes affected climate. And as expected we see the climate respond to volcanic eruptions (e.g., Mt. Pinatubo). And in the past we’ve seen climate change with greenhouse gases (GHGs). And as expected we are seeing indications that the climate is being affected by changing concentrations of GHGs, primarily CO2. In fact of the major climate drivers, the one changing most substantially over recent years is the greenhouse gas concentration. So what are the indications that climate is changing in response to forcing today as it has in the past? Here are a few:

1. Increasing concentrations of CO2 and other GHGs in the atmosphere

2. Rising temperatures at and near the surface

3. Cooling temperatures in the stratosphere (An expected effect of CO2-warming, but not other forcings)

4. Rising sea levels

5. Loss of Arctic sea ice, particularly multiyear ice

6. Loss of mass from the Greenland and Antarctic ice sheets

7. Recession of most mountain glaciers around the globe

8. Poleward expansion of plant and animal species

9. Ocean acidification (a result of some of the added CO2 being absorbed by the ocean)

It is possible that latter 8 points are completely unrelated to point 1, but I think one would be hard-pressed to say that the above argues against NH2.

My Comment: Saying “it is possible that the latter 8 points are completely unrelated to point 1” begs the question. It is possible that they are related, but that is the question at hand that we are trying to answer. If Dr. Meier thinks that they are related, he needs to establish causation, not just say it is “possible that [they] are completely unrelated”.

Whether his points argue for or against NH2 is not relevant, since NH2  is not falsifiable, and is a null hypothesis for a position no one is taking. In addition, they are presented as “indications that climate is changing in response to forcing today as it has in the past” … but it is a mix of statements about forcings, and responses to increasing warmth. So I don’t see how that applies to NH2 in any case.

Despite those problems, let me address them, one by one, starting with one without a number:

In the past we’ve seen climate change with greenhouse gases (GHGs): This cries out for a citation, but there is none. When did we see that, who showed it, what evidence is there to support it?

1. Increasing concentrations of CO2 and other GHGs in the atmosphere: Yes, GHGs are increasing. However, this says nothing either way about NH2.

2. Rising temperatures at and near the surface: Yes, temperatures generally have been rising, and they have been ever since the Little Ice Age in the mid 1600’s. But again, what does this have to do with NH2?

3. Cooling temperatures in the stratosphere (An expected effect of CO2-warming, but not other forcings): I would greatly appreciate a citation to the claim that this is an expected result of GHG forcing but not other forcings. Given our general lack of understanding of the climate, it would be a very difficult claim to establish.

For one of the reasons why it would be hard to establish, here is the actual change in the stratospheric temperatures:

Figure 1. UAH and RSS satellite measurements of stratospheric temperature. DATA SOURCES UAH, RSS

Now, how on earth (or off earth and in the stratosphere) is that an “expected effect” of increasing GHGs? Since recovering from the Pinatubo eruption stratospheric temperatures have been stable … which climate model projected that outcome? What theoretical calculations showed that flat-line response?

4. Rising sea levels: Sea levels have been rising since 1900. If GHGs were driving the rise, we would expect to see an acceleration in the rate of rise corresponding to the acceleration in the rise of GHGs. However, we have seen no such acceleration in the long-term, and we see deceleration in the short-term. Here are two long-term records. Fig.2 is from Church and White and Jevrejeva:

Figure 2. Church & White and Jevrejeva sea level records from tidal stations. Photo is of Dauphin Island Tidal Station. PHOTO SOURCE

There is good agreement between the Church & White and the Jevrejeva records. As they were calculated in different ways, this increases the confidence in the result. Note that, despite increasing CO2, there is no increase in the rate of sea level rise.

Next, we have a short-term but presumably more accurate sea level record from the TOPEX satellite. Fig. 3 shows that record:

Figure 3. Sea level record from the TOPEX satellite. Black line is the trend from 1993 to 2004, and is projected to 2007 in gray. Red line is the trend since 2004.

As you can see, rather than increasing, the rate of sea level rise has dropped in recent years. And while it may well start to rise again, it is certainly not accelerating as the AGW hypothesis requires.

5. Loss of Arctic sea ice, particularly multiyear ice: As Dr. Meier would agree, the satellite record of Arctic ice is quite short, much shorter than the long-term changes in Arctic temperatures. The Arctic was as warm or warmer in the 1930s, and many records from that time attest to greatly reduced ice conditions. Both the Polyakov and the NORDKLIM records [see Update 10] show the time around 1979 as being about the bottom of the Arctic temperature swing, so reducing Arctic sea ice is to be expected since 1979. In addition, I was surprised that Dr. Meier did not mention the last three years, which have seen both increasing Arctic sea ice and increasing multiyear sea ice.

6. Loss of mass from the Greenland and Antarctic ice sheets: NASA reports that the GRACE satellite data shows the Antarctic and Greenland ice sheets to be losing a total of ~ 1,700 cubic km of ice per year. While this sounds large, this is about 0.005% of the total ice in the two sheets … I hardly see this as indicating anything but a confirmation that the earth has been warming for centuries, and is generally continuing to do so with the usual fits and starts. Since at the current rate of loss it will take about two hundred years to lose 1% of the ice, I don’t see this as a critical issue.

7. Recession of most mountain glaciers around the globe: According to the NSIDC, the excellent organization that Dr. Meier works for, there are about 100,000 glaciers on the planet. Again according to the NSIDC, we have measured the mass balance on 300 of them, and we have continuous records since 1960 for only 60 of them … so we have at least one record on 0.3% of the glaciers, and decent (although short) records on 0.1% of the glaciers. Given those percentages, “recession of most mountain glaciers” seems to be a bit of an overstatement of what scientific research actually has shown …

It is true that many of the glaciers we have measured have receded since the colder period of the 1960s when the records started. It is also true that some are advancing. Many of the known glaciers have been generally receding since sometime after the Little Ice Age in the 1600’s. Before that, they were advancing, so much so that in 1678 the village of Aletsch in Switzerland made a formal church vow to live virtuously if only the nearby advancing glacier would not over-run their village … a vow which they are now trying to recant as the glacier recedes. That dratted climate never stops changing.

All this shows is that when the earth cools, glaciers generally advance, and when it warms, they generally retreat. Surprising, huh? It says nothing about whether or not GHGs control the temperature.

8. Poleward expansion of plant and animal species: Animals and plants advance and retreat with the seasons and with the climate. In a time of general warming, like the last 300 years, we would expect them to move slightly polewards. However, care is required, because climate change is blamed for everything. For example, in this South African study (subscription required), they say (emphasis mine):

Evidence from the Northern Hemisphere and simple theoretical models both predict that climate change could force southern African birds to undergo poleward range shifts. We document the chronology and habitat use of 18 regionally indigenous bird species that colonised the extreme south-western corner of Africa after the late 1940s. This incorporates a period of almost four decades of observed regional warming in the Western Cape, South Africa. Observations of these colonisation events concur with a ‘climate change’ explanation, assuming extrapolation of Northern Hemisphere results and simplistic application of theory. However, on individual inspection, all bar one may be more parsimoniously explained by direct anthropogenic changes to the landscape than by the indirect effects of climate change. Indeed, no a priori predictions relating to climate change, such as colonisers being small and/or originating in nearby arid shrublands, were upheld.

9. Ocean acidification (a result of some of the added CO2 being absorbed by the ocean): Again, this appears to be happening, although we have very little in the way of data. If verified, this would indicate that atmospheric CO2 levels are rising … but we knew that already.

Overall, Dr. Meier’s points show that when the world warms we are likely to see various phenomena related to that warming. But that says nothing about his null hypothesis NH2, nor about my null hypothesis. None of them either support NH2 nor falsify NH2, as NH2 is a straw null hypothesis that cannot be falsified. They also say nothing about whether GHGs are currently causing unusual warming.

Next, Dr. Meier addresses NH1, my null hypothesis:

Of course none of the above says anything about human influence, so let’s now move on to Willis’ null hypothesis, call it null hypothesis 1 (NH1). Willis notes that modern temperatures are within historical bounds before any possible human influence and therefore claims there is no “fingerprint” of human effects on climate. This seems to be a reasonable conclusion at first glance. However, because of NH2, one can’t just naively look at temperature ranges. We need to think about the changes in temperatures in light of changes in forcings because NH2 tells us we should expect the climate to respond in a similar way to forcings as it has in the past. So we need to look at what forcings are causing the temperature changes and then determine whether if humans are responsible for any of those forcings. We’re seeing increasing concentrations of CO2 and other GHGs in the atmosphere. We know that humans are causing an increase in atmospheric GHGs through the burning of fossil fuels and other practices (e.g., deforestation) – see Question 6 below for more detail. NH2 tells us that we should expect warming and indeed we do, though there is a lot of short-term variation in climate that can make it difficult to see the long-term trends.

So we’re left with two possibilities:

1. NH2 is no longer valid. The processes that have governed the earth’s climate throughout its history have suddenly starting working in a very different way than in the past.

Or

2. NH1 is no longer valid. Humans are indeed having an effect on climate.

Both of these things may seem difficult to believe. The question I would ask is: which is more unbelievable?

This is a false dichotomy, created by using my real null hypothesis NH1, and Dr. Meier’s straw man null hypothesis NH2. Yes, both CO2 and temperatures rose over the 20th century … but correlation is not causation, and CO2 does not correlate any better with temperature than a straight line correlates with temperature. Next, Dr. Meier seems to think that NH1 and NH2 are somehow related, so that one or the other must be false. But both could easily be true. It could be true that the climate variations are natural (NH1), and also true that the historical forcings still apply (NH2). So his “one true / one false” duality is not valid.

At the end of the day, as Dr. Meier says himself, none of what he has said falsifies the null hypothesis NH1 that the observed climate changes are natural variations rather than human-caused. Since it is not falsified, we have nothing for the AGW hypothesis to explain. This is an important conclusion.

Skipping over some questions where we generally agree, we come to:

Question 6: How are humans affecting the climate?

Willis mentions two things: land use and black carbon. These are indeed two ways humans are affecting climate. He mentions that our understanding of these two forcings is low. This is true. In fact the uncertainties are of the same order of as the possible effects, which make it quite difficult to tell what the ultimate impact on global climate these will have. However, Willis fails to directly mention the one forcing that we actually have good knowledge about and for which the uncertainties are much smaller (relative to the magnitude of the forcing): greenhouse gases (GHGs). This is because GHGs are, along with the sun and volcanoes, a primary component that regulates the earth’s climate on a global scale.

My Comment: First, despite the IPCC claims, our knowledge of the effects of the GHGs is not as good as our knowledge of the effects of black carbon or deforestation. This is because we can actually measure the effects on the temperature of chopping down a forest. We can actually measure an amount of black carbon on snow, and see what difference it makes to the melting rate of the snow, and the temperature above the snow.

But we cannot make any such measurements for CO2. All of our numbers for the GHG forcings are based on climate models rather than measurements. The IPCC, and many scientists, give them great credence. I, and a number of scientists, do not.

Dr. Meier again:

It might be worth reviewing a few things:

1. Greenhouse gases warm the planet. This comes out of pretty basic radiative properties of the gases and has been known for well over 100 years.

My Comment: This is one of the most widely held misconceptions in the field. Here’s an example of the identical incorrect logical jump, from another field:

It is clear from the basic radiative properties that solar radiation warms what it hits. Therefore, if I walk out into the sunshine, my core body temperature will rise.

Clearly, the mere fact that a source is radiating does not mean that it will necessarily cause whatever the radiation strikes to warm up …

This is a crucial point, and one which is either overlooked or ignored by AGW proponents. Here’s another example. If your house has an air conditioner on a thermostat, despite the sun getting warmer and warmer as the day goes on, the house does not warm up. Again, we have a radiation source which does not cause what it strikes to warm up.

So yes, we know that CO2 is a greenhouse gas. And we know it will increase the forcing, although the amount is not well established.

But we absolutely do not know if that will cause the earth to warm over time. This is why my Question 1 above, about whether the Earth has a thermostat, is so important. If the earth has a thermostat, there are many basic assumptions that need to be reconsidered. I discuss this issue in detail at “The Unbearable Complexity of Climate”.

The short version of that post is that “basic radiative properties” are far from enough to determine what will happen from increased forcing in a complex system such as the climate or the human body, or even in a simple system like an air-conditioned house.

2. Carbon dioxide is a greenhouse gas. This is has been also been known for well over 100 years. There are other greenhouse gases, e.g., methane, nitrous oxide, ozone, but carbon dioxide is the most widespread and longest-lived in the atmosphere so it is more relevant for long-term climate change.

My Comment: Agreed.

3. The concentration of CO2 is closely linked with temperature – CO2 and temperature rise or fall largely in concert with each other. This has been observed in ice cores from around the world with some records dating back over 800,000 years. Sometimes the CO2 rise lags the temperature rise, as seems to be the case in some of ice ages, but this simply means that CO2 didn’t initiate the rise (it is clear that solar forcing did) and was a feedback. But regardless, without CO2 you don’t get swings between ice ages and interglacial periods. To paraphrase Richard Alley, a colleague at Penn State: “the climate history of the earth makes no sense unless you consider CO2”.

My Comment: As temperatures warm and cool, the CO2 levels go up and down. We can see that in the ice core records. SInce CO2 lags temperature in the Vostok ice core records of these changes, this means that the CO2 is not the cause of temperature change. Instead, it is a result of the warming ocean giving off more CO2. So far, Dr. Meier and I totally agree.

He then says “sometimes the CO2 rise lags the temperature rise.” This is not borne out by the data, where the correlation with lagged CO2 is greater than with un-lagged CO2 for the entire dataset. This indicates that the lag is a phenomenon common to the entire time period of the data.

Then Dr. Meier makes the claim that the CO2 “was a feedback”. If this were true, once the CO2 started to rise or fall, we should see a change in the rate of temperature rise or fall. To my knowledge no one has ever mathematically demonstrated such a feedback-driven change in temperature rise or fall in the actual ice core data. In addition to searching the literature for such a demonstration, I have used a variety of mathematical methods to try to find such a lagged feedback effect in the data, without  any success. So why does Dr. Meier say that CO2 is operating as a feedback?

Dr. Meier may not even realize it, but he has totally conflated reality and models. What Dr. Meier is trying to say is that “without CO2 the models don’t get swings between ice ages and interglacial periods.” And what Richard Alley has shown is that “the modelled version of the climate history of the earth makes no sense unless you consider CO2”. Neither of them are talking about reality, they are discussing model-ice on Model-world, not ice on the Earth.

This blurring of the line between reality and models is a recurring and very frustrating feature of the climate discussion. I’m talking about reality, and meanwhile, without saying so, Dr. Meier is discussing model results. This habit of climate scientists, of talking about models as if they were discussing reality, is very frustrating and impedes communication.

4. The amount of carbon dioxide (and other GHGs) has been increasing. This has been directly observed for over 50 years now. There is essentially no doubt as to the accuracy of these measurements.

My Comment: Agreed

5. The increase in CO2 is due to human emissions. There are two ways we know this. First, we know this simply through accounting – we can estimate how much CO2 is being emitted by our cars, coal plants, etc. and see if matches the observed increase in the atmosphere; indeed it does (after accounting for uptake from the oceans and biomass). Second, the carbon emitted by humans has a distinct chemical signature from natural carbon and we see that it is carbon with that human signature that is increasing and not the natural carbon.

My Comment: Agreed.

6. Given the above points and NH2, one expects the observed temperature rise is largely due to CO2 and that increasing CO2 concentrations will cause temperatures to continue to rise over the long-term. This was first discussed well over 50 years ago.

My Comment: We have no evidence (not model results but evidence) that at the current general temperature equilibrium, changes in GHG forcing affect the temperature. We have no evidence that they affected temperature in the transitions between glacial and interglacial periods. We have no evidence that there is a linear relationship between temperature and forcing, it may well be temperature dependent and asymptotically approach zero at equilibrium. Yes, as Dr. Meier points out, forcings affect temperature in those situations (and all others) in the models. But I’ve been programming computers for almost fifty years now, and I’ve written too many computer models and I know too much about computers to trust untested, unverified models that are tuned to reproduce the past. Too many parameters, too many degrees of freedom, too much error propagation, too little understanding of important processes, they have, as Kipling said, been “twisted by knaves to make a trap for fools”.

Question 7: How much of the post-1980 temperature change is due to humans?

Here Willis says we get into murky waters and that there is little scientific agreement. And indeed this is true when discussing the factors he’s chosen to focus on: land use and soot. This is because, as mentioned above, the magnitudes of these forcings are small and the uncertainties relatively large. But there is broad scientific agreement that human-emitted CO2 has significantly contributed to the temperature change.

My Comment: Post 1980, the temperatures rose, peaked in 1998, and have been basically level since then. While there is broad agreement on something like “CO2 contributes significantly”, how significantly did it contribute to the post 1998 period of basically no temperature change? The answer, presumably, is unknown. Some scientists see CO2 as a second order forcing, after land use/land cover change (LULCC) and black/brown carbon forcing, particularly for the Arctic. My point is that there is still ongoing scientific discussion on the question of how much each forcing might affect the climate, particularly given that the temperature hasn’t risen in the last decade.

Question 8: Does the evidence from the climate models show that humans are responsible for changes in the climate?

Willis answers by claiming that climate models don’t provide evidence and that evidence is observable and measurable data about the real world. To me evidence is any type of information that helps one draw conclusions about a given question. In legal trials, it is not only hard physical evidence that is admitted, but information such as the state of mind of the defendant, motive, memories of eyewitnesses, etc. Such “evidence” may not have the same veracity as hard physical evidence, such as DNA, but nonetheless it can be useful.

My Comment: I fear this answer makes no sense. Dr. Meier says evidence is “any type of information that helps one draw conclusions”. Many people are helped to draw conclusions by astrology. Does that make astrology evidence? The conclusions of some scientists are shaped by their religious beliefs. Does that make religious beliefs evidence? Hunches and intuition help scientists draw all kinds of conclusions … are they evidence?

I don’t think Dr. Meier really believes what he is saying here. For example, said that above that I think that the earth has a thermostat. The first thing that Dr. Meier said in response to that was “Scientists need to look for evidence to support or refute any such initial beliefs.”

I don’t think he was referring to astrology, or my state of mind, or the memories of eyewitnesses. I think he was talking about data, observations, facts to support my hypothesis. And that is what I have provided at the citation listed above, for the same reason that he asked – because science is based on evidence, data, facts, measurements, and not on states of mind. The modeller’s mantra says “All models are wrong … but some models are useful.” Yes, they are often useful, but they don’t produce evidence.

Regardless, let me first say that I’m a data person, so I’ve always been a bit skeptical of models myself. We certainly can’t trust them to provide information with complete confidence. It may surprise some people, but most modelers recognize this. However, note that in my response to question 6 above, I never mention models in discussing the “evidence” for the influence of human-emitted CO2 on climate. So avoiding semantic issues, let me say that climate models are useful (though far from perfect) tools to help us understand the evidence for human and other influence on climate. And as imperfect as they may, they are the best tool we have to predict the future.

My Comment: As anyone who has looked at a weather forecast for next weekend knows, some models may be the best tool we have and still be no better than flipping a coin.

As to whether the models are useful, we have some simple ways to determine whether a model is useful. One is to see if they can make falsifiable predictions of the future states of a given system. To date, the models have failed miserably at this test. The current hiatus in warming was not predicted by a single model that I know of. Even if the GHG forcing were overwhelmed by natural variations, according to the models the stratosphere should have continued to cool. It did not do so. They have not been able to forecast the trend in the numbers of hurricanes, despite making a host of claims after the recent single-year peak in hurricane numbers. The claim is often made that the models are not accurate in the short-term, but they are accurate in the long-term. I’m still waiting … how long a term does it take until their accuracy starts to show up? Twenty-six years? Fifty-three years? Where is the theory that tells us when they will start to be right?

Another way to judge a model’s usefulness is if it can identify missing factors in a system. The classic example is the discovery of Neptune based on what was missing in models of the solar system. But the climate models are assumed to already contain all the important forcings, so they cannot discover any possible missing forcings. What verified new facts have the models told us about the operation of the climate that we did not already know?

Another way to judge the models is to see if the results of various models agree or not. Figure 4 shows the amount of clouds by latitude from a number of climate models:

Figure 4. Cloud cover of the Earth by latitude, as shown by 31 climate models, from the AMIP study (1999). Black line is the observed cloudiness by latitude.

Dr. Meier, if you think that any of those model results are evidence for the actual cloud cover by latitude, I fear that we have vastly different definitions of “evidence”. They are model results, and are not evidence of latitudinal cloud cover in even the most expansive conceivable definition of evidence. Models can be useful, but their results are not evidence of anything.

Question 9: Are the models capable of projecting climate changes for 100 years?

Based on Willis’ answer to Question 1, I’m surprised at his answer here. If the earth has a preferred temperature, which is actively maintained by the climate system, then it should be quite easy to project climate 100 years into the future. In Question 1, Willis proposed the type of well-behaved system that is well-suited for modeling.

My Comment: I see no theoretical reason that a complex chaotic system with a preferred temperature would be any simpler to model than a complex chaotic system without a preferred temperature. I have provided links in my Thermostat Hypothesis to two simple models of such a system, one by Bejan and one by Ou. However, I do not think that either of them produce evidence, or that either can project the climate a hundred years from now.

However, Willis claims that such a projection is not possible because climate must be more complex than weather. How can a more complex situation be modeled more easily and accurately than a simpler situation? Let me answer that with a couple more questions:

1. You are given the opportunity to bet on a coin flip. Heads you win a million dollars. Tails you die. You are assured that it is a completely fair and unbiased coin. Would you take the bet? I certainly wouldn’t, as much as it’d be nice to have a million dollars.

2. You are given the opportunity to bet on 10000 coin flips. If heads comes up between 4000 and 6000 times, you win a million dollars. If heads comes up less than 4000 or more than 6000 times, you die. Again, you are assured that the coin is completely fair and unbiased. Would you take this bet? I think I would.

But wait a minute? How is this possible? A single coin flip is far simpler than 10000 coin flips. …

I fear I don’t know where to start explaining the host of reasons why this doesn’t work as a metaphor for the difference between a weather model and a climate model, or as an explanation of how climate models could possibly project a hundred years out. But I’ll give it a shot.

Both weather and climate models are what are called “iterative models”. The model looks at the current state of the weather, and predicts what the weather will look like after the next time step (typically under an hour in modern models).

This type of model is very, very hard to get right, because the errors “propagate”. This means that if your calculation of the weather at one time step of the model is off a little, the next time step will likely be off a little more, and on ad infinitum. Error propagation of this type is an unavoidable feature of iterative models. It is one of the main reasons that weather models diverge from the actual weather over a very short period of time. This makes long-term forecasts very difficult.

Predicting the number of heads in 100 flips or a million flips, on the other hand, does not suffer from this problem. It is a simple and well understood statistical problem which can be solved with a single equation. In fact, the more flips, the less error you will find in the result. Ahhh, would that climate could so easily be reduced to a single equation …

Next, coin flips do not contain any variables. They are not affected by things such as humidity or temperature. It’s just the coin, period. That’s why we use them as a decision tool, because they are random, they are not dependent on variables. Weather models, by contrast, have a host of variables: temperature, humidity, barometric pressure, wind speed, wind direction, and many others. They are anything but random.

And while one coin flip has the same number of variables as a thousand coin flips (none), climate models must include a host of variables that can be neglected in weather models. These include variables like terrestrial biology, sea biology, ocean currents, variations in soil moisture, slow changes in ice cover, and lots of others. This makes climate models much more complex than weather models … and in iterative models, this means more sources of error.

Finally, both climate and weather are chaotic. This introduces a host of other problems into any attempt to model the climate or the weather.

As a result, the idea that climate models can project the climate a hundred years out because “a single coin flip is simpler than 10000 coin flips” is untrue, simplistic, and in no way a metaphor which would help us understand the problem with long-term climate model projections of the future.

Moving along, I find:

Question 13: Is the current peer-review system inadequate, and if so how can it be improved?

There is always room for improvement and Willis makes some good suggestions in this regard. Speaking only from my experience, the process works reasonably well (though not perfectly), quality papers eventually get published and bad papers that slip through the peer-review process and get published can be addressed by future papers.

My Comment: I love the idea that “quality papers eventually get published”. It just sounds so good. However, please read Ross McKitrick’s saga with his paper on Surface Temperatures,  and Bishop Hill’s post on Caspar and the Jesus Paper, before you become too enamoured of the idea that the system is self-correcting and works in the end. A review of the CRU emails in this regard is in order as well.

From my own experience, I wrote a paper explaining the problems with a study by Michael Mann that had been published in Geophysical Research Letters (GRL). His study claimed that the best way to extend a smooth (Gaussian or otherwise) to the end of a series was to pad the end of the series by reflecting it around both the x and y axes. (This results in forcing the smooth through the last point of the series, which is absolutely the last thing you want to do).

The paper was rejected by GRL because one reviewer said I was too hard on poor Mike. So I set off to re-write it.

Within a few months, Mann published a new paper in GRL on the subject, incorporating my ideas as his own. Coincidence? You be the judge … I threw up my hands, my paper never got published. I think the present peer-review system sucks. The CRU emails contain hosts of references to this kind of scientific malfeasance, stacking peer-review panels with people who will give papers an easy pass, circulating papers like mine to other scientists, blackballing journals, and pressuring editors. We know it is happening, we have their emailed confessions.

Yes, I understand that Dr. Meier’s personal experience is different, and I respect that. But only looking at his own experience is a very restricted view of the situation. The repeated refusal of many climate scientists to go outside their own experience and  honestly look at the scientific malfeasance going on in their own field is a constant source of amazement to me.

Question 14: Regarding climate, what action (if any) should we take at this point?

This is of course an economic and political question, not a scientific question, though the best scientific evidence we have can and should inform the answer. So far there isn’t any scientific evidence that refutes NH2 and we conclude that the processes that influenced climate in the past are doing so today and will continue to do so in the future. From this we conclude that humans are having an impact on climate and that this impact will become more significant in the future as we continue to increase GHGs in the atmosphere. Willis answers no and claims that the risks are too low to apply the precautionary principle. The basis for his answer, in practical terms, is his conclusion that NH2 is no longer valid because while GHGs have been a primary climate forcing throughout earth’s history, they are no longer having an impact. This could of course be true, but to me there doesn’t seem to be much evidence to support this idea. But then again, I’m a skeptic.

My Comment: First, NH2 is not falsifiable and is a straw null hypothesis. Second, I make no claim that the factors operating now did not operate in the past. I did not conclude that  “NH2 is no longer valid because while GHGs have been a primary climate forcing throughout earth’s history, they are no longer having an impact” as Dr. Meier claims, and I am mystified that my words could be misunderstood in that way.

Also, I did not say that “the risks are too low to apply the precautionary principle”. I said “I disagree with those who say that the “precautionary principle” means that we should act now. I detail my reasons for this assertion at “Climate, Caution, and Precaution”. And nothing at that link says that the issue is that the “risks are too low”, I have no clue where Dr. Meier got that claim.

Regrettably, after explaining why he thinks that I’m wrong about what action to take, Dr. Meier does not say what action (if any) he thinks we should take.

Final Conclusions, in no particular order

1. Reading Dr. Meier’s answers to the questions has been very interesting and very productive for me. It has helped to identify where the discussion goes off the rails.

2. Understanding how the guy on the other side of the table sees the situation is valuable for everyone concerned.

3. Dr. Meier’s answers were well thought out and well expressed. He obviously has considered these matters in detail, answered honestly and fully, and taken the time to lay them out clearly.

4. As I didn’t discuss most of the questions where Dr. Meier and I were in basic agreement, it likely appears that I disagreed on almost all points. This is absolutely not the case.

5. I wish that Dr. Meier had included citations for his assertions. Not having them makes it harder to discuss his ideas.

6. I sincerely hope that I have not offended Dr. Meier. I am a reformed cowboy, but despite going to the cowboy reform meetings and following the twelve steps,  sometimes the raw ranch kid shines through. I am passionate about these matters, and sometimes I overstep the bounds. I apologize for any sins of omission or sins of commission I may have committed, and I hope that Dr. Meier considers my words in the spirit of vigorous scientific debate.

7. Since the null hypothesis that the climate variations are natural has not been falsified, the AGW hypothesis is still a solution in search of a problem.

8. As I have found out more than once to my own cost, putting one’s ideas out on the web for people to find fault with is a daunting prospect, and one which may not always end well. I offer Dr. Meier my profound thanks and my respect for his courage and willingness to put his ideas on the firing line, as it is not an easy thing to do.

Advertisements

265 thoughts on “My Thanks and Comments for Dr. Walt Meier

  1. The best science is science that is fully open for change or critism. A closed science and closed mind only means that the science is incorrect somewhere and someone is afraid it may be found out.
    My own research incorporates too many areas that one category can’t follow so automatically it is seen a the old science is the best science. I have found many mistakes in the old science that is just increasing how bad our knowledge base really is.

  2. Willis,
    Again, a very impressive post. Both you and Dr. Meier are to be commended for furthering the cause of intelligent discussion and debate of the AGW issue. This series of posts will be hard to match…
    Just slightly off topic– one comment on your Earth’s Thermostat theory. What does you model say about clouds at night, and how they play into the overall energy balance? Leaving out the role of precipiation at night, just cloudiness at night is always net warming, isn’t it? Certainly in the tropics, the clouds dissipate during the night to some extent (at leat they don’t continue to build, generally) Clouds tend to trap the heat in at night, preventing escape. I guess since the role of clouds in generally is one of the areas of much discussion in climate research, and clouds can be both cooling or warming, I’m just wondering how your thermostat model accounts for night time cloudiness in terms of energy balance?

  3. I’m not greatly knowledgeable in the implications of Karl Poppers work. I think that it has limitations in areas like whether the sun will rise tomorrow.
    Willis says:
    >Dr. Meier’s NH2, on the other hand, extends into the future … how can we possibly >falsify that?
    This indicates to me that the the Poppers methodology might not be suitable for judging the veracity of climate science. Popper stated that Darwinism is not a scientific theory (Wikipedia), by the same measure perhaps climate science is not a valid Popperian science but useful for modeling nonetheless.

  4. “”As I didn’t discuss most of the questions where Dr. Meier and I were in basic agreement, it likely appears that I disagreed on almost all points. This is absolutely not the case.””
    Thank you Dr. Meier, if for no other reason, though there are many, for this quote from Willis. Where persons of reason agree to show the value of remarks and approach from the other side, I have found to almost always to be leading to positive outcomes. I find it most unfortunate when worthy conversation devolves into essentially a mud wrestling contest (with pigs, no less). Again thank you, Willis, and our host.

  5. “The increase in CO2 is due to human emissions.
    My Comment: Agreed.”
    Fossil fuel derived CO2 makes up about 2% of atmospheric CO2.
    Oceans emit and absorb 20 times as much CO2 as man.
    There are tens of thousands of chemical measurments in the early 20th C that show higher levels than currently.
    Where is the evidence that increases in atmospheric CO2 are caused by man, rather than changes in ocean temperature, circulation, some long term oscillation or some other unknown effect ?

  6. Well, I’m from Texas, 1/2 Apache cowboy, so I be direct and cowboy up,
    they have been telling one another that they are correct far to long,and got used to that.
    Time for them to learn to “cowboy up too.”

  7. In the field of climatology, Willis Eschenbach is the last defense against scientific nihilism or Lysenkoism. No one can construct a reasonable criticism of anything Eschenbach has said in this wonderful article.

  8. Well thought out post and I think you raise some interesting points. I think your point about his false dichotomy between each of your null hypotheses is right on. It’s not an either or situation, even if one is willing to accept his ‘narrative’.
    I do think there is some confusion with Dr. Meier’s example of 10000 coin flips versus one. Your final assessment of that section is
    ‘As a result, the idea that climate models can project the climate a hundred years out because “a single coin flip is simpler than 10000 coin flips” is untrue, simplistic, and in no way a metaphor which would help us understand the problem with long-term climate model projections of the future.’
    I think this misses the point. I don’t think he is saying that one coin flip is simpler than 10000. In fact, after reading his post, I gathered he was saying the exact opposite. That because we know given any amount of coin flips one should get heads half the time, so as you toss the coin more and more, it gets easier to predict how many times heads or tails will come up.
    I think your point concerning the internal variables in and iterative nature of climate models make them infinitely more complex than Dr. Meier’s example is quite valid.
    I think you get the point of his example in your argument. I just think the conclusion is a bit confusing though.
    Thanks again for the thoughtful and temperate response.

  9. …I think we need to also commend Dr. Meier’s on engaging in the discussion here. Thanks again for his time.

  10. Big difference between the loud alarmism screaming for rejection and the thoughtful exchange of points of view that deserve a discussion and make it the effort worthwhile and valuable.
    Great work!
    Thanks to all of you!

  11. Congratulations to Willis and Dr. Meier for a very interesting and respectful exchange of ideas!
    We need more of this type of scientific dialogue, and I thank Anthony and the moderators for providing this forum. Kudos to those who contribute through thoughtful comments.

  12. A very enlightening point-counterpoint! The exchange really gets to some of the fundamental questions about the climate change debate. If we (climate crisis skeptics) can hold an opponent to any of those questions long enough, I can predict where the debate will go:
    The AGW supporter will eventually be forced to retreat to the use of the precautionary principle, which is neither ‘precautionary’ nor a ‘principle’.
    This is predictable because the process of the argument is not complex or non-linear. There are no feedbacks or iterations, only a steady, linear progression of AGW supporters being forced to logically concede point after point until all they have left is the exclamation: “Hey, but what if you are wrong? We can’t afford the risk if you are wrong!” Thus they invoke the ‘precautionary principle’.
    Please note that logic does not lead to the precautionary principle. It is only called forth out of fear. We feared DDT, even though there was very little evidence that it caused harm, or ‘any’ measurable harm when used properly. Invoking the precautionary principle out of fear, we made DDT very difficult to use. This has resulted in more deaths than the Holocaust!
    Historical use of the precautionary principle has often resulted in more harm than good, making its very use a violation of the precautionary principle. Because it does not work, it is not precautionary, and because it is self contradictory, it is not a principle.
    While the AGW folks claim to be seeking a solution to a problem, they have no ability to support their scientific definition of the problem, ultimately retreating to the pronouncement of a ‘fear’. Out of fear, they call for a ‘solution’. History proves that such ‘solutions’, driven by fear alone, invariably cause much more harm than good.

  13. Fascinating. This is the kind of discussion that should have been going on for the past 30 years. Two points strike me:
    1. I’d like to see Dr. Meier’s own improved version of NH2, which probably varies somewhat from Mr. Eschenbach’s wording.
    2. Dr. Meier’s trial analogy is most apt:
    “In legal trials, it is not only hard physical evidence that is admitted, but information such as the state of mind of the defendant, motive, memories of eyewitnesses, etc. Such ‘evidence’ may not have the same veracity as hard physical evidence, such as DNA, but nonetheless it can be useful.”
    Such criminal trials require a verdict based on establishing the defendant’s guilt “beyond a reasonable doubt.” The case for AGW cannot be shown beyond reasonable doubt, and the Climategate papers (and other sources) reveal that climatologists were (1) attempting to suppress evidence of such doubt, (2) falsely denying the existence of doubt, (3) failing to include data in their own papers that might create doubt, (4) attempting to damage those who doubted and those who printed their papers, and (5) not doubt-free themselves. (“It’s a travesty.”)
    I contend that the criteria for social action to control AGW should be at least as high as those in criminal justice, and probably much, much higher, considering the mammoth economic consequences.
    To continue Dr, Meier’s analogy, remember that there is a jury empaneling process designed to prevent bias on the part of jurors. There are not even a handful of climatologists who could be empaneled in AGW matters. Too much blatant prejudice and too many conflicts of interest. I would consider Dr. Meier as an exception.

  14. On Feedback, Roy Spencer regarding his new paper (in press) comments:

    The main message of the paper is that feedbacks are, in general, not observable in the real climate system because natural variations in cloud cover obscure them. This is the cause-versus-effect issue I have been harping on for years: You cannot measure cloud FEEDBACK (temperature changes causing cloud changes) unless you can quantify and remove the effect of internal radiative FORCING (cloud changes causing temperature changes). Causation in one direction must be accounted for in order to measure causation in the other direction.

    As I recall, he previously had stated that the uncertainty in cloud measurements and consequent uncertainties variations in radiative forcing swamp any inferences from anthropogenic CO2.
    When the very foundation of cloud influences and climate feedbacks are not understood, models cannot accurately predict future climate. Per your fig. 4, when a few percent change in clouds influence dominates, having models that vary by 40% to 90% cloudiness give meaningless predictions, though they can be stated to ten significant figures.

  15. Dr A Burns (14:38:27) :
    Fossil fuel derived CO2 makes up about 2% of atmospheric CO2.

    I would love you to provide the source of the data that the statement (2%) rests on so I can help you reverberate that aspect into this discussion on climatology. The vast majority of people just assume all of the CO2 is coming totally from us humans or our influences. I know that cannot be true. Help me here, can’t seem to find the source of data to back that up.

  16. Actually I agree with Burns. The current rise in recently measured Co2 may be speck in a 1000’s years rise from natural variation. Could we label the human produced C02 to verify? Even a 12 month experiment would clear this up…

  17. Your post is a model of clarity. Dr meier’s post not as much. I wonder if he is OK with post-normal science – they don’t require falsifiable hypotheses to do their ‘business’
    Great post. Thanks

  18. I respect Dr. Meier for setting out his views here but remain concerned about his general lack of up to date climate knowledge.
    Could it be that administrative expertise acquired over many years can cause one to lose one’s grip on the cutting edge of research ?
    Could it also be the case that once one immerses oneself in administrative responsibilities for any length of time one becomes dependent on the good faith of those still in the ‘front line’ so to speak.
    Would (or could) Dr. Meier know if he was being fed a diet of misleading and politically motivated disinformation by those beneath him ?

  19. R. Gates (14:31:33)

    Willis,
    Again, a very impressive post. Both you and Dr. Meier are to be commended for furthering the cause of intelligent discussion and debate of the AGW issue. This series of posts will be hard to match…
    Just slightly off topic– one comment on your Earth’s Thermostat theory. What does you model say about clouds at night, and how they play into the overall energy balance? Leaving out the role of precipiation at night, just cloudiness at night is always net warming, isn’t it? Certainly in the tropics, the clouds dissipate during the night to some extent (at leat they don’t continue to build, generally) Clouds tend to trap the heat in at night, preventing escape. I guess since the role of clouds in generally is one of the areas of much discussion in climate research, and clouds can be both cooling or warming, I’m just wondering how your thermostat model accounts for night time cloudiness in terms of energy balance?

    I discussed this briefly in the post, viz:

    6. Enhanced night-time radiation. Unlike long-lived stratus clouds, cumulus and cumulonimbus generally die out and vanish as the night cools, leading to the typically clear skies at dawn. This allows greatly increased nighttime surface radiative cooling to space.

    w.

  20. Dr A Burns (14:38:27

    “The increase in CO2 is due to human emissions.

    My Comment: Agreed.”

    Fossil fuel derived CO2 makes up about 2% of atmospheric CO2.
    Oceans emit and absorb 20 times as much CO2 as man.
    There are tens of thousands of chemical measurments in the early 20th C that show higher levels than currently.
    Where is the evidence that increases in atmospheric CO2 are caused by man, rather than changes in ocean temperature, circulation, some long term oscillation or some other unknown effect ?

    As Dr. Meier said, there are separate lines of evidence regarding fossil-fuel CO2.
    First, the rise in atmospheric CO2 is shown by the ice core records to be of recent (last few centuries) origin. I have examined the records in question (Greenland and Vostok ice cores) and the increase is definitely recent.
    Second, there is an excellent fit (after accounting for sequestration) between the amount of CO2 emitted by human activities and the rise in atmospheric CO2. I have run the numbers on this myself, and the fit is exceptional.
    Third, the changes in the proportions of “light” and “heavy” CO2 (due to different carbon isotopes) have changed in a manner that supports the idea that the increase is due to fossil fuel rather than a change in the natural carbon cycle. I have not done the math on this myself, but the studies are quite convincing to me, and I’m a very skeptical guy.
    w.
    PS – the CO2 measurements taken in the early 20th century were taken at ground level, and often near cities. Just like modern measurements taken at ground level near cities, these are often higher than “background” CO2 measurements. These background measurements are taken from high towers at places like Barrow, Alaska, at the South Pole, and high up on Mauna Loa. These measure the level of CO2 where it is well mixed and unaffected by cities and the like.

  21. Just to pile on with a “commendable”. *This* is the quality of discussion we want to see.

  22. @ MikeA (14:35:53) :
    “I’m not greatly knowledgeable in the implications of Karl Poppers work. I think that it has limitations in areas like whether the sun will rise tomorrow.
    Willis says:
    >Dr. Meier’s NH2, on the other hand, extends into the future … how can we possibly >falsify that?
    This indicates to me that the the Poppers methodology might not be suitable for judging the veracity of climate science. Popper stated that Darwinism is not a scientific theory (Wikipedia), by the same measure perhaps climate science is not a valid Popperian science but useful for modeling nonetheless.”
    Here’s what Popper stated in his autobigraphy:
    “It is metaphysical because it is not testable. . . . For assume that we find life on Mars consisting of exactly three species of bacteria with a genetic outfit similar to that of terrestrial species. Is Darwinism refuted? By no means. We shall say that these three species were the only forms among the many mutants which were sufficiently well adjusted to survive. And we shall say the same if there is only one species (or none). Thus Darwinism does not really predict the evolution of variety. It therefore cannot really explain it. At best, it can predict the evolution of variety under “favorable conditions”. But it is hardly possible to describe in general terms what favorable conditions are—except that, in their presence, a variety of forms will emerge.
    Take “adaptation”. At first sight natural selection appears to explain it, and in a way it does, but it is hardly a scientific way. To say that a species now living is adapted to its environment is, in fact, almost tautological. Indeed we use the terms “adaptation” and “selection” in such a way that we can say that, if the species were not adapted, it would have been eliminated by natural selection. Similarly, if a species has been eliminated it must have been ill adapted to the conditions. Adaptation or fitness is defined by modern evolutionists as survival value, and can be measured by actual success in survival: there is hardly any possibility of testing a theory as feeble as this.”
    AGW theory has made a number of testable predictions, such as the existence of a hotspot in the equatorial troposphere, decreasing summer soil moisture in Asia where most of the warming has occurred, continuing increase in atmospheric methane etc. that have been falsified by observation. It behooves the AGW proponents to explain these difficulties adequately. They don’t; they claim these observations are “consistent with” the theory. Thus AGW is metaphysical, rather than hard science as an engineer, chemist, geologist etc would define their sciences.

  23. maxwell (15:05:48)

    Well thought out post and I think you raise some interesting points. I think your point about his false dichotomy between each of your null hypotheses is right on. It’s not an either or situation, even if one is willing to accept his ‘narrative’.
    I do think there is some confusion with Dr. Meier’s example of 10000 coin flips versus one. Your final assessment of that section is
    ‘As a result, the idea that climate models can project the climate a hundred years out because “a single coin flip is simpler than 10000 coin flips” is untrue, simplistic, and in no way a metaphor which would help us understand the problem with long-term climate model projections of the future.’
    I think this misses the point. I don’t think he is saying that one coin flip is simpler than 10000. In fact, after reading his post, I gathered he was saying the exact opposite. …

    The statement “a single coin flip is far simpler than 10000 coin flips” is a direct quote from Dr. Meier, so he is definitely saying that.

  24. wayne (15:24:27)

    Dr A Burns (14:38:27) :

    Fossil fuel derived CO2 makes up about 2% of atmospheric CO2.

    I would love you to provide the source of the data that the statement (2%) rests on so I can help you reverberate that aspect into this discussion on climatology. The vast majority of people just assume all of the CO2 is coming totally from us humans or our influences. I know that cannot be true. Help me here, can’t seem to find the source of data to back that up.

    Numbers for the carbon cycle are only approximate at best. Various sources give various amounts. Here’s one such, from the usual font of misinformation (Wikipedia), other estimates are not much different:
    Annual fluxes (Gtonnes C) to the atmosphere from …
    Trees and plants, 60
    Decaying vegetation, 1.6
    Soils, 60
    Ocean, 90
    Humans, 9
    Total, 220.6 gigatonnes of Carbon annually
    Human percentage of emissions, 4%
    However, this number is not really relevant to the discussion, as it demonstrates nothing about whether humans are affecting the climate.

  25. I find this to be a very well-reasoned response. Willis overlooks a few good points the first of which is that there has been no measurable change in the Ph of bodies of fresh water. Furthermore, all published studies detailing laboratory experiments investigating the effects of greatly increased CO2 content in aquaria show either no effect or a greatly beneficial effect on corals and bivalves. Ocean Acidification is the latest fall-back position in the Warmest retreat. Dr. Meier disappoints in this regard.
    The most rigorous studies to date show no significant migration of species towards higher latitudes due the the trivial amount of warming over the last century. A recent study conducted in the Italian Alps is an example: http://www.co2science.org/articles/V13/N14/B1.php .
    The NH2 hypothesis is, of course, sucker bait. In effect, it asserts that the climate system is deterministic. This is a preposterous null hypothesis.
    Dr. Meier should be commended for his civility and courage. He should also be gently scolded for knowing less about climate change research than many of the bright and inquisitive laymen who frequent this site.

  26. Impressive and extremely interresting post, Willis!
    I think this quote says it all:
    7. Since the null hypothesis that the climate variations are natural has not been falsified, the AGW hypothesis is still a solution in search of a problem.
    The theory of CO2 as a strong climate forcing is what the AGW community thinks has falsified your null hypothesis, but a theory can’t falsify anything, can it? The strongest and most conclusive support for the theory is found in the models and since, as we all seem to agree on, they can’t be used as evidence, then there are no evidence for the theory either.
    This means that your null hypothesis, Will, is not by far falsified.
    Therefore it is 100% logical to be an AGW sceptic. 🙂

  27. As the only really immediately fatal consequences would be from the Arctic Ice melting off….
    … What is the point ???
    NASA’s study showed CAP & TRADE, by “forgiving” Diesel Soot & 3rd World Coal Soot, has Poka-dotted the formerly bright white reflective ice, and with Cap & Trade of Sulphur, warms the Arctic an incredible 3 TIMES the contribution of General Global Warming, from all causes (e.g. Dr. Roy Spencer puts the Pacific Oscillation at 3/4 of Globaly-averaged warming = cutting it to just 7% of Arctic Warming –others say the PO is as little as 15% of GW , BUT …. what counts is that the “PPG’s” = the Political Phony Greens = REALLY: the “Poisons & Particulates Group” … have scammed the Media, HIJACKED the Environmental Movement in the Public’s eyes — yet ALL of the Topmost Scientists condemn their CAP & TRADE e.g. the TOP 3 on the Warming side: Hansen, “Gaia” Lovelock, Paul Crutzen (“Dr. Ozone”) , all use the word “SCAM” which alleges a CRIME — & would be sued in a trice if it were not true.
    … In short, AGW is unimportant, a fifth to a twelfth as important as the actions of the PHONY GREENS (“PPGs”) with their CAP & TRADE Scam.
    PS: Obama’s Cap & Trade, as the USA already enforces 93% soot reductions on Diesels since 1995, would NOT be as counter-productive as the Kyoto version.
    PSS : all the “official” Ice forecasters REFUSE to use El Ninos for their forecasts& as such, have NEVER predicted any of the BIG jumps more than 6 weeks ahead of time — and as such, ignore the present SUPER EL NINO’s danger of complete melt off — inducing 300 mph winds this upcoming mid-winter ? — Unlikely ? — but a HIDEOUS Risk ! — Remember the 3 AGW stalwarts above ? — 2 of the 3 , plus Obama’s own AGW Czar Holden, have been RABID for years for offsetting the SO2 reductions with 1/300 of the SO2 we have cut, squirted high up, at a cost of 6 cents per American (and contuinuing 99.7% of the SO2 cuts so far). Also there are “REAL” Greens who would also cool things, with a Seawater Spray — that costs about $1 per American (use our mothballed ships?). Now: to preserve MY LIFE, that’s a bet I’d take (and I’d do BOTH, at least THIS YEAR).
    … Forget how much of General Warming is AGW. Heed the ARCTIC WARMING.
    … So: Pay $1.06. PLEASE.

  28. Steve Goddard (16:18:16)

    The stratosphere has cooled by about 1C since the beginning of the satellite record. The volcanic glitches introduce noise, but don’t explain the overall downwards trend.
    http://wattsupwiththat.files.wordpress.com/2010/04/msu_global_stratospheric_temp_anomaly1.jpg

    True. The curiosity to me is this.
    Let’s assume for the moment that the decrease in stratospheric temperature is caused by the increasing CO2 forcing which is changing the radiation balance between the troposphere and the stratosphere. This change in radiation balance should be occurring whether or not the surface is warming.
    But the stratosphere has not cooled in more than 15 years … which makes that assumption kinda sketchy, in my opinion. That’s the oddity.
    I also, as I said above, have never seen anything demonstrating that the only reason that the stratosphere would be cooling is increased GHGs … gotta love the climate, mystery piled upon mystery.

  29. And so many critics say this isn’t a “science” blog. But then, they’re the ones who use the term “climate scientist”, so they may just be right.
    Commendable, Mr. Eschenbach, commendable. And thanks, Dr. Meier for this great exchange of ideas; how thought provoking! (Willis always does his best work when challenged.)

  30. Willis Eschenbach (16:32:25) :
    The appearance of the graph is a step function downwards after volcanic eruptions. Step functions are often indicative of some sort of metastable state which requires a catalyst to get past an activation energy. I don’t have any theories about what the mechanisms are behind it, but we might expect to see another step downwards after the next big eruption.
    http://wattsupwiththat.files.wordpress.com/2010/04/msu_global_stratospheric_temp_anomaly1.jpg

  31. I wish I had the brains to fully understand the points made by both Walter and Willis, but I’m probably 60% along the path.
    But I can’t understand why we can’t live in the real world that shows 90% of co2 levels will come from the developing world in the future and those of us in first world countries will have little influence ( 10% ) on that future outcome.
    Other than a new energy source there is nothing we can do about our dependence on fossil fuel use, so therefore we must use the tool of ADAPTATION to future CC, whether co2 proves to be a problem or not.
    BTW I don’t think it’s a problem but adaptation is a much, much cheaper alternative and we humans are very good at it.

  32. An increase in temp (caused by whatever) would cause an increase in water vapour (which is a powerful GHG) which in turn would make the stratosphere cool.
    Since we can be pretty sure that the temp has risen somewhat since 1979… Well, go figure. 😉

  33. Willis!
    Concerning coinflips….I think another metafor is relevant.The problem is that you dont know what youre flippin..f its a coin dice or a card from a deck or a spin on a wheel of fortune?

  34. So, it seems everyone agrees that predictions of future climate have the confidence range of a coin flip.
    Glad that is “settled”.
    Now, how do we stop the runaway taxation, spending and power grab, predicated on the EVIL (yet beneficial) CO2 molecule.

  35. Willis Eschenbach;
    Let’s assume for the moment that the decrease in stratospheric temperature is caused by the increasing CO2 forcing which is changing the radiation balance between the troposphere and the stratosphere. This change in radiation balance should be occurring whether or not the surface is warming.>>
    I think you are breaking your own rule here. As you have pointed out, the earth seems to have a thermostat. Until the way it works is understood, everything else is is just a bunch of unrelated observations. The liklihood that the earth thermostat is one single factor is small. There are mutiple processes that determine the over all temperature of the planet. Part of my house is heated with forced air, part with electric baseboard heaters. There is a thermostat for each, and the over all temperature of the house is not determined by either one, but a combination of the two.
    So, by extension, the stratosphere seems to have a thermostat that delivers unexpected results. How does it work? Without understanding that, surface warming is just an observation whose relationship is unknown.

  36. Mr Eschenbach you say
    However, this formulation has some serious problems. First, a null hypothesis must be capable of being falsified. My null hypothesis (NH1) could be falsified easily, by a showing that measurements of the modern climate are outside the historical values

    Dr. Meier’s NH2, on the other hand, extends into the future … how can we possibly falsify that?
    .
    How can NH1easily be falsified. Any proxy records have been consigned to the trash bin by skeptics. There is no time machine therefore there is no historical record before CET. (1659) – and this shows no warm periods greater than today.
    NH2 will be falsifiable in the future NH1 cannot be because you do not believe the proxies.

  37. I’m confused, and pardon me if someone else made the same point, but, given that the total carbon sink accepts (uses) both man made and “natural” carbon equally, won’t man made carbon ALWAYS be increasing in the atmosphere until manmade carbon reaches 50% of the total carbon?

  38. The peer review system does need reconsideration. In particular, since it is now easy to publish material online it would make much more sense for papers to be published then reviewed, rather than reviewed and then published. However any attempt to review the peer review system at this time is complicated by the existence of a substantial army camped out at the gates of science trying to force an entry.
    I am referring to the problem posed by “creation scientists”. These people are relentlessly trying to force entry into science classrooms in order to pervert the science curriculum and teach religion dressed up as science. Indeed their ultimate objective is clearly the suppression of one aspect of science – evolution – which they believe is incompatible with their religious conviction. The peer review system is part of the wall protecting us from these barbarians.
    It is very hard to demolish walls and open up your community in the middle of a siege.

  39. Harry Lu:

    Any proxy records have been consigned to the trash bin by skeptics.

    Any records? Which records are those?

  40. Willis, I tried to post this at the destination that you linked to ” The Thermostat Hypothesis.”, but the gods have an interdict on posting new info there. Thus, it’s here!
    Willis Eschenbach (22:15:07) : 26/08/2009.
    I took a link here from your guest post of ‘My Thanks and Comments for Dr. Walt Meier’ and I’m impressed with the way that you address and explain the resultant effect of the Clausius Clapyron relationship in this post.
    However, I also notice, in your last response to this post, that you question the temperature change in the Vostok ice record with regard to CO2 propensity. So do I, but, probably not for the same reason.
    We now know that microbial life continues to exist within ice following its compassion from precipitated snow. IMHO, this calls to question the validity of ice core gas analysis, as this may well be affected by the surviving microbes within the ice over the time-scale of the ice’s survival.
    Best regards, suricat.

  41. Willis,
    I have to admit, you’ve put far more effort into Dr. Meier’s response than I would have.
    I’d have simply said his response is a regurgitation of the same old refuted taking points which the government has been paying its scientists to repeat ad nauseum for decades.
    Though I have to admit I did enjoy watching someone playing the roll of Uncle Sam’s sock puppet trying to paint himself as a skeptic. 🙂

  42. IMHO this is the best of Willis’ articles to date, and that’s saying something.
    I also appreciate Dr Meier for being willing to take a position, which is far more than most on the AGW side will do.
    Michael Mann, Al Gore and the others only express opinions and grant interviews that are tightly controlled, and where they don’t have to answer specific questions from those skeptical of their hypothesis. So kudos to Dr Meier for stepping up to the plate.
    I hope Dr Meier continues this debate with Willis. I’m happy to sit back and listen to both sides defend their positions. Although I have to say that Willis scored all the pertinent points in this debate, Dr Meier can try to even the score if he’s willing. I look forward to his response. This is too good to stop here.
    ***
    wayne (15:24:27),
    I’ve posted this IPCC information about human CO2 contributions several times. Maybe it’s what you’re looking for: click

  43. I’m trying not to get verklempt here. This is the kind of dialogue I wish there was a lot more of on global warming and its causes. Well done, gentlemen. (Walt and Willis).

  44. Re: Willis Eschenbach (Apr 10 16:32),
    On stratospheric cooling, this recent paper delves into it. Roy Spencer postulated volcanoes have a long term effect on the stratosphere, but the authors in that paper don’t seem to be interested. Instead, they say ozone may possibly be overriding GHG (CO2).
    http://www.jstage.jst.go.jp/article/sola/5/0/53/_pdf
    I don’t know, its like every other AGW component; when observations don’t agree with the hypothesis, there is always a new explanation ready and waiting. Can anything AGW be falsified?

  45. The absorption of CO2 is posited to happen in 100 years in current Models.
    If this isn’t true, and it doesn’t seem to be, the doubling in 100 years will not happen, possibly it never will
    Here is a list of 36 peer reviewed studies which come up with a value of around 10 years. Where do the modelers get the 100 years ? Pure politics.
    I googled 6 of them at random to see if they checked out. They seem genuine.
    http://c3headlines.typepad.com/.a/6a010536b58035970c0120a5e507c9970c-pi
    As you know CO2 absorption is like filling a bathtub with a drain open. It is negative feedback because the higher the water gets the more pressure and the faster it drains. It might fill slowly or come to an equilibrium point where no further filling happens.
    With a faster absorption time the IPCC predictions of a doubling of CO2 in 100 years looks less possible. I think they posit geometrically rising CO2 which is nonsense too.
    I never see references to this hole in the CAGW arguments, why not ?

  46. suricat (17:51:18) :
    Mistake: “We now know that microbial life continues to exist within ice following its compassion from precipitated snow.”
    Should read: “We now know that microbial life continues to exist within ice following its compaction from precipitated snow.”
    Sorry!
    Best regards, suricat.

  47. jim (17:27:41) :
    How much fossil fuel reserves do you think there are? Right now, only three percent of CO2 emissions are man made. We would have to increase emissions by over 30X to reach 50% of the total.

  48. The appearance of the graph is a step function downwards after volcanic eruptions. Step functions are often indicative of some sort of metastable state which requires a catalyst to get past an activation energy.
    The mechanism is well understood, WMO chapter 4 Heterogeneous chemistry and Volcanics
    eg Volcanic eruptions and global ozone dynamics
    Abstract
    It is well known that heterogeneous chemistry plays a very important role in global ozone dynamics. A survey of new observations and numerical modelling results has been made to analyse the potential impact of volcanic eruptions on the stratospheric ozone layer in the context of the heterogeneous chemical reactions on surfaces of volcanically produced stratospheric aerosol particles. An important aspect of the problem is that volcanically induced ozone changes may lead to a substantial impact on the radiative forcing for the surface-troposphere system and, subsequently, on climate.

    http://www.informaworld.com/smpp/content~content=a777907097&db=all
    Further the WMO assessment is rather succinct on this aspect eg WMO 2006
    Ozone heats the stratosphere by absorbing incoming solar energy and outgoing infrared radiation from the Earth’s surface. A significant component of the observed stratospheric cooling (-0.17 °C/decade) can be attributed to ozone depletion, rather than being solely a radiative effect of climate change. Therefore, if ozone amounts were to increase in the future, this would tend to warm the stratosphere, diminishing the future cooling there due to increasinggreenhouse gases (GHG). The warming effect would aid further recovery of the ozone layer in polar regions where heterogeneous chemistry on ice crystals dominates the ozone loss processes,
    Subsequently this line of “evidence” ( Meier’s co2 cooling) is neither robust nor rigorous and is awarded a d-

  49. Neville (16:48:31)

    I wish I had the brains to fully understand the points made by both Walter and Willis, but I’m probably 60% along the path.
    But I can’t understand why we can’t live in the real world that shows 90% of co2 levels will come from the developing world in the future and those of us in first world countries will have little influence ( 10% ) on that future outcome.

    I agree, see my post here on the subject.

  50. Willis,
    thanks for replying to my comment. You are right that you cited a statement from Dr. Meier’s post. However, you forgot the rest of the quote.
    Dr. Meier states,
    ‘A single coin flip is far simpler than 10000 coin flips. The answer of course is that what is complex and very uncertain on the small scale can actually be predictable within fairly narrow uncertainty bounds at larger scales. To try to predict the outcome of a single coin flip beyond 50% uncertainty, you would need to model: the initial force of the flip, the precise air conditions (density, etc.), along with a host of other things far too complex to do reasonably because, like the weather, there are many factors and their interactions are too complex. However, none of this information is really needed for the 10000 toss case because the influence of these factors tend to cancel each other out over the 10000 tosses and you’re left with a probabilistic question that is relatively easy to model.’
    So the point is that a single coin is NOT easier to model and therefore, not simpler in the end. I don’t think that this fact really negates your argument in any way. I think your point that error propagate very quickly in climate models, versus a variable-less model for coin flips is quite valid in this context. I also think Dr. Meier’s example suffers as much from oversimplification as any I have seen in some time.
    I just think that it would fruitful for other readers who have also read Dr. Meier’s post to clear up your statement because, based on the rest of the quote containing that which you have cited, Dr. Meier is definitely not making the point that a single coin flip is simpler than 10000. He is making the opposite point.
    Thanks for your time.

  51. Walt Meier: 2. Carbon dioxide is a greenhouse gas. This is has been also been known for well over 100 years. There are other greenhouse gases, e.g., methane, nitrous oxide, ozone, but carbon dioxide is the most widespread and longest-lived in the atmosphere so it is more relevant for long-term climate change.
    My (Willis) Comment: Agreed.
    What?? No comment about not including water vapour in that list & how much more of a GHG it is as opposed to those other trace gases as well as the result of the water vapour transforming to a liquid/solid (cloud) & it’s effects on the temperature balance of the atmosphere? Hmmm…

  52. netdr (18:01:35)

    The absorption of CO2 is posited to happen in 100 years in current Models.
    If this isn’t true, and it doesn’t seem to be, the doubling in 100 years will not happen, possibly it never will
    Here is a list of 36 peer reviewed studies which come up with a value of around 10 years. Where do the modelers get the 100 years ? Pure politics.

    netdr, you are conflating two times here. One is the “residence time”, which is the amount of time a typical CO2 molecule spends in the atmosphere before getting absorbed by something else (plant, ocean, etc.). This can be calculated in a couple of ways, and is on the order of 8 years or so.
    The other is the “e-folding” time or the “half-life”, which is the amount of time that it takes for the concentration of a “pulse” of CO2 emitted into the atmosphere to decay to half (or to 1/e) of its original value. It is much more difficult to estimate. The IPCC, using the “Bern Carbon Model”, estimates this to be over 100 years. I calculate a much lower number, around 35 years or so. This is supported by the work of Jacobsen.
    There is a lot of confusion over this issue, you’re not the only one to confuse the two very different measurements.
    w.

  53. Willis Eschenbach (16:24:21) :
    Thanks Willis, a close estimate, two to four percent will do but explicit source beside Wikipedia would help.
    By the way, have enjoyed your interactions posted of late. Keep it coming. However, your problem others are bound to comment on are you not forcing the discussion to also include the big gorilla GHG, water molecules. With that exclusion I must disagree with much above from both you and Walt.
    For instance:
    1. Increasing concentrations of CO2 and other GHGs in the atmosphere: Yes, GHGs are increasing. However, this says nothing either way about NH2.
    To me, his statement and your agreement can only hold if you also include water vapor. The total GHGs in the atmosphere may not have been increasing. See my point.
    But more of my arguments came wait till later, this current post addresses yourself and Walt’s discussion on null theories.

  54. Smokey (17:54:08) :
    I think it’s info overload! 😉
    I do remember seeing that brown chart before. Thank you so much, that’s exactly what I needed.

  55. On sea level rise: The sea level rises with temp because the density goes down. But the ocean basins are not bounded by vertical walls, but by nearly horizontal beaches in many areas. So, an accelerating increase in volume per gram need not produce an accelerating increase in sea level. The water spreads out, it does not just pile up. (Glacier melting may add to the rate, but I don’t think this has been a major factor, although that could change in the future.)

  56. In my essay, “Precautionary Principle — Philosophical Implications
    Or “to many idiots, to few tigers”” (at: retreadresources.com/blog)
    I tackle the question. How well, is for others to judge. “…I still see many problems with all this. What most concerns me is the attempt to justify the reliance on ideology to inform political and economic processes, by demanding that science do what it can not, prove something. The framers of this ideology make bold unproven and undefined, statements about irrevocable harm, yet push the owns on others to “prove” that is not the case. They do this by making assumptions and developing questionable models. Just like AGW, it must work the other way. Gentlemen if we don’t know then best we get busy and find out instead of speculate; but if you desire the scientific method to provide answers, you best be prepared not to get the one you may be expecting.”

  57. I find this respectful discussion between proponents of different views very informative. I would really enjoy it if Anthony would host “debates” here with invited guest debaters.

  58. wayne (18:55:57)

    Willis Eschenbach (16:24:21) :
    Thanks Willis, a close estimate, two to four percent will do but explicit source beside Wikipedia would help.

    Wayne, check my post again, I put in a source beside Wikipedia.

  59. Meier’s null hypothesis is falsifiable without the future part.
    “NH2: The factors that controlled earth’s climate in the past are the same factors that control it today and will continue to do so into the future”
    All they need to falsify it is to show that there are factors today that are affecting climate that were not affecting it in the past.
    Thus it is neither a strawman nor non falsifiable
    Not an impossible task and time will tell.

  60. That which is true is provable.
    That which is not provable is false.
    All unprovable assertions are self contradictory.
    This is a fundamental epistemological axiom, duh.
    The class of entities which are designed to be unprovable are called ‘mysteries’.
    They are used for entertainment and beguilement.
    You will be able to prove heaven is everlasting life – when you’re dead.
    This is another use of mysteries – fraud.
    Post normal science is the explicitly the art of fraud. It’s hallmark is the negation of truth as an absolute. It’s purpose is the degradation of critical thinking in order to substitute authority for reason.
    Meier is a well schooled, professionally slick fraud.

  61. http://retreadresources.com/blog Dennis states “”What most concerns me is the attempt to justify the reliance on ideology to inform political and economic processes, by demanding that science do what it can not, prove something.”” This is not how risk analysis is atually done. The precautionary principle is a polemic to avoid the basic assumptions (read empowerment) of risk assessment. In this structures approach, one determines a matrix. At one end are the cheap, easy, effective, known to work solutions. At the other end are the expensive, hard, ineffective, unknown solutions. The precautionary principle is an attempt to undermine the accepted hierarchy of risk assessment. This is done because the basic conclusion of “don’t know” is “don’t know.” The precautionary principle proponents maintain that the “don’t know” is perhaps that if we know more than we think, the don’t know doesn’t apply, an unproveable a priori statement. This is a fallacy (either you know or don’t) or it is a non-sequitor “we will know in the future when we are shown to be correct even though we don’t know right now.” And note “we could be wrong but it has no costs or bearing on the discussion. ” Even though they (PP advocates) then spend all this time and effort proclaiming that the costs are acceptable.
    One must ask “acceptable top whom?”

  62. Willis – thanks for your post.
    I admire the thoroughness and thoughfullness of your analysis.
    I have not read through all the comments as I am busy with my own very small research on changes in reported Australian temperature, which is proving to be very interesting and informative.
    I hope to contact you personally when I have finished because you have an interest in that subject.
    In the meanwhile, congratulations.

  63. Perhaps it is time to clarify something:
    Insulators do not warm. They hold warmth.
    There is a huge difference.
    As to the Hypothesis of Dr. Meier, I think it must be pointed out very clearly that he is making misleading claims in his hypothesis.
    I would offer this as the basis to work from:
    CO2 is causing a world wide climate catastrophe.

  64. Question 13: Is the current peer-review system inadequate, and if so how can it be improved?
    The peer-review system has been blind-sided by advocacy. In the good old days, N% of the published papers were junk or unimportant (there is a consensus on this, but perhaps not on the value of N). They were in general ignored. Why bother refuting junk?
    There is still a percentage of junk papers published. Who knows, maybe it is even higher, but now there are also press releases to get the media attention, and in the climatology field, anything that supports AGW is given lots of publicity. The “peer-reviewed” brand is sold as a certificate of truth. I wish.

  65. OMG you used the handshake image—think Animal House parade. Please promise me you won’t use a jigsaw piece for another thread. Please/

  66. Mike (19:25:00)

    On sea level rise: The sea level rises with temp because the density goes down. But the ocean basins are not bounded by vertical walls, but by nearly horizontal beaches in many areas. So, an accelerating increase in volume per gram need not produce an accelerating increase in sea level. The water spreads out, it does not just pile up. (Glacier melting may add to the rate, but I don’t think this has been a major factor, although that could change in the future.)

    Although what you say is kinda true (most coastlines are not “nearly horizontal”), if you run the numbers you will see that the change in ocean area is trivial. I’ve never done the exercise, but I’m a seaman, and the sea is huge. Let me get some numbers … OK, here we go.
    Per the CIA World Factbook, the total coastline of the planet is ~ 3.6E+8 metres. The surface area of the ocean is ~ 3.6E+14 sq. m.
    If we added say a centimetre of water to the ocean and the sides of the ocean were perfectly vertical, we would of course get an ocean that is 1 cm deeper.
    If the sides sloped at say 1:100 (one vertical to 100 horizontal, most coastlines are nowhere near that flat), the surface area would increase by the length of the coastline times 1 metre, or 3.6E+8 square metres. This is an increase in surface area of 0.000099%, so including the diagonal, instead of one centimetre deeper, it would be 0.9999995 cm deeper …
    So for all practical purposes, the slope of the coastlines can be neglected for sea surface height excursions.
    w.

  67. Ian H. (17:41:43)
    I am not a believer in “creation science”. However, the best way to respond to the claims is to debate them, not block them. The logic and facts of ideas are best supported by airing them for all to see. There were other ideas previously given that the main proponents thought were wrong such as the continental plate concept. If the idea was blocked because it was thought wrong, the correct solution would not have come out. I am confident that the “creation” theories can’t overcome the facts, but to block any ideas from light is a mistake. Keep in mind that in fact the “skeptics” ideas are attempting to be blocked. Since I am convinced they are more likely right, think what would happen if they were blocked from the debate due to a majority vote.

  68. bob (20:04:20)

    Meier’s null hypothesis is falsifiable without the future part.
    “NH2: The factors that controlled earth’s climate in the past are the same factors that control it today and will continue to do so into the future”
    All they need to falsify it is to show that there are factors today that are affecting climate that were not affecting it in the past.
    Thus it is neither a strawman nor non falsifiable
    Not an impossible task and time will tell.

    Thanks for your comment, Bob, I had considered that.
    If we knew what the factors are that control the climate, it would be trivially easy. But since that’s what the debate is about (e.g. does CO2 control the global temperature?), obviously we don’t know what those factors are … so how can we say if they were present in the past and absent now?
    Additionally, if they are absent now, how could we show that they are present in the past? If they were absent now, how would we even know that they existed a million years ago, much less we show that they affected the temperature then?
    And if they are present now but absent in the past, how could we show that?
    So falsifying that null hypothesis is theoretically possible, I suppose, but practically, no way with our current state of knowledge.
    Finally, if that is the null hypothesis, the alternate hypothesis is that the factors that affect the climate now are not the same ones that affected it in the past. I know of no one making that claim, that the present climate runs on different principles than the past climate. So it is a null hypothesis for an argument that no one is making … which is why it is a “straw man” null hypothesis.

  69. Ron House (18:11:00)

    This makes climate models much more complex than climate models
    Shouldn’t that be “weather” models?

    Thanks, fixed.

  70. Like this discussion and prefer to look at the hypothesises.
    Will’s NH1 is compelling to me: “The climate variations are natural.”
    Period. As long as this is not falsified, there is no reason to act.
    (Never touch a running system!)
    “bob” has now suggested to cut the future tail of Meier’s NH2, to make it handier:
    “The factors that controlled earth’s climate in the past are the same factors that control it today …”
    As CO2 is what we watch, has been one of the factors for earth’s climate, this hypothesis is true so far.
    We now have to find out If CO2 as one factor
    – has been higher in the past, without rising temperatures or
    – has been lower in the past, with rising temperatures,
    then CO2 can be excluded as a controlling/governing factor.
    Think this can be confirmed, but don’t know where to find it.
    If yes, than NH2 even confirms NH1, is not an alternative.
    Makes sense?

  71. JeffK (18:41:27)

    Walt Meier:

    2. Carbon dioxide is a greenhouse gas. This is has been also been known for well over 100 years. There are other greenhouse gases, e.g., methane, nitrous oxide, ozone, but carbon dioxide is the most widespread and longest-lived in the atmosphere so it is more relevant for long-term climate change.
    My (Willis) Comment: Agreed.

    What?? No comment about not including water vapour in that list & how much more of a GHG it is as opposed to those other trace gases as well as the result of the water vapour transforming to a liquid/solid (cloud) & it’s effects on the temperature balance of the atmosphere? Hmmm…

    JeffK, thanks. I considered it, but I didn’t want to get sidetracked. Water vapor (rightly or wrongly) is generally considered a feedback rather than a forcing. In addition (as you point out) the question of the effects of water vapor are hugely complex (evaporation, condensation, sublimation, transpiration, deposition, clouds, hydrometeors, GHG potential). I try to choose my battles, and I limited the scope of my discussion mostly to what Dr. Meier had said rather than introduce new and extensive topics.

  72. Steve Goddard said (16:18:16) :
    “The stratosphere has cooled by about 1C since the beginning of the satellite record. The volcanic glitches introduce noise, but don’t explain the overall downwards trend.”
    ———-
    But AGWT does explain stratospheric cooling quite well. Place your hand on the outside of your nice warm blanket on a cold winter night and then it’s quite obvious…

  73. Sean Peake (20:57:41)

    OMG you used the handshake image—think Animal House parade. Please promise me you won’t use a jigsaw piece for another thread. Please/

    Well, if I knew what the “Animal House parade” was, I suppose that would make sense … and I also don’t understand what a “jigsaw piece” has to do with the image. But then I was born yesterday, what do I know?

  74. Without the scientific training of many of you I read Dr Meier’s letter, it seemed to be written by 2 people, one scientist and one conflicting supporter of AGW, it looked like a chummie whitewash!

  75. Re: JeffK (Apr 10 18:41),
    I agree. There is reason to claim that H2O should be treated differently, but its vapor is quite clearly a GHG and failing to mention it in a list of GHGs in the atmosphere just renders Dr. Meier’s point worthless. Even assuming we consider the vapor a (positive) feedback, the other two phases of H2O can’t be so considered. Ice and snow have to be considered negative feed-backs and liquids in the atmosphere (aka clouds) are not well quantified, but probably are negative feed-backs.
    The total H2O contribution to world temperature is almost certainly the equivalent of a thermostat from a priori logic. A planet with a basically stable climate, like the earth, is going to settle into a temperature well almost all the time. Any deviation, positive or negative, is going to result in a negative feed-back to the deviation. QED. [sure there’s a couple of assumptions there, but they’re a lot more believable than their opposites.

  76. Re: Willis Eschenbach (Apr 10 21:25),
    I expected that was what you were trying to do, but given the negative feedback Dr. Meier got on skipping water vapor as a GHG, you surely didn’t think you’d fail to get dinked for agreeing with him, did you? A simple sentence like, “Water vapor is also a GHG, but we’ll ignore it for now.”

  77. R. Gates (21:25:30) :
    “But AGWT [Anthropogenic Global Warming Thermogeddon] does explain stratospheric cooling quite well. Place your hand on the outside of your nice warm blanket on a cold winter night and then it’s quite obvious…”
    Obvious how? That the warm blanket radiates heat? That’s a real good ‘explanation’ of stratospheric cooling.

  78. Willis Eschenbach;
    I know of no one making that claim, that the present climate runs on different principles than the past climate.>>
    Sure you do. There’s these dudes with tree ring studies who discarded 50 years or so of data because “they suddenly stopped tracking temperature”. They don’t know why, but even though they don’t, they are quite certain that what ever it is they don’t know is causing it now, they “know” it never happened before.

  79. Willis (and anna V): on the topic of environmentalism and stewardship:
    Mr Higgins asks the question
    “My starting point was ‘how do we create a duty of care to the planet, a pre-emptive obligation to not harm the planet?'”
    Sounds reasonable? Let us see what that leads us to:
    http://www.guardian.co.uk/environment/2010/apr/09/ecocide-crime-genocide-un-environmental-damage
    More details here:
    http://www.thisisecocide.com/
    Apologies if this has been discussed.

  80. Question 9 under ‘My Comment’- error?
    “This makes climate models much more complex than climate models…”
    Great points by both sides- wonderful reading. Thanks to both.

  81. Smokey (21:45:02) :
    R. Gates (21:25:30) :
    “But AGWT [Anthropogenic Global Warming Thermogeddon] does explain stratospheric cooling quite well. Place your hand on the outside of your nice warm blanket on a cold winter night and then it’s quite obvious…”>>
    Obvious how? That the warm blanket radiates heat? That’s a real good ‘explanation’ of stratospheric cooling>>
    When you wrap insulation around something with a heat source, there is a temporary increase in temperature until a new steady state is achieved. If you measure the w/m2 being radiated out before the insulation was installed, and the watts/m2 radiated after the insulation was installed and the new steady state had been achieved, they would be the same. There is only a difference in the w/m2 radiated for that brief period of time when the insulation has been installed but the new equilibrium temperature has not yet been achieved.
    Injecting additional CO2 into the Troposhere would in theory reduce the amount of LW going through to to Stratosphere, but only for that brief period when a new equilibrium temperature was being achieved. Then it would be the same as before. I read a couple of articles recently showing that the amount of water vapour in the Stratosphere has dropped, and being a greenhouse gas, stratosphere temps have been heavily influenced by that. So the question is what process is driving the drop in water vapour, of particular insterest given that the postitive feedback from CO2 was supposed to be additional water vapour in the troposhpere at least.

  82. Dave Dardinger (21:44:00)

    Re: Willis Eschenbach (Apr 10 21:25),
    I expected that was what you were trying to do, but given the negative feedback Dr. Meier got on skipping water vapor as a GHG, you surely didn’t think you’d fail to get dinked for agreeing with him, did you? A simple sentence like, “Water vapor is also a GHG, but we’ll ignore it for now.”

    Dave, thanks. I get “dinked” no matter what I do, either for sins of omission as in this case, or for sins of commission, so I just do the best I can and trust in the commenters such as yourself to rectify my errors …

  83. davidmhoffer (21:45:51) :

    Willis Eschenbach;

    I know of no one making that claim, that the present climate runs on different principles than the past climate.>>

    Sure you do. There’s these dudes with tree ring studies who discarded 50 years or so of data because “they suddenly stopped tracking temperature”. They don’t know why, but even though they don’t, they are quite certain that what ever it is they don’t know is causing it now, they “know” it never happened before.

    Very good, I busted out laughing … well played.

  84. Willis Eschenbach (19:53:26) :
    Thanks, I got the info.
    Willis, I always try to got out of my way to be kind and congenial initially with any person, but I won’t be made a fool by someone and you shouldn’t either. I just wrote much that I am not going to post here, I just hate to see someone like you who I consider a friend to be jerked around. Re-read his null hypothesis two very carefully. Since it is miles from anything you would call science (absolutely no direct tie to the matter at hand) I see it very darkly as a direct slap to you and therefore everyone who enjoys WUWT’s openness. (Hint: he feels he has totally falsified it himself therefore all he said of AGW is proven true as if by a proper scientific papers. And what is the new ‘factor’, co2 spewing modern mankind) I had to read the previous post answering your questions twice before I felt the slap.

  85. The precautionary principle can be used to promote practically any type of action or inaction. For example, suppose the precautionary principle is applied to the world’s economy instead of the world’s climate. Then economists could warn — although they cannot of course be sure — that trying to reduce significantly mankind’s output of carbon dioxide might well lead to massive, impossible-to-fix disruption causing the death and impoverishment of hundreds of millions of innocent people. (Remember how just trying to increase the use of ethanol in the US drove up the price of grain worldwide, causing food shortages in many poor countries.) Therefore, by the precautionary principle, it is our duty not to hinder the world economy, not to interfere with mankind’s emission of carbon dioxide in any way. Indeed, the precautionary principle is really just a fancy way of saying “I’m scared of this, let’s do something about it.” To which, again using the precautionary principle, there is always the obvious reply “I’m scared to do something about it, let’s do nothing.”

  86. Willis,
    Your response to Meiers is excellent. I hope he takes the time to consider it and replies to your criticisms.
    On a trivial point, thanks for using the term “begs the question” correctly. Too many people are currently using the phrase to mean “prompts the question” rather than its true meaning – “Begging the question (or petitio principii, ‘assuming the initial point’) is a logical fallacy in which the proposition to be proved is assumed implicitly or explicitly in the premise.”
    Sadly I rarely hear it used correctly anymore.
    Sorry for the pedantic distraction from the discussion of your impressive deconstruction of Dr. Meier’s response.

  87. Willis,
    the coin flips might be, looked at like this,
    weather is one coin flipped once, climate is that coin plus 9,999 more in a pile,
    now, what will the pile look like? will it be 40/60? it may all be tails showing with the heads buried in the bottom.
    the weather/drivers add up but at random force by random events.
    so, he would be dead if i said we would use 10,000 coins and count the pile, faces showing!
    Tim L

  88. wayne (15:24:27) :
    >>Dr A Burns (14:38:27) :
    >>Fossil fuel derived CO2 makes up about 2% of atmospheric CO2.
    >I would love you to provide the source of the data that the statement (2%) rests
    Here’s one paper on the topic http://folk.uio.no/tomvs/esef/ESEF3VO2.htm which suggests a maximum of 4% fossil fuel derived CO2 in the atmosphere. 1% to 4% seems the range, with 2% being a likely figure.
    Willis Eschenbach (16:01:32) :
    >First, the rise in atmospheric CO2 is shown by the ice core records to be of recent (last few centuries) origin.
    I assume you are familiar with the ice core expert Jaworowski’s work, such as this paper where he points out the fallibility of ice core records :
    http://www.warwickhughes.com/icecore/zjmar07.pdf
    “More than a decade ago, it was demonstrated that these
    four basic assumptions are invalid, that the ice cores cannot
    be regarded as a closed system, and that low pre-industrial
    concentrations of CO2, and of other trace greenhouse gases,
    are an artifact, caused by more than 20 physical-chemical
    processes operating in situ in the polar snow and ice, ”
    >Second, there is an excellent fit …
    A correlation does not imply causation.
    >Third, the changes in the proportions of “light” and “heavy” CO2 …
    Carbon isotope analysis is the basis of the first paper I quoted.
    >CO2 measurements taken in the early 20th century were taken at ground level, and often near cities …
    Tens of thousands of such measurement were not. Alarmists have turned their backs on these measurements without justification, simply because they did not fit alarmist proaganda. Like Callendar, they cherry pick analyses that give the results desired. Here’s a detailed analysis:
    http://www.biomind.de/nogreenhouse/daten/EE%2018-2_Beck.pdf
    The drivers for ocean circulation are as poorly understood as the atmosphere, Different regions of the ocean absorb and desorb CO2 at different times of the year. Natural changes in these could easily account for recent increases in atmospheric CO2
    Atmospheric CO2 levels have been 7 times as high as at present, since mammals walked the earth. Recent increases are trivial by comparison and could easily be caused by the same drivers as in the past. There is no clear evidence that man has caused recent increases.

  89. Maxwell 15:05:48 says “That because we know given any amount of coin flips one should get heads half the time, so as you toss the coin more and more, it gets easier to predict how many times heads or tails will come up.”
    Wrong.
    At each new flip of the coin, the coin has no memory of its history. With a perfect coin, the probability on every flip is the same, 0.5.

  90. @ Aargh. (20:32:54) :
    “That which is true is provable.
    That which is not provable is false.
    All unprovable assertions are self contradictory.”
    Google “Gödel’s incompleteness theorems”, or read some basic philosophy of logic…

  91. davidmhoffer (21:45:51) :
    Willis Eschenbach;
    I know of no one making that claim, that the present climate runs on different principles than the past climate.>>
    Sure you do. There’s these dudes with tree ring studies who discarded 50 years or so of data because “they suddenly stopped tracking temperature”. They don’t know why, but even though they don’t, they are quite certain that what ever it is they don’t know is causing it now, they “know” it never happened before.
    ====
    Who knows what that means?

  92. As an earth scientist I have been aghast at some of the schenanigans played in the climate science game to date.
    I agree wholeheartedly with Willis about the difference between real observations and models – the first being evidence (however difficult it might be to interpret them) and the second being guesses/predictions. The predictions may get better when we can adjust models on the basis of real observations of their accuracy, but this will take time and evenso they remain predictions.
    I would like to commend both Mr Escherbach and Dr Meier on this exchange – it is wonderful to see such robust, but respectful debate and hope we can see more in the future. I would particularly like to thank Dr Meier for venturing into what could have been seen as hostile territory.
    cheers
    Theresa

  93. “Some scientists see CO2 as a second order forcing”
    Thanks for mentioning that. I felt nobody was listening.

  94. Well, I have a new null hypothesis, I’ll call it NH3.
    NH3: The factors that controlled earth’s climate models in the past are the same factors that control them today and will continue to do so into the future
    Unfortunately, I have already falsified the null hypothesis NH3, viz (emphasis mine):

    (PhysOrg.com) — Two very different forms of climate-influencing particles do not behave as expected, according to scientists at Pacific Northwest National Laboratory and Imre Consulting. The team’s measurements show that when the particles mix, they create new, layered particles, not the expected mixtures. The team also found that one form adsorbs onto the other, creating bigger particles that do not behave as predicted in climate models.

    Ah, well … it was a good null hypothesis while it lasted …

  95. This coin tossing stuff confuses me.
    Everyone knows that every time you toss a coin, the coin has no memory of last time result. Therefore next result is random, and you get a gaussian distribution. (If all else is perfectly ideal)
    Everyone knows that a model iteration uses the result of last iteration as input to the next.
    If this is the two cases that is compared, I dont see the point.
    Except of course as a proof that you cannot use many model-runs for anything. They will just aquire different errors as they process forwards, all in different directions, impossible to say which one is correct.

  96. “As a result, the idea that climate models can project the climate a hundred years out because “a single coin flip is simpler than 10000 coin flips” is untrue, simplistic, and in no way a metaphor which would help us understand the problem with long-term climate model projections of the future.”
    ——————–
    I think the coin-flip analogy was supposed to show that you can be more confident in the outcome of a 1,000 flips than a few flips. Even though the odds are against two successive heads or tails (only 0.25), it happens frequently. The odds against 1,000 successive heads or tails, however, are virtually nil.
    Similarly, you can be more confident in a projected trend line for a large number of years(e.g. 2010 – 2050) than you can be for an interpolated value for any individual year within that period. You can appreciate this by plotting a regression line through actual historical temperature records for a like number of years. You will find that line does not describe year to year changes very well but does describe the long-term trend.

  97. Thank you Willis for your ‘model’ response concerning the limitations of models as substitute evidence for data. And no evidence of the cowboy in your elegant and courteous argument.

  98. Willis Eschenbach (16:01:32) :
    First, the rise in atmospheric CO2 is shown by the ice core records to be of recent (last few centuries) origin. I have examined the records in question (Greenland and Vostok ice cores) and the increase is definitely recent.

    When you “examined the records in question,” what did they have to say about the quantitative losses of carbon dioxide in the ice core samples due to metabolization by microbial lifeforms?
    If the “records in question” are in such high agreement without taking into account the losses of carbon dioxide due to confounding factors such as metabolization by microbial lifeforms, improper sampling methods, and more; it would appear the true values would then falsify the claimed accuracy of the ice core reports.
    Personally, the ice core reports appear to be incorrect for a variety of reasons, ranging from a failure to account for confounding factors to inconsistency with other lines of evidence such as stomatol frequencies in plant life. I would encourage a sharpening of your skeptic’s razor with respect to the validity of the ice core analyses.

    Price, P. Buford; Sowers, Todd. Temperature dependence of metabolic rates for microbial growth, maintenance, and survival. Proc Natl Acad Sci U S A, v.101(13); 2004 March 30; 101(13): 4631–4636.
    doi: 10.1073/pnas.0400522101. PMCID: PMC384798.
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC384798/
    Abstract
    Our work was motivated by discoveries of prokaryotic communities that survive with little nutrient in ice and permafrost, with implications for past or present microbial life in Martian permafrost and Europan ice. We compared the temperature dependence of metabolic rates of microbial communities in permafrost, ice, snow, clouds, oceans, lakes, marine and freshwater sediments, and subsurface aquifer sediments. Metabolic rates per cell fall into three groupings: (i) a rate, μg(T), for growth, measured in the laboratory at in situ temperatures with minimal disturbance of the medium; (ii) a rate, μm(T), sufficient for maintenance of functions but for a nutrient level too low for growth; and (iii) a rate, μs(T), for survival of communities imprisoned in deep glacial ice, subsurface sediment, or ocean sediment, in which they can repair macromolecular damage but are probably largely dormant. The three groups have metabolic rates consistent with a single activation energy of ≈110 kJ and that scale as μg(T):μm(T):μs(T) ≈ 106:103:1. There is no evidence of a minimum temperature for metabolism. The rate at -40°C in ice corresponds to ≈10 turnovers of cellular carbon per billion years. Microbes in ice and permafrost have metabolic rates similar to those in water, soil, and sediment at the same temperature. This finding supports the view that, far below the freezing point, liquid water inside ice and permafrost is available for metabolism. The rate μs(T) for repairing molecular damage by means of DNA-repair enzymes and protein-repair enzymes such as methyltransferase is found to be comparable to the rate of spontaneous molecular damage.
    Price, P. Buford; Sowers, Todd. Temperature dependence of metabolic rates for microbial growth, maintenance, and survival. Proc Natl Acad Sci U S A, v.101(13); 2004 March 30; 101(13): 4631–4636.
    doi: 10.1073/pnas.0400522101. PMCID: PMC384798.
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC384798/
    Abstract
    Our work was motivated by discoveries of prokaryotic communities that survive with little nutrient in ice and permafrost, with implications for past or present microbial life in Martian permafrost and Europan ice. We compared the temperature dependence of metabolic rates of microbial communities in permafrost, ice, snow, clouds, oceans, lakes, marine and freshwater sediments, and subsurface aquifer sediments. Metabolic rates per cell fall into three groupings: (i) a rate, μg(T), for growth, measured in the laboratory at in situ temperatures with minimal disturbance of the medium; (ii) a rate, μm(T), sufficient for maintenance of functions but for a nutrient level too low for growth; and (iii) a rate, μs(T), for survival of communities imprisoned in deep glacial ice, subsurface sediment, or ocean sediment, in which they can repair macromolecular damage but are probably largely dormant. The three groups have metabolic rates consistent with a single activation energy of ≈110 kJ and that scale as μg(T):μm(T):μs(T) ≈ 106:103:1. There is no evidence of a minimum temperature for metabolism. The rate at -40°C in ice corresponds to ≈10 turnovers of cellular carbon per billion years. Microbes in ice and permafrost have metabolic rates similar to those in water, soil, and sediment at the same temperature. This finding supports the view that, far below the freezing point, liquid water inside ice and permafrost is available for metabolism. The rate μs(T) for repairing molecular damage by means of DNA-repair enzymes and protein-repair enzymes such as methyltransferase is found to be comparable to the rate of spontaneous molecular damage.
    Chapter 6 Atmospheric CO2 during the 13th century AD a case study based on ice core measurements and stomatal frequency analysis
    http://igitur-archive.library.uu.nl/dissertations/2004-1214-121238/c6.pdf
    A role for atmospheric CO2 in preindustrial climate forcing
    Thomas B. van Hoof,*†‡ Friederike Wagner-Cremer,† Wolfram M. Kürschner,† and Henk Visscher†
    *TNO Geological Survey of the Netherlands, Princetonlaan 6, 3584 CB Utrecht, The Netherlands; and
    †Palaeoecology, Institute of Environmental Biology, and Laboratory of Palaeobotany and Palynology, Utrecht University, Budapestlaan 4, 3584 CD Utrecht, The Netherlands
    Proc Natl Acad Sci U S A. 2008 October 14; 105(41): 15815–15818. Published online 2008 October 6. doi: 10.1073/pnas.0807624105. PMCID: PMC2562417
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2562417/
    ReferencesAbstract
    Complementary to measurements in Antarctic ice cores, stomatal frequency analysis of leaves of land plants preserved in peat and lake deposits can provide a proxy record of preindustrial atmospheric CO2 concentration. CO2 trends based on leaf remains of Quercus robur (English oak) from the Netherlands support the presence of significant CO2 variability during the first half of the last millennium. The amplitude of the reconstructed multidecadal fluctuations, up to 34 parts per million by volume, considerably exceeds maximum shifts measured in Antarctic ice. Inferred changes in CO2 radiative forcing are of a magnitude similar to variations ascribed to other mechanisms, particularly solar irradiance and volcanic activity, and may therefore call into question the concept of the Intergovernmental Panel on Climate Change, which assumes an insignificant role of CO2 as a preindustrial climate-forcing factor. The stomata-based CO2 trends correlate with coeval sea-surface temperature trends in the North Atlantic Ocean, suggesting the possibility of an oceanic source/sink mechanism for the recorded CO2 changes.

    Good exchange, Willis, and a thank you to Dr. Meier for participating.

  99. Excellent debate.
    I do enjoy Mr Watts writing style, but this exchange is one of the most important things available to read on this website.
    I truly hope the debate continues.

  100. [Apologies, the last post formatted incorrectly]

    Willis Eschenbach (16:01:32) :
    First, the rise in atmospheric CO2 is shown by the ice core records to be of recent (last few centuries) origin. I have examined the records in question (Greenland and Vostok ice cores) and the increase is definitely recent.

    When you “examined the records in question,” what did they have to say about the quantitative losses of carbon dioxide in the ice core samples due to metabolization by microbial lifeforms?
    If the “records in question” are in such high agreement without taking into account the losses of carbon dioxide due to confounding factors such as metabolization by microbial lifeforms, improper sampling methods, and more; it would appear the true values would then falsify the claimed accuracy of the ice core reports.
    Personally, the ice core reports appear to be incorrect for a variety of reasons, ranging from a failure to account for confounding factors to inconsistency with other lines of evidence such as stomatol frequencies in plant life. I would encourage a sharpening of your skeptic’s razor with respect to the validity of the ice core analyses.

    Price, P. Buford; Sowers, Todd. Temperature dependence of metabolic rates for microbial growth, maintenance, and survival. Proc Natl Acad Sci U S A, v.101(13); 2004 March 30; 101(13): 4631–4636.
    doi: 10.1073/pnas.0400522101. PMCID: PMC384798.
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC384798/
    Abstract
    Our work was motivated by discoveries of prokaryotic communities that survive with little nutrient in ice and permafrost, with implications for past or present microbial life in Martian permafrost and Europan ice. We compared the temperature dependence of metabolic rates of microbial communities in permafrost, ice, snow, clouds, oceans, lakes, marine and freshwater sediments, and subsurface aquifer sediments. Metabolic rates per cell fall into three groupings: (i) a rate, μg(T), for growth, measured in the laboratory at in situ temperatures with minimal disturbance of the medium; (ii) a rate, μm(T), sufficient for maintenance of functions but for a nutrient level too low for growth; and (iii) a rate, μs(T), for survival of communities imprisoned in deep glacial ice, subsurface sediment, or ocean sediment, in which they can repair macromolecular damage but are probably largely dormant. The three groups have metabolic rates consistent with a single activation energy of ≈110 kJ and that scale as μg(T):μm(T):μs(T) ≈ 106:103:1. There is no evidence of a minimum temperature for metabolism. The rate at -40°C in ice corresponds to ≈10 turnovers of cellular carbon per billion years. Microbes in ice and permafrost have metabolic rates similar to those in water, soil, and sediment at the same temperature. This finding supports the view that, far below the freezing point, liquid water inside ice and permafrost is available for metabolism. The rate μs(T) for repairing molecular damage by means of DNA-repair enzymes and protein-repair enzymes such as methyltransferase is found to be comparable to the rate of spontaneous molecular damage.
    Coupling between atmospheric CO2 and temperature during the onset of the Little Ice Age / Thomas Bastiaan van Hoof – [S.l.] : [s.n.], 2004 – Tekst. – Proefschrift Universiteit Utrecht
    Abstract:
    Present day global warming is primarily caused by the greenhouse effect of the increased CO2 emissions since the onset of the industrial revolution. A coupling between temperature and the greenhouse gas CO2 has also been observed in several ice-core records on a glacial-interglacial timescale as well as on a millennial timescale during the glacials. In marked contrast, no significant ice-derived CO2 fluctuations occur on centennial time scales contemporaneously with well-documented cooling events such as the Younger Dryas, Preboreal Oscillation, and the 8.2 kyr BP event. Intriguingly however, the Little Ice Age cooling event seems to be recorded in several Antarctic ice cores of the last millennium, showing changes in CO2 from 5-12 ppmv As CO2 fluctuations of these magnitudes only generate a minor temperature response the role of these small CO2 perturbations in climate forcing of the last millennium is considered to be non-significant.
    An alternative methodology to assess palaeo-atmospherical CO2 concentrations, is based on the inverse relationship between the number of leaf stomata and the atmospheric CO2 concentration. Compared to the ice cores this stomatal frequency based CO2 proxy in general observes a much more dynamic CO2 regime throughout the Holocene and therefore implies a much larger role for CO2 in Holocene climate forcing. In order to corroborate the concept of coupling between atmospheric CO2 and temperature during the onset of the Little Ice Age, a palaeo-atmospheric CO2 reconstruction for the first half of the past millennium was developed by studying a high resolution stomatal frequency record from fossil Quercus robur (oak) leaves from the Netherlands.
    The results of this study indicate that during the thirteenth century AD a 35 ppmv shift in atmospheric CO2 did occur. More evidence of this CO2 perturbation have previously been observed in one other stomatal frequency based record and two Antarctic ice core records. By applying a firn diffusion model on the in this study presented stomatal frequency based CO2 reconstruction, the lower amplitude of the recorded shift in the ice core D47 could be fully explained by smoothing of the ice core record due to the diffusion processes that take place in the firn layer of the ice. Radiative forcing calculations of the stomatal frequency based CO2 record suggest a more prominent role for CO2 in climate forcing during the onset of the Little Ice Age.
    Chapter 6 Atmospheric CO2 during the 13th century AD a case study based on ice core measurements and stomatal frequency analysis
    http://igitur-archive.library.uu.nl/dissertations/2004-1214-121238/c6.pdf
    A role for atmospheric CO2 in preindustrial climate forcing
    Thomas B. van Hoof,*†‡ Friederike Wagner-Cremer,† Wolfram M. Kürschner,† and Henk Visscher†
    *TNO Geological Survey of the Netherlands, Princetonlaan 6, 3584 CB Utrecht, The Netherlands; and
    †Palaeoecology, Institute of Environmental Biology, and Laboratory of Palaeobotany and Palynology, Utrecht University, Budapestlaan 4, 3584 CD Utrecht, The Netherlands
    Proc Natl Acad Sci U S A. 2008 October 14; 105(41): 15815–15818. Published online 2008 October 6. doi: 10.1073/pnas.0807624105. PMCID: PMC2562417
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2562417/
    ReferencesAbstract
    Complementary to measurements in Antarctic ice cores, stomatal frequency analysis of leaves of land plants preserved in peat and lake deposits can provide a proxy record of preindustrial atmospheric CO2 concentration. CO2 trends based on leaf remains of Quercus robur (English oak) from the Netherlands support the presence of significant CO2 variability during the first half of the last millennium. The amplitude of the reconstructed multidecadal fluctuations, up to 34 parts per million by volume, considerably exceeds maximum shifts measured in Antarctic ice. Inferred changes in CO2 radiative forcing are of a magnitude similar to variations ascribed to other mechanisms, particularly solar irradiance and volcanic activity, and may therefore call into question the concept of the Intergovernmental Panel on Climate Change, which assumes an insignificant role of CO2 as a preindustrial climate-forcing factor. The stomata-based CO2 trends correlate with coeval sea-surface temperature trends in the North Atlantic Ocean, suggesting the possibility of an oceanic source/sink mechanism for the recorded CO2 changes.

    Good exchange, Willis, and a thank you to Dr. Meier for participating.

  101. An excellent & reasoned debate.
    It’s a pity that Real Climate doesn’t permit such, neither by inviting “A Denier” to post their views nor to refrain from censoring those who don’t support their stance on AGW and to rein in the hecklers and insult throwers.

  102. I’m sure it will have been noticed that the Stratospheric Temperature Anomaly peaks seem to run close to recent solar maxima. True, the latest peak is relatively low but this might reflect a quieter sun in terms of magnetic field strength (Livingston & Bell). The trend is now upward again as we approach the next solar maximum.
    If the STA was strongly linked to CO2, then surely the trend should be steadily downward rather than the dynamic picture seen in the graph.

  103. I’ll have to get to both of these posts tomorrow. But, I will have to say that the responses of Dr Meier and Dr Eschenbach mean that the science isn’t settled and that the debate isn’t over.
    Thanks to both of you.

  104. Re: maxwell (15:05:48) :

    I think this misses the point. I don’t think he is saying that one coin flip is simpler than 10000. In fact, after reading his post, I gathered he was saying the exact opposite. That because we know given any amount of coin flips one should get heads half the time, so as you toss the coin more and more, it gets easier to predict how many times heads or tails will come up.

    Actually, the more times you flip the coin the more difficult it is to predict how many times heads will come up. As an example, if you flip the coin twice and predict heads will come up once your odds of being correct are 1 in 2 whereas if you flip it 10 times the odds of correctly predicting 5 heads is 63 in 256 or slightly more than 1 in 4.

  105. Willis — typo:

    “This makes climate models much more complex than climate weather models “

  106. Wren (00:33:53), thanks for your comments.

    “As a result, the idea that climate models can project the climate a hundred years out because “a single coin flip is simpler than 10000 coin flips” is untrue, simplistic, and in no way a metaphor which would help us understand the problem with long-term climate model projections of the future.”

    ——————–
    I think the coin-flip analogy was supposed to show that you can be more confident in the outcome of a 1,000 flips than a few flips. Even though the odds are against two successive heads or tails (only 0.25), it happens frequently. The odds against 1,000 successive heads or tails, however, are virtually nil.

    I understand this, but it has absolutely nothing to do with climate models. Coin flips are the repetition (however many times) of a single random event which is not affected by any variables such as humidity. I can’t think of too many things that are more different from climate models, which are a repeated iteration of multi-variable non-random events occurring in simulated 3D space … how are those similar by any measure?

    Similarly, you can be more confident in a projected trend line for a large number of years(e.g. 2010 – 2050) than you can be for an interpolated value for any individual year within that period. You can appreciate this by plotting a regression line through actual historical temperature records for a like number of years. You will find that line does not describe year to year changes very well but does describe the long-term trend.

    I’m not following this one. You say the trend line does not describe year to year changes but does describe the trend … this seems like a tautology.
    And the idea that we can simply project a trend line from now to 2050? Mark Twain saw your argument coming when he said:

    In the space of one hundred and seventy-six years the Lower Mississippi has shortened itself two hundred and forty-two miles. That is an average of a trifle over one mile and a third per year. Therefore, any calm person, who is not blind or idiotic, can see that in the Old Oolitic Silurian Period, just a million years ago next November, the Lower Mississippi River was upwards of one million three hundred thousand miles long, and stuck out over the Gulf of Mexico like a fishing-rod.
    And by the same token any person can see that seven hundred and forty-two years from now the Lower Mississippi will be only a mile and three-quarters long, and Cairo and New Orleans will have joined their streets together, and be plodding comfortably along under a single mayor and a mutual board of aldermen. There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.

    Take another look at Fig. 4 above to see why we can place no more confidence in the climate models than we can in Mark Twains trend line projections. Clouds are central to the projection of the evolution of the climate. The models can’t get the historical clouds anywhere near right, much less project the evolution of the future clouds …
    You say that the trend line is somehow easier to forecast than the individual years temperatures. What this claim neglects is that the models don’t forecast the trend line. The models forecast the temperature of the the individual years (actually, the individual hours). Once the hourly temperatures are forecast, then and only then can the trend line be calculated. If the hourly temperatures are wrong, so are the yearly temperatures, and so is the trend line … so we can place no more confidence in a projected trend line than we can in the individual hourly or yearly temperatures.
    I hope this is clear, if not, please clarify what you are saying.

  107. @ jorgekafkazar (15:20:24)
    “I contend that the criteria for social action to control AGW should be at least as high as those in criminal justice, and probably much, much higher, considering the mammoth economic consequences.”
    You are spot on.
    To take the analogy yet further:
    (1) Evidence from both sides should be taken under oath and on penalty of perjury.
    (2) The burden of proof must be on the prosecution (CAGW supporters), and may not be reversed e.g. by invoking the “precautionary principle”.
    (3) No individual should be punished (taxed) until he/she has been convicted by due process.
    Cheers,
    Neil

  108. wayne (22:19:05)

    Willis Eschenbach (19:53:26) :
    Thanks, I got the info.
    Willis, I always try to got out of my way to be kind and congenial initially with any person, but I won’t be made a fool by someone and you shouldn’t either. I just wrote much that I am not going to post here, I just hate to see someone like you who I consider a friend to be jerked around. Re-read his null hypothesis two very carefully. Since it is miles from anything you would call science (absolutely no direct tie to the matter at hand) I see it very darkly as a direct slap to you and therefore everyone who enjoys WUWT’s openness. (Hint: he feels he has totally falsified it himself therefore all he said of AGW is proven true as if by a proper scientific papers. And what is the new ‘factor’, co2 spewing modern mankind) I had to read the previous post answering your questions twice before I felt the slap.

    Wayne, I generally refrain from ascribing unpleasant or devious motives to peoples actions on the web. Additionally, in this case I do not think that Dr. Meier was in any way offering a “direct slap” to me or people on this site. I do not get the impression that he has done anything other than to honestly explain his ideas and beliefs to the best of his ability … and I think that he has done that well, and I can ask no more of anyone.
    Having been variously accused myself on the web of bad faith, of falsifying claims, of lying, and a host of other things when I knew that I was not guilty of any of them, I have become very sensitive on this subject. As a result, I do not ascribe to bad intent what is explainable by inattention, exhaustion, lack of clarity, honest error, or simple confusion. In addition to preventing me from making false accusations, it keeps the conversation from descending into a morass of ad hominems and personal attacks.

  109. I have to agree to some degree with Dr A Burns. The amount of CO2 in the atmosphere that is man-made is still not clearly defined. Even so, this is just one of many factors influencing the climate. The answer as to how much man is influencing the climate is even more vague than the amount of man-made CO2 therein. We appear to be going around in circles. Either we are facing a global warming catastrophe or we are not. Who is really going to solve this mystery honestly?

  110. To me, this sharing of perspectives between Willis and Dr. Meier is THE cutting edge of the climate debate. I can’t express how highly I value this thread.
    As I work my way through the article and comments (often re-reading several times to get my head around the science), I find myself asking two questions…
    1. Why is this informed and respectful debate only happening now? How much time, money, opportunities, column inches, lives and careers may have been wasted by our collective inability to ‘sit down’ and talk like this?
    2. Where does all this go? I’m keen not to head off piste here, but it’s widely acknowledged that our political class essentially shut their ears and say ‘we listen to the IPCC’.
    Is the way forward simply to find ways to engage with respected scientists in the field one at a time, in order to create a growing snowball of ‘scientists formerly known as believers’ (which I appreciate Dr. Meier isn’t – yet!), which leads to some kind of rationalising (shaming?) effect on the IPCC?
    Great work Willis – and, of course, thanks to Anthony for hosting this important discussion.

  111. Congratulations both to Willis and Dr. Meier for such a fruitful discussion. Long may it continue.
    One aspect of Dr Meier’s original piece caught my attention. It was about the ice cores. Dr Meier says that the CO2 ‘sometimes’ lags behind the temperature. In fact, as Willis points out, it always lags behind, and so cannot be the cause. Dr Meier also invokes feedback i.e. the idea that as the oceans get warmer they will emit more CO2, thus causing more warming etc etc.
    I’m reminded of a piece in New Scientist a year or so back. Basically they said, Okay, the CO2 lags behind temperature but because it’s a positive feedback we’re doomed after all.
    Well, really. If we’re trying to discover whether CO2 can drive climate then this becomes close to a circular argument. Arguments from the AGW side seem to start from the assumption that AGW is true, which certainly is a circular argument.
    If there is a positive feedback then it should be clearly visible in the ice core data. But, as Willis says, there’s no trace of any feedback. To me, this is a damning indictment of AGW. Earth has performed a perfect experiment over the last million years, with temperature and CO2 going up and down like yo-yo’s and it has taken the trouble to store the record of this in the ice cores. If the record shows no evidence whatsoever that CO2 can affect the climate then AGW is in *big* trouble.
    The sad thing is that climate scientists, Dr Meier included, simply seem to be looking the other way and ignoring this very inconvenient truth. I can only hope that Willis’ truly excellent responses will at least give Dr Meier much pause for thought….
    Chris

  112. I started to read it then got bogged down – I’m sorry = I don’t care a rat’s arse what he believes. He’s only a simple individual like the rest of us. – I only want to know that the figures he posts as part of his duty are accurate and without bias or tampering..

  113. The logic of Willis’s response is sharp as a razor and stands like a beacon against the obscurity of Mier’s muddled illogic. Yet, I stand aghast at this outcome. Here we have an amateur showing a professional scientist that he doesn’t understand the scientific method and can’t even follow the logic of his own arguments. Is this the problem with climate science today? I think it is.
    And Wayne has a point when he writes “but I won’t be made a fool by someone and you shouldn’t either.” Willis, when Professor Ravetz wrote his Post Normal science article, you said you would stamp all over him with your size 10 cowboy boots, and you did. Dr Mier is peddling junk science and nonsense. Does it not behove one to call a spade a spade?
    On second thoughts, your reasoned arguments do a lot more to deconstruct the junk than any amount of cowboy boots could.

  114. Friends:
    Dialogue of the kind being conducted here has been missing for two decades. Hence, this debate possibly provides an opportunity that many – including me – have sorely wanted. It would be a pity if the opportunity were wasted by the dialogue being side-tracked to debate of other issues than those presented by Eschenbach and Meier.
    The main issues they present are
    (a) what is the appropriate null hypothesis for AGW
    and
    (b) can that hypothesis be falsified using available data?
    Several here have attempted to get discussions of H2O being the major GHG, the politics of AGW, and the anthropogenic contribution to change in atmospheric CO2 concentration.
    I strongly dispute that the preponderence of evidence supports an assertion that change in recent atmospheric CO2 concentration has a predominantly anthropogenic cause. But I have refrained from debate of this matter in this thread and in the thread concerning Dr Meier’s post. And I shall continue to refrain from such discussion because I think Willis E is right to concentrate on the main issues.
    If those main issues can be resolved then it may be appropriate to discuss other matters. For the present, I would be grateful if the discussion could be constrained to the matters presented by Eschenbach and Meier.
    They boil down to
    1. attempt to agree the proper null hypothesis for climate change,
    2. the validity of the assumption that climate change is driven by radiative forcing,
    3. the probability that human activities are significantly affecting radiative forcing, and
    4. the validity of models (e.g. by ue of the ‘coin-toss analogy’) as tools to resolve points 1 to 3.
    I genuinely think this opportunity to start a long-needed process of dialogue is too valuable to waste on non-immediate matters however much individuals care about other issues.
    Richard

  115. A striking revelation in this dialogue is the discussion of models as evidence. It’s easy to put yourself in the mind of the modelers who’ve invested so much time and energy in their development and thereby realize that the first thing they want to do with their new baby is run it into the future to see what will transpire. Do this enough times and you’ll begin to think about your model output as evidence.
    I’ve built enough models to have experienced this urge myself. The problem though, is that the very act of attempting a model is to admit that what your modeling is too complex to lend itself to a simple set of static equations. Models are nothing if not a form of dynamic calculus. Model output is (and should always be viewed as) a hypothesis test, not evidence. Any divergence between output and real evidence is an indication that your math is wrong.
    Knowingly extending your incorrect math into the future is not evidence. It’s an artifact of incomplete knowledge. Using the output to effect policy is not science, it’s spin.
    Extracting the admission that models produce evidence is huge.

  116. Willis you mention the recession of mountain glaciers, but then spend your time looking at the glacier mass balance data set. Glacier terminus behavior is measured at a much larger number of glaciers. For example in 2009 if we just look at data from four programs in the US, Norway, Switzerland and Austria we have: 1) Of 93 glaciers surveyed by the Austrian Alpine Club in 2009, 85 receded, 7 were stationary and 1 advanced. 2) Norwegian Water Resources and Energy Directorate reported terminus fluctuation data from 27 glaciers for 2009 indicate, 22 retreating, 3 stable and 2 advancing. The average terminus change was -18 m, compared to -13 m in 2008. 3) North Cascade Glacier Climate Project reported all 42 glaciers observed retreating. http://glacierchange.wordpress.com/2009/09/04/north-cascade-glacier-climate-project-2009-field-season/4) Swiss Glacier Monitoring Network noted in 2009 that 81 glaciers retreated 5 were stations and 2 advanced. If we look at mass balance data the increase in mass loss is accelerating. The cumulative loss of the last 30 years is 11.6 m w.e. the equivalent of cutting a 13 m thick slice off of the average glacier (Figure 1). The trend is remarkably consistent from region to region (WGMS, 2009). WGMS mass balance based on 30 reference glaciers with 30 years of record are not appreciably different from the results for all reporting glaciers. The decadal mean annual mass balance was -198 mm in the 1980’s, -382 mm in the 1990’s and -.624 mm for 2000-2008. The declining mass balance trend during a period of retreat indicates alpine glaciers are not approaching equilibrium. http://www.nichols.edu/departments/Glacier/global%20glacier%20mass%20balance.htm
    the loss is consistent in North America
    http://glacierchange.wordpress.com/2009/12/19/helm-glacier-melting-away/

  117. I love the thermostat, especially the view from the sun. One of the great difficulties with global scale of anything is that people don’t deal with scale very well. We all tend to think of our own particular circumstance and point of view and lend more weight to our own spatial and temporal experience. Pretending to be the sun can break you out of your scale comfort zone.
    As neat as the thermostat hypothesis is, it too is limited in scale. It doesn’t for example, address the enormous mechanical engine that coexists with the heat engine. It doesn’t address the contribution of life to the atmospheric gas content or the color of the oceans and land.
    None of this is to say that the model is wrong or somehow corrupt, just that it is necessarily limited.
    I sometimes wonder of the CO2 centric modelers have simply gotten themselves too far into the weeds. If we don’t begin at a spatial and temporal scale worthy of the complexity there is no way to ever hope to model it all. Kind of like making the mistake of learning all about the coreallis effect by staring at your bathtub drain.

  118. That’s my point – We’ll never reach a time when overall concentrations of man made c02 aren’t increasing.

  119. Wren (00:33:53) : I think the coin-flip analogy was supposed to show that you can be more confident in the outcome of a 1,000 flips than a few flips. Even though the odds are against two successive heads or tails (only 0.25), it happens frequently. The odds against 1,000 successive heads or tails, however, are virtually nil.
    You have misunderstood Walt’s example. He doesn’t ask for a prediction of successive flips. He only asks for a prediction of the aggregate of the number of heads:
    You are given the opportunity to bet on 10000 coin flips. If heads comes up between 4000 and 6000 times, you win a million dollars. If heads comes up less than 4000 or more than 6000 times, you die. Again, you are assured that the coin is completely fair and unbiased. Would you take this bet? I think I would.
    The sequence of outcomes could easily be { … H,T,T,T,H,T,H,H… }. At the end, we count up the total number of heads to see if we have correctly predicted the number of heads.
    But when Walt gave himself a range of 4000 to 6000, he turned this into a near certainty that he could correctly predict the outcome. And that means he either doesn’t understand the point, or he was trying to mislead. I’d prefer to go with the former.
    When folks on the street hear these arguments from a figure of authority, there is a danger that a bad idea will be accepted as correct.
    MMGW proponents use this type of argument to claim that long-term climate averages are more predictable than weather. But they fail to make a like-for-like comparison: the statistics of a prediction of an average must be “more tightly consrtained” than a preduction of an individual event.
    If we correctly formulate the measure of the utility of different types of prediction, we will see that a prediction of an average will be no better than the prediction of an individual event.
    It is only fair to point this out to Walt to ensure he is in a better position to challenge his own assumptions.

  120. I may have missed it having been stated already in all the comments, but would like to add:
    Any warming effect from CO2 is NON-linear.
    The bent-stick proponents missed this, assuming a doubling of CO2 (nominally 200 to 400 ppm) produced a doubling of effect, whereas it’s more likely to produce an approximately 15% to 20% increase.

  121. I have posted the following question earlier to Walt Meier. I did not see him answering any questions here on this post. It is an omen that he is trying to steer clear of this debate. Pity. He seemed a bit apologetic about not knowing much of exactly what’s on both sides of the coins,. Anyway, here is the question again:
    We know that Svante Arrhenius’ formula has long been proven wrong. If it had been right earth should have been a lot warmer. So I am asking: what is the correct formula? I am sure Walt Meier can give me this formula or ask the scientists who work with this every day, at the NSIDC?

  122. I can,t wait for Dr Meiers response to Eschenbach.
    This is epic science debate, brilliant, both of you have earned my most sincere respect and trust, openess is the solution.

  123. Well argued Willis. If this was a tennis match you would be 6 -0 in the first set.
    Reference your point about models.
    When I graduated over forty years ago my physics tutor told me that I was “wasting my degree” when he heard I was going into manufacturing!
    I thought differently. My tutor had spent all his life in academia whilst I had already spent over a year in highly demanding manufacturing activities tackling stubborn technical problems. The thing about manufacturing is that it as real as you get. If it does not work it does not work, and no one is interested in a theory that does not improve performance. The other thing about manufacturing is that harbours all the errors inherent in the models used by the process designers. If the models were correct the plant would be capable of making zero defect product 100% of the time. In practice it takes decades to achieve this and in my experience the single biggest cause of this excessively long development period is the refusal of process scientists to believe that their concept of what is happening is wrong. This is not to say that these scientists are stupid. The vast majority of their models are accurate. The problem is that they are never completly accurate. Yet every time I came up with conflicting data they would argue that I had made some sort of mistake in my analysis or the operator had not documented things properly or had not carried out the process as instructed. Do ad hominem attacks ring a bell?
    My first job after graduating was very exciting. Having been blamed for everything that was going wrong the manufacturing personnel gave me tremendous support as I “took on” the development department. I eventually set up my own scientific experiments and although it took 5 years ( by which time I was running the department) I eventually proved that one of the fundamental assumptions used in the process design ( for which a nobel prize had been awarded!) was actually wrong. There were many other errors found by others the most common being that we were not measuring what we thought we were measuring. As Feynman once said of the new “sciences” like economics “I know how much effort it takes to really know something. These guys just haven’t done their homework”.
    Climate scientists seem to fall into this category. For the most part they are a disparate group of data gatherers and modelers. Unfortunately hardly any of them seem to be doing fundamental science which measures the real world in a way that tests the theories that abound. Those that are are to be applauded but it hard to discover who they are and what they are saying since so much is spun by the media and vested interests. If I found it hard to fight the scientests when it was theoretically their job to help me what chance have we when the ir main interest of the average climate scientis is maintaining their funding.

  124. Willis Eschenbach (02:18:26) :
    ‘You say that the trend line is somehow easier to forecast than the individual years temperatures. What this claim neglects is that the models don’t forecast the trend line. The models forecast the temperature of the the individual years (actually, the individual hours). Once the hourly temperatures are forecast, then and only then can the trend line be calculated. If the hourly temperatures are wrong, so are the yearly temperatures, and so is the trend line … so we can place no more confidence in a projected trend line than we can in the individual hourly or yearly temperatures.’
    My apologies if this is a very naive question, I come somewhat late (and probably very uninformed) to this fascinating discussion – but is it not possible to envision a model who’s hourly predictions are sensitive to initial conditions and model parameters, which would make it’s prediction for any given hour quite poor, but on repeated running of the model with variations in initial conditions and/or model parameters a similar trend could be observed through the data – allowing more confidence to be placed in the trend than in the individual hourly or yearly temperatures? I have no idea if this is the case as I am not familiar with the models in question, but is it not possible that there at least a class of models for which this could be true?

  125. Why does everyone so easily subscribe to the view that fossil fuel carbon differs chemically from “natural” carbon? It differs from inorganic carbon but not from organic carbon. The ratio of 13C to 12C – i.e. d13C – in the atmosphere has been used to try to characterise the origins of the changing CO2 concentration – to identify the sources and sinks. Plant photosynthetic biochemistry is remarkably specific preferring to fix 12C over 13C by a large margin. Thus during photosynthesis d13C rises due to the preferential abstraction of 12C. Respiration – being derived from very low 13C plant material preferentially adds 12C to the atmosphere and d13C falls. Fossil fuels are derived from carbon of photosynthetic origin and have similarly low 13C content. The rise and fall in d13C observed at the many monitoring stations such as Mauna Loa matches perfectly with the annual changing balance of respiration and photosynthesis. Because both respiration and combustion oxidise photosynthetic carbon the change in d13C cannot distinguish the origin of the added CO2.
    Note also that Respiration is Temperature dependent roughly doubling with a ten degree increase in temperature whereas photosynthesis is by and large not temperature dependent. This confounds attempts to correlate temperature with CO2 concentration which in any case cannot be done by simple regression as has been done recently in this discussion. This is because both are measured dependent variables. For regression to work properly one variable must be independent and known.

  126. Well, I have to abandon my lurking posture, since authors and commentators alike seem to be off-base on the coin flipping issue. The alternative heads/tails pattern of coin flipping is not an inherent property of the coin. The observed largely random results from coin flipping are completely attributable to the manner in which a person does the flipping. A clever lady mathematician at Stanford has created a coin flipping machine that produces heads (or tails if you wish) every trial.
    http://www.thebigmoney.com/articles/hey-wait-minute/2009/07/28/flipping-out?page=0,1&g=1
    In my miind, this information makes the coin flipping analogy all the more fitting for climatology discussions. We have drawn conclusions as to the random performance of the coin although the source of the randomness observed randomness comes entirely from the variable behavior of the coin flipper.

  127. At the end of the day scientists are tied to their research.
    But the crux of this matter, is measurment and theory.
    Theory to progress must address measurement.
    I found Mier’s arguments, mathematically specious. I found his reliance on past theories unfortunate to cover theory failure spectacular.
    To my mind Scientists with the best tools of humanity can provide cannot cover falsification as a more time issue. That is anti science.
    The only and correct answer is, we dont know, work in progress.
    Not protecting politics or business.
    This herd think is dangerous when conjoined with PNS or politicisation of science.
    Science is not belief.
    I am not one for Kumbayahs and lets cuddle. Does it work, if not why not.
    The question of trust is based in only two areas, expertise/abiltiies and truth.
    Trusted people can make mistakes, but they must admit them.
    I am not buying the other sides excuses in PNS or anti science.
    The models have failed.
    Their behavior remains the same, cover up and denial.
    For a theory to work, it works all the time.
    Mathematically in applied science we build error and downside assessments, we build failure into the models.
    This is what the standard normal distribution is about.
    But all these models have failed at 100%, in predictive capacity. You cannot spin that. They have failed at 100 per cent.

  128. Willis, I think the whole discussion is pointless.
    Why?
    Because everyone knows that the most important GHG is water vapour.
    Everyone knows that water wapour is created from evapouration just by the sun shining on the oceans.
    Everyone knows that in a short time enough water vapour is created to offset any cap and trade result.
    When enough water vapour is created an equillibrium is reached. We see it every day.
    So whats the point in discussing it?

  129. Wren (23:43:11) :
    davidmhoffer (21:45:51) :
    Willis Eschenbach;
    I know of no one making that claim, that the present climate runs on different principles than the past climate.>>
    Sure you do. There’s these dudes with tree ring studies who discarded 50 years or so of data because “they suddenly stopped tracking temperature”. They don’t know why, but even though they don’t, they are quite certain that what ever it is they don’t know is causing it now, they “know” it never happened before.
    ====
    Who knows what that means?>>
    I don’t know Wren. I only know that they have said that they don’t know. Until we know, what they don’t know, we don’t know what they don’t know means. When we do know, we can tell them. That won’t mean that they then know of course, as they can choose to not know what they have been told we now know.
    Uhm…. what was the question again?

  130. The earth’s temperature has risen approximately every 60 years since the invention of the thermometer. The periodic warming cycles are “normal” and have probably been going on forever but we only have records since 1860 or so. The rate and amount of the most recent cycle isn’t significantly different from the previous ones to cause us to believe there is a different cause or amplifying factor.
    Even Dr Phil Jones of CRU fame admits this.
    Question: Do you agree that according to the global temperature record used by the IPCC, the rates of global warming from 1860-1880, 1910-1940 and 1975-1998 were identical?
    Dr Phil Jones: An initial point to make is that in the responses to these questions I’ve assumed that when you talk about the global temperature record, you mean the record that combines the estimates from land regions with those from the marine regions of the world. CRU produces the land component, with the Met Office Hadley Centre producing the marine component.
    Temperature data for the period 1860-1880 are more uncertain, because of sparser coverage, than for later periods in the 20th Century. The 1860-1880 period is also only 21 years in length. As for the two periods 1910-40 and 1975-1998 the warming rates are not statistically significantly different (see numbers below) ”
    http://wattsupwiththat.com/2010/02/14/phil-jones-momentous-qa-with-bbc-reopens-the-science-is-settled-issues/#more-16418
    The first two warmings seem to be in sync with ocean cycles [PDO and ADO] ! Why should we ascribe a different cause to the third warming ?
    This cyclical warming seems to be added to a long slow warming caused by recovery from the little ice age [increasing sunspot activity] so we get a ratchet effect. Each cycle starts just a little higher than the last.
    http://sidc.oma.be/html/wolfaml.html
    There seems to me to be no reason to ascribe a different cause to the most recent warming cycle.

  131. Correlation is not causation, even when it’s 100%.
    Example: during testing, a particular piece of hardware of ours showed totally out-of-spec performance. In the process of failure investigation, I noticed that the aircraft navigation system appeared to not be integrating velocity correctly, i.e. velocity did not integrate to position. It turned out that during each of the performance assessment tests, our system was mounted on the same aircraft, even though the assessment was performed over a period of weeks.
    The natural thing to do would be to blame the aircraft, because (it turned out) it had a faulty navigation system. But because I had a good understanding of the underlying physics, I couldn’t connect the aircraft as cause to the system’s out of spec performance. Even though the correlation was 100%.
    It turned out to be a complete coincidence that the aircraft’s navigation system (which had to be replaced) was bad; the problem was an unanticipated one local to that particular piece of hardware.
    Point being: correlation is not enough. There has to be a causal relationship, and it’s up to the scientist to establish causality; it’s not up to the rest of the world to prove that no (or minimal) causality exists.

  132. Geoff,
    ‘Wrong.
    At each new flip of the coin, the coin has no memory of its history. With a perfect coin, the probability on every flip is the same, 0.5.’
    …but that’s the whole point. The model of coin flips, their independence, implicitly means that as the dynamic variables that determine the outcome of a single coin flip will average out over the course of ALL the flips, we get closer and closer to this idealized model. That’s Dr. Meier’s whole point. Given the parameters of a moment, it’s hard to model the outcome of a single coin flip due to complex dynamics going on (wind, precipitation, air pressure, etc), but those parameters average out for many, many coin flips, so we can fall back on our model of independence.
    It seems like many readers have misunderstood this point.
    TerryS writes,
    ‘Actually, the more times you flip the coin the more difficult it is to predict how many times heads will come up. As an example, if you flip the coin twice and predict heads will come up once your odds of being correct are 1 in 2 whereas if you flip it 10 times the odds of correctly predicting 5 heads is 63 in 256 or slightly more than 1 in 4.’
    Huh?
    You have quantified the probability of picking as many heads as one wants in more than one flip. You proved my point. Steven Goddard did it for 10000 flips 100000 different times and there was a very sharp distribution around heads coming up 5000 times.
    It seemed to me that Dr. Meier’s point, in the context of discussing the difference of modeling climate in the short term versus the long term, is that one can have a great deal more confidence in a model of the range of times heads comes up in 10000 flips versus confidence in a model determining whether heads comes up in a single flip. Your comment supports this idea as well, though I don’t know if that was your intent.
    Willis’ assessment, however, seems to make it as though Dr. Meier is saying the opposite on my assessment. It is obviously confusing readers who think they are somehow making an argument against me when they are not.
    Willis,
    would you mind cleaning up the end of that section to reflect Dr. Meier’s point correctly?
    Thanks again for your time and everyone for their comments.

  133. Beth Cooper (00:37:44),
    Who says cowboys aren’t elegant and courteous?
    Well, courteous anyway. ☺

  134. Maxwell,
    In a nutshell,
    The flip of a coin is a two event result based in a single variable, the coin.
    The distribution is 50 50.
    The Climate is not a coin.
    The climate is a multivariable system.
    Don’t talk probabilistic nonsense.

  135. I’m just a capenter/lurker and way over my head here but I think the coin flip analogy is faulty as are most analogies. Each coin flip has exactly the same odds, Each flip adds or subtracts one unit. As the total number of flips increase the effect of each flip decreases.
    For climate science I think a better analogy might be one we carpenters face all the time.
    We need a level line. We can eliminate the error if we know the two points are level and string a line. If we only know one point and project a line out using a level that has some unknown error the total rise/error after a meter or two is very small. Over a great distance obviously the rise/error will become very large. (I know , an analogy and probably faulted too) .I think Willis said something similar.

  136. Maxwell just so you can see it in your head.
    Consider the Climate, 15 dice and none with an equal number of sides or equal sized sides.
    The climate system is obviouusly not a coin toss, because if was on past scientific history no one would want to roll a serious ice age.
    Obviously not only dont they teach logic anymore they dont even teach probablility.

  137. The coin toss argument is not relevant, not in a discussion of multi variables.
    Different variables have different weights or progessive orders of magnitude.

  138. Re: Ian H (17:41:43)
    Due to this massive legion of “creation scientists” trying to get their “dissenting view” into the classrooms, now is not a good time to revisit the peer-review process?
    Really? That the best you got? Those challenges from “creation scientists” tend to wind up in court, which has shown it can ably decide to exclude such “dangerous” views from public education.
    Please point us to the legal process, perhaps describe the procedure, where challenges to a peer-review process and/or judgment may be brought, where evidence may be presented and considered, and legally-binding rulings are issued that force the retraction or acceptance for publication of scientific papers. From what I’ve seen, there is a demonstrable need for such a method of redress, and I would be pleased to know it already exists.

  139. This has been *the* most informative and instructive discussion I’ve met. May I warmly congratulate both participants. I hope I may look forward to Dr Meier’s response.

  140. Steve Goddard says:

    How much fossil fuel reserves do you think there are? Right now, only three percent of CO2 emissions are man made. We would have to increase emissions by over 30X to reach 50% of the total.

    This is a claim that is based on a complete misunderstanding of the carbon cycle. I recommend you read, for example, Section 2.4 of the book “Global Warming: The Hard Science” by L.D. Danny Harvey. The summary is this: The atmosphere + biota + soils + ocean mixed layer components form a tightly coupled subsystem that rapidly exchange carbon between them but exchange it only slowly with the deep ocean. Hence, any new slug of carbon introduced into this subsystem rapidly partitions between the different components. To confuse exchanges between components with the introduction of new carbon from outside this subsystem is to make a very elementary error in basic understanding of the carbon cycle.

  141. Richard S Courtney,
    “2. the validity of the assumption that climate change is driven by radiative forcing.”
    This is a crucial point, because so often the appeal of AGW falls on the CO2 is a GHG ==> radiative forcing ==> warming. This chain is taken as an axiomatic truth, but it should not be.
    The “Faint Sun Paradox” led to a the search for a CO2 solution to prove the radiative forcing theory, but such a theory would be classed as an “ugly” theory because it depends on “fine tuning” the change in CO2 with change is solar irradience. A beautiful theory is one that appears to be inevitable – it could not have been any other way.
    I believe (that unscientific word again) that the albedo theory could be the beautiful solution to the faint sun paradox. Broadly speaking, the early earth was saved from turning into a ball of ice due to a very low albedo caused by a lack of clouds. There was a recent paper that demonstrated the physical reasons why this could be correct. As the sun warmed up, cloud formation increased and increased the albedo. As one causes the other, there is an inevitability about the events and does not depend on fine tuning.
    Showing close similarity to Willis’ thermostat idea, whether it turns out to be right or wrong, I would say that at least it passes the test of being a beautiful theory.

  142. I suggest it is misleading to call CO2 in the atmosphere a greenhouse gas, as there is no evidence that it causes any significant warming. To say that it must add to “forcing” implies that we know the energy balance of CO2 with everything else, in particular the biosphere.

  143. Joel Shore,
    ““Global Warming: The Hard Science” by L.D. Danny Harvey.”
    Why would anyone seeking an objective description of science want to read a book whose very title contains within it the seeds of its own prejudice?

  144. Capn Jack,
    this is a portion of my second comment to Willis:
    ‘So the point is that a single coin is NOT easier to model and therefore, not simpler in the end. I don’t think that this fact really negates your argument in any way. I think your point that error propagate very quickly in climate models, versus a variable-less model for coin flips is quite valid in this context. I also think Dr. Meier’s example suffers as much from oversimplification as any I have seen in some time.’
    There is a conversation going on in this thread that necessitates looking at all of the information I have provided in order to get a appropriate context on what I am trying to say. I appreciate you trying to help me make this point more clearly, but I would appreciate it if you take a look at all the information I have provided before jumping on a portion of it undeservingly.

  145. I have to agree that the coin-toss discussion doesn’t have much of a place in a discussion of dynamical system, where the randomness is mostly apparent, rather than true.
    You can get chaotic-seeming behavior out of a system with no random inputs or processes. Pretty much everyone who’s worked with complex systems knows this, but you don’t need a complex system to demonstrate random-seeming behavior.

  146. Joel, 8:34:13
    Sounds like there is still a huge CO2 sink available in the deep oceans.
    ====================

  147. Willis: May I please take a copy of this brilliant discourse? I promise that I will not claim any of it to be my own work, aka Mann! Also, are/were you ever a cowboy or a fisherman? Sceptisim, coupled with the patience of a fisherman and the toughness of a cowboy, seem to me to be qualities every bit as necessary as academic achievement, in order to make a true scientist.

  148. Willis,
    Your post,
    Finally, if that is the null hypothesis, the alternate hypothesis is that the factors that affect the climate now are not the same ones that affected it in the past. I know of no one making that claim, that the present climate runs on different principles than the past climate. So it is a null hypothesis for an argument that no one is making … which is why it is a “straw man” null hypothesis.
    I think the case can be made that the IPCC is making the claim that there are greenhouses gases affecting climate now that were not affecting climate in the past, namely the chlorofluorocarbons. Which only requires assuming that chlorofluorocarbons are predominantly man-made.
    Same principles but different gases, and as for the important one that we can do something about, CO2, there are people making the claim that man-made CO2 is having an effect on the climate over and above the effect of naturally produced CO2.
    Hansen is making that claim isn’t he?
    And slightly off topic, but shouldn’t we as a society be trying to find something better to do and better people to work for, for the coal miners recently in the news?

  149. I wouldn’t get too hung up on the physical science of coin flipping and whether it can be manipulated. We could more productively look at the “fair coin” as a metaphor for a simple machine which can produce one of two outcomes, but can only be characterised in advanceby the probability of each outcome.
    TerryS (01:52:04) : “Actually, the more times you flip the coin the more difficult it is to predict how many times heads will come up. As an example, if you flip the coin twice and predict heads will come up once your odds of being correct are 1 in 2 whereas if you flip it 10 times the odds of correctly predicting 5 heads is 63 in 256 or slightly more than 1 in 4.”
    Well that’s just the binomial distribution for a small number of tests (flips). It is beside the point for what Walt was trying to argue, and that’s why he went as far as 10000 flips.
    The second bet should have been: what will come up more frequently, heads or tails? That leaves us with the same 50/50 chances for 10000 flips (well beyond the point where we need to concern ourself with the binomial).
    I would also discourage the argument that coin flipping is too simple to be relevant to climate forecasts. There is a general principle being asserted by Walt and others: that an average of outcomes is easier to predict than a single outcome. This claim is incorrect and that is what I’d prefer to focus on.

  150. 5. The increase in CO2 is due to human emissions. There are two ways we know this. First, we know this simply through accounting – we can estimate how much CO2 is being emitted by our cars, coal plants, etc. and see if matches the observed increase in the atmosphere; indeed it does (after accounting for uptake from the oceans and biomass). Second, the carbon emitted by humans has a distinct chemical signature from natural carbon and we see that it is carbon with that human signature that is increasing and not the natural carbon.
    My Comment: Agreed.

    My comment: Absolute BS
    You have no evidence of this.

  151. Willis Eschenbach (02:50:25) :
    Point well taken. Possibly I tend to read too deep in another persons words and what they really seem to be portraying at the very bottom layer of their words. I didn’t mean Dr. Meier was not being polite, he was, all of his words were soft, and I guess in the end it was myself who was not. On congeniality, I’ll try your approach though I’m rather new to posting on the web and sometimes lack the correct words to press a hard point softly.
    In the meantime, when I think some point really needs to be made and made at that very moment, I will probably still do what I have always done in life and press my point wrongly rather than wrongly not pressing my point at all. Too much for my descendent’s well being is at stake.

  152. Willis, my take is there may be a temperature control valve near the Tropopause. The relative humidity as measured by NOAA has been continuously decreasing since 1948 when measurements began. Increasing carbon dioxide concentration decreases the partial pressure of the water vapor thereby displacing it to outer space (since it is beyond the point where it can be attracted back to earth).
    Secondly, production of ground water from slow to recharge aquifers (fossil water) accounts almost perfectly with observed increase in ocean levels. Dropping water levels in the aquifers started decreasing total extraction rate in about 1995.
    Thirdly, Arctic sea ice variation is demonstrably controlled by variations in wind blown ice removal and variations in the Gulf Stream currents.

  153. Deepest graditute to Dr Meiers and Willis Eschenbach and the legion of commenters. I am engrossed and educated by reading your work.
    I feel like I came in mid way through the debate and without a program, however.
    Who are you Mewirers and Eschenbach? I would like to see your bios and learn of your earlier writings.
    And, I would like to encourage you to focus more on the core issue, as I see it. What proof is there of the forcing impact of CO2 in its greenhouse role and what is the counter to that argument.
    Second, what proof is there of warming beyond the long term natural warming of the interglacial period.
    I agree with all who say this is the most productrive debate, point-countger-point, scientific exchange to date in the 30 years of controversary.
    Thank you again to all.

  154. Vincent:

    Why would anyone seeking an objective description of science want to read a book whose very title contains within it the seeds of its own prejudice?

    If they prefer to remain ignorant of the science rather than confront scientific ideas that might be uncomfortable to their beliefs, then they wouldn’t.

  155. A comment on the conclusion that climate models are more complex than weather models.
    I have not seen anyone support this with any evidence.
    And although I am unqualified to answer this question one thought springs to mind.
    That is that both are currently limited by the amount of processing power available, and so on a first order approxiamation, both are of equal complexity.
    Though the grid cells of a weather model must be of orders of magnitude smaller than those of a climate model, and faster as we want real time results.
    So, anybody know the real answer?

  156. Mr. Pelto (is it Dr.?), {Willis you mention the recession of mountain glaciers, but then spend your time looking at the glacier mass balance data set……}
    Since you are a director for glacial studies, may I ask you a couple of questions…
    Is there any data from the previous warm period of 1930/40 that shows abatement of the glaciers comparable to the last 30 years?
    What is your personnal view about reports that as some of the European glaciers abate they are “unearthing” past roads/settlements/bodies, etc.
    Thanx for your involvment in this progressive discussion.

  157. Jordan,
    ‘I would also discourage the argument that coin flipping is too simple to be relevant to climate forecasts. There is a general principle being asserted by Walt and others: that an average of outcomes is easier to predict than a single outcome. This claim is incorrect and that is what I’d prefer to focus on.’
    Very well said. The coin flipping is ‘correct’ in the sense that it is self-consistent and show ‘correct mathematics. The problem is that this example is not pertinent, which I think we all agree on because even if dynamics average out for many flips, such averaging is problematic in climate models.
    Thanks.

  158. bob,
    since both are time iterative and climate is the long term average of weather, one had better get the weather model correct so that it provides the correct long term average. Otherwise, one is likely injecting what might appear as a systematic error that is going to compound the problem.
    Concerning a cooler stratosphere, it makes sense that the actual radiating upper portion of the atmosphere. Lower down there is little radiation that is not balanced by downward radiation from above. When one adds additional GHGs one is increases the ’emissivity’ as well as the ability to absorb IR. Combine this with the upper part radiating the same power outward as well as inward and one has the creation of an energy (power) imbalance because the increase in emissivity increases radiated power in both outward and downward directions while the absorption of energy is only from the increase in absorption from below. It would balance if space were at the same temperature as the lower atmospheric levels rather than being at almost 0 Kelvins. To balance the increase in absorbed power with the increase in power radiated at the top levels, the temperature must drop.

  159. Willis,
    re: ‘My null hypothesis NH1 is that the currently observed climate variations are the result of natural variation. The opposite of my null hypothesis is the alternate hypothesis, that currently observed climate variations are the result of human-caused GHG increases.’
    Your ‘opposite of the null hypothesis’ is not THE opposite, it is only , as you state, an alternative. The opposite must involve all possible ‘unnatural causes of variation’. One might substitute ‘new’, ‘extraordinary’, ‘unexpected’, ‘super-natural’, or even ‘human-caused’, but these are simply then alternate hypotheses, not really opposites.
    I think you are spot on to discuss the hypothesis and the null hypothesis — science is very much about defining the problem to be considered.
    It is my opinion that this is the crux of the ‘global warming’ controversy–the original problem statement has been malformed.
    I’ll be the first to admit that I am not knowledgeable enough to attempt to re-state the problem….I hope your efforts will lead to some better formulation.

  160. DeNihilist 10:16:11
    This is not data, just anecdotal, but in “Three Got Through” by the late Martin Lindsay, who was a member of the British Arctic Air-Route Expedition to Greenland and led the British Trans-Greenland Expedition, both in the early 1930’s, wrote of being told by the local Eskimos that the glaciers had retreated considerably since their fathers’ day. Best wishes, Dave.

  161. Pompous Git (23:42:38) :
    @ Aargh. (20:32:54) :
    “That which is true is provable.
    That which is not provable is false.
    All unprovable assertions are self contradictory.”
    Google “Gödel’s incompleteness theorems”, or read some basic philosophy of logic…
    I did. Now you go read Goedel and uncerstand that he wasn’t providing you with handy excuses, he was telling you to enlarge the context to solve your issue.

  162. Pompous Git (23:42:38) :
    And next, if you follow the usual script of the anti-intellectual, you’ll be on with ‘there’s no such thing as an absolute’, with no sense of hypocrisy at having just uttered one.
    I know all your tricks- the kids of the 60’s has been using them to talk each other out of their pants for half a hundred years now.

  163. Jordan (09:40:05)
    ‘I would also discourage the argument that coin flipping is too simple to be relevant to climate forecasts. There is a general principle being asserted by Walt and others: that an average of outcomes is easier to predict than a single outcome. This claim is incorrect and that is what I’d prefer to focus on.’
    maxwell (10:28:20) :
    ‘Very well said. The coin flipping is ‘correct’ in the sense that it is self-consistent and show ‘correct mathematics. The problem is that this example is not pertinent, which I think we all agree on because even if dynamics average out for many flips, such averaging is problematic in climate models. ‘
    I would be grateful if someone could supply some pointers to the literature so that I can learn more about why this claim is incorrect (that modelling trends requires first modelling the detail correctly). I have access to academic journals and some experience in dynamical systems modelling (not weather/climate though), so technical stuff is ok.
    It seems to me that modelling weather is in itself providing a ‘trend’ over a few hours – I doubt that anybody would claim that a weather model has to be able to predict correctly minute by minute, or over very small (a few cubic meters) spatial extents. And yet they can provide some useful information. It’s a just a question of the scalings (both time and space) which one is interested in.

  164. Another excellent,well written, coherent article.
    Great to have the discussions in the open on the best ‘site’ on the net.
    I look forward to the reply to Willis’s reply.Hopefully, other scientists who are proponents of AGW will join in presenting citations, data and evidence for their views, not just opinions, possibilities and computer generated facts.

  165. I meant of course to say:
    …why this claim is incorrect (that modelling trends *does not* require first modelling the detail correctly)

  166. björn (05:41:46) :
    I can,t wait for Dr Meiers response to Eschenbach.
    This is epic science debate, brilliant, both of you have earned my most sincere respect and trust, openess is the solution.

    I wouldn’t hold my breath. Meirs has lost, and surely must realize it. This is why the craftier, less naive Warmists refuse to debate – because they know they are at a distinct disadvantage. About all they can ever do really is hand-wave, try to change the subject, and the usual trickery (cherry-pickin’, appeals to authority, straw men, ad hominems, etc.), because, unfortunately for them, the facts just aren’t on their side.
    But, Meirs gets a gold star for his effort.

  167. “It has helped to identify where the discussion goes off the rails.”
    Brilliant. This is the most important point IMO.
    Now we can nail it all down to these particular areas instead of roaming the intellectual landscape in search of a battle site.
    And whenever I end up in discussion with warmists I can always ask ‘is this real world or model world we’re talking about?’

  168. DeNihilist (10:16:11) : Mr. Pelto (is it Dr.?), {Willis you mention the recession of mountain glaciers, but then spend your time looking at the glacier mass balance data set……}
    Is there any data from the previous warm period of 1930/40 that shows abatement of the glaciers comparable to the last 30 years?
    Yes:
    “The stupefying pace of glacier melt in the 1940s” (GRL 12/2009)
    The most recent studies by researchers at ETH Zurich show that in the 1940s Swiss glaciers were melting at an even-faster pace than at present. This is despite the fact that the temperatures in the 20th century were lower than in this century. Researchers see the main reason for this as the lower level of aerosol pollution in the atmosphere…
    Huss points out that the strong glacier melt in the 1940s puts into question the assumption that the rate of glacier decline in recent years “has never been seen before”. “Nevertheless”, says the glaciologist, “this should not lead people to conclude that the current period of global warming is not really as big of a problem for the glaciers as previously assumed”. This is because it is not only the pace at which the Alpine glaciers are currently melting that is unusual, but the fact that this sharp decline has been unabated for 25 years now.
    http://hockeyschtick.blogspot.com/2010/03/austrian-alps-glaciers-have-almost.html

  169. Will (05:53:25), thanks for your contribution.

    Willis Eschenbach (02:18:26) :

    ‘You say that the trend line is somehow easier to forecast than the individual years temperatures. What this claim neglects is that the models don’t forecast the trend line. The models forecast the temperature of the the individual years (actually, the individual hours). Once the hourly temperatures are forecast, then and only then can the trend line be calculated. If the hourly temperatures are wrong, so are the yearly temperatures, and so is the trend line … so we can place no more confidence in a projected trend line than we can in the individual hourly or yearly temperatures.’

    My apologies if this is a very naive question, I come somewhat late (and probably very uninformed) to this fascinating discussion – but is it not possible to envision a model who’s hourly predictions are sensitive to initial conditions and model parameters, which would make it’s prediction for any given hour quite poor, but on repeated running of the model with variations in initial conditions and/or model parameters a similar trend could be observed through the data – allowing more confidence to be placed in the trend than in the individual hourly or yearly temperatures? I have no idea if this is the case as I am not familiar with the models in question, but is it not possible that there at least a class of models for which this could be true?

    First, you say “My apologies if this is a very naive question.” For me, the only naive question is the one I don’t ask, because then I stay uninformed forever …
    Can we place more confidence in the average of a model run with different initial states and different parameters? These are actually two different questions.
    If we vary the initial states, we are investigating the sensitivity of the model’s responses to where the model starts. This gives us an idea of the variety of possible modeled outcomes.
    If we vary the parameters, on the other hand, we are performing a sensitivity analysis of the settings of the model itself. This will show us how much the outcomes depend on the settings of the model’s dozens and dozens of dials.
    Bear in mind that the parameters have been very, very carefully adjusted and tuned to get the model to agree with the historical temperature trend. Unfortunately, in a complex model this often throws other parts of the model way out of kilter. If you tune the model for temperature, some other aspect (say humidity or clouds or rainfall) will likely get worse.
    The key to answering your question, however, is to note that none of this tells us anything about the variability of the trend line or the hourly temperatures in the real world. All they do is show us the variability of the trend line and the hourly temperatures in the model. And as Fig. 4 above clearly demonstrates, the two (whether averaged or not) are very different. Model results are not evidence about anything but the model itself.
    Finally, I cannot emphasize enough that these models are tuned to reproduce the past. This means absolutely nothing about their predictive ability. It is very, very tempting to think that if a model is successful in reproducing past conditions it will be equally successful in forecasting future conditions. Nothing is further from the truth, particularly when modelling chaotic systems. A good example of such a system is the stock market.
    But the stock market is unlike the climate, in that unsuccessful models get squashed like bugs. How many are left after that squashing? Well … none. Nobody can predict the markets. The difference is that stock market models have to actually perform, whereas climate models only need to impress the credulous and kinda sorta fit historical temperature patterns.
    The climate modellers make the extraordinary claim that, although their models are no better than the stock market models in predicting tomorrows conditions, or next month’s conditions, they can forecast the climate a hundred years from now.
    I see no theoretical reason to believe that is true. The claim is that the general trend of the climate over time is more predictable than the individual hours and days and months of weather over time. But since the general trend is just the average of the individual hours and days and months, if those are bad, then the trend will be equally bad. So where do the models get the enhanced accuracy at long timescales?
    All the best, questions are good,
    w.

  170. Speaking of parameters, the models have dozens of them that they can adjust. This allows them to tune the models so that they can “hindcast” historical climate conditions. And they do a reasonable job of that hindcasting, as the modellers never tire of pointing out and demonstrating with impressive looking maps and graphs.
    But should that impress us? In this regard, the following story by Freeman Dyson bears repeating:

    By the spring of 1953, after heroic efforts, we had plotted theoretical graphs of meson–proton scattering. We joyfully observed that our calculated numbers agreed pretty well with Fermi’s measured numbers. So I made an appointment to meet with Fermi and show him our results.
    Proudly, I rode the Greyhound bus from Ithaca to Chicago with a package of our theoretical graphs to show to Fermi. When I arrived in Fermi’s office, I handed the graphs to Fermi, but he hardly glanced at them. He invited me to sit down, and asked me in a friendly way about the health of my wife and our newborn baby son, now fifty years old. Then he delivered his verdict in a quiet, even voice.
    “There are two ways of doing calculations in theoretical physics”, he said. “One way, and this is the way I prefer, is to have a clear physical picture of the process that you are calculating. The other way is to have a precise and self-consistent mathematical formalism. You have neither.”
    I was slightly stunned, but ventured to ask him why he did not consider the pseudoscalar meson theory to be a self-consistent mathematical formalism. He replied, “Quantum electrodynamics is a good theory because the forces are weak, and when the formalism is ambiguous we have a clear physical picture to guide us. With the pseudoscalar meson theory there is no physical picture, and the forces are so strong that nothing converges. To reach your calculated results, you had to introduce arbitrary cut-off procedures that are not based either on solid physics or on solid mathematics.” In desperation I asked Fermi whether he was not impressed by the agreement between our calculated numbers and his measured numbers.
    He replied, “How many arbitrary parameters did you use for your calculations?” I thought for a moment about our cut-off procedures and said, “Four.” He said, “I remember my friend Johnny von Neumann used to say, with four parameters I can fit an elephant, and with five I can make him wiggle his trunk.”
    With that, the conversation was over. I thanked Fermi for his time and trouble,and sadly took the next bus back to Ithaca to tell the bad news to the students.

    So yes, the models “work”, just like Dyson’s results “worked”. And the models can hindcast fairly well … but with dozens of parameter, it would be a surprise if they couldn’t make the elephant wiggle its trunk. However, as the stock brokerage investment ads always say, “Past success is no guarantee of future performance”

  171. will,
    I think the point, and I don’t even know if this has been posed as a research question, that if one makes an error in the calculations of a climate model, that error will propagate very quickly due to the huge amount of calculations involved in the simulations of climate models. Since the same equations are used for weather models and climate models, it’s good to know that we have the weather figured out for a few weeks before we start talking about years and years. This should be fairly clear if you have dealt with error analysis and propagation in your own work.
    It is also the reason I have hard time believe ‘hindcasting’ because the error in reconstructed data is ill-conditioned (no well-controlled error measurements made on it) and, therefore, it is almost impossible to track how the error propagates in simulations of past climate. This is very similar to the snowball effect.
    In seems that the standard for models of future climate is that they agree with each ‘fairly well’, which was one of the points that Dr. Meier made in his post.
    I think when it comes to your point concerning time and space scaling, one must be very careful with one’s data. If forecasters had data on a one mile by one mile grid of the USA on a basis of every 30 seconds, I think they could provide very good models for the weather on the time scale of minutes. But because weather data does not have that kind of resolution, longer time scales are modeled with more predictive power and the ‘sweet spot’ is a couple days.
    If you’re trying to model the climate with ill-conditioned proxy data, low resolution time series data, no data on feedbacks and ’empirical parameters’ based on some measurements of the real world that may or may not capture the variation of the desired quantity it can be hard going.
    Given the complexity of a physical model of the climate with its interactions between the oceans, land and atmosphere, there very well be some averaging that helps the simulations. There is, unfortunately, no reason to believe this would happen a priori for the whole model simulation however.
    The ultimate test is validity in the real world where, as Willis points out, weather models do pretty good for a couple days and climate models don’t work.
    I understand your confusion, as I am continually confused by all this, because simple statistics seems to tell us that the more one averages, the better. But, as is the case in almost all research fields which I’m sure you know, climate science is not simple statistics.
    Hope that helps.
    Cheers.

  172. Thanks for the reply, Willis. I should mention that I have only taken a real interest in this since reading ‘The Hockey Stick Illusion’ last week, and was utterly appalled by the abuse of the scientific method documented therein (although a mathematician by training, I currently work in neuroimaging – a field which is also prone to play fast and loose with statistics from time to time, which bothers me greatly).
    I’m still not convinced by your argument (although am quite willing to be corrected 🙂 The models which produce the weather forecast for the next several hours may predict ‘a likelihood of some showers’ and this will often turn out to be a good prediction (I need to take my umbrella with me) – however those same models are unlikely to be able to tell me if it is going to be raining from 2.07am until 2.13am, and then again at 2.37am for another 13 minutes. So they have produced a reasonably accurate model of the future weather over one time scale (hours) but not another (minutes). Could the same not be said to apply to climate models? Note, this is a different issue from whether model predictions should be treated as ‘facts’. But it seems to me that predicitive models can have their part to play in scientific enquiry – otherwise there would be little point in ever producing one in any branch of science.

  173. Will (12:17:38) : I would be grateful if someone could supply some pointers to the literature so that I can learn more about why this claim is incorrect (that modelling trends requires first modelling the detail correctly).
    It would only be necessary for you to refer to a decent statistics text book.
    I’ll explain my objections/concerns as follows.
    Suppose you have a random variable with standard deviation ‘Sd’. As it is a random variable, individual observations (the “random variates”) cannot be fully determined in advance. The future values of individula random variates can only be described in terms of probability distributions.
    One further condition: I’ll assume we have carefully sampled the random variable for maximum information in each random variate- this means the unpredicable components in the random variates are statistically independent (each sample provides no information about the random component in the neighbouring sample). [Things tend to get more complicated if this condition is not met, but the additional complications do not alter the conclusion.]
    Let’s say we’d like to predict futre values, but decide that individual random variates are too unpredictableide. This has been the case for “weather” predictions – for example, the Met Office withdrawing its quarterly forecasts.
    It is claimed that we can turn to averaging random variates to gain improvements in predictability (climate versus weather).
    For simplicity, lets say we choose to average 100 samples of the above random variable. On the conditions set out above and using only well established stastistical properties, any mean value we produce from 100 samples will also be random variate. The “standard error” (standard deviation) of this new random variate will be (Sd/10).
    Does that mean the average is more predictable in some useful sense? I say it doesnt.
    All we have done is to transform the problem from one of predicting a variable with standard deviation Sd, to predicting a different variable with standard deviation of (Sd/10). If Sd was intolerable for the underlying data, (Sd/10) will be no more tolerable for the 100-point mean.
    We have added no useful information in moving from one prediction to the other. All we have done is to give ourselves a more slowly moving random variable. But we will also discover that trying to predict the 100-point mean to better than (Sd/10) is no easier than predicting the underlying random variate to better than Sd.
    That’s the way I see it – on arguments of averaging, the climate can be no more predictiable than the weather, when we take into account the appropriate measures of variability.
    Happy to hear other views.

  174. maxwell (07:30:53)

    … It seemed to me that Dr. Meier’s point, in the context of discussing the difference of modeling climate in the short term versus the long term, is that one can have a great deal more confidence in a model of the range of times heads comes up in 10000 flips versus confidence in a model determining whether heads comes up in a single flip. Your comment supports this idea as well, though I don’t know if that was your intent.
    Willis’ assessment, however, seems to make it as though Dr. Meier is saying the opposite on my assessment. It is obviously confusing readers who think they are somehow making an argument against me when they are not.
    Willis,
    would you mind cleaning up the end of that section to reflect Dr. Meier’s point correctly?

    Well, I would if I understood what it had to do with climate models. You say his point is:

    … one can have a great deal more confidence in a model of the range of times heads comes up in 10000 flips versus confidence in a model determining whether heads comes up in a single flip.

    This is not true. The confidence intervals for coin flipping are well established. The standard error of the average for a coin that is flipped N times is 0.5 / sqrt (N-1).
    So we don’t have “more confidence in our model” if we flip a coin a hundred times rather than ten times. We have 100% confidence in our model in both cases. We know for sure that the standard error will be smaller if we flip more times. We have 100% confidence that our model can tell us what the standard error is if we tell it how many times we are flipping the coin.
    But what on earth does this have to do with climate models? The standard error of flipping coins, whether ten or a thousand times, is well encapsulated in a single equation. We know in advance that we will come closer and closer to 50% heads as the number of flips increases, in a manner which is exactly predictable by a well-understood mathematical equation.
    With climate models, unfortunately, we have no corresponding equation, understanding, or mathematical predictability. We don’t know whether it will get closer and closer to some value as the number of iterations increases. In fact, we don’t know if the process is stationary (wanders around some central value) or not, and indications are that it is not stationary.
    Here is the central problem with his analogy, which makes it useless for our present purposes:
    With coin flips, as the number of flips increases, our answer gets more accurate (smaller standard error).
    On the contrary, with iterative models like climate models, as our number of interations increases, our answer gets less accurate.
    That’s why predictions of weather models are no good beyond a few days, where predictions of coin flipping have less error the more times we flip.
    So whether Dr. Meier’s point is true or not, and whether I understand it or not, is immaterial, because coin flipping has absolutely nothing to do with climate models.

  175. Bob (Sceptical Redcoat) (09:25:12)

    Willis: May I please take a copy of this brilliant discourse? I promise that I will not claim any of it to be my own work, aka Mann! Also, are/were you ever a cowboy or a fisherman? Sceptisim, coupled with the patience of a fisherman and the toughness of a cowboy, seem to me to be qualities every bit as necessary as academic achievement, in order to make a true scientist.

    These days, I toss my ideas onto the electronic winds so that they may possibly take root wherever they fall. I’m not doing this to get credit (although I don’t mind getting credit). I’m doing this to get results. Use them as you wish.
    As to whether I have been a cowboy or a fisherman, I have been both, along with many other trades. My CV is here, it’s the record of a life lived foolishly and furiously at the edge of the envelope. My motto has always been “Retire early … and often”. Not a path I’d counsel for most folks, but it has served me well.

  176. bob (09:39:38)

    Willis,
    Your post,

    Finally, if that is the null hypothesis, the alternate hypothesis is that the factors that affect the climate now are not the same ones that affected it in the past. I know of no one making that claim, that the present climate runs on different principles than the past climate. So it is a null hypothesis for an argument that no one is making … which is why it is a “straw man” null hypothesis.

    I think the case can be made that the IPCC is making the claim that there are greenhouses gases affecting climate now that were not affecting climate in the past, namely the chlorofluorocarbons. Which only requires assuming that chlorofluorocarbons are predominantly man-made.
    Same principles but different gases, and as for the important one that we can do something about, CO2, there are people making the claim that man-made CO2 is having an effect on the climate over and above the effect of naturally produced CO2.

    My take on Dr. Meier’s NH2 was that he meant that the classes and kinds of factors that affect the climate (e.g., greenhouse gases, volcanic eruptions, changes in the sun) are the same now as they were in the past. I agree with that, with the proviso that we don’t know what many of those factors are/were.
    Yes, we’ve added a couple new very minor GHGs to the mix, but I don’t think that was what he was referring to.

  177. Steve in SC (09:40:19)

    5. The increase in CO2 is due to human emissions. There are two ways we know this. First, we know this simply through accounting – we can estimate how much CO2 is being emitted by our cars, coal plants, etc. and see if matches the observed increase in the atmosphere; indeed it does (after accounting for uptake from the oceans and biomass). Second, the carbon emitted by humans has a distinct chemical signature from natural carbon and we see that it is carbon with that human signature that is increasing and not the natural carbon.
    My Comment: Agreed.

    My comment: Absolute BS
    You have no evidence of this.

    Steve, without commenting on whether or not you are correct, your style of argumentation doesn’t add to the discussion. If you think Dr. Meier is wrong, it doesn’t do anyone any good to just call BS, even if you are right.
    Go out and read the scientific papers on the subject of the variations in carbon isotopes in fossil fuels and in the atmosphere. Familiarize yourself with the arguments both pro and con. Research the authorities. Run the numbers yourself, don’t trust anyone’s calculations. Find the citations that support your point of view. Use them to demonstrate, not claim but demonstrate, that what Dr. Meier said is wrong.
    Until you do that, I fear you have not added anything to the discussion other than your vote … but science is not settled by a vote (no matter how many times we’re told that there is a “consensus”). It is settled by evidence and logic and math.
    All the best,
    w.

  178. Willis (or anybody else with the skill),
    like a true skeptic I went back to the source data and plotted my own graph for Figure 3. It raised a few questions I hope can be answered.
    1) When I plot that data and add a linear trend to the whole data set I don’t see such a divergence in trends as is obvious from the article graph.
    I’m using Excel so I don’t have Gaussian average (in fact I don’t know what a Gaussin average does to the data). The website the data is derived from tends to use smoothed data. What is the strength in using this averaging?
    2) I’m also unsure that there is any strength in comparing the trend from 93-04 with the trend from 05-09. The trend is constantly changing over different time periods. For example 92-97 would show a lower trend than for the whole data set. I guess on the basis of that argument you could also ask the strength of the trend from 93-09. My question is are these short term flutuations in data really relevant?
    3) Finally the colorado website seems to agree with Willis. They plot a graph with a linear trend and don’t mention anything about an increasing rate of sea level rise. I was wondering should the predicted increase in the rate of sea level rise be showing up in the present observed data or is this something that will only start appearing down the line? What do the models predict about significant, observable rate change rise?

  179. I’ll answer my own question 3!!
    This section of the IPCC contains the relevant info.
    Climate Change 2007: Working Group I: The Physical Science Basis
    Section 5.5.2.4 Interannual and Decadal Variability and Long-Term Changes in Sea Level
    Quoting
    “Interannual or longer variability is a major reason why no long-term acceleration of sea level has been identified using 20th-century data alone”
    It seems nobody expects to see the increasing rate of sea level rise at the moment. Willis have you set up a straw man?

  180. John Coleman (10:08:57)

    Deepest graditute to Dr Meiers and Willis Eschenbach and the legion of commenters. I am engrossed and educated by reading your work.
    I feel like I came in mid way through the debate and without a program, however.
    Who are you Meiers and Eschenbach? I would like to see your bios and learn of your earlier writings.

    Dr. Meier is a research scientist at the National Snow and Ice Data Center. His bio and publication list is here.
    I am (inter alia, and lots of alia at that) an amateur scientist and independent climate researcher. My bio is here.
    Some of my earlier writings are:
    Underground Problems with Mann-Holes
    An Analysis of the Topex Sea Level Record
    Data Smoothing and Spurious Correlation
    Can’t See the Signal For the Trees
    The Thermostat Hypothesis
    Tropical Tropospheric Amplification, an invitation to review this new paper
    Why Copenhagen Will Achieve Nothing
    The Steel Greenhouse
    The people -vs- the CRU: Freedom of information, my okole…
    When Results Go Bad …
    The Smoking Gun At Darwin Zero
    Willis: Reply to the Economist
    Darwin Zero Before and After
    The Unbearable Complexity of Climate
    Climate, Caution, and Precaution
    Where Are The Corpses?
    Floating Islands
    Congenital Climate Abnormalities
    Fudged Fevers in the Frozen North
    Judith, I love ya, but you’re way wrong …
    Sense and Sensitivity
    Himalayan Hijinks
    Another Look at Climate Sensitivity
    More on the National Geographic Decline
    GISScapades
    Skating on the Other Side of the Ice
    Carbon Emissionaries
    Trust and Mistrust
    Conservamentalism
    Global Radiation/Conduction/Evaporation Climate Model (Excel Spreadsheet)
    Nature Magazine “Communications Arising” on Lake Tanganyika
    E&E article on Tuvalu
    E&E article on Svalbard

    And, I would like to encourage you to focus more on the core issue, as I see it. What proof is there of the forcing impact of CO2 in its greenhouse role and what is the counter to that argument.
    Second, what proof is there of warming beyond the long term natural warming of the interglacial period.

    I have discussed these questions in a number of my writings. I would note in passing that there is almost never “proof” in science, other than in a mathematical sense. All we can do is falsify claims, we can never prove them.

    I agree with all who say this is the most productive debate, point-counter-point, scientific exchange to date in the 30 years of controversy.

    Exaggeration, but I love it.
    w.

  181. Regarding CO2 and water and reradiation:
    http://pubs.acs.org/subscribe/journals/ci/31/i11/html/11box.html
    which references (worth your time to read):
    http://pubs.acs.org/subscribe/journals/ci/31/special/may01_viewpoint.html
    Dr. Robert Essenhigh later worked out the calculations for absorption and published here:
    http://pubs.acs.org/doi/abs/10.1021/ef050276y
    subscription required. (Libraries may have subscriptions. If anyone can slog through this and post a summary, I would appreciate it. I’d be willing to buy the paper myself, but I’m not sure my calculus skills are still up to the task.)
    The main point he makes is that H2O gas accounts for ~80% of what we call the greenhouse effect, and CO2 accounts for essentially the rest.
    Apparently RealClimate beat on Dr. E back in the day.
    Dr. E had an MS student do this:
    http://etd.ohiolink.edu/view.cgi?acc_num=osu1259613805
    Study of Energy Balance Between Lower and Upper Atmosphere
    I haven’t read the thesis yet.

  182. @ Willis Eschenbach (13:13:39) :
    I see no theoretical reason to believe that is true. The claim is that the general trend of the climate over time is more predictable than the individual hours and days and months of weather over time. But since the general trend is just the average of the individual hours and days and months, if those are bad, then the trend will be equally bad. So where do the models get the enhanced accuracy at long timescales?
    They do not get enhanced accuracy over longer timescales, they only get a higher probability of showing a result that is similar to reality, which some people seem to confuse with ‘correct’.

  183. Aargh. (15:45:20), you raise a good question:

    How do you propose to prove that there is no proof, Willis

    Proof? Well, first I establish the null hypothesis:
    There is proof in science.
    Then since this is climate science, to falsify the null hypothesis, I merely need to quote from that well known scientific work, “The Treasure of the Sierra Madre”, viz:

    Proof, to god-damned hell with proof! We have no proof. In fact, we don’t need proof. I don’t have to show you any stinking proof, you god-damned [snip] and [snip]!

    Then I just say “Q.E.D.”, and go back to collecting my government grant to study the effect of climate change on Man-In-The-Moon marigolds …

  184. Willis Eschenbach (21:28:38) :
    “Well, if I knew what the “Animal House parade” was, I suppose that would make sense … and I also don’t understand what a “jigsaw piece” has to do with the image. But then I was born yesterday, what do I know?”
    Sorry for the obscure reference Willis. The reference to the jig saw piece was more of a flashback to all those PowerPoint presentations I sat through that dealt with team building or management solutions. (Time actually stood still during those sessions.) It stems from my natural aversion to Microsoft Clipart. My therapist is helping me deal with it.

  185. On the Willis Eschenbach and Walt Meyer climate debate
    by Arno Arrak
    I found the point counter point questions of Willis Eschenbach (March 31) and Walt Meyer (April 8) interesting and illuminating. Eschenbach says he wants to “…detail my own beliefs about the climate and how it works.” Meyer then provides, as best he can, “…the current thinking of most scientists working in the various aspects of climate science.” Reviewing these questions in my mind I realized how much misinformation is out there and how far from the truth current thinking is. Below I will try to correct that with comments as needed and will stay with the numbering of the questions. By their nature some questions require a direct response from me. My information is based on research I did for “What Warming?” now available on Amazon.com. Al Gore made me do it. Here we go.
    Question 1: Does the earth have a preferred temperature which is actively maintained by the climate system?
    This question smacks of Lovelock’s “Gaia” hypothesis. It is true that the “faint young sun” paradox has yet to be explained but I am not ready to go with either Willis or Lovelock on this. Walt Meyer is right that it is not relevant to the existence or otherwise of AGW.
    Question 2: Regarding human effects on climate, what is the null hypothesis?
    Both agree to a null hypothesis that any changes in climate are due to natural variations. But Walt Meyer adds null hypothesis 2 (NH2) which brings in a historical perspective, that the past is the key to the future.
    Question 3: What observations tend to support or reject the null hypothesis?
    Willis can’t find anything in the record that is in any way unusual or anomalous and rightly points out that the Medieval Warm period was both widespread and warmer than the present. I agree with him. But Walt Meyer steps up and gives us a grab-bag of climate lore to overthrow the null hypothesis. Some of these things are just plain wrong: Pinatubo, for instance, did not depress global temperature as he and others claim because its cooling was restricted to the stratosphere and never reached ground level. But James Hansen came out with a GISS climate model just for Pinatubo and promised to “…estimate the predicted global cooling on such practical matters as the severity of the coming Soviet winter and the dates of cherry blossoming next spring…” Intended, no doubt, for that year’s issue of the “Collective Farmers’ Almanac.” Walt goes on from there and states that “…we are seeing indications that the climate is being affected by changing concentrations of greenhouse gases, primarily CO2.” That is complete nonsense: concentration of a gas and its effect on temperature are two different things. To prove existence of greenhouse effect not just in the laboratory but in nature you must be able to demonstrate that partial pressure of carbon dioxide and air temperature change in parallel. If you cannot do that your hypothesis is false. And you must prove that it is actually happening now, today, and not in some geologic past. That is your obligation because IPCC says it is true and makes recommendations based upon it. I see no sign that it has happened anytime since 1978 when satellites first began to measure temperature. Nevertheless, believing IPCC to be correct, the EPA has declared CO2 to be a pollutant and Congress has gone crazy and passed a cap and trade law. I can demonstrate the complete falsity of this belief from climate data. We know from satellite and other sources that from 2002 to 2007 world temperatures did not change while CO2 partial pressure kept on increasing. On top of that a substantial cooling followed in 2007 which bottomed out in 2008. It is totally impossible to explain any of this by the greenhouse effect. Knowing that this spells trouble for their theory Keenlyside has come out with damage control. He claims that natural factors like the Gulf Stream may have interfered with warming but not to worry, warming will be back in fifteen years. Fifteen years? If non-carboniferous factors have been in charge of world temperature since 2002 and will continue to be so for the next fifteen years then that greenhouse theory of yours sure as hell isn’t working and you might as well get rid of it. Keenlyside is just fighting a rearguard action, promising that warming will be back to keep the faithful happy. But Walt still has eight more points to make after his greenhouse claims and he lists them. As he himself says, they may be completely unrelated to the GHG’s in the atmosphere and they are.
    Question 4: Is the globe warming?
    The question is overly broad because a time frame is not specified. I choose to limit the observational time frame to the satellite era for which reliable temperature measurements exist. Satellites measure Oxygen microwave emission line intensities from the lower troposphere which are thermally excited and hence a proxy for local air temperature. These are polar orbiting satellites that sample the entire globe uniformly and are not subject to site selection bias as all ground-based measurements are. The most recent global warming before the beginning of the satellite era lasted a century and brought us out of the Little Ice Age. It ended with the start of World War II. From that point on until 1998 the climate was stable or even cooled a bit and some speculation about a coming ice age was published. For the last twenty years of this period satellite records are also available. But those same twenty years are shown by NASA, NOAA, and the Met Office as a period of rising temperatures. How is this possible? Satellite record shows only a multi-year temperature oscillation, up and down by half a degree, but no rise until 1998 (Figure1.)
    Figure 1. Data from two satellite systems – UAH (University of Alabama Huntsville) and RSS (Radiation Sensing Systems) – plotted on a common curve.
    There are five such cycles within that twenty year period and they are not random but correlate with the ENSO system in the Pacific. This is very different from a steady warming. It is clear that there was no warming and that all three curves showing it are cooked. As in falsified. It is pure scientific fraud but in the middle of this period James Hansen gets up in front of the Senate and testifies that global warming has started and that carbon dioxide we are putting into the air is its cause. Their fake warming starts suddenly about 1977 but checking the Mauna Loa carbon dioxide curve does not show a corresponding increase in carbon dioxide that laws of physics require. Hence, both claims made by Hansen in 1988 are false. Nevertheless, his testimony sparked the establishment of the IPCC, which was in the planning stages, as well as the Kyoto and Copenhagen agendas that were to follow. And now consider this: without Hansen’s warming there was no greenhouse warming at all in the twentieth or the twenty-first centuries. Greenhouse warming has simply never been observed, then or since. This does not mean that there has not been any warming. Real warming did start when the 1998 super El Nino showed up. But its cause was not carbon dioxide in the air but a storm surge in the Indo-Pacific region that brought warm water of the Indo-Pacific Warm Pool to the start of the equatorial countercurrent near New Guinea. The countercurrent then carried it to South America where it spread out and caused the super El Nino we observed. That is the regular route of all El Ninos we get. The one-time global temperature increase from that super El Nino was a full degree Celsius, more than the recorded temperature increase of the entire twentieth century. Its left-over warm water was responsible for the twenty-first century high, a run of six warm years from 2002 to 2007. Most of them were among the top ten and collectively they made that decade the warmest on record. But temperature stagnated during this period while carbon dioxide kept going up at the rate it has been for the last fifty years. This drove the model-makers nuts who were feeding carbon dioxide into their models and expecting temperature increase from that. All that came to an end with a La Nina cooling in 2008 that Kevin Trenberth of CRU could not understand. It is actually simple: the La Nina signifies the resumption of the oscillatory climate of the eighties and nineties that NASA, NOAA, and the Met Office obliterated with their deception to get that late twentieth century warming on the books. From now on, look for an alternation of warm El Nino and cool La Nina periods like those that existed before 1998.
    Question 5: Are humans responsible for global warming?
    Both are wrong – humans are not warming the planet. As I pointed out above there is a criminal conspiracy pushing that lie that needs to be brought to justice.
    Question 6: How are humans affecting the climate?
    Both list greenhouse gases and other nonsense from literature, none of which are implicated for the simple reason that humans are not affecting the climate.
    Question 7: How much of the post 1980 temperature change is due to human activities?
    Walt: “…there is broad agreement that human-emitted CO2 has significantly contributed to temperature change.” Willis: ”…there is no indication that post 1980 temperature rise is in any way unusual.” I have already come down on this question but it is so important that we need to look at the details. You can find out what is going on by directly comparing satellite and land-based temperature curves. Let’s take, say, HadCRUT3 from the Met Office, and compare it to UAH MSU satellite data as in Figure 2.
    Figure 2. UAH MSU satellite (red) and Met Office HadCRUT3 (blue) temperature curves compared. Observe how Met Office rising trend is achieved.
    You notice right away that they start by cherry picking the El Nino peaks and then raising up the low La Nina temperatures in between. But this only works with the first four El Ninos. The fifth one is too low so it gets raised up bodily. The super El Nino is next and is gratefully incorporated even though it is non-carbonaceous in origin. The twenty-first century high, a run of warm years near the El Nino maximum, follows. But this is just not high enough for them so the entire right side of their graph gets raised up and floats above the satellite curve. NOAA is worse: while HadCRUT3 at least retains the greatly reduced La Nina valleys they stay with the peaks, jettison all low values in between, and raise up the twenty-first century high as well. NASA (Land-Ocean, from Hansen 2006) starts out exactly like Hadcrut3. But they don’t have the nerve to raise up peaks so these are all in place and so is the twenty-first century high. Only the super El Nino is off (too low) and their data point for 2005 is too high. In all cases the warming in the eighties and nineties has been manufactured by distortion from an originally horizontal and oscillating curve. What real warming there is comes from the super El Nino and its aftermath and is not carbonaceous. The answer to question 7 is thus: NONE AT ALL! This sustained and coordinated scientific fraud by guardians of global climate data is the only thing that presently holds up the claim that AGW is happening. Since it involves three organizations it is also a criminal conspiracy and should be internationally investigated. A climate Nuremberg perhaps?
    Question 8: Does the evidence from climate models show that humans are responsible for changes in climate?
    Walt has faith in models, Willis does not. Willis is correct – models are a Trojan horse for introducing secret bias and passing it off as a “scientific” fact. They consume huge amounts of supercomputer time that Uncle Sam pays for and produce worthless predictions expressing the prejudices of the modelers. But it did not start with climate models – the Club of Rome was first. They were concerned with the “predicament of mankind” and commissioned a world model from MIT to predict the future. It was published as “The Limits to Growth” by Meadows and Meadows in 1972. They predicted that the world would run out of oil in the nineties and that civilization would collapse in the twenty-first century. When this did not happen their true believers started to “update” it every ten years and may still be doing it for all I know. Pilkey and Pilkey-Jarvis have surveyed many such failed attempts to model natural processes, from predicting cod fishery yields, to environmental impact statements, climate forecasts, beach erosion problems, Yucca Mountain drainage, sea level rise, and more, and list their failure modes. Such models often involve approximations or guesstimates. They also may depend strongly on initial values which are poorly known but may have a strong influence on the outcome. Plus “adjustments” may be required to make them correspond to reality and these adjustments are nothing more than fudge factors. They are opaque to users and political pressure can be, and has been, exerted to get the “right” answer which is then passed off as a “scientific” fact. They have two words to describe all this: “Useless Arithmetic.”
    Question 9: Are the models capable of projecting climate changes for 100 years?
    Willis says “No.” but Walt fudges it for more than a page. His argument is statistical and verges on, but does not actually utilize, statistical mechanics. Statistical mechanics says that if there are millions of identical molecules the properties of the assembly converge on macroscopic observables. I cannot see this level of input uniformity in averaging climate data and have to reject his theory. And I don’t want to wait a hundred years either. But that won’t be necessary because the current models are all guaranteed to fail. For a starter, they are programmed to use the fake warming of the eighties and nineties as input to be extrapolated. And secondly, they all utilize atmospheric carbon dioxide to compute the greenhouse effect. Since there is no observable greenhouse effect they all spectacularly failed when the twenty-first century high and the 2008 cooling appeared. And since our future climate will be an oscillating climate similar to that of the eighties and nineties there is no hope that any of them can make any meaningful predictions from now on.
    Question 10: Are current climate theories capable of explaining the observations?
    Willis says no and he is right. Walt says yes if one includes the greenhouse gases. “Increasing greenhouse gases should result in increasing temperatures and that is what we have observed” he amplifies. This is incorrect. Only if temperature increases in step with CO2 concentration can we postulate a causal relationship, and this is not happening. Not only is the temperature not cooperating, it is positively contrarian when it decides to sit still as it did from 2002 to 2007, or decrease as it did in 2008 while CO2 kept increasing on its metalled ways of time past and time future.
    Question 11: Is the science settled?
    I have to say yes: science proves the absence of AGW.
    Question 12: Is climate science a physical science?
    Does it have to be a science? Why not call it a religion which it really is?
    Question 13: Is the current peer-review system inadequate, and if so, how can it be improved?
    The peer-review system is simply broken. No papers questioning the existence of anthropogenic global warming can be published. I know, for a year ago I submitted one to Science, Nature and PNAS and was turned down. I did not fail peer review, they simply did not dignify me with sending it out for peer review. No criticism, no explanation, no nothing. Wegman did a network analysis for the U.S. Congress and unearthed a tightly knit group of scientist that control publication. Climategate shows that they block publication of papers they do not like and threaten editors of journals who do not toe the line. The effectiveness of their grip on the system is shown by Naomi Oreskes who could not find any opposition to the “consensus” view in over nine hundred papers she surveyed.
    Question 14: Regarding climate, what action (if any) should we take at this point?
    The answer is simple: stop the insanity of fighting an imaginary danger. It is costly, irrational, and designed to destroy civilization as we know it. Let me give you some examples. The EU has declared that ten percent of its transport should use biofuels by 2020. I bet you did not know that this requires that seventy percent of its cropland should be devoted to producing nothing but biofuel. Grain prices are already sky high thanks to current projects and there have been food riots in poor countries as a result. For England this means using their entire present grain harvest for biofuel and requires her to import as much grain as she now grows. No one has any idea of where that grain will come from. Or take the windmills. Denmark is ahead of everyone in windmill land. But in 2002 they declared a moratorium on new windmill projects. Why? Because the wind does not blow steadily. They found that when the wind was light they had to buy expensive electricity from Germany. And when it was strong they had an excess on their hands. You cannot store it so they ended up selling it to Norway below their own cost. It was a lose-lose proposition and the Danish people are now paying the highest electricity rates in Europe. And if you think that windmills are carbon free think again. It turned out that because of the uncertainty of wind speeds the outputs of individual systems kept fluctuating and it was necessary to keep a conventional “spinning reserve” on hand to take up that unpredictable slack at a moment’s notice. They get approximately eight percent of their electricity from these windmills. The same amount of power could be supplied at far lower cost by just one conventional coal-fired power station. If the Waxman-Markey ever becomes law we are in for a whole lot of such irrational actions, all to fight a non-existent warming.
    [Sorry – the figures did not come through in this format. AA]

  186. 7. Recession of most mountain glaciers around the globe
    I know in recent years that it has been portrayed that nearly all glaciers are in retreat.
    I came across The World Glacier Monitoring Service’s preliminary mass balance data for 2007/08 (the most recent year). It shows continuing overall mass loss from all monitored glaciers but interestingly about 1/3 advanced in that particular year.
    http://www.geo.unizh.ch/wgms/mbb/sum08.html

  187. I’ll ask you straight up Willis.
    Do you believe in the skill of RTE, radiative transfer equations, and what they tell us about how concentrations of various GHGs in the atmosphere govern the propagation of radiation.
    Simple yes or no.

  188. Willis Eischenbach says: So we don’t have “more confidence in our model” if we flip a coin a hundred times rather than ten times. We have 100% confidence in our model in both cases. We know for sure that the standard error will be smaller if we flip more times. We have 100% confidence that our model can tell us what the standard error is if we tell it how many times we are flipping the coin.
    But what on earth does this have to do with climate models?

    I may be way off here, but when I read Meier’s post, through the coin-toss stuff, into his “predictability feel-good” stuff, I assumed he was trying to build confidence, ala The Particle Physicists who use probability to define where an electron is at a point in space-time. He used a very simple, tangible, everyday system (coin-flip),showed how we can “feel good” about confidence intervals (indirectly), and then moved on to his climate models, begging us to “feel good” about their potential for predictability…the more we learn, the more likely they are to be taken seriously.
    So, I don’t think he made a stretch at all, bridging the coin-flip with climate modeling.
    I still happen to disagree with his conclusion, lol, based on everything else already commented here and elsewhere (climate models will NEVER be reliable predictors!), but I wish everyone would stop getting all “puffy-chested” about the coin-flip scenario. He used it as a thought-experiment, nothing more or less, get over it already.
    I’ve gotten to the point that I don’t even bother reading posts when I find “Coin” in the initial scan. It’s boring. Move on.

  189. Willis Eschenbach (02:18:26) :
    Wren (00:33:53), thanks for your comments.
    “As a result, the idea that climate models can project the climate a hundred years out because “a single coin flip is simpler than 10000 coin flips” is untrue, simplistic, and in no way a metaphor which would help us understand the problem with long-term climate model projections of the future.”
    ——————–
    I think the coin-flip analogy was supposed to show that you can be more confident in the outcome of a 1,000 flips than a few flips. Even though the odds are against two successive heads or tails (only 0.25), it happens frequently. The odds against 1,000 successive heads or tails, however, are virtually nil.
    I understand this, but it has absolutely nothing to do with climate models. Coin flips are the repetition (however many times) of a single random event which is not affected by any variables such as humidity. I can’t think of too many things that are more different from climate models, which are a repeated iteration of multi-variable non-random events occurring in simulated 3D space … how are those similar by any measure?
    Similarly, you can be more confident in a projected trend line for a large number of years(e.g. 2010 – 2050) than you can be for an interpolated value for any individual year within that period. You can appreciate this by plotting a regression line through actual historical temperature records for a like number of years. You will find that line does not describe year to year changes very well but does describe the long-term trend.
    I’m not following this one. You say the trend line does not describe year to year changes but does describe the trend … this seems like a tautology.
    And the idea that we can simply project a trend line from now to 2050? Mark Twain saw your argument coming when he said:
    In the space of one hundred and seventy-six years the Lower Mississippi has shortened itself two hundred and forty-two miles. That is an average of a trifle over one mile and a third per year. Therefore, any calm person, who is not blind or idiotic, can see that in the Old Oolitic Silurian Period, just a million years ago next November, the Lower Mississippi River was upwards of one million three hundred thousand miles long, and stuck out over the Gulf of Mexico like a fishing-rod.
    And by the same token any person can see that seven hundred and forty-two years from now the Lower Mississippi will be only a mile and three-quarters long, and Cairo and New Orleans will have joined their streets together, and be plodding comfortably along under a single mayor and a mutual board of aldermen. There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.
    Take another look at Fig. 4 above to see why we can place no more confidence in the climate models than we can in Mark Twains trend line projections. Clouds are central to the projection of the evolution of the climate. The models can’t get the historical clouds anywhere near right, much less project the evolution of the future clouds …
    You say that the trend line is somehow easier to forecast than the individual years temperatures. What this claim neglects is that the models don’t forecast the trend line. The models forecast the temperature of the the individual years (actually, the individual hours). Once the hourly temperatures are forecast, then and only then can the trend line be calculated. If the hourly temperatures are wrong, so are the yearly temperatures, and so is the trend line … so we can place no more confidence in a projected trend line than we can in the individual hourly or yearly temperatures.
    I hope this is clear, if not, please clarify what you are saying.
    ======
    I will try to clarify what I am saying by using a recent graph from Steve Goddard as a visual aid. You will need to refer to this graph to follow my comments. It is the top one in the following link:
    http://docs.google.com/View?id=ddw82wws_5998xrxhzhc
    The blue line in the graph is the actual temperature records from 1890 to 2009 by year(3-year intervals). As you can see, it is jagged with lots of peaks and valleys, reflecting the short-term fluctuations in temperature. In contrast, the red line in the graph is a straight line describing the long-term (120-year) trend in temperature.
    You can see the straight red line does a good job of representing the 1890-2009 trend. However, by interpolating along this line, you also can see it does not do well at representing the temperature for the points in time within the 120 years.
    As a simple exercise, we could extrapolate the 1890-2009 trend out to the year 2100, and that naive projection might seem to most a more reasonable than a no-change extrapolation(i.e., 2100 having the same temperature as 2009). As in our historical series, however, we would not expect our straight line to do a good job of intersecting with the actual future yearly values.
    Now, how about overlaying that saw-tooth looking 1890-2009 line of actual values over our 2009-2100 trend line? Would that be a good idea? I don’t think so. Even if the pattern repeated itself, if we get it a little to far to the left or right, we will be wrong for lots and lots of years. It will zig when we said zag, and zag when we said zig. The long-term trend, however, we can look at with more confidence. What it’s been doing, it probably will keep doing unless, something unforeseen comes along.
    Real projections of temperature into the future are trend projections, and usually look like that red line in that they are straight or smooth. I have never seen one that looked like the blue line with all those peaks and valleys. I think you said you have. I would like to see it. Please provide a link .

  190. Holy peer review Batman!
    This is what scientific thinking and debate is supposed to look like! Thank you very much for an excellent discussion, including all the other ‘peer review’ in the comments. It is all so educational, and reveals the complexities involved, some of the known unknowns, and hints at the unknown unknowns. The real story is so much more complex than the simple CO2 story, but the latter is of course far easier and more convenient to sell.
    Willis your comments about ‘models and reality’ reminded me of an 1980s paper by the same name that applied to the field of bear biology. It was obviously ignored because they too have moved from messy reality and gone into programmable model bears and model bear habitats with similar effect and similar convenient results.
    Finally, Theo Goodwin (14:54:56) wrote: “In the field of climatology, Willis Eschenbach is the last defense against scientific nihilism or Lysenkoism. No one can construct a reasonable criticism of anything Eschenbach has said in this wonderful article.”
    Oh yeah!!! 2 + 2 = 5!!! Denier, denier, denier!!! Throw him in jail Comrade!!!
    I’m sure some of the more zealous Watermelons would find that to be a perfectly reasonable criticism. They might even publish a peer reviewed paper supporting that in their distinguished Journal of PseudoScientific Mob Fundamentalism.
    And you question Lysenko? The consensus agreed that his methods were correct, didn’t they?

  191. steven mosher (20:19:13)

    I’ll ask you straight up Willis.
    Do you believe in the skill of RTE, radiative transfer equations, and what they tell us about how concentrations of various GHGs in the atmosphere govern the propagation of radiation.
    Simple yes or no.

    You’ll have to refine the question, Mosh, I don’t understand it. What do the RTEs tell us about how GHGs govern radiation propagation? That’s a complex field. An example might help.
    I do think that (generally) changes in GHGs do change the radiative forcing. However, I think it is a third order effect. Total radiation striking the surface of the earth is ~ 500 W/m2 (~ 170 W/m2 solar and ~ 330 W/m2 from GHGs). So a doubling of CO2, if it actually does change the forcing by 3.7 W/m2, is less than a 1% change in the forcing. I divide forcings into more than 10% of the total (first order), 1% to 10% of the total (second order), and less than 1% of the total (third order).
    And that’s the change from a doubling. The change year to year in theoretical CO2 forcing is ~ 0.03 W/m2. That’s lost in the noise.
    When a cloud comes over the midday sun in the tropics, the change in forcing is on the order of several hundred W/m2. The average change in the tropics from the change in clouds between 10:30 and 11:30 AM is on the order of 60 W/m2. These totally swamp the change from a doubling of CO2. Clouds to me are not a “negative feedback” or a “positive feedback”. They are an active, first-order forcing. In the tropical mornings, no clouds, you get full sun. At mid-day, as soon as a critical temperature is crossed, clouds form, cutting down the sun. If it is warm, clouds form earlier, cooling the planet. If it is cool, clouds form later, warming the planet. That’s not a feedback. That’s an active control system, warming the planet when it is cold, and cooling it when it is warm. But I digress …
    I know that doesn’t answer your question, so if you could make your question clearer, I may be able to give a yes/no answer.
    Thanks,
    w.

  192. Wren (20:54:08)

    I will try to clarify what I am saying by using a recent graph from Steve Goddard as a visual aid. You will need to refer to this graph to follow my comments. It is the top one in the following link:
    http://docs.google.com/View?id=ddw82wws_5998xrxhzhc
    The blue line in the graph is the actual temperature records from 1890 to 2009 by year(3-year intervals). As you can see, it is jagged with lots of peaks and valleys, reflecting the short-term fluctuations in temperature. In contrast, the red line in the graph is a straight line describing the long-term (120-year) trend in temperature.
    You can see the straight red line does a good job of representing the 1890-2009 trend. However, by interpolating along this line, you also can see it does not do well at representing the temperature for the points in time within the 120 years.
    As a simple exercise, we could extrapolate the 1890-2009 trend out to the year 2100, and that naive projection might seem to most a more reasonable than a no-change extrapolation(i.e., 2100 having the same temperature as 2009). As in our historical series, however, we would not expect our straight line to do a good job of intersecting with the actual future yearly values.
    Now, how about overlaying that saw-tooth looking 1890-2009 line of actual values over our 2009-2100 trend line? Would that be a good idea? I don’t think so. Even if the pattern repeated itself, if we get it a little to far to the left or right, we will be wrong for lots and lots of years. It will zig when we said zag, and zag when we said zig. The long-term trend, however, we can look at with more confidence. What it’s been doing, it probably will keep doing unless, something unforeseen comes along.
    Real projections of temperature into the future are trend projections, and usually look like that red line in that they are straight or smooth. I have never seen one that looked like the blue line with all those peaks and valleys. I think you said you have. I would like to see it. Please provide a link .

    Climate models provide the blue line with all those peaks, that’s what the output of their runs look like. They do not just project a trend line. The trend line is created afterwards. They create a projection with all those peaks. Likely the most famous one is Hansen’s original 1988 climate model projection, viz:

    All of those except the red line are climate model projections of the future, using different assumptions. No trend lines, just annual projected temperatures.
    w.

  193. *applause*
    You got style, man, no question about that.
    “there is no proof in science’ is a self contradictory statement and therefore false.
    this is how science dies – from the failure to properly define the nature of identity. inversion of the law of identity is the hallmark of post normalism and the logical result is substitution of consensus or revelation or even whim.
    when you can’t tell them apart, you can’t do science.

  194. @ Willis Eschenbach (22:04:44) :
    Sorry, but when I see phrases like ‘Estimated temperature in (insert million or billion year old era)’ I often wonder what the level of significance was in creating the underlying work. I accept that it is possible to recreate these temperatures, but as Steve McIntyre has shown, there can be issues with the use of proxies. Is there a proxy that we can compare to current temperatures? Besides, of course, tree rings?

  195. Dave F (22:44:59)

    @ Willis Eschenbach (22:04:44) :
    Sorry, but when I see phrases like ‘Estimated temperature in (insert million or billion year old era)’ I often wonder what the level of significance was in creating the underlying work. I accept that it is possible to recreate these temperatures, but as Steve McIntyre has shown, there can be issues with the use of proxies. Is there a proxy that we can compare to current temperatures? Besides, of course, tree rings?

    I looked through my post that you cited [Willis Eschenbach (22:04:44)] and found nothing about estimated temperature. In fact you are the only person to use the term in this whole thread, so it’s not clear what you are referring to.
    In any case, certainly there are problems with the use of any and all proxies, but they’re all we have to understand the past. I take them all cum grano salis …

  196. @ Willis Eschenbach (23:41:15) :
    Sorry, it is in the graph in the post I referenced.
    “Estimated temperatures in altithermal and eemian times.”
    Isn’t Scenario C the ‘Drastic reduction in CO2’ scenario? That looks like where we are tracking, so I think that it is interesting.

  197. Willis,
    You know what RTE’s are. But if you want a practical application of the physics and how one actually uses it. here:
    www-ee.ccny.cuny.edu/www/web/eebmg/ee330/05_rte.pdf
    So. Do you accept that the physics of RTE is sound science?
    Simple question.

  198. HR (19:05:24)

    7. Recession of most mountain glaciers around the globe
    I know in recent years that it has been portrayed that nearly all glaciers are in retreat.
    I came across The World Glacier Monitoring Service’s preliminary mass balance data for 2007/08 (the most recent year). It shows continuing overall mass loss from all monitored glaciers but interestingly about 1/3 advanced in that particular year.
    http://www.geo.unizh.ch/wgms/mbb/sum08.html

    Good find, HR. To me, “recession of most mountain glaciers” means 80% or 90% are receding, not that a third are advancing. AGW folks do love their exaggeration.
    Also, the mass loss has been overestimated as well, viz:

    Research team breaks the ice with new estimate of glacier melt
    FLAGSTAFF, Ariz. (March 1, 2010) –The melting of glaciers is well documented, but when looking at the rate at which they have been retreating, a team of international researchers steps back and says not so fast.
    Previous studies have largely overestimated mass loss from Alaskan glaciers over the past 40-plus years, according to Erik Schiefer, a Northern Arizona University geographer who coauthored a paper in the February issue of Nature Geoscience that recalculates glacier melt in Alaska.
    The research team, led by Étienne Berthier of the Laboratory for Space Studies in Geophysics and Oceanography at the Université de Toulouse in France, says that glacier melt in Alaska between 1962 and 2006 contributed about one-third less to sea-level rise than previously estimated. SOURCE

    The authors have only measured the Alaskan glaciers, but it appears that they think the rest of the data has the same problem … so it is likely that the global mass loss is only about two thirds of previous estimates. But the science is settled.

  199. I forgot Hansen did the year-by-year projections. I also thought the projections to 2100 in the most recent IPCC report were just base year to target year trend lines, but on closer examination they are year-to-year (Figure 10.2). I don’t see the need for this level of detail going out 90 years. Why would would anyone care if, for example, the projected temperature for 2098 was a little higher than that for 2099?
    BTW the GISS globe temp change for the first Quarter of 2010 is about .7, which is approximately one-half way between his Scenario B and Scenario C projections. That’s pretty good performance.
    A measure of a projection’s accuracy is how it fares against a no-change extrapolation. All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.
    I think Hansen’s projections in the chart are consistent with what I said about a trend being easier to project than values for each and every year in the horizon. If you simply extend lines from the 1988 base year to the target year points for the three projected scenarios, and interpolate for the implied 2010 values, overall accuracy is improved, albeit not by a lot.

  200. Dave F (23:52:02)

    @ Willis Eschenbach (23:41:15) :
    Sorry, it is in the graph in the post I referenced.
    “Estimated temperatures in altithermal and eemian times.”

    Ah, yes. Sorry I missed it. The “Altithermal” is the warmest part of the Holocene (this interglacial period). It used to be called the Holocene Optimum, until warmth got a bad name. The Eemian is the previous interglacial.
    I assume that Hansen used the Vostok Ice Core data, which does give about those numbers. The Greenland Ice Core data, on the other hand, give a larger temperature for the Altithermal. Vostok says the Altithermal was .7°C warmer than 1950, Greenland says almost three degrees … go figure. I’d go with Greenland, better resolution.

    Isn’t Scenario C the ‘Drastic reduction in CO2′ scenario? That looks like where we are tracking, so I think that it is interesting.

    Yes, it is. Hansen said that Scenario B was the most probable, but we’re way, way below that one. Scenario C assumed no growth in CO2 emissions after 2000, and we’re below that one as well. Climate model “evidence” at its finest …

  201. wow, nice post: I was already quite partial to the thermostat hypothesis, a regulation mechanism linked to liquid water seems the best explanation to tropical temperature stability and faint sun paradox…
    Now it seems I share most (if not all) W. Eschenbach views on climate, and background too (Active in numerical models for about 15 years, finite elements mostly (the pole problem of finite difference GCM is quite funny from a FE/FV background :-)). I think I have to read all your previous posts Willis, clearly you are one of the best contributors here 🙂

  202. After reading most of the comments here I was especially interested in the post referencing the total amount of CO2 that enters the atmosphere on an annual basis. The source mentioned above estimates a total of 220 gigatons are emitted annually of which 4% is caused by humans. Yet, Dr. James Lovelock, noted UK atmospheric scientist stated recently that 550 gigatons of CO2 enters the atmosphere naturally and only 40 gigatons are attributed to humans on an annual basis. With numbers this far apart is it any wonder that people are confused and have little confidence in climate scientists?

  203. Someone indulge me with a silly question: why do we even have a theory of AGW? When did it become the fashion to create theories out of the clear blue sky when there is no data suggesting a theory is needed? How about a theory that microwave ovens are distorting the earth’s magnetic field, with the obvious political ramification that microwave ovens should be banned? It seems there are possible an infinite number of calamitous theories requiring immediate political action if we simply remove the data requirement.
    Since Willis has shown great ability in cutting through to the heart of these kinds of things, perhaps he can help me understand this apparent insanity…

  204. Ok. Let us get this model business resolved. Discussion of coin tosses will not do it.
    I considered the matter in a paper I published nearly a decade ago:
    Ref. Courtney RS, ‘Crystal balls, virtual realities and ‘storylines’ ‘ E&E (2001).
    Please note that the paper reports that models of climate response to radiative forcing provide no indication of future AGW in the absence of models that project emissions from human activities that are used to project atmospheric GHG concentration.
    The following are extracts from that paper which considers statements in Chapter 2 of the report from Working Group III in the IPC’s TAR (2001).
    “They report model projections based on “scenarios”.
    Chapter 2 of Working Group III describes the origin and nature of these “scenarios”.
    A sub-committee of IPCC Working Group III produced the scenarios with no input from the climate scientists (IPCC Working Group I) who were invited to comment on the TAR. Most of the scenario authors involved are economists and “futurologists”, and many of those invited to comment on their work were “activists”. “
    And
    “Also, the scenario authors do not place any probability levels on their scenarios. This means that they – and everybody else – are forced to assume that even the most improbable – some say ridiculous – scenarios are just as likely as those that agree with reality.
    Additionally, the word “scenario” may be thought to be ambiguous. The problem arises because the scenario authors use a method that does not permit distinct separation of scenario types.
    The method has the following stages.
    a) “Storylines” of future human activity changing over time are created (i.e. social/technology change scenarios).
    b) For each “storyline”, the GHG emissions anticipated in future years are estimated (i.e. emissions modelling).
    c) The changes to mean global temperature in future years resulting from the anticipated future GHG emissions are estimated (i.e. climate modelling).
    The complete scenario contains all three stages; (a), (b) and (c). Hence, in each complete scenario, accumulating effects resulting from social/technology changes alter extrapolations from existing social/technology systems, existing GHG emissions, and existing climate.
    The technology, wealth and population growth assumptions that go into the “storylines” and the political and social engineering required for the “storylines” are not published. Therefore, they cannot be challenged. However, it is possible to consider if each scenario projects a change to GHG emissions or climate conditions (e.g. atmospheric CO2 concentration) that is reasonable in the immediate future.
    The scenario authors say the “scenarios deal with the future, so they cannot be compared with observations”. On face value this seems reasonable, but it is not true because the scenarios project from the present and some of them project from disagreement with observations of the actual present climate data. The following examples illustrate this.
    • All the scenarios set their CO2 emissions for year 2000 at 7.9 to 8.1GtC. The likely fossil fuel figure is 6.2GtC but the scenarios add to this 0.7-1.2.GtC for “deforestation” and this means that their figure for fossil fuels’ emissions of 6.8GtC in year 2000 is too high by 10%.
    • All the scenarios except B1 have atmospheric CO2 concentrations starting to rise in year 2000 as a result of exponential increase (i.e. 0.4% p.a.), but the recent trend in the concentration has been a declining rate of increase.
    • All the scenarios except B1 assume that the downward trend of methane rate-of-increase of the past 16 years will suddenly reverse in the year 2000, and in two cases (A1F1 and A2) it rises dramatically. But nobody knows why the existing downward trend exists, so how can this trend be projected to reverse as a result of social/technology change ?
    • All the scenarios assume that SO2 emissions will reduce even if fossil fuel consumption increases.
    • All the scenarios except B1 project huge increases to fossil fuel usage by 2100 starting from now. The increase is a factor of 6.3 for scenario A1B. There is a staggering projected increase to coal production by a factor of 12 for scenario A2 rising to a factor of 14 for the most extreme scenario. (While SO2 emissions decrease !)
    Importantly, the probabilities of these scenarios are not assessed and if they cannot be compared to observations they are not science; they are guesses. Chapter 2 of Working Group III admits this.”
    And
    “Put another way, the “storylines” are a selection made using personal preference of 6 untypical models from 126 models that were chosen from a list of 519 quantitative models of one particular type, and other types of quantitative model also exist. The Chapter does not state the simple truth that such selection permits almost any storylines that could generate almost any preferred projections of the future.
    If that seems like pseudo-science, then the Chapter contains worse. The Chapter states that, “Most generally, it is clear that mitigation scenarios and mitigation policies are strongly related to their baseline scenarios, but no systematic analysis has published on the relationship between mitigation and baseline scenarios”. This statement is in the middle of the Chapter and is not included in the Chapter’s Conclusions. Failure to list this statement as a conclusion is strange because this statement is an admission that the assessed models do not provide useful predictions of effects of mitigation policies. How could the predictions be useful if the relationship between mitigation and baseline is not known ? ”
    The paper concludes
    “And the Chapter concludes: “Perhaps the most powerful conclusion emerging from both the post-SRES analyses and the review of the general futures literature is that it may be possible to very significantly reduce GHG emissions through integration of climate policies with general socio-economic policies, which are not customarily as climate policies at all.”
    Simply, this conclusion of Chapter 2 of WG III TAR calls for changes to socio-economic policies that are not climate policies (at very least, this conclusion provides an excuse for such changes). And the Chapter’s Introduction states that these changes are intended to achieve “a more desirable future state” based on “societal visions of the future”.
    This conclusion derived by the method that generated it for the purpose stated in the Chapter is an abuse of science. Indeed, it is not science to make predictions of how to change the future by use of selected scenarios when “no systematic analysis has published on the relationship between mitigation and baseline scenarios”: this is pseudo-science of precisely the same type as astrology”
    Richard

  205. @ Jordan (13:50:52) :
    I think the issue may have been clouded by the introduction of the coin-flip analogy. I am assuming (although I may be wrong) that the climate models are not stochastic, but in fact deterministic (and probably chaotic).
    If for example we go back to the granddaddy of all chaotic systems, the ultra simplified 3 state model of convection rolls in the atmosphere first examined by Ed Lorenz back in 1963, we could imagine that we have some plausible range of values for the different model parameters (and indeed initial conditions), and we wish to investigate how different choices of these parameters affect the outcome of the system. We might be interested in putting some bounds on the kinds of behaviour generated, for instance the distribution of intervals between peaks, or the typical amplitude of those peaks. For every different run of the model we do, the individual trajectories of the states will be completely different – useless for predicting where peaks and troughs will actually occur for the correct (but unknown) model parameters. Nonetheless we can get some reasonably robust estimates of e.g. distributions between peaks. Not the same as climate models and temperature trends, but you get the idea.
    Apologies if I am teaching you to suck eggs (so hard to know what everyones background is) , but it seems to me that this suggests that it may not be necessary to model weather accurately from one year to the next in order to make some useful predictions about more general parameters of the climate. (And I know that there are many other issues to address in this discussion, but the dynamical modelling is the only one I have any small understanding of. )

  206. @ Dave F (22:44:59) :
    “@ Willis Eschenbach (22:04:44) :
    Sorry, but when I see phrases like ‘Estimated temperature in (insert million or billion year old era)’ I often wonder what the level of significance was in creating the underlying work. I accept that it is possible to recreate these temperatures, but as Steve McIntyre has shown, there can be issues with the use of proxies. Is there a proxy that we can compare to current temperatures? Besides, of course, tree rings?”
    A common proxy for geologically recent times such as the Eocene and Holocene is tree pollen. Different tree species such as hazel, birch, oak and pine have differing temperature and moisture requirements. Pollen grains are long lived and can be found and counted in pond sediments. The sediments are layered and can often be dated to within a few years.
    If you find, for example, oak pollen in a sediment, you can be fairly certain there were oak trees growing nearby. We know the climatic requirements for oaks growing today, so we infer that oaks growing at the time the pollen grains were trapped in the sediment were growing under the same climatic conditions.
    We know from the work of the palynologists that the northern tree limit in Asia and Canada was much further north 4,000 years ago than today. Trees then permafrost today.
    The commencement of this work dates back to Lennart von Post who laid out the foundations, but the term playnology only dates from the 1940s. My favourite paper on the topic is “Climate, vegetation and forest limits in early civilized times by Hubert Lamb. (Phil Trans A 195-230 1974).
    Hope that helps.

  207. steven mosher (00:17:43) : Re
    www-ee.ccny.cuny.edu/www/web/eebmg/ee330/05_rte.pdf
    Rather a lot of Assumptions in that paper, starting with the Earth being a black body, is the whole of the Earth a black body?

  208. steven mosher (00:17:43) : Re
    www-ee.ccny.cuny.edu/www/web/eebmg/ee330/05_rte.pdf
    Your example does not answer Willis’ question “What do the RTEs tell us about how GHGs govern radiation propagation?”
    Only how the measurements and calculations are made, not how it varies with GHG, concentration although figure 5.2 may have some bearing on it.

  209. Willis continues to confuse recession with mass balance. As the person who reported to the WGMS many of the glaciers that had a positive mass balance in 2007, this is not the same as recession. In fact all of the glaciers were still receding. http://www.nichols.edu/departments/Glacier/mb.htm
    They just gained a little thickness that one year. This is like saying a stock is increasing in value based on one week, when the quarter report is still down. Recession is a long term response to mass balance which is a noisy record. Melt was quite rapid in the 1930’s and 1940’s, the glaciers were in general larger, extending to lower elevations which tends to increase melt a bit. For many glaciers we are seeing continued mass losses and retreat even though they have shed their lowest elevation sections.
    http://glacierchange.wordpress.com/2010/02/18/brady-glacier-poised-for-substantial-retreat/
    It is safe to say the mass loss rates were comparable, but the difference was that this period featured lower snowfall. Today we are seeing some areas with good winter snowfall still losing much mass because of high melt rates.

  210. @W. Eschenbach
    Brilliant post.
    “Since the null hypothesis that the climate variations are natural has not been falsified, the AGW hypothesis is still a solution in search of a problem.”
    This is what every serious discussion about AGW should be centered on. We are too easily sidetracked by minutiae.

  211. Re: Shub Niggurath (Apr 10 21:48),
    Willis (and anna V): on the topic of environmentalism and stewardship:
    Mr Higgins asks the question
    “My starting point was ‘how do we create a duty of care to the planet, a pre-emptive obligation to not harm the planet?’”
    Sounds reasonable? Let us see what that leads us to:

    http://www.guardian.co.uk/environment/2010/apr/09/ecocide-crime-genocide-un-environmental-damage
    More details here:
    http://www.thisisecocide.com/
    Apologies if this has been discussed.

    There exists a separate thread for the ecocide nonsense here
    Now on the content let me ask the following: you must be aware that there exist what we call in Greece , the Medeas, women who kill their children in order to save them from a cruel world/father/whatever or for revenge (according to the play by Euripides).
    Does this mean that mother love is dangerous and should be avoided?
    To be a Medea it is necessary to be a mother, but not sufficient. Fortunately very few mothers become Medeas.

  212. for Arno Arrak:
    I read with great interest your critique or should I say evaluation of the Willis-Walt discussion and may I say in my humble opinion you are spot on. Although I am not in the scientific field I read everything I can find on the supposed connection between CO2 and AGW and have not found one scintilla of credible evidence to support their alarmist view.
    I am more interested in the political ramifications of this virulent AGW movement than anything else. Luckily, Congress has not yet passed any “cap and trade” legislation since the Senate has not acted on the House bill. Keep up the good work. We need your common sense approach.

  213. Patrik (16:52:47) :
    An increase in temp (caused by whatever) would cause an increase in water vapour (which is a powerful GHG) which in turn would make the stratosphere cool.
    Since we can be pretty sure that the temp has risen somewhat since 1979… Well, go figure. 😉

    That is only examining the GHG aspect of H20. We have been discussing this at length on the Meier thread, but you cannot ignore the other physical aspect of atmospheric water, and that is that in a given gas (and all other things being equal) water vapor has an inverse relationship to temperature. This is another way of saying that given the same amount of energy in a given volume of gas, the one with more water will exhibit a lower temperature than the one with less water – and vice versa.
    The correlation between increasing CO2, water vapor, and temps has been broken for the past few years. Overall atmospheric water content has continued to rise alongside CO2 and other GHGs – while temperatures (sat records) have trended slightly down.
    Also, there is research stating that mid-upper tropospheric water content has decreased of late and that is implied, at least by the NYT, as explaining the recent lack of warming. That is complete crap IMO… lower troposphere water content (which because of pressure accounts for more mass in the troposphere) has continued an up trend, which means overall water vapor content has continued to trend up. Unless someone can explain to me how atmospheric distribution variances in water vapor somehow explain a significant loss in Greenhouse Effect, this seems to imply that the view regarding greenhouse effect as being *the* climate driver is massively overstated and that there is something big and unaccounted for that drove the changes between 1980-2000.

  214. Patrik
    After rereading your comment I think we might not disagree so much. You were implying that H2O is a more likely culprit regarding stratospheric cooling (which is really more of a step change) than CO2… I don’t think that’s unreasonable. In fact, it’s a much better explanation than CO2 IMO because 99+% of atmospheric H2O is in the troposphere whereas CO2 is well mixed. With CO2 well mixed in the stratosphere why would it not be assumed to warm also?
    I think what threw me off is the conjecture about why atmospheric water vapor should be expected to increase (which in fairness was reasonably stated, I read too fast and thought you were restating the greenhouse-centric worldview) – the increase in atmospheric water content has been measured and has continued since around 1980. I’m not sure if that’s just the starting point everyone decided on or if it’s because these measurements couldn’t be done before then without satellites.

  215. Wren (00:42:36) :
    A measure of a projection’s accuracy is how it fares against a no-change extrapolation. All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.
    Can you describe your no-change extrapolation? He made his projections in 88, so a trend extrapolation from 1979-1988 would probably be the proper test case wouldn’t it?

  216. Wren (00:42:36) :

    I forgot Hansen did the year-by-year projections. I also thought the projections to 2100 in the most recent IPCC report were just base year to target year trend lines, but on closer examination they are year-to-year (Figure 10.2). I don’t see the need for this level of detail going out 90 years. Why would would anyone care if, for example, the projected temperature for 2098 was a little higher than that for 2099?

    It’s not a question of “need for this level of detail”. It’s the only way to get the trends. First the computer model forecasts each year’s temperature. Then, and only then, can they calculate the trend line.

    BTW the GISS globe temp change for the first Quarter of 2010 is about .7, which is approximately one-half way between his Scenario B and Scenario C projections. That’s pretty good performance.

    Not sure what you mean by the “temp change for the first Quarter of 2010”. But since Hansen is in charge of manufacturing the GISS numbers, don’t be surprised if they match his predictions. You should look at the satellite data instead
    In any case, his projections did poorly . Do your homework, you’re not the first person to consider these questions.

    A measure of a projection’s accuracy is how it fares against a no-change extrapolation. All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.

    You really should do some research before being so forceful about your claims. In forecasting what is used is a “naive” projection. A naive projectction is not, as you claim, a “no-change extrapolation”. It is the extension of the previous trend. And beating that does not get a projection gold stars. It just means it’s better than nothing.

    I think Hansen’s projections in the chart are consistent with what I said about a trend being easier to project than values for each and every year in the horizon. If you simply extend lines from the 1988 base year to the target year points for the three projected scenarios, and interpolate for the implied 2010 values, overall accuracy is improved, albeit not by a lot.

    Is a trend “easier to project”? Sure, just draw a line, what could be simpler, even Mark Twain could do it. Is that a valid method of saying what the climate will be like in 2100? Hardly.

  217. A C Osborn (04:49:55) :
    I’m well aware of that. I provided the link to show an example of how the calculations are used in successful engineering. And my question still stands.
    are these models correct? Very simply, will radiation leaving the earth follow the predictions made by RTE codes? or are the experiments verifying this science all wrong? This isnt a question about the climate. It’s a simple question about the accuracy of RTE. what that implies for the climate is an entirely different question. Will raditation coming into the planet propogate in accordance with those models. Simple questions.

  218. Will (02:28:57) : I think the issue may have been clouded by the introduction of the coin-flip analogy.
    Sure, but people in the AGW camp run this kind of argument time after time. They claim to be “talking science” and appear to speak with some authority due to position and academic titles.
    Walt made a bad point about the coin flip analogy. Like others, he is being blinded by the convergence of “sample probability” towards the “population probability”. This says absolutely nothbing about improvement in our ability to predict outcomes (individual outcomes or averages of outcome).
    That’s why I gave an example of a random variable: if we conclude that we cannot predict a single variate because the standard deviation is too great (prediction error compared to standard deviation), we would draw exactly the same conclusion about our ability to predict an n-point mean of the same variable (prediction error compared to standard error).
    The crucial thing to do is to use the standard error to assess whether a prediction of an n-point mean is “accurate”, not the standard deviation of the individual variates.
    On the basis of a stochastic process, it can be shown that climate (“average weather”) is no more predictable than weather.
    Will: I am assuming (although I may be wrong) that the climate models are not stochastic, but in fact deterministic (and probably chaotic).
    Probably true for individual “realisations”. But I gather it is normal practice to run multiple “realisations” to obtain “guestimates” (my choice of words) of the distribution of the outcome. That’s why you’ll hear people playing-down individual realisations and emphasising the more general clustering of outcomes.
    Will: Nonetheless we can get some reasonably robust estimates of e.g. distributions between peaks.
    This is a question of how tightly clustered the outcomes are. And that comes down to the quality of the information we have in the first instance. If we are being honest with ourselves about what we don’t know, the distribution of future climate events should be widely spread. That would be a “robust” forecast, but not much use for drawing conclusions.
    In fact, that’s what you get with confidence intevals – demanding high confidence when you don’t have very good quality data will always leave you with such a wide range of outcomes as to be meaningless.

  219. Pops (02:16:29)

    Someone indulge me with a silly question: why do we even have a theory of AGW? When did it become the fashion to create theories out of the clear blue sky when there is no data suggesting a theory is needed? How about a theory that microwave ovens are distorting the earth’s magnetic field, with the obvious political ramification that microwave ovens should be banned? It seems there are possible an infinite number of calamitous theories requiring immediate political action if we simply remove the data requirement.
    Since Willis has shown great ability in cutting through to the heart of these kinds of things, perhaps he can help me understand this apparent insanity…

    Pops, this is the reason I put the null hypothesis question so high on the list. As you point out, until we identify what unusual or inconsistent phenomenon we are trying to explain, explanations are useless. I cover the lack of such phenomena at Congenital Climate Abnormalities. I keep adding updates to this, so even if folks have read it, it might be worth another look. The short version is that there is very, very little unusual climate going on. I probably should re-write, condense, and re-post it … in my spare time …

  220. Will (02:28:57)


    … Apologies if I am teaching you to suck eggs (so hard to know what everyones background is) , but it seems to me that this suggests that it may not be necessary to model weather accurately from one year to the next in order to make some useful predictions about more general parameters of the climate. (And I know that there are many other issues to address in this discussion, but the dynamical modelling is the only one I have any small understanding of. )

    Jordan, I have great trouble with your claim that ” it may not be necessary to model weather accurately from one year to the next in order to make some useful predictions about more general parameters of the climate.” This is the claim that all of the modeler’s make, but I don’t understand the theoretical basis of it at all.
    RealClimate uses a different example. They say that although we can’t predict the temperature for say next July, we can predict that (in the Northern Hemisphere) it will be warmer than next February. And therefore (lots of hand-waving here) it’s easier to predict climate than weather.
    I suppose if we understood the fundamentals of climate the way we understand the fundamentals of coin-flipping, that might be true. We can’t predict one coin flip, but we can take a very good guess at what the average of a million flips will look like.
    My problem is that I have never seen any mathematical, or even theoretical, justification of the assumption that we understand the fundamentals of climate well enough to use an iterative model for more than a very short time period. Heck, we don’t even know if the earth has a thermostat. We discover new forcings on a weekly basis. We have very little information on the interaction of electrical fields and the climate, although lightning reminds us that electricity is a major player in the game. We just found out recently that the main cloud nuclei over land are not aerosols or dust particles as we had thought, but microbial life … life raining down on us. How do the microbes make it up so high? The answer seems to be … electrical fields. We don’t know the effect of cosmic rays on the climate. The depth and breadth of our lack of climate knowledge is staggering.
    The past decade is an excellent example. We have not seen any warming … but none of the models predicted that. To me, this means that the models are missing something, and likely many somethings, very fundamental to climate. Why did the last decade not heat up? WE DON’T KNOW!!! Sure, the modelers mumble “natural variations” and “well, we see decade-long times of no warming in a few of the model runs”, as if that explained things.
    But until we do know why the Earth hasn’t heated up over a decade as all of the models predicted, why should we believe their projections a century out?
    And when you don’t understand the fundamentals of climate, there is absolutely no reason to think that your climate predictions will be any better than your weather predictions.

  221. Willis Eschenbach (15:04:25) : Jordan, I have great trouble with your claim …
    I take it your comments were addressed to Will, not me.
    My comment is that the n-point average of a random variable is no more precictable than the underlying random variable.
    The key point is to refer to the standard error of the n-point mean when assessing its prediction error. (We cannot justifiably argue that the mean is accurate by using the standard deviation of the unlerlying data – if that is indeed what is being implied in the climate versus weather debate.)
    If we cannot bear standard devaition of the underlying random variable, there is no reason to suggest the prediction errors for an averahe of the same random variable will be any more acceptable, if they are compared to the standard error. This is quite easy to show theoretically for a random variable.
    So if those unpredictable weather fluctuations are modelled as random variables, there is a theoretcial basis to challenge the view that predictions of climate (“average weather”) would be any more acceptable.

  222. So given that CO2 is what drives the climate and global warming; while H2O is just a fleeting passerby, and really doesn’t have anything to do with anything; at least not so Dr Meier would sit up and take notice of it.
    Has anyone run a computer simulation for two cases.
    Case one, we have oceans and atmosphere pretty much like today; BUT ; There is not a single molecule of CO2 anywhere; including in the atmosphere.
    Case two, we have all the natural carbon and CO2 sources and amounts that we now have; BUT, there are no oceans, and not a molecule of H2O anywhere.
    So roughly what would earth’s temperature range look like in those two scenarios.
    Would that enable us to rank CO2 and H2O as to climate importance ?
    Just asking.

  223. mspelto (05:00:44)

    Willis continues to confuse recession with mass balance. As the person who reported to the WGMS many of the glaciers that had a positive mass balance in 2007, this is not the same as recession. In fact all of the glaciers were still receding. http://www.nichols.edu/departments/Glacier/mb.htm
    They just gained a little thickness that one year. This is like saying a stock is increasing in value based on one week, when the quarter report is still down. Recession is a long term response to mass balance which is a noisy record. Melt was quite rapid in the 1930’s and 1940’s, the glaciers were in general larger, extending to lower elevations which tends to increase melt a bit. For many glaciers we are seeing continued mass losses and retreat even though they have shed their lowest elevation sections.
    http://glacierchange.wordpress.com/2010/02/18/brady-glacier-poised-for-substantial-retreat/
    It is safe to say the mass loss rates were comparable, but the difference was that this period featured lower snowfall. Today we are seeing some areas with good winter snowfall still losing much mass because of high melt rates.

    mspelto, thanks for your comments. First, I do not confuse mass balance with recession. Recession measures the length of the tongue of the glacier. Mass balance measures the mass of the glacier. I just consider mass balance to be the more important measure. It considers the total amount of ice in the glacier, which seems to me to be a more significant indicator than the length of the tongue of the glacier, which may contain only a small fraction of the total ice.
    Next, you say “All of the glaciers were still receding”. The WGMS dataset is here. It gives the following data counts for advance/retreat:
    Retreating: 8449
    Stationary or slight advance or slight retreat: 28655
    Advancing: 1122
    “All of the glaciers were still receding”??? Perhaps the few you looked at were. Most of the glaciers are not doing much at all, they’re either stationary or have a slight advance or retreat.
    Finally, I’m not sure what your point is here. I stated quite clearly above that in warmer times glaciers on average get smaller, and in cooler times they get larger. The earth has been warming for the last three hundred years. You may see it as significant that glaciers are getting smaller. I see it as expected and ordinary, and not unusual or anomalous in any way.

  224. steven mosher (00:17:43)

    Willis,
    You know what RTE’s are. But if you want a practical application of the physics and how one actually uses it. here:
    www-ee.ccny.cuny.edu/www/web/eebmg/ee330/05_rte.pdf
    So. Do you accept that the physics of RTE is sound science?
    Simple question.

    From the cited paper:

    There is no unique solution for the detailed vertical profile of temperature or an absorbing constituent because (a) the outgoing radiances arise from relatively deep layers of the atmosphere, (b) the radiances observed within various spectral channels come from overlapping layers of the atmosphere and are not vertically independent of each other, and (c) measurements of outgoing radiance possess errors. As a consequence, there are a large number of analytical approaches to the profile retrieval problem. The approaches differ both in the procedure for solving the set of spectrally independent radiative transfer equations (e.g., matrix inversion, numerical iteration) and in the type of ancillary data used to constrain the solution to insure a meteorologically meaningful result (e.g., the use of atmospheric covariance statistics as opposed to the use of an a priori estimate of the profile structure).

    So I would say that the physics of the RTE is sound science … but the application of that physics to a given physical situation may or may not be sound.

  225. steve mosher
    what does it matter what the radiation transfer accuracy is when it’s for clear skies and that amounts to less than 50% of the globe at any one time?

  226. wayne (15:24:27) :
    Dr A Burns (14:38:27) :
    Fossil fuel derived CO2 makes up about 2% of atmospheric CO2.
    I would love you to provide the source of the data
    REPLY:
    Here is the link: http://wattsupwiththat.files.wordpress.com/2009/02/eia_co2_contributions_table3.png
    Here are other useful links:
    photos of plant response to CO2 http://i32.tinypic.com/nwix4x.png
    Graph of Palo CO2 http://i224.photobucket.com/albums/dd137/gorebot/311s08x.jpg
    CO2: Ice Cores vs. Plant Stomata http://debunkhouse.wordpress.com/2010/03/28/co2-ice-cores-vs-plant-stomata/

  227. Stephan (15:29:03) :
    Actually I agree with Burns. The current rise in recently measured Co2 may be speck in a 1000’s years rise from natural variation. Could we label the human produced C02 to verify? Even a 12 month experiment would clear this up…
    If you go to Dr Meir’s responds to Willis, there is a long discussion about this topic starts some where around here
    Julian Flood (20:04:38) :
    …..
    quote
    Second, the carbon emitted by humans has a distinct chemical signature from natural carbon and we see that it is carbon with that human signature that is increasing and not the natural carbon.
    unquote…

    Link:
    http://wattsupwiththat.com/2010/04/08/nsidcs-walt-meier-responds-to-willis/#comment-364252

  228. Steve Goddard (16:18:16) :
    The stratosphere has cooled by about 1C since the beginning of the satellite record. The volcanic glitches introduce noise, but don’t explain the overall downwards trend.
    http://wattsupwiththat.files.wordpress.com/2010/04/msu_global_stratospheric_temp_anomaly1.jpg
    Yes it has but it looks like a “step change” due to the volcanos similar to the “step change” seen in Ocean Heat Content in response to El Ninos. See Bob Tisdale’s Can El Nino Events Explain All of the Global Warming Since 1976? – Part 2
    http://wattsupwiththat.com/2009/01/12/can-el-nino-events-explain-all-of-the-global-warming-since-1976-%E2%80%93-part-2/

  229. Regarding statement above “A measure of a projection’s accuracy is how it fares against a no-change extrapolation. All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value”
    This has already been published – A “no change” climate “model” outperforms the IPCC models by a factor of 7:
    http://hockeyschtick.blogspot.com/2010/03/paper-no-change-climate-model-is-7_02.html

  230. I completely agree with Willis. I especially agree with his statement :
    “This blurring of the line between reality and models is a recurring and very frustrating feature of the climate discussion. I’m talking about reality, and meanwhile, without saying so, Dr. Meier is discussing model results. This habit of climate scientists, of talking about models as if they were discussing reality, is very frustrating and impedes communication.”
    For me this has been the key problem with AGW. It’s based on proxy data and on models, not reality.
    I learned back in P-chem that PV=nRT is actually a very poor model over the real-world range of those variables, not to mention across various gases. It’s a fun excercise for students, but in the real world if you want to know how gases behave, you have to resort to empirical measurements. Why would we need steam tables if the “ideal” gas law was all that accurate?

  231. Re: Gail Combs (Apr 13 00:01),
    This so called ability to label carbon as of fossil fuel origin and therefore human induced depends on ratios of isotopes of carbon.
    See this thread, back in 2008.
    Also I remember that this ratio is also affected by some type of plankton or algae that exists in large numbers, so cannot really be a signature for the human induced presence of CO2.

  232. cba (17:01:53) :
    steve mosher
    what does it matter what the radiation transfer accuracy is when it’s for clear skies and that amounts to less than 50% of the globe at any one time?
    Well, I see that Willis has also agreed that radiation transfer equations are in fact sound science.
    The point of that is to understand that the arguments against AGW based on the notion that C02 is merely a trace gas, are clearly refuted. Note, this says nothing about FEEDBACKS. But without feedbacks considered its clear what the physics of RTE predict. So man skeptics deny even the truth of RTE I thought it important to clarify. BTW, the first tenet of being a Luke warmer is believing in the physics of radiative transfer.

  233. NickB. (11:41:54) :
    Wren (00:42:36) :
    A measure of a projection’s accuracy is how it fares against a no-change extrapolation. All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.
    Can you describe your no-change extrapolation? He made his projections in 88, so a trend extrapolation from 1979-1988 would probably be the proper test case wouldn’t it?
    —-
    A no-change extrapolation simply means projecting the base-year value to the target year. In other words, the 1988 temp would be the projected temperature for the target year, which as I recall in this case was the year 2020. Or put simply, no change in temperature would be projected over the 1988-2020 period.

  234. Willis Eschenbach (12:43:13) :

    Wren (00:42:36) :
    I forgot Hansen did the year-by-year projections. I also thought the projections to 2100 in the most recent IPCC report were just base year to target year trend lines, but on closer examination they are year-to-year (Figure 10.2). I don’t see the need for this level of detail going out 90 years. Why would would anyone care if, for example, the projected temperature for 2098 was a little higher than that for 2099?

    It’s not a question of “need for this level of detail”. It’s the only way to get the trends. First the computer model forecasts each year’s temperature. Then, and only then, can they calculate the trend line.

    BTW the GISS globe temp change for the first Quarter of 2010 is about .7, which is approximately one-half way between his Scenario B and Scenario C projections. That’s pretty good performance.

    Not sure what you mean by the “temp change for the first Quarter of 2010″. But since Hansen is in charge of manufacturing the GISS numbers, don’t be surprised if they match his predictions. You should look at the satellite data instead …
    In any case, his projections did poorly . Do your homework, you’re not the first person to consider these questions.

    A measure of a projection’s accuracy is how it fares against a no-change extrapolation. All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.

    You really should do some research before being so forceful about your claims. In forecasting what is used is a “naive” projection. A naive projectction is not, as you claim, a “no-change extrapolation”. It is the extension of the previous trend. And beating that does not get a projection gold stars. It just means it’s better than nothing.

    I think Hansen’s projections in the chart are consistent with what I said about a trend being easier to project than values for each and every year in the horizon. If you simply extend lines from the 1988 base year to the target year points for the three projected scenarios, and interpolate for the implied 2010 values, overall accuracy is improved, albeit not by a lot.

    Is a trend “easier to project”? Sure, just draw a line, what could be simpler, even Mark Twain could do it. Is that a valid method of saying what the climate will be like in 2100? Hardly.

    ——
    I’m glad you finally agree with me that a long-term trend is easier to project than year-to-year fluctuations. I think we generally can have more confidence in a projection of a long-term trend too. Look at it this way, you can be wrong on yearly values for 99 years of a 100-year projection and hit the target on the 100th year.
    You said: “You really should do some research before being so forceful about your claims. In forecasting what is used is a “naive” projection. A naive projectction is not, as you claim, a “no-change extrapolation”. It is the extension of the previous trend. And beating that does not get a projection gold stars. It just means it’s better than nothing.”
    You should know I know a naive projection is an extension of a previous trend. Here is you quoting me in your post of 22:04:44.
    “As a simple exercise, we could extrapolate the 1890-2009 trend out to the year 2100, and that naive projection might seem to most a more reasonable than a no-change extrapolation(i.e., 2100 having the same temperature as 2009).”
    Do you quote people and then forget what they said?
    Oh well, my short-term memory isn’t so good either.
    Both “no-change extrapolations” and “naive projections” provide ways to evaluate sophisticated projections. The no-change extrapolation may seem odd, but it addresses the question of whether the projection being evaluated is turning out or has turned out to be better than no projection at all.
    Hansen’s 1988-2020 projections clearly are turning out to be more accurate than a no-change extrapolation. With 8 years left in the projection period, I wouldn’t bet against his middle scenario projection outperforming a naive projection.

  235. Wren, thanks for your reply. I have reformatted it to make it easier to read, I hope that’s OK with you.
    Yes, a long-term trend is easier to project than individual years. You just draw a line. But as I keep saying, no climate model ever does that. They calculate the individual years. Then they calculate the trend. So I don’t see why that is relevant.
    My apologies for the misunderstanding about naive projections. You had said:

    All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.

    I interpreted that to mean that the “no-change” extrapolation merely kept the 1978 value. You are correct that you had said differently about a “naive projection” before. I can only say in my defence that I am currently tending five active threads, with a total of 1510 comments and counting …
    In order for temperatures to get back to Hansen’s middle scenario projection, they would have to skyrocket … very doubtful.
    My best to you, sorry if I was out of line,
    w.

  236. Steve Mosher,
    rather than luke warm, how about realistic analysis? Just because someone can identify hysteria and exageration and reject it as such – with data doesn’t require applying political labels.
    your massive positive feedbacks are refutable in so many ways, one could say it’s fully robust. My favorite – and feel free to provide the math to refute it if you can – is that the water vapor feedbackcannot provide even as much as the miniscule co2 warming. It’s based simply on the possible increase in absolute humidity with constant relative humidity (a standard climatology assumption) and upon its radiative characteristics.
    besides, radiative absorption is only half the science. The other half is radiation emission. Whether an atmospheric slab will absorb or emit spectral lines in the context of a continuum source behind the slab is a matter of relative temperature and the temperature of the slab is an energy (power) balance problem.

  237. Willis Eschenbach (01:43:12) :
    Wren, thanks for your reply. I have reformatted it to make it easier to read, I hope that’s OK with you.
    Yes, a long-term trend is easier to project than individual years. You just draw a line. But as I keep saying, no climate model ever does that. They calculate the individual years. Then they calculate the trend. So I don’t see why that is relevant.
    My apologies for the misunderstanding about naive projections. You had said:
    All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.
    I interpreted that to mean that the “no-change” extrapolation merely kept the 1978 value. You are correct that you had said differently about a “naive projection” before. I can only say in my defence that I am currently tending five active threads, with a total of 1510 comments and counting …
    In order for temperatures to get back to Hansen’s middle scenario projection, they would have to skyrocket … very doubtful.
    My best to you, sorry if I was out of line,
    w.
    —–
    Willis, I appreciate your courtesy, but no apology is necessary. You weren’t out of line.
    Hansen’s middle scenario is better than I had thought. I was wrong about the GISS global temperature anomaly averaging 70 in the first quarter of 2010. The average of the Jan., Feb., and Mar. anomalies is 95.
    http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts.txt
    I think the 95 would be .95 on the graph of Hansen’s 1988-2020 projections, which is almost identical to his middle projection for 2010.
    I would be happy for you to check on this. I do sometimes make mistakes.

  238. anna v (10:45:33) :
    “This so called ability to label carbon as of fossil fuel origin and therefore human induced depends on ratios of isotopes of carbon.
    See this thread, back in 2008…..”

    Thank you for the reference.
    In looking at the subject I also found that coal has various levels of C14 despite the theory that all the C14 has decayed due to age. One theory is that this is because there is microbe activity deep within the rocks! Another is the formation of new C14 due to nearby radioactive isotopes. The biggest problem I have found in the AGW discussions is the “selected evidence” problem. Thank goodness for the collected wisdom of WUWT.

  239. Wren: A measure of a projection’s accuracy is how it fares against a no-change extrapolation. All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.
    Well, obviously a for prediction made in 1988, the naive, no-change extrapolation should be from 1988, not 1978. So how did Hansen do compared to the naive prediction, based on an independent measure of global temperature? From the graphs, it looks to me like Hansen projected an increase from 1988 to 2010 of about 0.65 degrees for scenarios A and B, and about 0.40 degrees for scenario C. According to RSS, the average of the 12 monthly TLT anomalies from 1988 was 0.067 and the average for the last complete year, 2009, was 0.260, for an increase of 0.193. By this measure, the naive forecast wins for every scenario. Using the last 12 months, which includes more El Nino months, the average is 0.353, for an increase of 0.286. So Hansen’s unrealistic scenario C outperforms the naive forecast, but the others don’t.
    I noticed in your previous comment, you use the average GISS anomaly 0f 0.95 for the first three months as evidence for the accuracy of Scenario B (“, which is almost identical to his middle projection for 2010”). Unless Hansen specifically forecast an strong El Nino for those months, that’s a ridiculous claim, and actually evidence against the accuracy of the projection.

  240. MJW (22:18:44) :
    Wren: A measure of a projection’s accuracy is how it fares against a no-change extrapolation. All three of Hansen’s projections beat a no-change extrapolation(i.e., are now closer to the actual value than is the 1978 base-year value.
    Well, obviously a for prediction made in 1988, the naive, no-change extrapolation should be from 1988, not 1978. So how did Hansen do compared to the naive prediction, based on an independent measure of global temperature? From the graphs, it looks to me like Hansen projected an increase from 1988 to 2010 of about 0.65 degrees for scenarios A and B, and about 0.40 degrees for scenario C. According to RSS, the average of the 12 monthly TLT anomalies from 1988 was 0.067 and the average for the last complete year, 2009, was 0.260, for an increase of 0.193. By this measure, the naive forecast wins for every scenario. Using the last 12 months, which includes more El Nino months, the average is 0.353, for an increase of 0.286. So Hansen’s unrealistic scenario C outperforms the naive forecast, but the others don’t.
    I noticed in your previous comment, you use the average GISS anomaly 0f 0.95 for the first three months as evidence for the accuracy of Scenario B (“, which is almost identical to his middle projection for 2010″). Unless Hansen specifically forecast an strong El Nino for those months, that’s a ridiculous claim, and actually evidence against the accuracy of the projection.
    ———-
    You are correct about the base year for Hansen’s temperature projection. It was 1988, not 1978.
    Hansen project the GISS series, not the RSS series, so I’m puzzled as to why you would want to use RSS to evaluate his projection for 2010.
    If you do want to compare GISS and RSS anomalies, they would have to be on a common base. I didn’t check your numbers so I’m not sure what you did.
    No,you don’t take out the cooling effect of La Nina or the warming effect of El Nino when evaluating temperature projections. That would be like alternately moving the goal post closer than farther.
    Although almost one-third of the projection period is remaining, it looks like all of Hansen’s projections will beat a no-change extrapolation. As for beating a naive projection, I think we will have to wait and see.
    BTW, you said “Hansen projected an increase from 1988 to 2010 of about 0.65 degrees for scenarios A and B, and about 0.40 degrees for scenario C.” A and B don’t look the same in 2010 to me. Are we looking at the same projections?

  241. I used RSS because I wanted an independent measure of global temperature. Frankly, I don’t trust GISS. Hansen’s in charge of the GISS record, and I’m not willing to let him grade himself. Though it’s true GISS and RSS don’t measure exactly the same thing, not only would it be surprising if the change in the anomalies of the surface and lower troposphere temperatures weren’t essentially equal, it would also raise questions as to the meaningfulness of the surface measurement. Since I compared the change of the RSS values from 1988 to the present to Hansen’s forecast changes for the same period, I see nothing wrong with using RSS instead of GISS.
    No,you don’t take out the cooling effect of La Nina or the warming effect of El Nino when evaluating temperature projections.
    But you can’t choose three months during a strong El Nino as confirmation of Hansen’s projection. What will you do when the El Nino ends and the temperature anomaly drops? Will you declare Hansen’s prediction falsified? I doubt it. You probably joined in criticizing those who claimed a cooling trend by picking 1998 as the starting point, and now you want to do the same thing.
    A and B don’t look the same in 2010 to me. Are we looking at the same projections? They also don’t look the same in 1988 (which is odd for a prediction made in 1988). I used the change between the 1988 and the 2010 values as the predicted temperature increase, which are nearly the same for scenarios A and B.

  242. MJW, the RSS baseline is Jan 1979-Dec 1998, but the GISS baseline is Jan 1951-Dec 1980, so the two anomalies cannot be compared unless converted to a common baseline. This has been done at
    http://www.woodfortrees.org/notes#wti
    If you look at the next to last chart you will see RSS and GISS follow the same path and show about the same increase from 1988 to 2009. Why would you trust RSS but not GISS when their anomalies, as well as those of UAH and HADCRUT, all show about the same temperature increase? Hansen’s 2010 projections are looking good regardless of which of these metrics you use.
    You can use any point you like for evaluating Hansen’s projections. He didn’t condition his projections on an absence of natural warming and cooling influences.

  243. By comparing the projected increase ot the actual increase, I avoid the problem of what baseline is used. If RSS and GISS follow the same problem, you should have no objection to my choice to use RSS. How can you claim Hansen’s projection look good, when I clearly demonstrated that scenarios A and B underperform the a naive, no change projection, and scenario C — which is based on a completely unrealistic assumption for GHG emissions — only outperforms the naive forecast when the last three El Nino months are included? If that’s “looking good,” I wonder what you’d consider to be looking bad.

  244. By comparing the projected increase to the actual increase, I avoid the problem of what baseline is used. If RSS and GISS follow the same path, you should have no objection to my choice to use RSS. How can you claim Hansen’s projection look good, when I clearly demonstrated that scenarios A and B underperform the naive, no-change projection, and scenario C — which is based on a completely unrealistic assumption for GHG emissions — only outperforms the naive forecast when the last three El Nino months are included? If that’s “looking good,” I wonder what you’d consider to be looking bad.
    (I accidentally hit “submit comment” too soon, so the first version was a bit garbled.)

  245. I hit submit to soon myself sometimes so I know the problem.
    A no-change extrapolation and a naive projection are different things. A no-change extrapolation would use the base value(1988) for the target year(2010). On the other hand, a naive projection would extrapolate the historical trend to the target year.
    While Hansen’s projections didn’t perform well in the early year’s of the projection horizon, they are doing well now. Actually, for 2010, they beat both a no-change extrapolation and a naive projection, if you start the extrapolation from pre-1900.
    In an evaluation of each year’s projections, the errors are fairly large over-projections in the early years. However, in the remaining years (2010-2020) they may be under-projections, which would offset. We will have to wait and see.

  246. I don’t believe there’s a clear definition of what a naive prediction is, and it would seem to depend on the underlying assumption; for example (and this is just an example, not a hypothesis), one could assume the any previous trend was simply the result of a random walk, in which case the no-change prediction would be appropriate. In any event, the no-change forecast is the most disadvantageous choice, and it still beat Hansen’s scenarios A and B.
    I’m sure you’ll disagree, but I’m not willing to use any global temperature data prior to the satellite data, because I think it’s guesswork. If I get a chance, I’ll figure out a naive GISS projection based on the years from 1979 to 1988. I’m pretty sure it will beat both the no-change projection and Hansen’s projections.
    When you say, “In an evaluation of each year’s projections, the errors are fairly large over-projections in the early years. However, in the remaining years (2010-2020) they may be under-projections, which would offset,” the old saw about one bird in the hand being worth two in the bush comes to mind. But, as you say, we’ll have to wait and see.

  247. Sorry, I meant I’ll figure out a naive RSS projection based on the years from 1979 to 1988. (By which I meant a simple least-squares fit to the 1979 to 1988 RSS TLT data.)

  248. Maybe somebody out there watching this thread clarify.
    I am translating the excellent willis/meier exchange of positions
    into German. Being a non native English speaker nore being
    proficient in statistics I don’t understand the implications of
    the cited willis statement:
    quote:
    From my own experience, I wrote a paper explaining the problems with a study by Michael Mann that had been published in Geophysical Research Letters (GRL). His study claimed that the best way to extend a smooth (Gaussian or otherwise) to the end of a series was to pad the end of the series by reflecting it around both the x and y axes. (This results in forcing the smooth through the last point of the series, which is absolutely the last thing you want to do).
    unquote
    My question ist: did M.M. really mean that the best way … was … to pad … by reflecting … around both the x and y axes?
    Doesn’t that mean that one would inverse the positive/negative signs of the values, and by doing so the resulting extended smooth would be completely spurious?
    Could M.M. really have suggested such an obviously misleading method?
    Sorry for asking this silly question in your high flying discussion forum.
    But clarification would prevent me from wrongly translating an excellent
    discussion.
    Thanks to everybody who will care and answer!
    ORION

  249. Maybe somebody out there watching this thread clarify.
    I am translating the excellent willis/meier exchange of positions into German. Being a non native English speaker nore being proficient in statistics I don’t understand the implications of the cited willis statement:
    quote:
    From my own experience, I wrote a paper explaining the problems with a study by Michael Mann that had been published in Geophysical Research Letters (GRL). His study claimed that the best way to extend a smooth (Gaussian or otherwise) to the end of a series was to pad the end of the series by reflecting it around both the x and y axes. (This results in forcing the smooth through the last point of the series, which is absolutely the last thing you want to do).
    unquote
    My question ist: did M.M. really mean that the best way … was … to pad … by reflecting … around both the x and y axes?
    Doesn’t that mean that one would inverse the positive/negative signs of the values, and by doing so the resulting extended smooth would be completely spurious?
    Could M.M. really have suggested such an obviously misleading method?
    Sorry for asking this silly question in your high flying discussion forum.
    But clarification would prevent me from wrongly translating an excellent discussion.
    Thanks to everybody who will care and answer!
    ORION

    First, ORION, your English is good, and your understanding is correct. Mann in fact did claim that the best way to deal with the ends of a smooth is to pad the end of the data by reflecting it around both the X and Y axes.
    Of course, since it is then symmetrical around the final point, the smooth (regardless of type) must pass through the last point. This is absolutely not the right answer.
    My paper (rejected by GRL) is here, it explains the whole thing. It was rejected for being too hard on poor Mann …
    w.

  250. hello willis,
    thanks for the immediate reply and the
    link to the original paper.
    what a surprise reading an answer
    directly from you.
    i didn’t want to waste your time because
    its better spent on your your
    excellent posts. thanks a lot!
    you render a great service to everybody
    who doesn’t want to be fooled and tricked
    by the warm mongers.
    just for your info: i have already translated some of your former pieces.
    they were published on the leading german skepics-site
    http://www.eike-klima-energie.eu/
    i esteem your work greatly.
    making your ideas public in german
    is very important, because there seems
    to be no other society in the west which
    is so perfectly misled by politicians
    and the msm as ours into believing in the agw-thesis.
    the roaring discussion about climagate is passing almost unheard here,
    where it not for WUT and the like.
    these sites, however, are in english and need translation in order to inject them as antidote against the co2 poison legend.
    best wishes to you, hope you feel well
    on TUVALU and continue to do so.
    i am sure you are not building an ark!
    helmut jaeger (my clear name)
    munich, germany

Comments are closed.