RW (14:34:08) :

[Snark snipped]

Well, it’s your choice, I suppose.

And my choice is to stop feeding the local troll.

You also appear not to understand that many factors affect climate. You appear to think that unless there is a 100% correlation between CO2 and temperatures, then CO2 does not affect climate. I guess I’ll have to spell the reductio ad absurdum out to you – in your beloved phanerozoic graph, there is only a weak correlation between temperature and any variable you care to consider. Your premise would lead to the conclusion that nothing affects the climate. The conclusion is absurd; that’s because the premise is absurd.

You keep on using the word ’secular’ in ways that don’t make sense. Secular in this context simply means non-cyclic.

You don’t seem to understand that the Sun has been increasing in luminosity throughout its main sequence evolution, and thus its output changed significantly over the period covered by your favourite graph.

You don’t seem to understand just how approximate the temperature line is on that graph of yours. You clearly haven’t read the literature widely enough to know that much more realistic reconstructions exist.

Please do explain what physical basis there is in a 12th order polynomial fit. I look forward to that very much.

I don’t think you’re incapable of understanding these things, because most 16-year olds could grasp them quite easily. Your lack of knowledge is, I think, self-inflicted. You clearly want to believe that CO2 is not a greenhouse gas. Unfortunately, to maintain that belief, you have to pretend not to understand simple physics and maths. Well, it’s your choice, I suppose.

oms (10:01:45) :

Dave Middleton (08:19:24) :

[…]

Then don’t ever become a seismic data processor or a sequence stratigrapher…because the name of both of those games is cyclical resolution of physical phenomena….Polynomial trend-line functions are analogous to seismic wavelets. The higher the polynomial – the higher the frequency – the higher the resolution. A densely sampled data set, like the UAH monthly temperature anomaly series can return meaningful cyclical information with polynomial trend-line functions.

In seismology, wavelet transforms have a clear physical analogy to wave packets. Does high order polynomial fitting have a similar interpretation?

Sort of…

In seismic reflection profiling, the thinnest bed that can be resolved is determined by frequency and seismic velocity. “Over-fitting” the frequency content in wavelet processing is desirable so long as the frequency is below Nyquist and the signal-to-noise ratio can be managed.

If I simply had the raw data that went into this Relative Changes of Sea Level chart, a third or sixth order polynomial might give me the First Order curve; whereas a higher polynomial might give me the Second Order curve. A linear trend-line wouldn’t reveal anything about the cyclical nature of past sea level changes…In the same manner that linear trend-lines through the obviously non-linear temperature data do not reveal anything about modern cyclical changes.

Isn’t the point of polynomial fitting for these temp series actually interpolation and data reduction (opposite to the point about Nyquist)? If so, it makes more sense to present “trends” produced from a low-order polynomial fit.

Yes. You are correct. In this graph of UAH Lower Trop. Temp’s The 12-month moving average captures the high frequency cycles and the 6th order polynomial captures the low frequency cycle…In a similar fashion as the second and first order sea level changes mentioned above.

I’m essentially using the sixth order polynomial as a low frequency filter to image the low frequency cycle. Ideally, I should take the data into the frequency domain and apply various filters. But, that would be turning a hobby into a job.

If you think you can meaningfully fit a 12th-order polynomial to any physical phenomenon, I’m afraid you’re horribly mistaken.
Then don’t ever become a seismic data processor or a sequence stratigrapher…because the name of both of those games is cyclical resolution of physical phenomena….Polynomial trend-line functions are analogous to seismic wavelets. The higher the polynomial – the higher the frequency – the higher the resolution. A densely sampled data set, like the UAH monthly temperature anomaly series can return meaningful cyclical information with polynomial trend-line functions.

In seismology, wavelet transforms have a clear physical analogy to wave packets. Does high order polynomial fitting have a similar interpretation?

Isn’t the point of polynomial fitting for these temp series actually interpolation and data reduction (opposite to the point about Nyquist)? If so, it makes more sense to present “trends” produced from a low-order polynomial fit.

RW (13:52:53) :
“I have never denied that CO2 is a greenhouse gas.”
You said “the greenhouse effect would not significantly vary with variations in the atmospheric concentration of CO2“. If CO2 is a greenhouse gas, that would be impossible.
“However the greenhouse effect of CO2 is a logarithmic function of diminishing returns. Hence, there is no secular relationship between CO2 and temperature.”
That’s interesting. You’re saying that there’s a logarithmic relationship, therefore there’s no relationship. Let me give you the corrected version of what you said: the greenhouse effect of CO2 is a logarithmic function. Hence, there is a relationship between CO2 and temperature.

That’s because the relationship is one of “diminishing returns”…A logarithmic relationship. Each additional unit of CO2 yields less greenhouse warming than the prior unit. So there is no linear or secular relationship.

If CO2 provides 25% of Earth’s greenhouse warming (~30 C)…285ppm of CO2 was yielding 7.5C of warming. A doubling of that CO2 level to 580ppm could not raise the Earth’s temperature by more than 1.5 C to 2.0 C (without positive feedback mechanisms).

0 to 285ppm -> 7.5 C
285 to 570ppm -> 1.75 C

Remarkably similar to Delta-T = ln(CO2). The first 285ppm yields four times as much warming as the next 285ppm. If you extrapolate this relationship to 4000ppm, you’ll understand why CO2 and temperature had no secular relationship throughout the Phanerozoic Eon (the last ~600 million years).

Now…Since the Earth has been cooling while CO2 has climbed from 375 to 385ppm and since the Earth was warmer in the Sangamon Interglacial with only about 315ppm CO2…In fact the Earth was warmed in each of the last four interglacials with less CO2 (if the ice cores are right – a big IF)…It should be apparent that there are no significant positive feedback mechanisms operating. In fact, the feedback mechanisms pretty well have to be negative.

“I should have said that an asymptotically increasing function never reaches infinity in the same way that a logarithmic function never vanishes.”
If you’d said that, you’d still be wrong. y=1/x is an asymptotic function. as
x tends to infinity, y tends to a fixed value – in this case, zero. y=log(x) is
not an asymptotic function. As x tends to infinity, y also tends to infinity,
and not a fixed value.

In an asymptotically increasing function delta-y increases exponentially with linear increases in delta-x. In a logarithmic function (like y=ln(x)) delta-y decreases exponentially with linear increases in delta-x.

Hence a logarithmic function vanishes in exactly the same way that an asymptotically increasing function never reaches infinity.

Think Douglas Adams…”The Vogon Constructor Ship hung in the sky exactly like a rock doesn’t.”

“irrespective of plate tectonics and all of the other massive changes apart from CO2, Earth’s hothouse phases had a remarkably consistent average temperature of about 22 C. And three of the four icehouse phases pretty well settled in with ~12 C average temperatures.”

Ah, I see the problem. Just how accurate do you believe that 22C figure is? Do you know how it was estimated? Do you know what other estimates are available? If not, why not? If so, why do you prefer this one?

It’s as accurate as an estimate can be for something that happened so long ago and has been averaged over such a long time period. Past temperatures are largely reconstructed from oxygen isotope ratios and from paleoclimatological/paleogeographical reconstructions.

One “alternative” reconstruction that I am aware of is Royer (2004). He used a pH modifier to force the temperature data to “fit” the CO2 levels. I believe that the pH modifier was based on CO2 levels. Nir Shaviv and Jan Veizer concluded that…

The analysis of Royer et al. (2004) assumes an unrealistically high pH correction. First, it neglects the ice-volume effect, which changes the relation between sigma18O and dT. Second, this large pH correction implies high temperatures for seawater even during times of extensive glaciations. Moreover, the analysis of Royer et al. (2004) consists of bootstrapping, by introducing a correction to dT that is an implicit function of RCO2. It is then not surprising that a correlation between dT and RCO2 is obtained. This would be the case irrespective of the RCO2 model utilized.

This model of the Phanerozoic climate has been the standard since at least the 1970s… LINK. It’s what I was taught way back in the Upper Pleistocene.

Could this paradigm one day be shifted? Sure it could. But force-fitting the stable isotope temperature data to the CO2 levels using a CO2-derived pH flux adjustment, is not really very valid.

“The icehouse vs. hothouse phases may have been dictated by plate tectonics or astrophysical phenomena. But CO2 appears to have played no role in that cycle.”
The Sun was fainter when it was younger, and liquid water could not have existed on Earth without a much greater greenhouse effect than exists today. This is a simple fact but one which you appear to be entirely unaware of.

Maybe so…But you are talking about the pre-Phanerozoic. Very little is known about the pre-Phanerozoic atmosphere. The general consensus is that the Earth’s atmosphere evolved through three main phases.

The Earth’s first atmosphere formed during the Hadean Period and is thought to have been mostly water vapor and CO2, with lesser amounts hydrogen-sulfur compounds, methane, CO and other gases. The Hadean gets its name from Hades…A generally warm place. Much of the Earth’s surface was molten during the Hadean. Molten rocks are kind of hot; even under a faint sun.

From 3.8 to 3.4 billion years ago a second atmosphere formed during the Achaean Period…

“The atmosphere was very different from what we breathe today; at that time, it was likely a reducing atmosphere of methane, ammonia, and other gases which would be toxic to most life on our planet today. Also during this time, the Earth’s crust cooled enough that rocks and continental plates began to form.”

Some studies have suggested that the Achaean was more dominated by CO2 than CH4. I suppose that’s possible. Even with a logarithmically declining greenhouse effect, CO2 could yield an awful lot of atmospheric warming with concentrations measured in 100’s of thousands of ppm…

ln(30) = 3.4, ln(90) = 4.5, ln(180) = 5.2, ln(285) = 5.7, ln(570) = 6.3, ln(4,000) = 8.2, ln(800,000) = 13.6

So, an 80% CO2 atmosphere should be significantly warmer than any atmosphere with less than 0.001% CO2. But variations in CO2 in the 100’s of ppm will not yeld much variation in temperature.

The Earth’s current atmosphere was generally in place about 600 million years ago. Trace gases have varied over the last 600 million years; but the Nitrogen-Oxygen atmosphere that The Phanerozoic Eon, it the geological time period in which life as we know it came to be. Phanerozoic means “visible life”. It was during this period that eukaryotic life literally exploded into evolution.

It is during the Phanerozoic Eon that CO2 and temperature demonstrate no secular relationship because the very minor greenhouse-effect differences between 185ppm and 4,000ppm CO2 are totally obliterated by geological, solar and weather processes.

“A third or sixth order polynomial would fit the First Order quite well. A twelfth order polynomial might just fit the Second Order quite well. A linear trend-line would be meaningless on either.”
If you think you can meaningfully fit a 12th-order polynomial to any physical phenomenon, I’m afraid you’re horribly mistaken.

Then don’t ever become a seismic data processor or a sequence stratigrapher…because the name of both of those games is cyclical resolution of physical phenomena.

As long as the sample rate of your data is at least twice the highest frequency you are trying to resolve, you can obtain meaningful cyclical resolution without aliasing. The maximum resolvable frequency based on sample rate is known as the Nyquist Frequency.

Polynomial trend-line functions are analogous to seismic wavelets. The higher the polynomial – the higher the frequency – the higher the resolution. A densely sampled data set, like the UAH monthly temperature anomaly series can return meaningful cyclical information with polynomial trend-line functions.

You said “the greenhouse effect would not significantly vary with variations in the atmospheric concentration of CO2“. If CO2 is a greenhouse gas, that would be impossible.

“However the greenhouse effect of CO2 is a logarithmic function of diminishing returns. Hence, there is no secular relationship between CO2 and temperature.”

That’s interesting. You’re saying that there’s a logarithmic relationship, therefore there’s no relationship. Let me give you the corrected version of what you said: the greenhouse effect of CO2 is a logarithmic function. Hence, there is a relationship between CO2 and temperature.

“I should have said that an asymptotically increasing function never reaches infinity in the same way that a logarithmic function never vanishes.”

If you’d said that, you’d still be wrong. y=1/x is an asymptotic function. as
x tends to infinity, y tends to a fixed value – in this case, zero. y=log(x) is
not an asymptotic function. As x tends to infinity, y also tends to infinity,
and not a fixed value.

“irrespective of plate tectonics and all of the other massive changes apart from CO2, Earth’s hothouse phases had a remarkably consistent a
verage temperature of about 22 C. And three of the four icehouse phases pretty well settled in with ~12 C average temperatures.”

Ah, I see the problem. Just how accurate do you believe that 22C figure is? Do you know how it was estimated? Do you know what other estimates are available? If not, why not? If so, why do you prefer this one?

“The icehouse vs. hothouse phases may have been dictated by plate tectonics or astrophysical phenomena. But CO2 appears to have played no role in that cycle.”

The Sun was fainter when it was younger, and liquid water could not have existed on Earth without a much greater greenhouse effect than exists today. This is a simple fact but one which you appear to be entirely unaware of.

“A third or sixth order polynomial would fit the First Order quite well. A twelfth order polynomial might just fit the Second Order quite well. A linear trend-line would be meaningless on either.”

If you think you can meaningfully fit a 12th-order polynomial to any physical phenomenon, I’m afraid you’re horribly mistaken.

“You are just not sufficiently familiar enough with geology and geophysics to be able to look at climate change from the same perspective that I do. I fail to see why that elicits such obnoxiousness on your part. Oh well, C’est la vie”

You’re not sufficiently familiar with basic physics or statistics to come to any
sensible conclusion about climate change, as we can see from the litany of basic misconceptions that appear throughout your posts.

Smokey (15:24:49) :

Dave Middleton,

You are right, of course. But you’re feeding a troll. The only reason his snarky attitude isn’t insufferable is because his beliefs are so easy to deconstruct. But on general principles, it’s best not to feed trolls.

You are correct…And I hope I am not diminishing the collegiality of WUWT…But he (or she) keeps hanging curveballs…Swinging away is just so irresistible…;)

You are right, of course. But you’re feeding a troll. The only reason his snarky attitude isn’t insufferable is because his beliefs are so easy to deconstruct. But on general principles, it’s best not to feed trolls.

RW (12:23:38) :

By the way, I’ve asked you twice and you haven’t answered: if there were no greenhouse gases in the atmosphere, what would be heating the upper atmosphere to make it warmer than it is today?

Because the Sun would still be warming the Earth. A heat source will still warm air that is devoid of greenhouse gases.

The amount of incoming UV radiation would be exactly the same as it is now. The same amount would reach the Earth’s surface and warm it. The Earth would radiate back the same amount of IR radiation. Since less of the IR would be warming the lower atmosphere, more of it would be available to warm the upper atmosphere on its way back into space.

From Understanding the Earth by Geoff Brown (a geology text book)…

“Stratospheric aerosols alter the global radiation budget mainly by absorbing and backscattering incoming solar radiation, and they also absorb some outgoing infrared radiation.” Aerosols have a double-warming effect on the upper atmosphere. Incoming UV and outgoing IR both warm stratospheric aerosols.

In the absence of a greenhouse effect…The same amount of incoming UV would warm the Stratosphere and more outgoing IR would warm the Stratosphere.

Am I certain that this would result in a warmer Stratosphere? No. Too many variables to consider.

Now, perhaps you will return the favor an explain how an increase of “well mixed greenhouse gases” cool the Stratosphere and an absence of “well mixed greenhouse gases” also cool the Stratosphere.

RW (12:22:16) :

You seem to be struggling with a number of basic misconceptions:

1. You are denying that CO2 is a greenhouse gas, which is absurd; that has been known for 150 years.

No. I have never denied that CO2 is a greenhouse gas. I have clearly stated on numerous occasions that the only facet of the AGW hypothesis that is correct, is the assertion that CO2 is a greenhouse gas.

However the greenhouse effect of CO2 is a logarithmic function of diminishing returns. Hence, there is no secular relationship between CO2 and temperature.

Most of the greenhouse effect comes from water vapor…A little bit of CO2 is enough to yield a disproportionately large percentage of CO2’s entire greenhouse warming potential. Each additional unit of CO2 yields less greenhouse warming than the unit preceding it.

2. You don’t seem to understand a logarithmic relationship. If y ∝ log(x), then doubling x always results in the same increase in y. Always. This relationship never ‘vanishes’ and it has nothing in common with asymptotic functions.

If “x” was CO2 and “y” was delta (temp) in the equation y=ln(x)…

+0 CO2 -> +0 T
+1 CO2 -> +2 T
+2 CO2 -> +3 T
+8 CO2 -> +4 T

Of course the formula is not so simple as T=ln(CO2).

I should have said that an asymptotically increasing function never reaches infinity in the same way that a logarithmic function never vanishes.

3. When you say “If there was a secular or linear relationship between CO2 and temperature there would be a long-term correlation between CO2 and temperature. There is no such correlation at the Phanerozoic scale”, you would only be right if CO2 was the only thing changing. How could you be unaware that over those timescales, that is a disastrously wrong assumption?

Because irrespective of plate tectonics and all of the other massive changes apart from CO2…Earth’s hothouse phases had a remarkably consistent average temperature of about 22 C…And three of the four icehouse phases pretty well settled in with ~12 C average temperatures.

The continents moved all over the place…building and tearing down mountain ranges several times over…CO2 more or less declined in a generally secular manner…Yet the temperatures generally oscillated from 12 C to 22 C.

The icehouse vs. hothouse phases may have been dictated by plate tectonics or astrophysical phenomena…But CO2 appears to have played no role in that cycle.

4. You don’t understand how to calculate trends. Drawing a line joining two arbitrary points in the data is not a valid method. Using a 6th order polynomial is absurd over-fitting. Presumably you are unaware that if you were to do a 12th order polynomial, you’d get an even better fit – and even less meaningful.

No. I understand that quite well. But I also know how to recognize non-linear “nested” cycles. Like the First Order Relative Changes of Sea Level and the Second Order Relative Changes of Sea Level on this chart.

A third or sixth order polynomial would fit the First Order quite well. A twelfth order polynomial might just fit the Second Order quite well. A linear trend-line would be meaningless on either.

I would be amazed if you really genuinely couldn’t understand these four points. You seem to be affecting a sort of pseudo-ignorance, for reasons I can’t begin to imagine.

I’m amazed that I’ve put up with your insults for so long…Maybe my temper works on geological time.

You are just not sufficiently familiar enough with geology and geophysics to be able to look at climate change from the same perspective that I do. I fail to see why that elicits such obnoxiousness on your part. Oh well… C’est la vie.

1. You are denying that CO2 is a greenhouse gas, which is absurd; that has been known for 150 years.

2. You don’t seem to understand a logarithmic relationship. If y ∝ log(x), then doubling x always results in the same increase in y. Always. This relationship never ‘vanishes’ and it has nothing in common with asymptotic functions.

3. When you say “If there was a secular or linear relationship between CO2 and temperature there would be a long-term correlation between CO2 and temperature. There is no such correlation at the Phanerozoic scale”, you would only be right if CO2 was the only thing changing. How could you be unaware that over those timescales, that is a disastrously wrong assumption?

4. You don’t understand how to calculate trends. Drawing a line joining two arbitrary points in the data is not a valid method. Using a 6th order polynomial is absurd over-fitting. Presumably you are unaware that if you were to do a 12th order polynomial, you’d get an even better fit – and even less meaningful.

I would be amazed if you really genuinely couldn’t understand these four points. You seem to be affecting a sort of pseudo-ignorance, for reasons I can’t begin to imagine.

RW (04:33:27) :
“So an increase in greenhouse gases cools the Stratosphere and a decrease in greenhouse gases cools the Stratosphere?”
Why should there be one fixed linear relation between CO2 concentrations and stratospheric temperatures that applies for all possible values of CO2? I invite you again to think about the case of no greenhouse gases. You said that the stratosphere would be warmer in that case than it is today. What would be heating it?

Then, why does an increase of “well mixed greenhouse gases” cool the Stratosphere and an absence of “well mixed greenhouse gases” also cool the Stratosphere.

You said “There’s an apparent secular positive trend in three peaks and two troughs of any symmetrical oscillating function”, but then later “It’s an apparent negative secular trend. The point was that a symmetrical oscillating function can impart an apparent secular trend where there is no true secular trend.” Given
that you claimed that hadcrut temperature data could be represented by an oscillating function, and that there were three peaks and two troughs, how do you reconcile the observed upward trend with the mathematically inevitable downward trend that would actually be derived for an oscillating function over 2.5 oscillations containing three peaks and two troughs?

I also said that asymmetric oscillating functions can have apparent, but insignificant, linear secular trends and that an underlying lower frequency/higher amplitude oscillating function can also impart an apparent secular trend.

“…there is no secular relationship between CO2 and temperature”
Now this is simply denialism. If this were true, there would be no greenhouse effect.

No. It just means that the greenhouse effect would not significantly vary with variations in the atmospheric concentration of CO2.

“If the postulated feed back mechanisms are negated, CO2’s greenhouse effect is
logarithmic and not linear. So any secular relationship would vanish with increasing CO2 concentrations.”
Non sequitur. Logarithmic relationships don’t ever vanish.

It’s not a no sequitor; it’s an effective approximation. Logarithmic functions don’t vanish in exactly the same way asymptotic functions never reach infinity.

“So…Let’s make sure I understand. The secular relationship between CO2 and temperature is only operative over intermediate periods of time…Is that correct?”
I’m very surprised that you seem to be struggling with very simple concepts. There is, of course, always and inevitably a relationship between the concentration of a greenhouse gas and global temperatures. If you expect as a result of that to see a 100% correlation between the two variables over any and all timescales, then your science intuition is woeful. So, let me make sure I understand: do you think there should be a 100% correlation between CO2 concentrations and temperatures over all timescales?

If there was a secular or linear relationship between CO2 and temperature there would be a long-term correlation between CO2 and temperature. There is no such correlation at the Phanerozoic scale. The Pleistocene cyclical relationship went bust somewhere around 1939 (if the ice core CO2 are correct). The correlation between CO2 and ~0.5 C of warming from 1908-1942 is different from the _0.5 C of warming from 1978-2003. And CO2 concentrations increased faster during the
~0.2 C of cooling from 1942-1978 than they increased during the 1908-1942 warming…And CO2 concentrations have increased at the same rate (if not faster) during the ~0.4 C of cooling since 2003 than they did during the 1978-2003 warming.

“That steady decrease should be apparent in the observed data before, between and after the two volcanic events. It is not.”
I already told you that in all three periods not affected by volcanoes, the tren in stratospheric temperatures was negative. You only got a different result by wrongly excluding data that was not affected by the eruptions, such as the entire year before Pinatubo erupted.

No. The trend was not negative between El Chicon and Pinatubo; nor is it negative since Pinatubo. It was slightly negative before El Chicon. The Stratosphere warmed slightly from Jan. 1995 to Jan 2002…As did the Troposphere. Since Jan. 2002, both the Stratosphere and the Troposphere have linear cooling trends. Essentially, the Stratosphere warmed during both volcanic events and the cooled rapidly in a “steplike” transition; assuming a generally neutral trend after the passing of each volcanic event.

From Ramaswamy et. al.: ”Observations reveal that the substantial cooling of the global lower stratosphere over 1979–2003 occurred in two pronounced steplike transitions. These arose in the aftermath of two major volcanic eruptions, with each cooling transition being followed by a period of relatively steady temperatures.”

“When did the climate start paying attention to the human-derived numerical values of decades?”
That’s a rather hilariously odd comment. Are you saying that no natural phenomenon can be measured using human-invented units? How do you propose to measure climate trends? Tell me the units you want to use and I’ll tell you the answer,
and we can pretend decades don’t exist if you like. It won’t change the underlying physics.

To the climate “decades” are “pretend.” The underlying physics doesn’t cause climate variations to fit into human decades.

“According to the UAH Lower Troposphere global temperature anomalies, the Earth
warmed 0.96 C from Dec. 1978 to Apr. 1998 and then cooled -0.67 C from Apr. 1998 to Apr. 2009. That works out to a warming rate of 0.62 C per decade from Dec.
1979 to Apr. 1998 and a cooling rate of -0.73 per decade from Apr. 1998 to the present.”
Drawing a line between two points is not how trends are calculated.

So…April 2009 was warmer than April 1998? I would have guessed that +0.76 was a larger number than +0.09.

Drawing a linear trend-line through a non-linear function is also another way to not properly calculate a trend. A 6th order polynomial fit of the UAH LT data yields an “S” shaped curve that starts just below 0 in Dec 1978 and finishes just above 0 in Apr. 2009. It looks a lot like an oscillation. And it starts and ends very close to phase shifts in the PDO.

A linear trend through the UAH LT data has an R^2 of 0.2792…The sixth order polynomial has an R^2 of 0.3513.

It’s been a while since I took statistics…Isn’t R^2 a measure of how well a trend fits a function?

“Let me guess…Dec. 1978 to Apr. 1998 is significant and Apr. 1998 to Apr. 2009 is “internal variation”…Right?”
If you want to find out the answer to this you need first of all to calculate the trends properly, using all the data, and then to choose an appropriate statistical model for the noise pattern in the data to calculate the error on the trend. If the error is larger than the trend for a given period, then the trend is not significant.

Use all of the data? If I use all of the data, “geological and solar” processes overwhelm variations in CO2. If I don’t use enough data, “internal variation” and natural variability overwhelm the variations in CO2.

So…I have to use just the right bits of data that fit the AGW narrative? Sounds like cherry-picking to me.

Why should there be one fixed linear relation between CO2 concentrations and stratospheric temperatures that applies for all possible values of CO2? I invite you again to think about the case of no greenhouse gases. You said that the stratosphere would be warmer in that case than it is today. What would be heating it?

You said “There’s an apparent secular positive trend in three peaks and two troughs of any symmetrical oscillating function”, but then later “It’s an apparent negative secular trend. The point was that a symmetrical oscillating function can impart an apparent secular trend where there is no true secular trend.” Given
that you claimed that hadcrut temperature data could be represented by an oscillating function, and that there were three peaks and two troughs, how do you reconcile the observed upward trend with the mathematically inevitable downward trend that would actually be derived for an oscillating function over 2.5 oscillations containing three peaks and two troughs?

“…there is no secular relationship between CO2 and temperature”

Now this is simply denialism. If this were true, there would be no greenhouse effect.

“If the postulated feed back mechanisms are negated, CO2’s greenhouse effect is
logarithmic and not linear. So any secular relationship would vanish with increasing CO2 concentrations.”

Non sequitur. Logarithmic relationships don’t ever vanish.

“So…Let’s make sure I understand. The secular relationship between CO2 and temperature is only operative over intermediate periods of time…Is that correct?”

I’m very surprised that you seem to be struggling with very simple concepts. There is, of course, always and inevitably a relationship between the concentration of a greenhouse gas and global temperatures. If you expect as a result of that to see a 100% correlation between the two variables over any and all timescales, then your science intuition is woeful. So, let me make sure I understand: do you think there should be a 100% correlation between CO2 concentrations and temperatures over all timescales?

“That steady decrease should be apparent in the observed data before, between and after the two volcanic events. It is not.”

I already told you that in all three periods not affected by volcanoes, the tren in stratospheric temperatures was negative. You only got a different result by wrongly excluding data that was not affected by the eruptions, such as the entire year before Pinatubo erupted.

“You got me again…I am a weather denialist. It’s all climate; there’s no such thing as weather.”

OK…

“When did the climate start paying attention to the human-derived numerical values of decades?”

That’s a rather hilariously odd comment. Are you saying that no natural phenomenon can be measured using human-invented units? How do you propose to measure climate trends? Tell me the units you want to use and I’ll tell you the answer,
and we can pretend decades don’t exist if you like. It won’t change the underlying physics.

“According to the UAH Lower Troposphere global temperature anomalies, the Earth
warmed 0.96 C from Dec. 1978 to Apr. 1998 and then cooled -0.67 C from Apr. 1998 to Apr. 2009. That works out to a warming rate of 0.62 C per decade from Dec.
1979 to Apr. 1998 and a cooling rate of -0.73 per decade from Apr. 1998 to the present.”

Drawing a line between two points is not how trends are calculated.

“Let me guess…Dec. 1978 to Apr. 1998 is significant and Apr. 1998 to Apr. 2009 is “internal variation”…Right?”

If you want to find out the answer to this you need first of all to calculate the trends properly, using all the data, and then to choose an appropriate statistical model for the noise pattern in the data to calculate the error on the trend. If the error is larger than the trend for a given period, then the trend is not significant.

RW (08:34:12) :

“If Earth’s atmosphere lacked greenhouse gases…The lower atmosphere would be colder and the upper atmosphere would be warmer.”

No. Both would be colder, in that case. Think about it. What would be heating the upper atmosphere?

So an increase in greenhouse gases cools the Stratosphere and a decrease in greenhouse gases cools the Stratosphere?

How is stratospheric cooling evidence of tropospheric greenhouse warming if the Stratosphere cools in response to any change in GHG’s?

“There’s an apparent secular positive trend in three peaks and two troughs of any symmetrical oscillating function.”

Nope.

You got me there. I was doing the “math” in by head. It’s an apparent negative secular trend. The point was that a symmetrical oscillating function can impart an apparent secular trend where there is no true secular trend.

“However, there is no “secular” relationship between CO2 and temperature.”

That is an absurd claim.

It may sound absurd to you; but there is no secular relationship between CO2 and temperature. Pleistocene ice core data suggest that there was a cyclical relationship…However, if the ice core CO2 data are correct, that relationship has changed… LINK …If the relationship was truly secular, it would be quite a bit hotter today than it was ~130,000 years ago during the Sangamon interglacial. But the Sangamon was considerably warmer than it is now (several degrees C) and sea level was 5-8 meters higher…Yet CO2 was at most ~300ppm (if the ice core data are correct).

If the postulated feed back mechanisms are negated, CO2’s greenhouse effect is logarithmic and not linear. So any secular relationship would vanish with increasing CO2 concentrations.

“If there was, this chart would look a lot different than it does…Phanerozoic CO2 vs. Temp.”

Presumably you were unaware that geological and solar processes dominate over the longest timescales.

That’s mind boggling. I must’ve wasted the last 30+ years of my life as a geophysicist…All this time I thought geological process dominated over short periods of time…(slaps own forehead).

So…Let’s make sure I understand. The secular relationship between CO2 and temperature is only operative over intermediate periods of time…Is that correct?

The cooling from 1942-1978 and from 2003 to the present are such short periods of time that “internal variation” and natural variability overwhelm the greenhouse effect of increasing CO2. And the Phanerozoic lack of correlation is due to “geological and solar processes” overwhelming the anti-greenhouse effect of decreasing CO2. So…I guess that this interglacial lacks the same cyclical relationship between CO2 and temperature because ““geological and solar processes” suddenly changed during the Holocene transgression? Or is it just “internal variation” and natural variability?

“However, if he removed the volcanic effects, the model predicted a steady decline in stratospheric temperatures; which was inconsistent with the observed temperature trends.”

Yes, if you remove events which actually occurred, it’s hardly a surprise that you can’t replicate what actually occurred.

That steady decrease should be apparent in the observed data before, between and after the two volcanic events. It is not.

“Since almost all of the stratospheric cooling over the satellite record is due to volcanic events and stratospheric ozone depletion…There’s little (if any) stratospheric cooling left over to account for much enhancement of tropospheric greenhouse warming over the past 30 years.”

Try reading the Ramaswamy paper again. Here’s a small bit, which you already directly quoted but don’t seem to have noticed: “the overall lower stratospheric temperature decline is driven primarily by the depletion of ozone, and to a lesser extent by the increase in well-mixed greenhouse gases.”

I noticed the words “primarily” and “lesser.” That’s why I quoted that particular passage.

Ozone depletion (if it is really occurring) cools the Stratosphere because the ozone-depleted upper atmosphere absorbs less incoming UV radiation than an ozone-enriched upper atmosphere would absorb. Well mixed greenhouse gasses would add to that cooling by retaining more outgoing IR radiation in the lower atmosphere.

Ozone depletion-driven stratospheric cooling wouldn’t have to be concurrent with tropospheric warming. GHG-driven tropospheric warming would have to be complemented by a concurrent stratospheric cooling.

“I picked the dates to fully isolate the effects of the volcanic events and the 1997-1998 ENSO.”

Temperatures are not influenced by volcanoes until after the eruption. You chose periods that ended three months and one year before the two major eruptions. Why?

I picked them to put the “bumps” in the middle of the segments that included the bumps. I could pick the segments in almost any configuration and the curves almost never approach one-another when the Stratosphere curve is placed above the Troposphere curve.

Ideally it should look something like a Neutron-Density curve on a well log through a gas bearing sandstone. The two curves should approach one another, if not cross over, during periods in which there is an enhancement of greenhouse warming of the lower atmosphere.

“As far as your comment about “internal variation”….I’m certain that advocates of the Ptolemaic solar system invoked “internal variation” and natural variability quite often in dismissing criticisms of the epicycles hypothesis to explain away retrograde motion.”

‘Internal variation’ in a climate context is also known as ‘weather’. Are you denying that weather exists?

You got me again…I am a weather denialist. It’s all climate; there’s no such thing as weather.

“The Earth’s climate did not behave in a manner consistent with CO2-driven global warming from the dawn of the Cambrian until 1978″

At no time in the Earth’s history has CO2 not absorbed infrared radiation. At no time in the Earth’s history has its radiative balance been independent of the CO2 concentration.

That explains the lack of correlation between CO2 and temperature quite well.

“…And since 1998 it has not behaved in a manner consistent with CO2-driven global warming.”

The average temperature in the 2000s is almost 0.2°C higher than the average temperature in the 1990s.

And?

When did the climate start paying attention to the human-derived numerical values of decades?
According to the UAH Lower Troposphere global temperature anomalies, the Earth warmed 0.96 C from Dec. 1978 to Apr. 1998 and then cooled -0.67 C from Apr. 1998 to Apr. 2009.

That works out to a warming rate of 0.62 C per decade from Dec. 1979 to Apr. 1998 and a cooling rate of -0.73 per decade from Apr. 1998 to the present.

Let me guess…Dec. 1978 to Apr. 1998 is significant and Apr. 1998 to Apr. 2009 is “internal variation”…Right?

“And since the Stratosphere did not concurrently cool with the warming of the Troposphere, it really didn’t behave a manner consistent with CO2-driven global warming from 1979-1998 either.”

The stratosphere did concurrently cool. Try reading the Ramaswamy paper again.

I ‘ve got a copy of it right in front of me.

The Stratosphere cooled in the immediate aftermath of the major volcanic events. Before, between and after those events stratospheric temperatures remained relatively flat, irrespective of what the Troposphere was doing…”Observations reveal that the substantial cooling of the global lower stratosphere over 1979–2003 occurred in two pronounced steplike transitions. These arose in the aftermath of two major volcanic eruptions, with each cooling transition being followed by a period of relatively steady temperatures.”

No. Both would be colder, in that case. Think about it. What would be heating the upper atmosphere?

“There’s an apparent secular positive trend in three peaks and two troughs of any symmetrical oscillating function.”

Nope.

“However, there is no “secular” relationship between CO2 and temperature.”

That is an absurd claim.

“If there was, this chart would look a lot different than it does…Phanerozoic CO2 vs. Temp.”

Presumably you were unaware that geological and solar processes dominate over the longest timescales.

“However, if he removed the volcanic effects, the model predicted a steady decline in stratospheric temperatures; which was inconsistent with the observed temperature trends.”

Yes, if you remove events which actually occurred, it’s hardly a surprise that you can’t replicate what actually occurred.

“Since almost all of the stratospheric cooling over the satellite record is due to volcanic events and stratospheric ozone depletion…There’s little (if any) stratospheric cooling left over to account for much enhancement of tropospheric greenhouse warming over the past 30 years.”

Try reading the Ramaswamy paper again. Here’s a small bit, which you already directly quoted but don’t seem to have noticed: “the overall lower stratospheric temperature decline is driven primarily by the depletion of ozone, and to a lesser extent by the increase in well-mixed greenhouse gases.”

“I picked the dates to fully isolate the effects of the volcanic events and the 1997-1998 ENSO.”

Temperatures are not influenced by volcanoes until after the eruption. You chose periods that ended three months and one year before the two major eruptions. Why?

“As far as your comment about “internal variation”….I’m certain that advocates of the Ptolemaic solar system invoked “internal variation” and natural variability quite often in dismissing criticisms of the epicycles hypothesis to explain away retrograde motion.”

‘Internal variation’ in a climate context is also known as ‘weather’. Are you denying that weather exists?

“The Earth’s climate did not behave in a manner consistent with CO2-driven global warming from the dawn of the Cambrian until 1978″

At no time in the Earth’s history has CO2 not absorbed infrared radiation. At no time in the Earth’s history has its radiative balance been independent of the CO2 concentration.

“…And since 1998 it has not behaved in a manner consistent with CO2-driven global warming.”

The average temperature in the 2000s is almost 0.2°C higher than the average temperature in the 1990s.

“And since the Stratosphere did not concurrently cool with the warming of the Troposphere, it really didn’t behave a manner consistent with CO2-driven global warming from 1979-1998 either.”

The stratosphere did concurrently cool. Try reading the Ramaswamy paper again.

RW (19:02:09) :

Dave Middleton: you can call whatever you’re seeing a fluctuation if you like. There is still a secular upward trend, and it’s rather weird to pretend that there isn’t.

There’s an apparent secular positive trend in three peaks and two troughs of any symmetrical oscillating function. There is also an apparent secular trend in any asymmetrical oscillating function. A complex oscillating function will also appear to have secular trends due to the underlying lower frequency function.

However, there is no “secular” relationship between CO2 and temperature. If there was, this chart would look a lot different than it does…Phanerozoic CO2 vs. Temp.

As far as the Stratosphere goes, it is a “step” function; see Ramaswamy et. al. (2006)…

Observations reveal that the substantial cooling of the global lower stratosphere over 1979–2003 occurred in two pronounced steplike transitions. These arose in the aftermath of two major volcanic eruptions, with each cooling transition being followed by a period of relatively steady temperatures…

[...]

[T]he overall lower stratospheric temperature decline is driven primarily by the depletion of ozone, and to a lesser extent by the increase in well-mixed greenhouse gases.

[…]

If the solar and volcanic aerosol forcing were entirely absent, the temperature evolution would have comprised a steady decrease driven by [anthropogenic forcing], but this is inconsistent with the observations.

Ramaswamy was able to replicate the step function with a combination of natural and anthropogenic effects. The bulk of the anthropogenic cooling was the result of ozone depletion (Isn’t stratospheric cooling the primary evidence for stratospheric ozone depletion?). However, if he removed the volcanic effects, the model predicted a steady decline in stratospheric temperatures; which was inconsistent with the observed temperature trends.

The greenhouse effect does not simply redistribute heat. Heat that would otherwise escape into space is retained in the atmosphere. This is pretty basic stuff, and I’m surprised you don’t know it.

The result of the greenhouse effect is a redistribution of energy (heat). CO2 is generally “transparent” to incoming high frequency (UV) solar radiation and “opaque” to certain bandwidths of low frequency (IR) radiation. After the solar radiation heats the Earth’s surface, some of this energy is radiated back into the atmosphere. Certain bandwidths of the radiated IR are “absorbed” or “retained” by C2. This “warms” the lower atmosphere. The greenhouse effect results in the lower atmosphere effectively retaining a higher percentage of the outgoing IR. Since heat is retained in the lower atmosphere, less Earth-radiated IR reaches the upper atmosphere…leading to cooling of the upper atmosphere.

Hence the ultimate effect is a redistribution of heat within the atmosphere.

Since almost all of the stratospheric cooling over the satellite record is due to volcanic events and stratospheric ozone depletion…There’s little (if any) stratospheric cooling left over to account for much enhancement of tropospheric greenhouse warming over the past 30 years.

As for your graphs, well, in any period of just a few years, internal variation dominates and long term trends are not apparent. I said that before, you made no comment so I presume you were unaware of that as well. Notwithstanding that they are over periods too short to be meaningful, your graphs stop and start at very strange places. Pinatubo erupted in June 1991, not June 1990. El Chichón erupted in March 1982, not January 1982. The effects of these eruptions on stratospheric temperatures were over by January 1984 and June 1993, not June 1984 and June 1994. Why did you pick the dates you did?

I picked the dates to fully isolate the effects of the volcanic events and the 1997-1998 ENSO.

As far as your comment about “internal variation”….I’m certain that advocates of the Ptolemaic solar system invoked “internal variation” and natural variability quite often in dismissing criticisms of the epicycles hypothesis to explain away retrograde motion.

The Earth’s climate did not behave in a manner consistent with CO2-driven global warming from the dawn of the Cambrian until 1978…And since 1998 it has not behaved in a manner consistent with CO2-driven global warming. And since the Stratosphere did not concurrently cool with the warming of the Troposphere, it really didn’t behave a manner consistent with CO2-driven global warming from 1979-1998 either.

oms (10:23:26) :

[...]

Dave Middleton’s original statement, “Greenhouse warming, by definition, requires a redistribution of heat from the upper atmosphere to the lower atmosphere,” doesn’t seem 100% correct (in that I don’t believe it to be quite definitional) but it doesn’t seem to be as wrong as you are making it out to be either.

It is a simplification. But heat redistribution is the effective result. If Earth’s atmosphere lacked greenhouse gases…The lower atmosphere would be colder and the upper atmosphere would be warmer.

Dave Middleton’s original statement, “Greenhouse warming, by definition, requires a redistribution of heat from the upper atmosphere to the lower atmosphere,” doesn’t seem 100% correct (in that I don’t believe it to be quite definitional) but it doesn’t seem to be as wrong as you are making it out to be either.