Error Theory

Crossposted at WUWT

A recent article by Fox News reporter Maxim Lott highlights a sentence from leaked draft of AR5 that seems to acknowledge a larger solar influence on climate than previously estimated:

“[Results] do suggest the possibility of a much larger impact of solar variations on the stratosphere than previously thought, and some studies have suggested that this may lead to significant regional impacts on climate,” reads a draft copy of a major, upcoming report from the U.N.’s Intergovernmental Panel on Climate Change (IPCC).

The bracketed "[results]" are the post-AR4 findings of "much greater than expected reduction at UV wavelengths in the recent declining solar cycle phase." (AR5 SOD page 11-57.) Lott describes the vaguely referenced stratospheric and climate impacts of this larger-than-expected UV-shift as an admission by the IPCC that "[h]eat from the sun may play a larger role than previously thought." The Fox headline writers go even further, describing the IPCC as "admitting solar activity may play significant role in global warming."

This reporting is a bit messy. No, the UV-shift is not a shift in heat from the sun. Total solar irradiance (TSI) changes very little as solar activity ramps up and down. The UV shift just alters where in the electromagnetic spectrum the heat comes through. Yes this UV shift could affect the climate system in ways that have a substantial impact on global temperature but no, the IPCC has not admitted it. They actually belittle the possibility. That's what the reference to "regional impacts on climate" is about. They mean regional as opposed to global. They are denying any significant impact on global temperature.

Still, the gist of the story—that the draft report admits stronger solar effects on global temperature than previously estimated—is correct, only this admission comes not in chapter 11 but in chapter 7, where it is followed by a crude trick of evasion. The UV-shift discussion in chapter 11 repeats this same trick of evasion. To catch this trick in chapter 11 you have to first catch it in chapter 7.

Dodging the evidence

This is the third article Lott has written on the Second Order Draft of AR5 (which I leaked to public in mid-December). All have been pretty good.

One was about Donna Laframboise's finding that the IPCC, despite being taken to task for relying on non peer-reviewed literature from political advocacy groups in its last assessment report (AR4), is continuing the practice in AR5. Lott also wrote an article on what Anthony calls as "the real IPCC AR5 draft bombshell": the chapter 1 graph showing current temperatures falling substantially below the entire range of projections from every previous assessment report.

To follow up, Maxim emailed that he wanted to cover my take on the draft IPCC report: that its admission of strong evidence for some mechanism of solar forcing beyond what is included in the IPCC models destroys all confidence in projections based on those models.

Excellent. Thanks Maxim. It was just a little awkward that he wanted to start with the chapter 11 sentence about the new much higher estimate of the shift in solar UV, but we can make it right in the end. There is space here to put that UV-shift discussion on its proper foundation (and from the length of this post, let's just say I'm not surprised that Maxim declined my advice to provide the necessary context).

I used my leak of the draft report to publicize an important new admission by the IPCC: that solar-climate correlations found in the geologic record seem to imply a substantially stronger solar effect on climate than can be accounted by the very slight variation in TSI (the only solar forcing that is included in current IPCC models). Chapter 7, page 7-43:

The forcing from changes in total solar irradiance alone does not seem to account for these observations, implying the existence of an amplifying mechanism such as the hypothesized GCR-cloud link.

This is huge. The authors admit broad evidence ("many results") indicating that some substantial mechanism of enhanced solar forcing must be at work. They mention one theory of what that mechanism might be (GCR-cloud), and then comes the crude trick of evasion.

They proceed to judge (very prematurely) that the evidence regarding the GCR-cloud mechanism indicates a weak effect, and they use this as an excuse to ignore the already admitted evidence for some substantial mechanism of solar amplification. The evidence is never mentioned again and it is never taken into account anywhere else in the report. As Maxim Lott quotes me in his Fox News article:

“Even after the IPCC acknowledges extensive evidence for ... solar forcing beyond what they included in their models, they still make no attempt to account for this omission in their predictions. ... It's insane,” he told FoxNews.com.

For a response, Lott quotes UCS press flack Aaron Huertas:

“I see climate contrarians try this trick almost every time a big new solar study comes out. They somehow tend to neglect mentioning that solar variation is smaller than the heat-trapping power of carbon dioxide,” Aaron Huertas of the Union of Concerned Scientists told FoxNews.com.

No doofus, I am not neglecting to mention what a tiny effect solar variation has in the current climate models. I am pointing specifically to it and noting how even the IPCC now admits that the paleo-evidence points to a substantially larger solar effect than its models factor in.

The IPCC's continued refusal to take into account the evidence for solar amplification (even after they have admitted it) is anti-scientific. For those who have read my earlier posts on the leaked AR5, this is familiar territory. I have been accusing the IPCC of inverting the scientific method: of using theory (their dismissive assessment of the GCR-cloud theory) as a grounds for ignoring evidence (the paleo-evidence for some substantial mechanism of solar amplification).

To fully support this charge it is necessary to look beyond chapter 7 to the other parts of the report where possible mechanisms of solar amplification might be addressed. Most importantly, this means chapter 11, where the UV shift is discussed.


UV effects on ozone and atmospheric oscillation: are they only "regional"?

UV in the shorter wavelengths (UVB and UVC) gets absorbed in stratosphere where it gets caught up in the creation and destruction of stratospheric ozone (net creation). More UV means more absorption which warms the stratosphere, with possible consequences for atmospheric circulation.

Total solar irradiance also increases when solar activity is high, but only very slightly, so that with the new five-times-higher estimate of the UV shift (Haigh et al. 2010), the amount of visible light reaching the planet's surface is now thought to decrease slightly. Thus in addition to having temperature effects that impinge on the troposphere from above, the UV shift could also have temperature effects than come up from the floor of the troposphere.

A 2011 modeling study (also co-authored by Joanna Haigh) claims that these UV effects can explain the observed correlation between solar activity and atmospheric oscillations in the North Atlantic (the Arctic Oscillation and the North Atlantic Oscillation or AO/NAO):

In years of low UV activity unusually cold air forms over the tropics in the stratosphere, about 50km up. This is balanced by more easterly flow of air over the mid latitudes – a pattern which then ‘burrows’ its way down to the surface, bringing easterly winds and cold winters to northern Europe. 

When solar UV output is higher than usual, the opposite occurs and there are strong westerlies which bring warm air and hence milder winters to Europe.

But don't worry, Haigh and her co-authors assure the public, this research poses no threat to the "consensus" claim that late 20th century warming can only have been caused by CO2. The described effects are only regional, not global:

Sarah Ineson, who performed the experiments, said: “What we’re seeing is UV levels affecting the distribution of air masses around the Atlantic basin. This causes a redistribution of heat – so while Europe and the US may be cooler, Canada and the Mediterranean will be warmer, and there is little direct impact on global temperatures.”

But a redistribution of heat does not preclude secular effects as well, and there is good reason to think that the AO/NAO oscillation (and similar effects in the Pacific), would affect global temperature. Notice how much bigger the Rossby waves in the polar jet stream are when AO/NAO is in its weak phase (from Joseph D'Aleo):



The jet stream forms where temperature differentials are at their maximum, and because surface air from the two sides are converging at this point (the northern end of the mid-latitude Hadley cell in the graphic at the top of the post), this is a main source of cloud formation. Thus as Stephen Wilde has been pointing out for several years, anything that causes the polar jets to move equator-ward, or increases the amplitude of the meanderings of the polar jets, is likely to cause a net increase in cloud formation.

Clouds reflect away more heat than they trap so almost any significant change in net cloudiness could easily have an effect on global temperature that is large compared to the effects of CO2 variation or TSI variation. Stephen argues (p. 7) that the polar jet seems to have been further south during the Little Ice Age (when solar activity was low) and further north during the Medieval Warm Period (when solar activity was high). These are known to have been global termperature events, not just regional, providing at least anecdotal evidence for a solar effect on global temperature through atmospheric circulation patterns.

Are Haigh and the IPCC taking such possibilities into account? It does not seem so. Their logic is just that there is a redistribution of heat taking place and therefore there is no effect on global temperature, a clear non-sequitur. This interpretation is reinforced by remarks from NCAR's 2012 workshop: "The Effects of Solar Variability on Earth’s Climate."


Conference chairman: We don't even know HOW to model that cloud-formation stuff

NCAR's conference report (p. 20) claims that "top down" effects of solar activity on stratospheric ozone would not have a significant effect on global temperature:

If borne out by future studies and shown to be of sufficient magnitude, this mechanism could be an important pathway in the Sun-climate connection, particularly in terms of regional impacts. However, it is important to realize that, unlike the bottom-up mechanism, it can in itself contribute very little to global temperature variations.

An interview with conference chair Gerald North, however, offers a quite different view:

For instance, solar energetic particles and cosmic rays could reduce ozone levels in the stratosphere. This in turn alters the behavior of the atmosphere below it, perhaps even pushing storms on the surface off course.

"In the lower stratosphere, the presence of ozone causes a local warming because of the breakup of ozone molecules by ultraviolet light," climate scientist Jerry North at Texas A&M University told SPACE.com.

When the ozone is removed, "the stratosphere there becomes cooler, increasing the temperature contrast between the tropics and the polar region. The contrast in temperatures in the stratosphere and the upper troposphere leads to instabilities in the atmospheric flow west to east. The instabilities make for eddies or irregular motions." 

These eddies feed the strength of jet streams, ultimately altering flows in the upper troposphere, the layer of atmosphere closest to Earth's surface. "The geographical positioning of the jets aloft can alter the distribution of storms over the middle latitudes," North said. "So the sun might have a role to play in this kind of process. I would have to say this would be a very difficult mechanism to prove in climate models. That does not mean it may not exist — just hard to prove."

North seems here to be describing the Wilde-thing. His reference to the effects of jet stream shifts on "the distribution of storms" sounds like a reference to cloud formation and cloud effects, and North's insight on this subject? That the models are not good enough yet to evaluate it.

North also throws some more complexity into the mix. In addition to the UV shift, there are a couple of other avenues by which solar activity can affect ozone: "solar energetic particles and cosmic rays."

There is evidence that the dreaded ozone hole is mainly caused by galactic cosmic radiation (GCR), not by the interaction of human produced CFCs and solar UV. A weak solar wind (low solar activity) allows more GCR to reach our planet, decreasing ozone, especially at the poles.

"Solar energetic particles" work in the opposite direction. Ozone is destroyed by the solar protons that are produced by solar flares (high solar activity), again mostly at the poles.

The whole subject is far from sorted out but it seems that in some way the variety of different solar effects on ozone conspire to have a strongly differential effect on stratospheric temperature between the equator and the poles and it is this differential that causes the jet stream to move.


Steering wheel vs. gas pedal

Is it implausible to think that the climate system could be controlled by top down effects? The bottom-up energy flow is so much stronger. Of the solar radiation that is not reflected away from the planet all-together, only a small amount is absorbed in the stratosphere, a bigger chunk is absorbed in the troposphere, and most is absorbed by the planet's surface.

How could the relatively tiny stratospheric effects compete? That's like thinking a single finger on the steering wheel of car could determine the direction in which the car travels. Pshaw. It has to be the drive train that determines the direction of the car. That's where all the horsepower is.

Yeah I'm joking. Larger forces obviously can be steered by smaller forces. The "consensus" uses a version of this argument when they call CO2 "the gas pedal" (or "the main dial" or "the biggest control knob"). "Gas pedal" is not quite analogous to "steering wheel." They are talking about modulation of the drive system, but they are still looking at a multiplier effect. It just turns out that the mechanism they see as doing the multiplying (climate sensitivity) doesn't seem to actually do much multiplying.

The climate modelers themselves provide an interesting anecdote in favor of the UV-shift as steering-wheel possibility (if you accept model results as evidence). Just as with a car, no matter what you do with the gas pedal in the "consensus" climate models, the steering wheel still works the same way.

Start with this quote from the Haigh co-authored paper above (2011):

Adam Scaife, one of the Met Office scientists involved in the research, said that while some studies have observed a link between solar variability and winter climate, our research establishes this as more than just coincidence. 

He said: “We’ve been able to reproduce a consistent climate pattern, confirm how it works, and quantify it using a computer model based on the laws of physics."

In this model the UV shift is so great that visible light reaching the surface goes down when solar activity goes up. Thus the bottom-up effect here is gas-pedal-off.

Compare this to the model results from Shindel, Schmidt et. al 2001, framed when it was thought that the UV shift was smaller. Their UV shift only eats up a fraction of the increase in TSI so surface warming increases when solar activity is high. The bottom-up effect in their model is gas-pedal-on, yet they get the exact same result, with high solar activity causing a strong AO/NAO:

Our previous studies have demonstrated how external forcings can excite the AO/ NAO in the GISS GCM (22, 25). Briefly, the mechanism works as follows, using a shift toward the high-index AO/NAO as an example: (i) tropical and subtropical SSTs warm, leading to (ii) a warmer tropical and subtropical upper troposphere via moist convective processes. This results in ... [a LONG train of causality, leading to] ... a high AO/NAO index.

Set aside cynicism for the moment and suppose that, instead of demonstrating how easily climate models can be tweaked to achieve a sought-after result, both models really do (in some degree) reveal the emergent properties of "the laws of physics." Whether the gas pedal is pushed down or the gas is let off, both models turn the same direction when UV is high. Sounds like a steering wheel to me.


Recent developments show a field in flux

Turns out the Met Office completely screwed up the stratospheric temperature data that all the atmospheric circulation modelers have been back-testing to since the 1980s. Thompson et al. 2012:

A new data set of middle- and upper-stratospheric temperatures based on reprocessing of satellite radiances provides a view of stratospheric climate change during the period 1979–2005 that is strikingly different from that provided by earlier data sets. The new data call into question our understanding of observed stratospheric temperature trends and our ability to test simulations of the stratospheric response to emissions of greenhouse gases and ozone-depleting substances.

What was wrong with the original processing of the satellite data? Nobody knows:

The methodology used to generate the original Met Office SSU data remains undocumented and so the climate community are unable to explain the large discrepancies between the original Met Office and NOAA SSU products highlighted here.

Shades of the Climategate "Harry read me" file, and this was only just discovered. All the modelers have been using bogus data to try to calibrate their models. Go back to square one guys.

Another recent paper gives a taste of just how preliminary present theorizing about ozone and climate is. "Ozone depletion trumps greenhouse gas increase in jet-stream shift," say researchers at Penn State. Okay, so they think ozone (and hence the sun) is a player, but wow, look at their methodology. It's pure data sifting in search of relationships to examine:

Lee and her colleague, Steven Feldstein, professor of meteorology, developed a new method to distinguish between the effects of the two forcings. The method uses a cluster analysis to investigate the effects of ozone and greenhouse gas on several different observed wind patterns.

Nothing wrong with data sifting. This is how scientists get leads, but it is the mark of a field that is in its infancy, at least in terms of figuring out the more complex relationships. In such a circumstance it is grossly inappropriate for the IPCC and NCAR to be making extravagant claims about what can't be happening, especially when they are claiming to exclude something that simple logic says probably will happen (that jet stream shifts will affect net cloud formation), and even more especially when the paleo-evidence indicates that some such thing is happening.


It's anti-science all the way down

In sum, the IPCC authors do pretty much exactly the same thing with solar UV in chapter 11 as they do with GCR in chapter 7. They look at one snapshot of how the UV shift might affect climate then very prematurely declare that it would only have a negligible effect on global temperature. In the process there is no mention of the IPCC's earlier admission of strong paleo-evidence for some substantial mechanism of solar amplification and they proceed as if this evidence does not exist, making no attempt to give it any weight in their predictive scheme.

Theory (their truncated theory of UV-shift effects) is used to dismiss evidence. That is not science, it is anti-science, and the IPCC has been very consistent about it.


UPDATE: on my characterization of the IPCC as admitting to "extensive evidence" for some "substantial" mechanism of solar amplification

The draft report does not use these words, but they are implied, which is why I use them to describe the IPCC's admission. I also quote the IPCC's exact words so nobody is being misled, but if I'm going to characterize the admission in stronger terms I really should explain why.

On the extent of the evidence, their exact wording is "many results."  Does "many results" mean "extensive evidence"? Absolutely. They would not admit "many results" unless there were MANY results. I also know just what results they are referring to because the added sentence about the evidence pointing to some mechanism of solar amplification is almost certainly in response to my comment on the First Order Draft, which ripped the authors up and down precisely for their failure to include such a sentence.

The second section of my FOD comment lists 2 dozen studies (and there are many more), typically finding a correlation of about .5 between solar activity and global temperature going back many thousands of years. That is, solar activity "explains" in the statistical sense something like half of all past temperature change, which is far more than the tiny variation in total solar irradiance (or TSI) can begin to explain. The evidence is extensive, and I think it is fair to say that the IPCC is admitting it.

Do they admit that the evidence points to a "substantial" mechanism of solar amplification? They don't use that word, but yes, this too is implicit, and not just because the IPCC would never admit that the evidence points to some mechanism of solar forcing beyond what they include in their models unless the evidenced effect were substantial. There is also a particular implication that the indicated mechanism must not just be substantial, but must be large, compared to the very slight forcing from variation in TSI.

TSI variation produces only a tiny forcing. By the IPCC's estimate it has had 1/70th the forcing effect of CO2 variation over the last 260 years ( 0.04 vs 2.83 w/m^2, table 8.7 on p. 8-39 of the draft report, and this is over a period when solar variance was at its maximum). CO2 itself only has a very modest temperature forcing effect--to be at all dangerous it has to be multiplied up several times by water vapor feedback effects, as the IPCC assumes--so for TSI variation to have only 1/70th the forcing effect of CO2, that really is a negligible forcing effect.

TSI variation is essentially a zero in the IPCC's estimation. It can't explain anything. That's their claim. Thus when they say that there seems to be some solar forcing larger than what can be accounted for by TSI what they are implicitly saying is that it is large compared to TSI. If it were not large compared to TSI it would itself be a nothing and there would be nothing for the IPCC to admit to. If it's too big for them to ignore then it has to be large compared to TSI, which is the only solar effect they have so far been taking into account.

So that's actually a pretty big implicit admission: that total solar effects are large compared to what the IPCC has been assuming. I use "substantial" as a shorthand, which is still on the weak side. "Substantial" is still a minimum concept. It just means "not insubstantial," or "not negligible." What the IPCC is actually admitting is that the evidenced solar amplification is large compared to what they include in their models, and if it is comparatively large then it has to be, in absolute terms,  at least not-insubstantial. N'cest pas?