Click images for more details



Recent comments
Recent posts
Currently discussing

A few sites I've stumbled across recently....

Powered by Squarespace

Discussion > Sense and Sensitivity, part two.


> The only way I could conceive of the kettle having
positive feedbacks was if the spout was blocked and the
lid sealed, when the increased pressure would allow the
temperature to rise above 100C.

Another one who doesn't understand feedback. I'm beginning to wonder if anyone here apart from MartinA understand what feedback is.

As I said in another thread, I marvel at your ability to 'know' that the standard view of climate science is wrong despite lacking an understanding of something as basic as feedback. How can you have a view on the correctness of climate sensitivity estimates without such an understanding? I also wonder why a population that says that its position is 'all about the science' would fail to correct your errors before I do.

Mar 22, 2014 at 2:16 PM | Unregistered CommenterChandra

"I disagree with this analogy because actually we know that the "natural" temperature increase without positive feedbacks is 1.2C from the S-B equation"

But unfortunately the simplistic radiation balance doesn't represent the Earth's atmosphere (Heat rises - so a natural negative feedback). Also to me it seems obvious the water cycle acts as an obvious negative feedback to heating in the Earth's atmosphere. Water vapour/clouds stop the highs getting as high and the lows as low. Just look at the sahara and Singapore.

Mar 23, 2014 at 10:06 AM | Unregistered CommenterRob Burton

And on the point of climate sensitivity I also believe it is a meaningless concept. You can linearise any problem to give a first order straight line solution. In fact it tends to be necessary to solve lots of problems analytically, but obviously not important if you want to iterate our problem using a computer model.

Mar 23, 2014 at 11:12 AM | Unregistered CommenterRob Burton

Rob, not sure whether you're taking issue with me or not, but there's little in what you've said I disagree with. There appear to be a dichotomy with "climate sensitivity". On the one hand it is reasonable to assume a warmer world will give rise to more water vapour, and from what I've read everyone agrees that it will cause more heat to be retained in the atmosphere - the so called positive feedback. They do take account of clouds but really have not the faintest idea what the effects of cloud albedo will be as there is no real understanding of the cloud "life cycle" for want of a better term.

In short there must be some response to increased water vapour but it seems to me that if the scientists wanted to forecast a cooler world they could point out that the water vapour would cause more clouds and the earth with cool dramatically with the same authority as they're forecasting an ECS of 3C. In other words I don't believe there is any foundation to the forecast ECS and it appears there's a growing number of scientists who, while not exactly agreeing with me, are having doubts about the numbers they've been using to make their "predictions".

Mar 24, 2014 at 9:02 AM | Unregistered Commentergeronimo

Rob Burton said:

> Heat rises - so a natural negative feedback

Is convection in and of itself a negative feedback as Rob suggests? To act as a feeback, increased convection would either have to reduce absorbed solar radiation or increase outgoing IR radiation. Clearly convection plays a part in cloud formation and aerosol distribution which affect radiation flows but these are side effects (and possibly +ve or -ve feedbacks) of convection; convection itself doesn't seem to be acting as a -ve feedback. So can convection lead directly to increased IR radiation to space?

Mar 24, 2014 at 1:19 PM | Unregistered CommenterChandra

My understanding is as follows.

Climate Sensitivity λ is the infinitesimal change in surface temperature DT to an (infinitesimal) increase in forcing DS.

So DT = λDS units of λ are deg.cm2/watts

The temperature response of the climate is non linear and λ does indeed also depend on T. We know this is the case because of Stefan Boltzman's law. If there are no climate feedbacks then

S = εσT^4

So DS/DT = 4εσT^3 and λ = 1/4εσT^3 = 3.5W/m2 for ε = 0.6

This tells us that if the earth's temperature rises by 1 degree C the surface will on average radiate an extra 3.5 W/m2.

The greenhouse gas effect from CO2 is due to the absorption of IR photons that fall into hundreds of very narrow wavelength lines centered mainly around 15 microns. Even at 280ppm many of these lines are already saturated and increasing concentration has no effect. It is only the lines at the side of the bands where absorption increases. This raises the altitude in the atmosphere where photons in these narrow frequencies can escape to space. These photons have less energy than before because they now originate from colder air due to the lapse rate. Therefore the final result of increasing slightly the CO2 concentration is that there is a tiny reduction in outgoing IR resulting in a tiny energy imbalance with the incoming solar radiation. As a result of this imbalance the surface warms up a little so that the same OLR as before is restored.

The calculation of this reduction in OLR (or energy imbalance) can be done - for example using MODTRAN under the assumption that nothing else changes and in particular that the lapse rate remains constant. The result of these calculations is a net "forcing" from increased CO2 of S = 5.3ln(C/C0), where C0 is the start concentration and C is the final concentration of CO2.

Now it is easy to calculate what the most people mean by "climate sensitivity", which is the temperature rise caused by a doubling of CO2. The reduction in OLR (called forcing by IPCC) is simply 5.3ln(2) = 3.7 W/m2. Therefore the change in temperature of the surface needed to restore energy balance is 3.7/3.5 or about 1.1C

This is Equilibrium Climate Sensitivity (ECS) because it is the temperature change that restores energy balance. Climate models predict higher values of ECS because they also have built in feedbacks for H2O, clouds and lapse rate changes. They also use aerosols to fine tune their outputs to match observed temperature data. The effects of aerosols is still rather uncertain.

Transient Climate Response (TCR) was introduced because the oceans take many tens of years to reach an energy balance. If you try to boil a pan of water with a candle then it takes a long time to warm up! The oceans have a huge heat capacity and it could even be that the oceans will never reach equilibrium but instead just change currents and heat flow rates to the poles. GCMs however do model the ocean in vertical layers and predict how energy eventually balances. These models give a "relaxation" time of about 60 years to reach an equilibrium.

The IPCC define TCR as “Transient climate response (TCR) is defined as the average temperature response over a twenty-year period centered at CO2 doubling in a transient simulation with CO2 increasing at 1% per year". TCR is a result of pure model simulations and different climate models predict different values of TCR. TCR is calculated in the models by holding all other external forcing constant while drip feeding in CO2. That is why IPCC give a spread of values for TCR from 1.5 to 4.5C. This spread is really due to the different handling of feedbacks in each model.

This definition of TCR cannot be measured without using a model. Nic Lewis tries to measure TCR directly but he still has to rely on CMIP5 forcing to let him untangle a "measurement" from the temperature data. This is because it is not just CO2 which is changing but also CH4, NO2 etc and the IPCC definition excludes them. The modelers are then able to criticise Tim because his result according to them, doesn't handle aerosols or the other GHGs properly. Only their models are a correct description of the complex phenomena involved. This is a catch-22 situation we are in. Climate science has become a closed shop where only an elite few with access to complex black-box models are qualified to tell the world how CO2 effects climate. Don't forget also that any large software code always has bugs.

The only way out of this situation is to redefine TCR so that it is independent of models. That is why I proposed a different definition as follows. “Transient climate response (TCR) is defined as the measured average temperature response over a twenty-year period centered on the observed CO2 doubling.” This then would simply be the observed temperature upon CO2 levels reaching 560ppm. It represents the full anthropogenic effect of human activity resulting in CO2 doubling including man's effect on aerosols, CH4, NO2 etc.

The remarkable fact is that there is a near perfect cancelation in AR5 between aerosol cooling and the "other" GHGs. CMIP5 net forcing is near as damn it a perfect fit to a pure CO2 forcing ! - see graph here

You don't need the models at all to measure TCR !

Mar 24, 2014 at 5:06 PM | Unregistered CommenterClive Best

Clive Best:

> TCR is a result of pure model simulations and different
> climate models predict different values of TCR. TCR is
> calculated in the models by holding all other external
> forcing constant while drip feeding in CO2. That is why
> IPCC give a spread of values for TCR from 1.5 to 4.5C.

> This spread is really due to the different handling of
> feedbacks in each model.

Is that so? The IPCC assesses many different estimates of sensitivity and GCMs are only one source of estimates. You ignore estimates resulting from paleo studies.

Mar 24, 2014 at 5:37 PM | Unregistered CommenterChandra

Clive Best
Many thanks for that.
I'm not sure whether I will end up any wiser but at least I am now much better informed.

Do pay attention. The IPCC's own definition is “... the average temperature response over a twenty-year period centered at CO2 doubling in a transient simulation with CO2 increasing at 1% per year". Where do paleo studies come into it?

Mar 24, 2014 at 6:15 PM | Registered CommenterMike Jackson


Paleo estimates are for ECS rather than TCR.
They too also heavily dependant on climate models.

Mar 24, 2014 at 6:40 PM | Unregistered CommenterClive Best

geronimo - "In the post nyb suggested putting heat under a kettle of water would cause the temperature of the water to increase and that the pre-boiling point of the kettle was TCR while ECS was the boiling point."

Nope - read my comment again! No boiling in my comment:

For the sake of simplicity, lets assume you are running a tight ship and only put it on a very low gas which is insufficient to induce boiling.
Mar 12, 2014 at 11:55 PM | Unregistered Commenter not banned yet
Thread "Myles out of line"

Mar 24, 2014 at 7:22 PM | Unregistered Commenternot banned yet

Clive Best - thanks for sharing the info. on TCR definition. I hadn't seen that so I searched for a reference which pulled up Nic Lewis's reply to you at Climate Audit:


I agree with your sentiments, but the problem is that in reality there are many drivers of surface temperature change – isolating just the effect of CO2 is not practical. However, if one converts the total effects of all greenhouse gases, aerosols, etc. into an equivalent increase in CO2 concentration (by reference to their effective radiative forcing RF, that from a doubling of CO2 being F2xCO2), then what you suggest would be pretty much in line with the generic definition of TCR in Section 10.8.1 of AR5 WGI:

“TCR was originally defined as the warming at the time of CO2 doubling (i.e., after 70 years) in a 1% yr–1 increasing CO2 experiment (see Hegerl et al., 2007b), but like ECS, it can also be thought of as a generic property of the climate system that determines the global temperature response ΔT to any gradual increase in RF, ΔF, taking place over an approximately 70-year time scale, normalized by the ratio of the forcing change to the forcing due to doubling CO2, F2×CO2: TCR = F2×CO2 ΔT/ΔF”
niclewis Posted Mar 11, 2014 at 6:34 AM
Thread "Does “Inhomogeneous forcing and transient climate sensitivity” by Drew Shindell make sense?"

Mar 24, 2014 at 7:35 PM | Unregistered Commenternot banned yet

Mar 24, 2014 at 1:19 PM | Unregistered CommenterChandra

All I as saying there is that if say the sun heats up the ground the hot air rises to be replaced by cooler air from elsewhere. A Sea Breeze is a very good example of cooler sea air coming in to cool the coast on an otherwise hot sunny day.

Mar 24, 2014 at 7:39 PM | Unregistered CommenterRob Burton

Mar 24, 2014 at 1:19 PM | Unregistered CommenterChandra

The whole convection cycle is the main driver of how energy moves around the atmosphere and eventually leaves it. It is involved in both moving energy from the Tropics to the Poles and also in moving energy from the ground to the Tropopause. It then involves the important role of gas/liquid H2O and CO2 in particular which emit in the IR and are effective in emitting this radiation high up in the atmosphere into space to remove the energy that was initially absorbed by the Earth primarily by the ground and sea.

Mar 24, 2014 at 7:47 PM | Unregistered CommenterRob Burton

@not banned yet,

and my reply to him was....:


I don’t think you quite understood what I was suggesting. You cannot “measure” effective forcing. You can measure CO2 levels in the atmosphere and you can measure the rise in surface temperatures. For everything else you need models. CO2 levels are also an indirect measure of methane emissions and aerosol emissions since they are all related to fossil fuel usage and to the human activity enabled by fossil fuels (such as intensive farming, chemical industry etc.) So why not aim for a simple separation between theoretical physics (models) and experimental physics (measurements) ? Otherwise climate science models are basically unfalsifiable.

The IPCC definition of TCR as it stands is an unmeasurable quantity. Natural variation acts over periods of decades whether it is due to PDO/AMO, due to solar variability or both. It is likely that natural variability is currently in a negative phase resulting in the hiatus in warming.

What we need to do is place experiment back in the prime role like any normal branch of physics and NOT the inverse. TCR should therefore be defined as the measured response to CO2 which includes both anthropogenic and natural effects. Models should then be developed whose goal is to explain the measurements and then make predictions which can then be tested by expriment.

I have been told by Ed Hawkins that climate science is an observational science like astronomy. This week we have experimental evidence for a theoretical prediction of gravity waves caused by inflation after the big bang.

That is the way science should work !

Mar 24, 2014 at 7:59 PM | Unregistered CommenterClive Best

Clive - thanks for the additional comment. Just to clarify - my interest was the specific, referenced IPCC definiton of TCR.

Mar 24, 2014 at 8:45 PM | Unregistered Commenternot banned yet

Lord Monckton has an article on WUWT "When will climate feedbacks fully function? Not for millennia".

In his article he says:

"...Experiment and line-by-line radiative transfer analysis have demonstrated that the CO2 radiative forcing ΔFt is reasonably approximated by the logarithmic relation (2),

delta Ft = k ln (Ct / C0) Watts (sic) per square meter, (2)

where (Ct/C0) is a proportionate change in CO2 concentration over t years, with C0 the unperturbed value. Myhre et al. (1998), followed by IPCC (2001), give the coefficient k as 5.35, so that, for example, the CO2 forcing that arises from doubled concentration is 5.35 ln 2, or 3.708 W m–2. ..."

(As an aside, (2) seems to be EM's favourite equation, which he seems to regard as being meaningful to at least three figures accuracy.)

I posted a couple of comments:

Martin A says:
March 26, 2014 at 1:35 am

That 'climate sensitivity' can be meaningfully calculated is something that even skeptics seem to accept as true.

Martin A says:
March 26, 2014 at 1:56 am

"Experiment and line-by-line radiative transfer analysis have demonstrated that the CO2 radiative forcing (delta)Ft is reasonably approximated by the logarithmic relation (2)…"

I am eager to learn about experiments that confirm calculations of 'radiative forcing' . Is there a reference to such experiments, please?

Lord Monckton kindly posted the following response:

Monckton of Brenchley says:
March 26, 2014 at 3:05 am

The uninterestingly hemionymous "Martin A" says that "even skeptics accept" that climate sensitivity can be meaningfully calculated. No: we accept that if we try to determine it we may perhaps be able to constrain it somewhat. My own attempts published in the literature suggest that 1.5 K per CO2 doubling may prove to be nearer an upper than a lower bound, but the uncertainties are so great that the best one can say is that the evidence for the proposition that on business as usual a warming substantial enough to do more harm than good will occur is, at present, wholly insufficient.

"Martin A" also asks about the experiments that have assisted the line-by-line modelers in determining the logarithmicity of the CO2 forcing equation. Well, of course, in order to determine the line-by-line spectral characteristics of an atmosphere influenced by changes in the concentration of CO2 and other greenhouse gases it is necessary to determine by laboratory measurement the peak wavelengths at which radiation will interact with molecules of each greenhouse gas, and it is also necessary to measure from space the changes in emission of radiation from the Earth's surface as greenhouse-gas concentrations change. The latter experiments are performed by the Earth Radiation Budget Experiment and by the Clouds and the Earth's Radiation Experiment Satellite.

Mar 26, 2014 at 1:37 PM | Registered CommenterMartin A

"The uninterestingly hemionymous "Martin A" " - anyone know what that means?

More to the point, why would you be interested in what Monckton says?

Mar 26, 2014 at 3:15 PM | Unregistered CommenterChandra

Well it's obviously a Moncktonism meaning 'half anonymous'. And presumably implying that no wit is apparent in my handle.

Why would I be interested in what he says? Well he's obviously not daft (he's pretty hot on maths) and a good number of people pay attention to what he says. What he says is at very least interesting - in the sense of being thought-provoking. We don't have to take what he says at face value but we can take a look and see for ourselves whether or not it makes sense.

The theme of Rhoda's original posting is that computing 'climate sensitivity' depends on numerous assumptions that have not, so far as anyone here can make out, been verified. I was interested that he does not seem to share that view.

I was also interested because the quaintly named Monckton of Brenchley's comment suggested that radiative forcing due to CO2 had been measured experimentally which, if true, is news to me.

Does my justification for being interested in what he says meet your approval?

Mar 26, 2014 at 4:29 PM | Registered CommenterMartin A

It's a very interesting post if you've been puzzled, as I have about some of then problems with TCR and ECS. Basically what I got from it is that in the equation for ECS:

TΔ = 5.35 x λ x ln (2)

The value of λ is time dependent, in other word it changes over time, which had puzzled because the TCR also has the term λ in it and was clearly a different number. The rest of it kinda solves other issues I have not come to grips with associated with the timescales. What I couldn't and still don't understand is the process of getting from 1.2C to 3C increase in global temperature, over what timescales this was supposed to happen, because it can't be instantaneous as it requires a build up of water vapour in physical terms. and none of the terms in the equation for ECS deal seemed to deal with this issue (if you assume, as I did, that λ is a constant). What he's saying, I think, is that λ changes over time and it will take 1000 -3000 years to reach the 3C ECS value of λ = 0..8, and the IPCC know it and have kept it buried in WG1.

If that's true then it's scandalous.

A propos of Monckton I believe he's no fool, but I don't like his abrasive faux intellectual abuse, like referring to someone who asks a polite question as "uninterestingly hemionymous" because he detects you might be challenging his sermon on the question of transubstantiation or some other secret of faith.

Mar 26, 2014 at 5:01 PM | Unregistered Commentergeronimo

notyet banned: "Nope - read my comment again! No boiling in my comment:"

I apologise unreservedly I did it from memory and believe I said so, but still I should have checked.

Mar 26, 2014 at 5:25 PM | Unregistered Commentergeronimo

Yes Martin, I stand corrected. But I'd be surprised if you learnt anything new from his response.

Mar 26, 2014 at 5:32 PM | Unregistered CommenterChandra

No, geronimo, it was inappropriate in a scientific discussion to mock someone's name whether or not it is their full name. To me it implied an Eschenbachish oversensitivity to any criticism, no matter how mild and he went down a few millimetres in my opinion of him. I did reply as follows:

Martin A says:
March 26, 2014 at 7:31 am

I'd like to thank the quaintly named Monckton of Brenchley for his kind reply (3:05 am) to my comment and my question.

Interesting to notice that another commenter followed up with:

Matthew R Marler says:
March 26, 2014 at 9:32 am

Martin A.That 'climate sensitivity' can be meaningfully calculated is something that even skeptics seem to accept as true.

Nobody doubts that a derivation from assumptions can be made. Lord Monckton takes a bunch of common alarmist assumptions and derives conclusions from them, which seems to show that if the alarmists are right, then there is nothing to be alarmed about.

However, neither the assumption that the climate sensitivity is constant, nor that the "equilibrium" approximation is accurate about something on Earth, has a strong justification or has been tested. So any conclusion following from those assumptions may not be "meaningful", and not all skeptics accept as true that they are meaningful.

That pretty well echoes my own view.

I do think that redoing climate sensitivity (CS) calculations to expose errors in previous CS estimations and to show that CS has routinely been overstated has some value

However, I also think that the results should still be treated as useless (ie they have to be regarded as bollocks) for any practical purpose, so long as they depend on unverified assumptions.

In publicising the results of CS recalculations, I think there is a risk of maintaining the credibility that CS have fraudulently (in my opinion) acquired.

Mar 26, 2014 at 8:02 PM | Registered CommenterMartin A