Global Warming and Chaos Theory

[Bored Wombat]

The whole discussion turns around *how much* an increase in CO2 affects the greenhouse effect, in other words, is CO2 a strong or a weak greenhouse gas. The bare bone optics tells us it is 0.77 K for a doubling from 280 ppm to 560 ppm, which means it is rather weak. In order to change that into 1.5 - 6 K, one needs to show me where the trick is. I don't say that this is impossible because there can be feedback effects. Those can be positive, negative, or irrelevant. But I would like to see - and with all the "scientific certainty claims" there MUST be - an almost airtight logical reasoning that proves us that a doubling of CO2 brings us within a 1.5 - 6 K range, *so that it could eventually be recognized as an important source of heating*, and I haven't seen that.

Well, I've pointed you to a few papers that estimate this very parameter.

You may or may nor recall the relevant part of the recent IPCC report (Section 6 of chapter 9). It cites about 20 papers, that address this same parameter. So if you are interested in understanding the ways in which extimates of climate sensitivity and what independent estimates that there are, then you're lucky because there's certainly not a paucity of studies on this aspect.

What I have seen are a lot of studies that *take it for granted that the observed heating is due to CO2* and then go on estimating what must be the feedback effect SO THAT the heating can be explained by the CO2 change.

I haven't seen any studies like that. Do you have citations for a handful or so of them? (Or links to the papers)?

In this respect, Mars is an interesting "reality check" on the bare bones effect of CO2: with 15 times more CO2 in the atmosphere, its "optical" pure-CO2 greenhouse effect is tiny. The main difference is that the "feedback effects" are switched off, or are different. Don't forget that this is nevertheless the "drive" of the current AGW theory. What we see on Mars is that the CO2 doesn't play a role of significance in any changes of temperature: albedo is then a strong forcing agent.

No, Mars has no combustion of fossil fuels, no oceans and no biosphere. So the CO₂ concentration will be dead constant, yielding no change in temperature.

And then the question is: imagine that those feedback effects really exist. That means, imagine that a tiny drive (such as the all-optical CO2 drive, but also a change in ocean currents, or a small change in albedo - think use of coal and black particles which is correlated of course with CO2 exhaust - or whatever) can and will indeed be amplified to large proportions (and hence, in the linear domain, that small temperature decreases can also be amplified into large coolings). That would then mean 2 things: 1) big instabilities and 2) *any* small initial temperature forcing would be amplified.

The main mechanisms are water vapour, ice albedo, and in the longer term oceanic outgassing. Yes, it affects any change in radiative forcing.

I may of course have it all backwards, and maybe you can reassure me with a logical proof that CO2 IS a strong greenhouse gas, and that I can hope that AGW is here to stay (indeed, my greatest fear is that it might not be there as strongly as one claims).

We need to get a couple of things straight. 150,000 people per year are dying of global warming, and a great many more are suffering serious illness. Ecologists are estimating from range changes that we are looking down the barrel of 35% of species committed to extinction by 2050. Conservative estimates of cost are an ongoing 5% of world GDP. Realistic ones are 20%. If you care about the lives of people or the world economy or the world biosphere, then your fear should be that it is as strong as median estimates.

But there's a significant possibility that it might be around 6K per doubling, and if that doesn't have you garbageting yourself you're not understanding the consequences.

So what's the logical argument that leads us to conclude, without hesitation, that CO2 cannot be anything else but a strong greenhouse gas ?

I'm not really familiar with these terms "strong" and "weak" greenhouse gas. I presume these are defined in terms of increase in radiative forcing per doubling of concentration?

CO₂ is about 3.7 W/m² for a doubling. If the line between "strong" and "weak" is more than that, then it is weak. If it's less, then it is strong. If you want to reduce it to a one bit step function.

Again, don't understand me wrong. I'm more kind of attacking the logic of the whole thing than discussing whether or not AGW is there. It's more like the mathematician who is verifying the proof of a theorem and finds logical leaps that don't seem to be sound, than discussing of whether the theorem is true. It's more about the scientific method than any actual climate science.

Proofs in Mathematics are very different from ones in any applied science. Different theories of gravitation, and different ways to unify it with other forces is under constant and prolific investigation. But if you drop a carton of eggs, it will fall to the ground.

Climate sensitivity is probably about 3.

 

[Andre]

Well, I've pointed you to a few papers that estimate this very parameter.

The netiquette would be that you quote or at least point to the specific sections where this occurs, so that the method of estimating is clear and can be evaluated. You can't expect from a busy scientist like Vanesch to wade through the deluge of papers. Just make your case in detail within the thread.

No, Mars has no combustion of fossil fuels, no oceans and no biosphere. So the CO₂ concentration will be dead constant, yielding no change in temperature.

I would expect from any climate alarmist to understand the basic idea / hypothesis of greenhouse effect being the difference in the theoretical black / gray body temperature and the actual atmospheric temperature at the surface. The data are ambiguous about any difference between the two, despite the much higher concentration of CO2 compared to Earth. No need for fossil fuels, oceans etc.

We need to get a couple of things straight. 150,000 people per year are dying of global warming, and a great many more are suffering serious illness.

Would you care to substantiate that and also demonstrate that this is caused by the increase of CO2 and also how this could be mitigated / prevented if we put our own economy under great strain to get a tiny reduction in CO2 production?

Ecologists are estimating from range changes that we are looking down the barrel of 35% of species committed to extinction by 2050.

That's a study of several years ago. Could you point to a list of the, no doubt hundreds of species that went extinct since then, to match that number and also demonstrate that those extinctions are caused by higher temperatures due to increased concentrations of CO2?

 

[vanesch]

Well, I've pointed you to a few papers that estimate this very parameter.
You may or may nor recall the relevant part of the recent IPCC report (Section 6 of chapter 9). It cites about 20 papers, that address this same parameter. So if you are interested in understanding the ways in which extimates of climate sensitivity and what independent estimates that there are, then you're lucky because there's certainly not a paucity of studies on this aspect.

I will have a look at those again, in as much as they are available. But to get this straight, what I'm looking for are papers that demonstrate logically by falsifying all other possibilities that the source of heating is due to CO2. As such, they must start with all reasonable hypotheses where CO2 is NOT a serious source of the warming, but that something unknown is, and arrive at an undeniable conclusion that this leads to a contradiction with known laws or observations. I must indeed have overlooked the 20 something papers that do this. I'll look again.

No, Mars has no combustion of fossil fuels, no oceans and no biosphere. So the CO₂ concentration will be dead constant, yielding no change in temperature.

I must have expressed myself badly. The *total* greenhouse effect on Mars is almost neglegible. But of course there is also no variation of it. The effect of albedo is of the same order of magnitude or bigger than the *total* greenhouse effect.

The main mechanisms are water vapour, ice albedo, and in the longer term oceanic outgassing. Yes, it affects any change in radiative forcing.

There are myriads of other possibilities that could change the surface temperature: changing ocean currents, ocean convection, changing wind patterns, changing cloud formation, changing precipitation, convection and I don't know what. All these mechanisms and probably many more can change the surface temperature by re-distributing the thermal energy, by changing the average temperature at equal radiant flux, and eventually by changing the outward radiant flux. The way thermal energy is re-distributed over the surface of the Earth can have an important effect. It is only after having demonstrated that all these effects are NOT the main source of the observed temperature variations that one could eventually, by elimination and falsification of all alternatives, arrive at pointing at CO2. I haven't seen that, but I take it that it has been conclusively done - that's what I'm in fact asking for. It is only from that point on that it starts to be reasonable to suspect a stronger heating by CO2 than given by its optical effect.

We need to get a couple of things straight. 150,000 people per year are dying of global warming, and a great many more are suffering serious illness. Ecologists are estimating from range changes that we are looking down the barrel of 35% of species committed to extinction by 2050. Conservative estimates of cost are an ongoing 5% of world GDP. Realistic ones are 20%. If you care about the lives of people or the world economy or the world biosphere, then your fear should be that it is as strong as median estimates.

Howdy ! Now we aren't even sure about the temperature predictions of global warming, we can even give estimations of the number of victims. Aren't we pushing things a bit here ?

But as I told you, my main fear about the possibility of global warming not being true (or at least, *dramatic* AGW not being true) is that it would backfire on nuclear power. (It would also seriously backfire on all of science. All of science would have lost entirely its credibility, due to the bet of the AGW people to take suggestive evidence as final proof, and to fight by all means any critical analysis - hence the denigration of the "AGW deniers" or the "AGW sceptics"... the "non-believers" in other words. But that's another point) After all - and that's why I like AGW so much - this is the biggest joke that environmentalists ever played on themselves: if it is true, their opposition to nuclear power for more than 40 years is then mainly responsible for the biggest ecological crisis in recent history, and that is a joke that I find terribly funny (I am entitled to my strange sense of humor, no ?). It would honestly annoy me if it turned out not to be true so that we can replace again nuclear power stations by coal-fired plants, and that I cannot have my good laugh anymore. So it is that fear that drives me to want to find out for real if the possibility exists that AGW might finally not be so dramatic, because I want to prepare myself for a big deception. It is also the exceptional quality of environmentalists as fearmongers (which they used against nuclear power) and their wholehearted embracing of AGW which makes me suspicious: if the quality of the inquiry for AGW is of the same nature as the quality of their inquiry into nuclear affairs, then that could tell a lot about the factual truth of AGW.

The estimates of 150 000 people dying because of global warming (I really really wonder where that comes from, but never mind - does this include all the people who also enjoy the better weather ?) is insignificant compared to what we accept in any case as a reasonable risk: car traffic kills 1.2 million people a year and we use it to go on a holiday. So this is a minor issue. I hope you can do better than that ! I want more drama !

However, you seem to forget that the atrocity of the crime is not an argument of guilt for the accused. It is not because the crime is terrible, that the accused is more likely to be guilty.

But there's a significant possibility that it might be around 6K per doubling, and if that doesn't have you garbageting yourself you're not understanding the consequences.

I will be dead by the time that it manifests itself, so for me this remains a purely academic discussion in any case.

I'm not really familiar with these terms "strong" and "weak" greenhouse gas. I presume these are defined in terms of increase in radiative forcing per doubling of concentration?

It's of course not a technical definition. Take it that a weak greenhouse gas is one that doesn't matter much (that isn't any source of *dramatic* heating). I consider 0.77 K per doubling not dramatic. It wouldn't, in that case, be the principal drive behind global temperature changes.

A strong greenhouse gas is one that matters. One that becomes by far the main source of temperature change. As I told you - but apparently I must have overlooked them - I've never seen a conclusive proof that CO2 is the principal drive in global temperature change. Nevertheless, that's claimed as being *scientifically certain*. It is strange to claim that it is scientifically certain, but not to provide an air-tight proof.

CO₂ is about 3.7 W/m² for a doubling. If the line between "strong" and "weak" is more than that, then it is weak. If it's less, then it is strong. If you want to reduce it to a one bit step function.

Yes, MODTRAN gives me about that. For a change from 280 ppm to 560 ppm, constant water vapor pressure and tropical atmosphere, I find a change from 289.2W to 286W per square meter. This corresponds to a ground temperature change of 0.9 K.
For the standard 1976 US atmosphere, the same change brings me from 260.0 to 257.2 W per square meter, so here it is 2.8 W. Which corresponds to a ground temperature change of 0.8 K.

Given that a rise of 0.8 or 0.9 K isn't dramatic and that we can expect about a doubling of the CO2 content in the atmosphere in the 21st century, I hence classify CO2 as a weak greenhouse gas.

This means, if I understand well, that a doubling of CO2 is equivalent to any other physical mechanism that would change the global temperature by a bit less than 1 K, like for instance a change in albedo of 1%. So to demonstrate that CO2 IS the main driving factor, we first must have conclusively shown that there cannot exist any other mechanism that could influence global temperature for about 1K. One has to be able to control ALL other thinkable mechanisms to much better than 1K and demonstrate that they don't play a role. Only in that case can one conclude that the initial drive was 1K, and that there are then feedback mechanisms which must turn this into 1.5K or even 6 K. Only then it is demonstrated.

If you want to include water vapor feedback for instance, which adds of course a greenhouse gas, and which is one of those potential positive feedback parameters (anything else equal, like cloud formation), then that feedback is just as well valid for, say, albedo change, or any other process that changes global temperature with 1K (at constant water vapor). In general, anything that amplifies the effect of the radiative forcing of CO2, will also amplify the effect of anything else that has the same effect as that radiative forcing.

Let us take albedo just again as a thought experiment. We get about a 1 K increase if the albedo changes by 1% (because solar flux is about 350 W/m2 and hence the radiative forcing is 3.5 W for an albedo change of 1%, which corresponds grossly to about 1 K change). If now, feedback included, the CO2 drive of 3.2 W (doubling) gives rise to a global temperature increase of 6 K (fear monger's preference), then so would then a 1% change in albedo, no ? A 1% change in albedo would then give rise to a 6K change in global temperature, wouldn't it ?

Proofs in Mathematics are very different from ones in any applied science. Different theories of gravitation, and different ways to unify it with other forces is under constant and prolific investigation. But if you drop a carton of eggs, it will fall to the ground.

The scientific method is the same everywhere. There are 2 kinds of proofs: by direct deduction from known facts, using strictly logical reasoning, and falsification of alternatives (reductio ad absurdum). Truth is not easier to establish when proof is more difficult to obtain. This is why one should remain modest with one's conclusions, and remain critical of one's theories. This is the cornerstone of scientific inquiry, it is what Feynman pointed out, and it is what I find lacking too much in the whole attitude towards AGW. Which is a pity, because it is an interesting scientific adventure and a challenging problem.

Again, all my attacks may make you think that I'm an AGW denier. I'm not - as should be clear. I'm just trying to find out how certain it has been established and up to now, I'm disappointed: my extremely elementary remarks should have been dealt with clearly because it should have been the first inquiries the AGW people should have done for themselves, if they had a truly scientific agenda. I suspect a big difference between the actual certainty and the displayed certainty. And that worries me.

 

[LURCH]

Just checking in with the OP;

There has been much controversy concerning Global Warming: sun spot activity, only heat radiation at 15 microns being trapped by CO2, vast amounts of greenhouse gases going intot he atmosphere from exploding volcanos and the general cyclical nature of climate change have attempted to explain the insignificance on human interaction on the environment as having a significant effect on GM.

I came across documentary which made reference to how 'Chaos Theory' explained the significant effect of human interaction on climate change. Can someone give some elaboration on this?

McHeathen (if you're still reading this), do you feel that this question has been answered?

 

[vanesch]

Just checking in with the OP;


McHeathen (if you're still reading this), do you feel that this question has been answered?

It may sound as if the OP's question hasn't been addressed and that we (I being part of that) are guilty of leading this thread off-topic. But in fact, that's not true. By questioning the standard AGW theory's basic hypothesis, namely, the principal drive behind climate change is human-produced CO2 (mainly fossil-fuel burning), and by looking at what the AGW proponents can (or cannot) give us as an answer, this opens (or closes) the case for climate being driven by a more complex dynamics (which might, or might not, have a chaotic behavior).

Indeed, if there is a clearly established causal and deterministic link between human-pumped CO2 and global warming, then the case for chaos is small. We have a simple, almost linear, cause-effect relationship: the more we pump CO2, the more it will warm. Period. That doesn't mean that there cannot be small oscillations and corrections and vibrations, driven by a simple, or a complex and even chaotic system, but the overall dynamics is relatively simple.
If, on the other hand, the case for this simple human-pumped CO2 --> delta-T is less secure than some AGW proponents would like to present it, then this allows us to look more into the detailled mechanics of the entire dynamics of global temperature and climate. We could then find *another* simple relationship, or we could find a more complex dynamics, in which human-induced CO2 might, or might not, play an important (but more complex) role. This dynamics might, or might not, display elements of chaotic behavior.

 

[vanesch]

Well, I've pointed you to a few papers that estimate this very parameter.

I re-read chapter 9, I admit I didn't bother looking up the papers. What's done in chapter 9 to estimate the climate sensitivity of human-caused greenhouse gasses ? Different models with free parameters are taken, and using, I suppose, Maximum Likelihood estimators, or even better, Bayesian estimators with a priori probabilities for the parameters, fitted to data that the models were supposed to predict as a function of input functions, probability densities were estimated for the model parameters in the sense, I suppose of the maximum likelihood method.

That's exactly what I said. You *assume* a certain behavior, and then you fit this to the data. For sure it is a good way of tuning your model to the data. However, one has to be extremely careful with this technique: it doesn't prove at all the physical hypotheses that are part of the model. Especially, this technique doesn't separate correlations from causal relationships.

Of course, in as much as this technique, when applied to a variety of data and a variety of settings (ocean temperatures, paleoclimate, recent climate, special events such as volcanic eruptions, etc...) all come up with a similar and narrow parameter probability distribution (say, with an error of 5% or so), and in as much that using the parameter estimations on one kind of data predict correctly the behavior of another set of data, one can start gaining confidence in at least the predictive value of the model - at least within the range of data for which it has been tested, and hence in its conceptual principles (in the theory that was behind the model).

It gives confidence, true. It is certainly suggestive. But there are dissonances. For instance, and the funny thing is that I didn't pick that out (honestly) because it is dissonant, but because it comes very close to the thought experiment I proposed earlier (with the albedo), a genuine test of sensitivity to radiative forcing is given by totally different events which introduce such a forcing, like volcanic eruptions. And there, we find "In contrast, an analysis by Douglass and Knox (2005) based on a box model suggests a very low climate sensitivity (under 1 degree) ..."
Of course, there has been an anti-publication to this, that suggests that this method underestimates sensitivity. Very well possible.

Nevertheless, I'm left with my original feeling. This is parameter fitting of models to data. That doesn't prove beyond reasonable doubt that the hypotheses of the models are correct, that the models don't miss anything or that the causal relationships in the models are correct. It could be. Or it could be that they contain enough flexibility to fit the data. That they contain enough liberty and nevertheless are constrained enough to distill correct correlations between input and output as in the data. BTW, that's exactly what you do with neural network models (which are not based upon any causal relationship or any physical property). Neural networks contain enough liberty to fit a rather large dynamical space, and "educating them" (fitting their parameters to time series) often distills a more or less correct input-output relationship, at least within the domain of the training set. But nobody considers the fitted parameters as representing anything physical.
Now, of course if the model is physics-inspired, and if all relevant relationships are implemented, then these fitted parameters DO have a physical meaning.

That's however not a *proof* of the correctness of the assumptions that lie at the basis of the model. It is a sensible thing to do. It's maybe the only sensible thing one can do. But my former professor of theoretical mechanics once told us: give me any 12-parameter model, and I'll fit an elephant. Add a 13th parameter, no matter which one, and I'll let his trump swing.

EDIT: to add to this: what is annoying in this way of doing things - although I can understand that one can't do anything else yet - is that the climate sensitivity is set as a free parameter, and is not calculated ab initio, by using all the physics we know. If it were calculated ab initio, had a fixed value, and it turned out that the data could be explained with this value, then this would have been a stronger case. But now that it is a free parameter, and hence can be "misused" by the model to try to accommodate for *other* effects which have not been implemented, it is harder to accept the value of this fit as both a proof of the causality on which the model is based, AND of the value of this parameter.

A simplistic example might illustrate this. Suppose that I have input data X, and output data Y. Now, suppose that in reality, there's a relationship between X and Y which goes as, say, Y = a X^2 + b X. But suppose now that we THINK that the relationship is Y = c X and that we ignore the a X^2 component (we didn't think of it, say). Now suppose that there is a physical reason, a causal relationship that has a meaning, and that is responsible for the b X term. If we are going to fit the real datacouples to this model, we will of course find a value of c. If the range of X is small enough, c will be close to (a X_av + b), and the performance of our model will not be too bad. We think it is an estimate of b. We will think that we have not only established the "law" Y = c X, but we think that we have also identified the value of whatever was this causal relationship. In other words, we think we've estimated a physical constant. We can apply, IMO, very similar comments to what is actually done in chapter 9, and we will find similar "confirmations". But we have it all backward, and although our model will do good work in the small range of X (around X_av) where it has been "trained", for, say, a doubling of X, it will be totally off. Our estimation of the "physical effect" given by b is also totally off.

This is of course a simplistic case, but it illustrates the kind of problem I see with the approach presented there when it is used to prove a causal relationship.

 

[Andre]

Talking about models and limitations

http://www.sciencedaily.com/releases/2007/12/071211101623.htm

A new study comparing the composite output of 22 leading global climate models with actual climate data finds that the models do an unsatisfactory job of mimicking climate change in key portions of the atmosphere...cont.

 

[vanesch]

Just for fun - I don't think it is really accurate. I used MODTRAN (using the web interface http://geosci.uchicago.edu/~archer/cgimodels/radiation.html ) to mimick the case of Mars. This was just a very quick thing, but if you put every greenhouse gas to 0 (including water vapor) and you put a ground offset of -60 degrees then you find (using tropical atmosphere - that's meaningless of course here) 173.14 W/m2. Now, if you add 7000 ppm (7 mbar) of CO2, then this falls to 163.7 W/m2, and you need to increase the ground temperature with 4.5K to compensate: at -55.5 you find again 173.3 W/m2.

Now, this is of course wrong for different reasons: first of all, the line widths will be different due to different Doppler broadening, and second, there's no oxygen and nitrogen. Also the assumptions of "tropical atmosphere" are wrong of course. But it gives a rough idea, and that rough idea is correct.

I had read somewhere that the greenhouse effect on Mars is around 6K (but with a large margin of error, something like 2 - 10 K), so that seems to be compatible with the purely optical effect given by modtran.

EDIT: ah, here it was: http://www.ess.uci.edu/~yu/class/ess200a/lecture.2.global.energy_cycle.pdf

EDIT 2: another fun experiment is the following. Take the current tropical atmosphere. We find, at 300 K an outgoing radiant flux of 287 W/m2. Next, put all greenhouse gasses to 0. For 300 K we now find an outward radiant flux of 398 W/m2. If you now decrease the temperature by 25 K, you find an outward flux again of 288 W/m2. So the purely optical effect, as calculated by MODTRAN, can explain a greenhouse effect of 25 K. That's pretty good, because the real greenhouse effect is 33 K. We only need "feedback amplification" of 33%, not the 200% to 800% as given by climate sensitivity.
I know, I know, this is waaaay too simple. I know. It is no proof that amplification factors cannot be stronger. But it indicates why one needs solid proof of it.

 

[Bored Wombat]

That's pretty good, because the real greenhouse effect is 33 K. We only need "feedback amplification" of 33%, not the 200% to 800% as given by climate sensitivity.

Water vapour is the lion's share of your 200% to 800% feedback, which is already included in MODTRAN.

 

[vanesch]

Water vapour is the lion's share of your 200% to 800% feedback, which is already included in MODTRAN.

You seem to forget that modtran gives only about 0.8 K for the radiative forcing of a doubling of CO2 then. Maybe you meant that the main greenhouse effect comes from water vapor. I agree with that. I wasn't implying that.

The 200% to 800% amplification is the parameter-fitted 1.5 to 6 K climate sensitivity versus the modtran-given 0.8K increase. What I meant was that modtran doesn't do so badly, for Mars, as well as for the earth. Of course I know this is too simplistic: modtran is simply calculating the optical effect in a 1-dim atmosphere in which the only transport mechanism is radiation and for which certain equilibrium conditions are taken. But given that it is ab initio (that means, it is all known physics, there are no "free fitting parameters"), and given the good performance (right ballpark for Earth greenhouse effect and for Mars greenhouse effect), anything that claims to *seriously* deviate from it needs substantial proof. And fitting a free-parameter model is less physical to me than having ab initio physics-based calculations, which, given the simplicity of the setup, give remarkably accurate results.

 

[vanesch]

Water vapour is the lion's share of your 200% to 800% feedback, which is already included in MODTRAN.

Mmm, maybe I misunderstood you. Maybe you are implying that the modest amount of CO2 is the reason why there is a certain amount of water vapor in the atmosphere. If we would remove all CO2, then all or a large fraction of the water vapor would also disappear (for an amplification factor close to 8). Well, the simple case is implemented in MODTRAN where relative humidity is used instead of constant water vapor.
For our tropical atmosphere, we go from 280 ppm and 289.2 W/m2 to 560 ppm and 286 W/m2 at 300 K (that's our famous 3.2 W/m2 forcing for a CO2 doubling), and we now have to increase the surface temperature not by 0.8K but by 1.5K (because heating up gives more water vapor in the atmosphere and hence more greenhouse effect).

So we get about a 87% extra feedback this way, true. That's not the 800%. It comes close to the 200%, if this relative humidity doesn't do anything else. But increasing the absolute water vapor pressure increases things like convection (it is the working principle of a cooling tower!). So it is not even said that this feedback is really there.

So using this feature, we get a cllimate sensitivity of not 0.8K but of 1.5K. And that's the highest modtran can do.

 

[Andre]

Mmm, maybe I misunderstood you. Maybe you are implying that the modest amount of CO2 is the reason why there is a certain amount of water vapor in the atmosphere. If we would remove all CO2, then all or a large fraction of the water vapor would also disappear (for an amplification factor close to 8). Well, the simple case is implemented in MODTRAN where relative humidity is used instead of constant water vapor.
For our tropical atmosphere, we go from 280 ppm and 289.2 W/m2 to 560 ppm and 286 W/m2 at 300 K (that's our famous 3.2 W/m2 forcing for a CO2 doubling), and we now have to increase the surface temperature not by 0.8K but by 1.5K (because heating up gives more water vapor in the atmosphere and hence more greenhouse effect).

So we get about a 87% extra feedback this way, true. That's not the 800%. It comes close to the 200%, if this relative humidity doesn't do anything else. But increasing the absolute water vapor pressure increases things like convection (it is the working principle of a cooling tower!). So it is not even said that this feedback is really there.

So using this feature, we get a cllimate sensitivity of not 0.8K but of 1.5K. And that's the highest modtran can do.

Would you please substract the energy, required for the enhanced evaporation rate to maintain the higher humidity, from the energy available to heat the surface?

 

[vanesch]

Would you please substract the energy, required for the enhanced evaporation rate to maintain the higher humidity, from the energy available to heat the surface?

Why would I ? Once the quantity of water is in the atmosphere, the latent heat of evaporation will equal the latent heat gained back during precipitation, no ? If a closed bottle containing water and air is brought to a higher temperature, then of course during the initial heating, you have to take into account the heat capacity and the latent heat of evaporation, but once equilibrium is reached, there's no need anymore for a power influx to maintain a higher partial pressure of water vapor.

 

[Andre]

Why would I ? Once the quantity of water is in the atmosphere, the latent heat of evaporation will equal the latent heat gained back during precipitation, no ?

No, you basically accellerate the water cycle. Precipitation is colder than the ambient temperature during falling into lower warmer levels (lapse rate). So it doesn't bring energy back to the surface. The increased latent heat released at altitude is emitted and partially escapes into space, not helping the required higher evaporation rate to sustain the quicker water cycle.

Also think in daily cycles. In many parts of the world, there is cumulus type clouds at daytime and no clouds at night time, During the day the decreasing albedo is restricting the temperature rise at the Earth surface. If the water cycle is acceralerated, the increase in cloudiness will limit the temperature rise even more as a substantial negative feedback, while at night time out radiation is changed only very little(assuming a small decrease in optical depth for IR radiation) It's just as if you turn the air conditioner to more cooling.

 

[vanesch]

No, you basically accellerate the water cycle. Precipitation is colder than the ambient temperature during falling into lower warmer levels (lapse rate). So it doesn't bring energy back to the surface. The increased latent heat released at altitude is emitted and partially escapes into space, not helping the required higher evaporation rate to sustain the quicker water cycle.

Also think in daily cycles. In many parts of the world, there is cumulus type clouds at daytime and no clouds at night time, During the day the decreasing albedo is restricting the temperature rise at the Earth surface. If the water cycle is acceralerated, the increase in cloudiness will limit the temperature rise even more as a substantial negative feedback, while at night time out radiation is changed only very little(assuming a small decrease in optical depth for IR radiation) It's just as if you turn the air conditioner to more cooling.

Yes, but these are more complicated mechanisms, which are part of more involved models taking into account cloud formation, convection, albedo changes etc...

I was talking about the simple model of a static (no air movement, no convection) monophase (no clouds) optical model as done in modtran. In such a model, one doesn't need to subtract the evaporation latent heat once equilibrium with the new radiative forcing is established.

 

[Robert Elliso]

‘The climate system is particularly challenging since it is known that components in the system are inherently chaotic; there are feedbacks that could potentially switch sign, and there are central processes that affect the system in a complicated, non-linear manner. These complex, chaotic, non-linear dynamics are an inherent aspect of the climate system.’ (IPCC TAR s14.2.2.1 - http://www.ipcc.ch/ipccreports/tar/wg1/504.htm)

‘Modern climate records include abrupt changes that are smaller and briefer than in paleoclimate records but show that abrupt climate change is not restricted to the distant past.’ (Abrupt Climate Change: Inevitable Surprises, 2002, NAP, p19 - http://books.nap.edu/openbook.php?record_id=10136&page=19).

A dynamic theory of ocean/climate states (http://www.nosams.whoi.edu/PDFs/papers/tsonis-grl_newtheoryforclimateshifts.pdf ) confirms that climate on decadal timescales is an emergent property of complex and dynamic Earth systems. "You go from a cooling regime to a warming regime or a warming regime to a cooling regime. This way we were able to explain all the fluctuations in the global temperature trend in the past century," Anastasios Tsonis said.