This forum is about wrong numbers in science, politics and the media. It respects good science and good English.
Oops, meant to reply to Brad below.
I wonder if this is the explication that Gavin Schmidt would have given if he had cleared this matter up properly?
Link-JEB's latest post,18.3.07 "what happens when they do"
Schmidt was awarding points for errors,cherry-picking,
"every time you hear there's a lag between CO2 and temperature in the ice cores, give yourself 2 points because that's a real doosy"
p12 of 33.
The debate was interesting to read but not much meat in it.
The logic is impeccable but it applies to a simple cause and effect relationship whilst the atmospheric CO2 concentration is a temperature controlled equilibrium between what are effectively infinite sinks.
The easiest way to illustrate what's going on is to treat the biological effects as being net fixed (i.e release = absorption) restricting the dynamic changes in CO2 to that absorbed in the hydrosphere and that in the atmosphere. As both atmosphere and hydrosphere are far from saturation they can be treated as infinite sinks so the relative concentrations at the interface are set purely by the energy levels involved.
Stable situations tend towards the lowest energy level so obviously the relative concentrations of CO2 on each side of the water-air interface must be such that there is no net energy reduction possible by moving CO2 from one reservoir to the other. Increasing temperatures alters the balance point towards a higher level of atmospheric CO2 whilst reducing temperatures lowers the atmospheric CO2. Under this simple model human produced CO2 which, despite the impressive numbers in absolute terms is small compared to the overall volume of CO2 available for dynamic interchange, is simply absorbed in the hydrosphere. Theoretically this increases the ocean acidity but there is a similar concentration gradient involved causing CO2 levels to increase as you get closer to the surface so the main effect is to increase the depth at which dissolved CO2 peters out.
In practice the biological effect cannot be treated as fixed because a small rise in temperature coupled with a rise in available CO2 increases biological activity fixing more carbon into plant tissue. It all gets very complicated very quickly.
Its important to remember that CO2 in the atmospheric concentration range achievable on planet earth, say 100 to 1,000 parts per million, is not a very strong greenhouse gas. The CO2 contribution to greenhouse temperature rise being about 1°C per 100 parts per million atmospheric concentration which is insufficient temperature gain for positive feedback. The extra CO2 released into the atmosphere by a solar radiation generated temperature increase does generate a small greenhouse temperature rise which does in turn release a smaller extra amount of CO2 generating a smaller extra temperature increase and so on with each mutual increase getting progressively smaller until the whole thing peters out. Xenos had the right of it with his paradox of Achilles never overtaking the tortoise.
The thing that the Ant-CO2 movement ignore with depressing regularity is that complex equilibia have a whole range of stable states and require serious positive feedback to drive them out of the stable range. Strange really as most Greens profess to support the Gia hypothesis which effectively states exactly that and try to twist it into supporting convergence to a single ideal state, presumably defined as AD 1970-5 when most of them appear to have been born.
Hope that helped.
The problem with water/air CO2 equilibria is not just simple Henry's Law. CO2 forms compounds in solution, carbonic acid, bicarbonates and carbonates for which the Law of Mass Action muct be applied. Each and every compound has its own equilibrium constant which of course can be combined, but what happens in most situations where equilibrium conditions are not appropriate? We're now into rate constants - another can of worms. I think the notion that increased CO2 atmospheric levels give rise to malign effects underestimates the oceans' ability to buffer and the oceans' reservoir of dissolved materials.
"is not just simple Henry's Law"
Aaargh. I was hoping to sort of slide round such extra levels of complexity by sticking to an energy level description.
As you so rightly say there are all sorts of equilibria involved in the take up of CO2 by water, especially sea water, besides the simple gas in solution one. The chemical ones being particularly complicated due to the hysteresis effects involved whereby the temperature has to be raised significantly before compounds start breaking down to return the concentrations to the nominal equilibrium point.
The way I see it the hysteresis effects mean that the chemical side can be pretty much ignored for small temperature changes so simple CO2 moving into and out of solution is the only dynamic capable of tracking temperature change. Hence, in principle, the "true" equilibrium points for all the reactions involved can be calculated for any chosen sets of baseline temperatures and the point to point changes in CO2 considered on straightforward relative solubility grounds. Obviously its not ever going to be truly accurate but the hysteresis involved in chemical absorption and release will offset things from the true, lowest overall energy, equilibrium state anyway. If I had to deal with the problem for real I'd be inclined to take a set of isothermals, calculate the mean concentrations of all the CO2 involved chemicals for each interval and track the relative shifts of the isothermals using the mean concentrations to get an estimate of the total CO2 take up. At least that way rate constants can be avoided. I'm a physicist not a chemist so thats one area I really don't want to get involved with.
However I do wonder how much change in chemical take up will actually occur given the small temperature shifts involved.
Many thanks Clive.
Could you kindly clear up one thong you wrote ?
"The CO2 contribution to greenhouse temperature rise being about 1°C per 100 parts per million atmospheric concentration which is insufficient temperature gain for positive feedback. The extra CO2 released into the atmosphere by a solar radiation generated temperature increase does generate a small greenhouse temperature rise which does in turn release a smaller extra amount of CO2 generating a smaller extra temperature increase and so on with each mutual increase getting progressively smaller until the whole thing peters out. Xenos had the right of it with his paradox of Achilles never overtaking the tortoise."
One thing that I am trying to come to terms with is the difference between the pro and anti-Co2 (as we call them for simplicity) sides on this very thing.
The "pro" side say that the effect of extra CO2 on temperature is logarithmic, and hence gets us into the Xenos situation. Surely, however, as soon as there is enough CO2 to completely absorb all incoming IR radiation at those wavelengths that Co2 absorbs then Co2 can do no more?
The climate models all appear to have some interaction between Co2 and water vapour, so that this diminution does not occur. Is there any scientific evidence for this effect and if so what is it ?
The pro CO2 folks assume that the CO2 greenhouse effect continues to increase in some manner proportionate to the CO2 concentration however large. Seem to be various versions of proportionate ranging from liner to logarithmic to something approaching hyperbolic, usually with an added multiplication feature to cope with rising baseline temperature.
We anti CO2 folks say such things are physically impossible because once the CO2 concentration rises sufficiently the absorbing band transmission drops near enough to zero and no further greenhouse temperature rise can be generated. We also say CO2 concentration increases being associated with temperature rise does not mean a direct causative link. I think a lot of the problem arises from people unfamiliar with the optics and thermodynamics involved trying to force a causative link on, which the physics really won't support. Some of the stuff looks to be lifted from CO2 laser physics!
I spent over a decade working on long range thermal imaging systems with considerable involvement in trying to reconcile predicted from lab work performance with actual field results and atmospheric transmission models. If I remember correctly the CO2 absorption band is over 95% opaque over a 10 km path at 300 parts per million concentration. So, to all intents and purposes, direct CO2 greenhouse effects cease to increase at greater concentrations. In practice higher levels of CO2 will concentrate the greenhouse rise lower in the atmosphere which is warmer anyway so there will still be a small rise. There are also other, lesser absorption bands which come into play as the concentration rises so there will always be some greenhouse contribution but never enough for positive feedback to start.
There are all sorts of potential interaction with water vapour and other things but nothing hugely significant in absolute terms. In any case most such effects are already factored in by the empirical temperature measurements. I guess the biggest effect is direct cooling of the CO2 by water vapour when it mixes but even that's smaller than you'd expect because radiation absorption doesn't warm up CO2 to any great extent. Pouring electromagnetic radiation into a gas has to be about the most inefficient method known for raising its temperature.
When I was professionally involved we used the LoTran atmospheric transmission model which was being actively developed to include a lot of the clever stuff. Frankly I never found the improvements worth a row of beans so, as soon as I got my hands on an Apple II with Visicalc, I wrote a simpler spreadsheet model which gave me more accurate results with faster turnround 'cos I didn't have to wait for the mainframe boys to run LoTran.
Like JB my professional career has given me plenty of reason to be hugely suspicious of big, all the effects, global models. They invariably evolve to the point where no one understands them and usually vastly overstate the import of second, third and higher order effects. Common error is to assume that all effects go to completion when, in practice, they get tangled up and the whole is much lower than the potential sum of the parts.
"The pro CO2 folks assume that the CO2 greenhouse effect continues to increase in some manner proportionate to the CO2 concentration however large." If only I could be sure where this idea comes from and what evidence they have to back it up!
If only I could be sure where this idea comes from and what evidence they have to back it up!
Evidence to back it up? Simple. None, nada, zilch.
It arises because the people writing the models and deciding what goes into them don't understand the physics involved and ignore the fact that every process has limits. Most models start out as small change approximations usually, whether implicitly or explicitly, including an assumption of the form "for a small change the finish conditions are sufficiently similar to the starting conditions that the rate or magnitude of the effect doesn't vary during the course of the change or reaction". Perfectly safe assumption so long as you know that you are doing it and so long as all the parameters are well off the maximum physical limit values.
However if you have a positive feedback situation like the pro CO2 types postulate the model obviously goes straight into runaway increase unless the loop gain is too small for positive feedback whereupon Xenos paradox applies. For CO2 the assumption is that the temperature increase is proportionate to the starting temperature and to the overall mass of CO2 so any increase in CO2 gives a temperature rise which in turn gives a second order rise due to the starting temperature now being higher.
Any careful person writing a model keeps careful track of such things ensuring that there is proper treatment of limits and and rate changes as things approach these limits. The chemical analogy is simplest. Consider two liquid substances A and B which react in equal proportions to form a solid C generating heat in the process. Assume that the whole thing goes on with everything in solution in a lot of water. Obviously if we start with lots of A and add a little B the reaction goes quickly and easily. If we continue to add B the reaction carries on but eventually slows as the amount of B approaches the amount of A and will eventually stop once there are equal amounts of A and B (tho' in practice it usually takes a slight excess of B before things grind to a halt). The progress can be tracked by monitoring temperature, turbidity or viscosity of the mixed liquids. Its obvious that any attempt to link temperature simply to the amount of B will be acceptable for small concentrations when there is plenty of A to react but will be grievously in error when things come towards completion with most of the A used up. If the model carries on allowing the amount of B to become greater than that of A using the same constant of proportionality linking temperature rise to quantity of B that was appropriate in the beginning the model output is clearly nonsense.
From bitter experience I can assure you that its quite a job to ensure that you've not left any such sillies behind even in simple models when you know exactly what's going on and have the benefit of everything being computable cause and effect. Sorting out a perturbed equilibrium situation is a heck of a lot more difficult. For starters you begin in the middle not at the beginning, in many cases such as climate the system probably doesn't even have a proper beginning having grown up like Topsy from something rather different. Being an equilibrium all sorts of rate constants and magnitude constants are already balanced so simple experiments aren't of great help sorting WTHIGO. Climate change is clearly a perturbed equilibrium with multiple drivers, multiple time constants and variable rate constants as well.
In general any model of changes in such equilibria postulating continuous change, especially increasing rate change, is by definition nonsense unless the driver is overwhelmingly large. Temperature changes associated with CO2 concentration changes are considerably over an order of magnitude too small to be overwhelming.
This is probably somewhat out of context, but can anyone answer a simple question?
Namely, we're told that some 50million years ago CO2 concentratopns were of the order of 1400ppm compared to today's 380ish... So why, assuming the "CO2 = heat death" theories are correct, didn't the climate enter a phase of runaway heating? Assuming also the multiplier effects of positive feedback from water vapour and other greenhouse gases why don't we have a climtae more akin to Venus?
Basically the CO2 thermal runaway heat death theories are incorrect.
The phenomena postulated have no possible physical mechanism.
Basis of the theory is that if a naked black body object in free space is illuminated with black body radiation its temperature will rise until the amount of energy being emitted as black body radiation equals the incoming radiation where upon it stabilises. For various reasons in our universe the temperature of the object will be lower than that of the original radiation source so the spectrum of its emitted radiation will be different to that of the incoming radiation being biased to the infra red.
If the object is now wrapped in an atmosphere of CO2 the amount of incoming radiation reaching its surface will be reduced due to the CO2 absorption bands so the stabilisation temperature will fall. However the CO2 absorption bands are in the infra red part of the electromagnetic spectrum and the object emission is proportionally greater than the incoming radiation in the infra red region due to its lower temperature. Therefore more of the outgoing radiation is "stopped" so the object temperature has to rise in order to bring things into energy balance. The difference between the balance temperatures for a naked object and one with an atmosphere is loosely called the green house effect. In reality its lots more complex because the "stopped" radiation warms up the atmosphere. Also the gas blanket takes heat from the object by convection and conduction which makes a huge difference to things seriously moderating the effects.
Returning to our black body with its CO2 atmosphere its obvious that as the amount of CO2 increases from the zero, no atmosphere case, the amount of "stopped" radiation increases so the temperature of the object must likewise increase to remain in radiation balance. Feedback exists whereby increasing CO2 raises the temperature. However at some point the amount of CO2 becomes so great that all the absorption bands become opaque. Further increases in CO2 will not increase the absorbed radiation so the feedback between increasing CO2 and temperature is broken. Conventional "heat death" theories ignore this limit. In practice the absorption bands become 95% or thereabouts opaque at around 300 parts per million of CO2 so the, by my maths, the feedback has pretty much stopped.
The Venus case is rather different. Firstly its lots closer to the sun so its hotter anyway and secondly it has lots more atmosphere. At this level of argument we can treat the atmosphere as an insulator reducing the rate at which heat is lost. So the much greater thickness of the Venusian atmosphere allows the surface to be much hotter before the rate of total heat loss from surface and atmosphere matches the incoming solar radiation.
As you know the effect of unit thicknesses of extra insulation falls off as the insulation gets thicker. Once the CO2 absorption bands have become opaque the only effect of extra CO2 is to add a bit more to the atmosphere "insulation blanket". But its only a little bit more, an extra 1,000 parts per million is still only 0.1% of the atmosphere, so the increase in surface temperatures its pretty much negligible.
Bottom line is that the absorption band effects are sexy science but its the insulating blanket atmosphere and temperature moderating hydrosphere that do the heavy lifting.
Serious food for thought there...