Comments 1 to 26:

1. MattJ at 10:05 AM on 3 September, 2010
   40%! Perhaps I heard that figure before, but to see it stated so precisely here, and with mounds of data to back it up, that certainly makes the situation sound dire. 40% is a lot, especially when the increase takes place so fast. 150 years is a blink of the eye on a geological time scale -- which is precisely the time scale that is most relevant.
   0 0
2. CBW at 11:05 AM on 3 September, 2010
   A very, very good article that lays out the facts in clear, logical terms. A suggestion and a nitpick: I would drop the final paragraph regarding the conservative estimate of the sensitivity -- it doesn't really fit the article and, as you correctly point out, the low sensitivities have been pretty much discredited. If you are going to keep the paragraph, however (and this is the nitpick), one would expect that the warming delay caused by thermal inertia would apply to this estimate as well, so CO2 under the low sensitivity model would account for less than 0.5 degree C of the total increase over the past 150 years. Again, though, really good article.
   0 0
3. dana1981 at 12:02 PM on 3 September, 2010
   Thanks CBW. I had thought of that, but then again the smaller the forcing, the less of an energy imbalance, and the less of a role the ocean thermal inertia plays. I felt it was worthwhile to point out that even with unrealistically conservative assumptions, the majority of the warming is still anthropogenic.
   0 0
4. HumanityRules at 13:26 PM on 3 September, 2010
   I think the big one here is climate sensitivity, it's seems this is the most controversial (if you're willing to accept anything is controversial in climate science). As usual I've got heaps of questions You state "the temperature change is proportional to the change in the amount of energy reaching the Earth's surface". Is that linear/exponential? As we progress further and further away from a 'normal' level of CO2 do the feedbacks change so the relation becomes more complex/non linear? With regard to the expected surface temperature change by 2010 it strikes me that the sums demand not only do we have to have "heating in the pipeline" but we also need nearly all the warming we've measured so far to be from AGW, there is little room for natural variation to have contributed much. I have read several papers recently (and in fact hassled a couple of the authors with emails) that assign, at least on a regional basis, large proportions of the measured warming to natural variation. Particularly these were the northern polar region, which are contributing disproportionally to the global average. In this region these authors were assigning 50-60% of recent warming to natural variability. I wonder if we start assigning part of the 20th century warming to nature how much we have to scale back the future catastrophe? From memory the IPCC only consider the impact of solar intensity and volcanos when assigning most of the late 20C warming to AGW. Do they consider that multi-decade climate variation (oscilations/seesaw and the like) to contribute anything to the trend we have measured in global mean temperature?
   0 0
5. dana1981 at 13:47 PM on 3 September, 2010
   Yes, climate sensitivity is the key factor here, which is why I gave a range of possible values. I'll be working on the advanced climate sensitivity rebuttal in the near future. The temperature change is directly proportional to the radiative forcing (energy change), as you can see in the formulas above. But keep in mind, we're only looking at the temperature change in response to increasing CO2. If various feedbacks kick in, like a large methane release, that's a completely different forcing which this calculation doesn't account for. It's just that conveniently right now, the non-CO2 forcings happen to add up to roughly zero, so we can get away with focusing on CO2 for the time being. I would have a hard time believing that 50+% of the Arctic warming is natural, but I haven't seen the particular papers you're referring to. Rapid Arctic warming is a projected consequence of AGW, because the melting ice creates a positive feedback by decreasing local reflectivity. Oscillations like PDO don't contribute to long-term warming. They're just that - oscillations which alternate between positive and negative states, moving heat around from oceans to air, and vice-versa. They neither create nor retain heat, so they don't have the capacity to warm the planet as a whole. In the short term they can warm or cool the surface, but they also have the opposite effect on ocean temperatures. Whereas right now we're observing both surface and oceans warming.
   0 0
6. Jim Meador at 14:08 PM on 3 September, 2010
   I found this posting to be very informative. I would be very curious to see how the feedback from water vapor affects this calculation. Is that something that can be estimated with any precision at this point?
   0 0
7. dana1981 at 14:32 PM on 3 September, 2010
   Well, the water vapor feedback in response to increasing CO2 should be contained within the climate sensitivity parameter.
   0 0
8. CBW at 14:55 PM on 3 September, 2010
   #4, HR, keep in mind that on a fine enough granularity, everything is linear. The difference between a climate system at 300 degrees K, and 300.8 degrees K can be pretty well approximated by linear functions. In some places this isn't true, for instance at the freezing point of water -- there is a big discontinuity between a system that is at 0 degrees C and frozen, and one at 0 degrees C and thawed. That is why the arctic is such a big concern for climate scientists. Other aspects of the system may need to include non-linear effects, like the huge jump in CO2 in the past century or so. There are a large number of interacting factors that we are only beginning to understand. One can speculate, and one can extrapolate, and one can choose to believe whomever one wants to believe, but the reality is that we're running a gigantic uncontrolled experiment on the environment upon which we depend for our survival, and everything we understand about the science tells us that the situation isn't going to slow, stop, or reverse unless we change what we're doing. #6, Jim Meador, the water vapor feedback is included in the total climate sensitivity. It is a factor in the uncertainty in sensitivity that dana1981 cites.
   0 0
9. RSVP at 17:56 PM on 3 September, 2010
   dana1981 I like the article as it explains things very clearly, yet I have a comment similar to what HR brought up. Putting on my skeptic hat now, the precision of scalar 5.35 gets my attention when dF does not appear to depend on temperature itself. CO2 back radiation (if it is even real) is supposedly a function of temperature (in this case that of CO2) and that of changes incident solar radiation. Basing the change of radiative forcing on a ratio such as C over Co implies "all things being equal" (including the amount of incident solar radiation). However while the overall model assumes natural changes in solar irradiance, the formula assumes this to be constant, which is fine for a first order approximation, however, what then is the basis of the (+- .005 W/m2) precision based on? Put another way, while dF is theoretically not independent from changes in average global temperature nor incident solar irradiance, the formula reflects this assumption in its simplicity, while at the same time includes a factor with a precision out to two decimals places. How real is all this for the minor changes in temperature we are talking about?
   0 0
10. Arkadiusz Semczyszak at 18:29 PM on 3 September, 2010
    In that case, I also make a very short balance ... Of course, this has already been spoken, not once but worth it (to organize discussion) in a condensed recalled again (absolutely essential) and most real skeptical arguments: - Assessment - assessing the feedback resulting from the increase of CO2 made by different researchers are up to several hundred percent different from the above presented - They often get even a negative balance of feedback. Moreover, also the IPCC in its report (full version) draws attention to the considerable range for uncertainty in their estimation. - changes in the value of some RF far exceeds the size of RF CO2 during the analyzed period. For example, I recall: here: “... the total global cloud cover reached a maximum of about 69 percent in 1987 and a minimum of about 64 percent in 2000s ... ... a decrease of about 5 percent. This decrease roughly corresponds to a radiative net change of about 0.9 W/m2 ...” and albedo change - by (generally recognized in the scientific world), more recent reconstructions from ice cores with higher resolution (not to mention other than the "core" reconstructions) during the last millennium, the level of CO2 in the atmosphere has changed much more than that in Figure 1 (Max. 60 ppmv between circa 1150 to 1300 yr A.D.), often reaching 320 ppmv (remember this is based on ice cores), ... and as says F. Engelbeen: “A causes B and shows a good correlation. A causes C and shows a good corelation. Thus B causes C, because there is a good correlation between the two. But that correlation is completely spurious, as there is not the slightest physical connection between B and C.” (...) - The effect of the sun at the Earth's climate is not only the TSI, and the Sun is out in the absolute most of their superposition cycles and cycles associated with Sun.
    0 0
11. HumanityRules at 18:49 PM on 3 September, 2010
    Dana, If by the range you mean 2-4.5°C then my understanding is that this range comes from the water vapour feedback. There seems good concensus that there would be around 1.5oC climate sensitivity directly from the properties of the CO2 molecule. The range comes from the extent to which water vapour plays a part. My question was really around whether water vapour has a linear response to the increase in CO2. I can accept that oscillations don't create heat but in terms of how we measure warming then I think an oscillation can 'retain heat' at least in terms of the measurements we make. If you catch an oscillation on the rise, and fail to measure the whole system (primarily the ocean) then it can lead to a trend in the measurements. Give it another 100 years (or a few decades of good ARGO data) and we'll probably have enough data to draw a line under this but neither you or me are happy with that answer. So I'm just looking for direction that the oscillations are recognised, measured and put to bed. One of the papers is this. You can infer from the results that ~2/3 of the recent warming is from the upstroke in a natural oscillation. The 50% estimate was in a private email from a climate scientist who says the work is due for publication. I'd be happy to send that mail to John Cook for him to confirm it, without naming the scientist, but I think that's probably pointless given we wouldn't be able to discuss the mechanisms involved. My point would be there are climate scientist out there who are looking at oscillations in the climate and think it is contributing to recent trends at least on the regional level. When the Basic version of polar amplification appears we can probably discuss this further.
    0 0
12. Arkadiusz Semczyszak at 18:49 PM on 3 September, 2010
    ... from the point 3 of my comment, follows that the natural balance of flux - sink of CO2 - in the preindustrial era - natural balance can not be - not exist - unless the scale of n x thousand years.
    0 0
13. TOP at 23:57 PM on 3 September, 2010
    Looks nice until you put it in context including H2O.
    
    CO2 closes off the right hand side of the H2O absorption band and CH4 contributes to the left hand side. One thing that is clear from this graph is that climate sensitivity changes with absolute temperature.
    The total spectrum of all atmospheric gases is given in the bottom plot. This shows a "window" between 0.3 and 0.8 microns (the visible window), which allows solar radiation (without the lethal UV component) to reach the earth's surface. "Earth radiation", the upwelling infrared radiation emitted by the earth's surface, has a maximum near 10 microns. The total atmosphere plot shows that a narrow window (except for an oxygen spike) exists in the range of wavelengths near 10 microns.Iowa State
    The effect of CO2 on absorption of IR and re-radiation is not linear as the equations above suggest. Once a certain level is reached adding another molecule has much less effect. Think of window shades. Once they are shut, closing them more has little or no effect. Nor does a decrease in CO2 below 280ppm cause a negative forcing which the equation for dF suggests. The absorption bands (wavelength regions) for carbon dioxide are nearly saturated, but those for other gases are not, so one additional molecule makes a larger impact.
    Iowa State
    
    CO2 is about saturated so adding more has little effect compared to other GHG which are not saturated.
    
    Iowa State
    0 0
14. Ned at 00:21 AM on 4 September, 2010
    TOP writes: CO2 is about saturated so adding more has little effect compared to other GHG which are not saturated. Sorry, nope. This is a common misconception, however. See Is the CO2 effect saturated? on this site, or (for more detail) A saturated gassy argument (by Spencer Weart) ... or see the really excellent explanation of this issue over at Science of Doom: CO2 – An Insignificant Trace Gas? – Part Eight – Saturation
    0 0
15. dana1981 at 01:34 AM on 4 September, 2010
    Lots of comments to respond to overnight. RSVP - I'm not sure why you think the CO2 radiative forcing is dependent upon the atmospheric temperature or solar irradiance. The solar forcing is a different calculation entirely. The CO2 forcing should just depend on the increase in atmospheric CO2 concentration, as the formula shows. HumanityRules - yes, the warming from a doubling of CO2 alone should be about 1.2°C. However, climate sensitivity takes into account all variables which change in response to increasing CO2, not just water vapor. I'll be working on a rebuttal on "climate sensitivity is weak" this weekend which will discuss this to some degree, but estimates of the value are based on empirical observations, among other things. It's not just the water vapor feedback which determines the sensitivity value. As for the Arctic study you reference, unfortunately I don't have access to the full paper. However, they suggest that when the Arctic warms, the Antarctic should cool. Currently both are warming, which would seem to throw a wrench in the spokes of this natural warming theory. TOP - as Ned shows, CO2 is not saturated. Additionally, if you check the formulas in this post again, you'll see that it's not a linear relationship between the radiative forcing and CO2 concentration, but a logarithmic one.
    0 0
16. RSVP at 07:52 AM on 4 September, 2010
    dana1981 "I'm not sure why you think the CO2 radiative forcing is dependent upon the atmospheric temperature or solar irradiance." CO2 does not generate energy on its own. Even according to AGW it is only a reflector of IR. The amount of energy assumed to come from CO2 according to your ln function only holds IF solar irradiance patterns do not change. But if it does what? CO2 would contributes nothing in a case where the sun shuts down for instance even if concentrations triples or quadrouples.
    0 0
17. dana1981 at 08:32 AM on 4 September, 2010
    Okay I see, your point is that the greenhouse effect is dependent upon the thermal radiation from the Earth's surface, which in turn is dependent upon the incoming solar radiation. My initial reaction to that is that total solar irradiance generally only varies by a fraction of a percent, so treating it as constant is a safe assumption.
    0 0
18. johnd at 08:39 AM on 4 September, 2010
    TOP at 23:57 PM, the absorption band chart illustrates very well the role each component plays. It is no coincidence that humans radiate at about 10um which is right in the transmission window. It is also illustrated very clearly that if there was no CO2,then the water vapour alone would allow more thermal energy to be radiated off as the transmission window gradually closed. With CO2 added, the transmission window closes off earlier, so the loss of heat is halted earlier. However it does not overlap the point at which water vapour allows maximum transmission. The only way for CO2 to overlap that portion of the spectrum is for it to change it's properties, or for the properties of water vapour to be altered. The chart represents very well the properties of H2O and the band occupied by water vapour reflects the points at which the changes of state occurs. For the concept of CO2 saturation to apply, the width of the CO2 window would have to either widen or the position occupied on the spectrum shift, something the known properties of CO2 do not allow for. If an analogy helps, water vapour is a sliding window that opens once solar radiation streaming through closed windows heats the room. As the heat increases, the window begins to open becoming almost fully open at about 10um which by coincidence allows us within the room to enjoy a comfortable environment, then as the loss of heat going out the window begins to overtake the heat coming in, the window begins to slide close. However with CO2 present, the final few um of the window opening has already been blocked off so the loss off heat is curtailed a little earlier than it otherwise would have been.
    0 0
19. Chris Colose at 08:58 AM on 5 September, 2010
    There's been some interesting questions and a few off-target remarks here, which touch on several topics. Hopefully I can provide some focus below. The radiative forcing for a change in CO2 is highly dependent on the the temperature structure of the atmosphere, cloud cover, and water vapor. You can boil down the physics to a simple statement about forcing going like the logarithm of concentration, but this cannot be derived (or the 5.35 W/m2 coefficient) without spectrally resolved calculations. This is in part because the other gases/clouds "steal" some of the radiation that would otherwise. In fact, the power of a specific greenhouse gas is maximized when it is acting by itself and does not need to compete with the spectral overlap by other constituents. This is discussed in an in press paper by Schmidt et al 2010 which attempts to partition the total greenhouse effect by contribution to individual gases, and I discuss it here. What's more, the logarithmic dependence breaks down under situations not too far from modern-day Earth-like conditions, such as at the very high CO2 concentrations thought to exist at the termination of a snowball planet. To touch on RSVP's point in 9, he is correct that there is still some uncertainty in the forcing for a doubling of CO2 (The central value is 3.7 W/m2, though with a range of about +/- 0.3 W/m2, see Forster and Taylor, 2006 depending on method used) but becomes much larger for different atmospheres (like past Mars or early Earth). The vertical temperature profile is also relevant because the greenhouse effect requires some sort of lapse rate to allow colder atmosphere aloft to radiate to space at a temperature colder than the surface. Even though the concentration of CO2 is pretty uniform over the globe, the forcing does have some variation over the planet due to changed tropopause location and lapse rate effects. The solar radiation of course does matter for the greenhouse effect to be relevant at all (and even the shortwave absorption by water vapor and CO2 is not completely zero, though much smaller than the longwave part). If the Earth had no incoming sunlight, the greenhouse effect would not support any temperature higher than the 'cosmic background' temperature, although it would take the planet a bit longer to cool off to that point than a planet where you turned off the sun and had no atmosphere at all. This is obviously quite removed from reality though and Dima is correct that the small variations in solar radiation do not matter for the CO2 greenhouse effect. In fact, the true no-feedback sensitivity parameter (in the article, 0.27 K/(W/m2))is also dependent on the finite absorption of the atmosphere, and so becomes more on the order of about 0.3 to 0.31 K/(W/m2). This so-called Planck feedback response is pretty robust across various models; see for example Table 1 in Soden and Held, 2006 [PDF]). This table gives an estimate of the magnitude of the Planck feedback amongst various models (you need to take one divided by these numbers to be consistent with my units). Water vapor acts to enhance this sensitivity by making a plot of the outgoing longwave radiation vs. temperature more linear than the T^4 dependence that a blackbody has. This enhances the sensitivty by a factor of about two, to the extent that the upper tropospheric moisture content scales with the Clausius-Clapeyron relation. To humanity rules (#11), the vapor content should go up nearly exponential with T, but the absorptivity goes up nearly like the logarithm (though not as nicely as CO2, and some have argued more of a square root dependence), so the feedback of water vapor as T increases should be somewhat linear. See my article on feedbacks here for more. Also keep in mind that the typical sensitivity-forcing equations in this article apply only at equilibrium and the timescale to reach equilibrium depends on the climate sensitivity, so the "heating in the pipeline" becomes larger as sensitivity increases. To HumanityRules (#11 regarding the natural oscillations)-- I don't think anyone argues that the polar regions are highly variable and exhibit significant amounts of influence from ocean circulation. In particular, windiness and advection is a large part for estimating year to year 'minimum' in sea ice extent, but the preconditioning of ice loss (not just extent but thickness) is clearly due to albedo feedback which is in part related to rising temperatures, and these temperatures are rising almost everywhere on the globe, it's not a redistribution, so this is a signature of external forcing. Just from the abstract, the Chylek paper uses detrended data to see a see-saw effect, so it's unclear to me how you can make statements about the trend from this (but I have not looked at the whole paper and cannot for some time), but the abstract itself says that natural variability in the Arctic as well as the trend are both at play here. Finally, the canonical 2 to 4.5 C estimate of equilibrium temperature change per doubling of CO2 (and the feedback parameter lambda itself) does encompass water vapor feedbacks but also the other effects (lapse rate, clouds, ice-albedo) required to get the full range.
    0 0
20. Eric (skeptic) at 11:12 AM on 5 September, 2010
    How does the C-C relation govern upper tropospheric water vapor? Isn't is governed by convection, cloud formation and other local factors? It seems like we are presupposing that the local variations are all somehow averaged out and the C-C relation holds on average. But it only holds for some (possibly changing) percentage of the upper troposphere and in the rest there is greater and lesser water vapor depending on local conditions.
    0 0
21. Chris Colose at 14:23 PM on 5 September, 2010
    Eric, this is a good question, and a lot of popular accounts of the water vapor feedback get this wrong. I'm trying to work on more posts on feedbacks on my own site and possibly for guest contributions to others to clarify such matters. First of all, C-C only provides an upper bound on the water vapor content in a given parcel of air at some temperature. It is therefore reasonable to assume that if the vapor pressure reaches or exceeds the saturation vapor pressure, you will get condensation. However, C-C does not, in itself, tell you how water vapor will change as the climate warms. Secondly, just because warmer air has a greater capacity to "hold" more water vapor, it is not at all self-evident that it will. It is like saying that a larger bucket will hold more water than a small one, neglecting how leaky either one of them is. However, if relative humidity (the percentage of vapor the air is actually holding relative to saturation) stays roughly constant, then the C-C relation is a great starting point for describing the fractional change in vapor content per unit warming (and like CO2, it's the fractional change rather than the absolute change which is important for the radiative transfer). And to the extent relative humidity is conserved in the large scale, as wasexpected going back to at least Manabe and Wetherald in the '60's water vapor must provide a strong positive feedback. There are solid energetic constraints for why relative humidity should remain roughly conserved in the lower atmosphere, but these arguments do not hold a lot of weight in the upper free atmosphere, and it is higher up where air is cold and where water vapor has the strongest contribution to feedback (the lower level air is more important though for precipitation responses). There are several physically plausible mechanisms for getting a negative water vapor feedback. The Held and Soden (2000) review paper on water vapor feedback provides some examples. Before the IRIS cloud hypothesis, Lindzen's idea in the early 1990's was that the mean detrainment altitude of deep convection will be both higher and cooler in a warmer climate. The water vapor deposited into the free atmosphere depends on the saturation vapor pressure at the temperature of cloud detrainment, and so this would in principle lead to less vapor left behind at higher altitudes. There is not really a theoretical basis (as far as I know) as to why this should be wrong, but like IRIS, there was really no supporting evidence for this. Many observational-based approaches and more sophisticated GCM's have evolved, and responses to ENSO, Pinatubo, and satellite observations show that models are pretty much getting things right. The Dessler and Sherwood paper in Science last year describes the evidence rather well. Furthermore, the developed thinking now is that the water vapor feedback is controlled by the large-scale dynamics and the saturation specific humidity in the outflow regions of tropical deep convective systems, and is not particularly sensitive to the detailed microphysics (e.g., Sherwood and Meyer 2006).
    0 0
22. Rob Painting at 18:51 PM on 5 September, 2010
    Chris Colose @ 21 - A Matter of Humidity Sherwood & Dessler 2009. "Thus, although there continues to be some uncertainty about its exact magnitude, the water vapor feedback is virtually certain to be strongly positive, with most evidence supporting a magnitude of 1.5 to 2.0 W/m2/K, sufficient to roughly double the warming that would otherwise occur. To date, observational records are too short to pin down the exact size of the water vapor feedback in response to long-term warming from anthropogenic greenhouse gases. However, it seems unlikely that the water vapor feedback in response to long-term warming would behave differently from that observed in response to shorter-time scale climate variations. There remain many uncertainties in our simulations of the climate, but evidence for the water vapor feedback—and the large future climate warming it implies—is now strong."
    0 0
23. Eric (skeptic) at 22:23 PM on 5 September, 2010
    Thanks for the answer Chris. In a nutshell my view is "all climate is local". Here's one particular location where the seasonal variations were examined and could not be explained by a temperature-SVP link. http://journals.ametsoc.org/doi/abs/10.1175/JHM484.1 But your answer is that the WV feedback is controlled by the aggregate large scale dynamics in tropical convection. If I have that right, do you also add the likelihood of an increased tropical region size as the world gets warmer? Are there estimates for that? Also I can not see how an energy flow equilibrium can govern water vapor. The ocean temperature buffering means that it takes years for an increase in CO2 forcing to raise global average temperature without feedback. But water vapor feedback takes hours or days at the most. The time scales are off. Just for the record, there is no such thing as conservation of energy other than over the long run at the top of the atmosphere. When there is concentrated heat somewhere like we saw this summer, there is not lack of heat elsewhere to make up for it.
    0 0
24. The Inconvenient Skeptic at 15:21 PM on 14 October, 2010
    One thing I have never been able to find is, what is the claimed current amount of energy that CO2 is absorbing from the surface? I have my own estimates, but the only thing I have been able to find on this site (or other sources) is the change in warming as a result of increasing CO2 levels. I am not asking for super precise information, but a range or magnitude. Is it currently claimed to be 5, 50, 100, 150 W/m2. If I plug 0.1 into dF = 5.35 ln(C/Co) using the 290 pre-industrial I get -42.6 W/m2. There is so much variation in results as the limit of C approaches zero that the whole thing is useless though. My guess is that the current claimed forcing of CO2 is on the order of 50-100 W/m2, but I have never been able to find a value for what it is. I am hoping that someone will be willing to help me with this, even if I am a skeptic. Thanks in advance. John
    0 0
25. scaddenp at 15:39 PM on 14 October, 2010
    John - the IPCC WG1 report. You have read this I hope. However, you comment is somewhat amnbiguous. A 50-100W/m2 FORCING would be deadly. More like 3.7 I think. Are you asking though about the amount of change in radiation from surface being absorbed by CO2 or the change in the amount of back radiation received by the surface from the amount of change in CO2 alone? You might want to look at Philipona 2004 or 0 0