Robert Brown on Ken Caldeira resigns = respect
Robert Brown says:
Maybe his German (she sounds like that) Postdoc explained the hydrological cycle to him.
A very interesting post. I wonder if Caldiera is changing his stance because he suddenly realized that his own research has just proven that the climate sensitivity is much smaller than has been claimed by e.g. Hansen et al.
She speaks of “irrigation” or “reforestation” as mechanisms that keep (relatively small) areas of the land wet. Water is then evaporated from this wet land (or from the pores of trees during respiration) and yes, it carries away at least the latent heat of vaporization as it does so, locally cooling the surfaces.
This water vapor — water containing heat that was picked up from the ground — then is transported up. H_2O molecules are lighter than O_2, N_2 and CO_2 molecules and — unlike CO_2 that tends to blanket the ground because it is literally “heavier (more dense) than air” — diffuses generally upward, especially when it is warmer than the ambient air, which, at least initially, it is.
Air in the lower atmosphere cools fairly predictably with height. As the water vapor transfers some of its surplus heat to the surrounding air (heating it), pressure instabilities are created that cause updrafts that carry the warmer wetter air up not through diffusion, but via active transport that rather quickly lifts it up to much cooler air. Water is a polar molecule and strongly interacts with just about everything, and as it encounters ever cooler air (and constantly transfers some of the heat it picked up on the ground to it) it locally adsorbs to e.g. neutral air molecules, creating a more dense (if short lived) complex that gradually experiences no net lift. Air molecules continue to carry the heat on up, however, warming air that is ever less dense very slightly.
As the density of water in atmospheric layers continues to increase and the water itself cools, it becomes unstable. It doesn’t bond well to (strongly covalently bonded) O_2 and N_2, although even those molecules experience short range dipole-induced dipole attraction, enough to actually help hold water molecules apart, but it forms relatively strong and long lasting bonds to nearly anything else, especially molecules with any sort of ionic or polar character. Neutral air molecules act as a dielectric to reduce the strength of the surrounding dipolar field of the solitary water molecules (which already drops off relatively quickly with distance), but bare ions exert a very strong force of attraction over a comparatively long range. Sunlight itself provides a continuous source of ions in the upper atmosphere. Various pollutants, particles of dust (including the continuous rain of microscopic meteoric dust that is constantly drifting down from overhead), ozone produced by sunlight, density waves caused by the passage of aircraft and yes, ionizing cosmic radiation that produces e.g. pair production cascades shotgun style as very high energy massive particles strike air molecules and are suddenly slowed down — all of these things nucleate water droplets in the unstable supersaturated air.
This generally happens where the droplets formed are not water droplets — they are cold enough to immediately become ice. Although the air may not be saturated for the formation of liquid droplets, it is often supersaturated for the formation of ice. Ice itself is a fascinating substance, because water is so very interesting and structured. The growth process tends to favor the creation of sharp points on the surface, and sharp points augment the local field strength, which cause those points to more strongly attract the surrounding water molecules and hence grow, becoming sharper still (and spawning new sharp points in a fractal but structured way as the underlying ice fleck grows). This is known as the Bergeron process after its original discoverer, and is one of the primary ways that clouds grow. “Seeding” clouds consists of dumping nucleation points into not-quite saturated air so that the air becomes supersaturated with respect to them and relatively rapid nucleation and growth cascades to form clouds.
Clouds do indeed have a high albedo. On the sunny side of the Earth, they contribute to net cooling by reflecting visible light, shading the surface and air underneath. Cloudy days are generally markedly cooler than clear days, all things being equal. On the dark side, they often contribute to net heating by reflecting infrared (and visible) back down. Cloudy nights are generally warmer than clear nights, all things being equal. It should be carefully noted, however, that the daytime reflection of energy involves a lot more net energy, energy that does not ever make it to the Earth, where the trapping of heat at night both requires there to be heat delivered during the day to trap and blocks only one of many cooling mechanisms.
Overall, clouds have a cooling effect on global climate, period — if the Earth were completely enshrouded in clouds (perhaps because of an oceanic strike of a small asteroid that dumped a few hundred cubic miles of water vapor and nucleating directly into the stratosphere) it would produce catastrophic cooling. See, for example, the hypothesized “nuclear winter” that could follow a global thermonuclear war, note the coincidence between the onset of above-ground nuclear testing (that put substantial amounts of radioactive dust into the upper atmosphere) and the cooling period from 1945 to 1970. Yes, only a hypothesis, but again, nobody doubts that clouds are net cooling so uniformly increasing cloud cover on average decreases global temperatures. It is why the CAGW enthusiasts hate and fear the GCR hypothesis and any other evidence of enhanced cloud related cooling driven by anything at all.
The mechanism described above has one more very important and, as far as I can tell, ignored mechanism for cooling present in it. The process as described has the net effect of taking heat from the Earth’s surface, where it was primarily (on average) received from the Sun and transporting it upward some 20-40 km into the atmosphere! As clouds form out of supersaturated air, they warm the surround air by depositing the heat of vaporization and the heat of fusion (as the molecules freeze directly onto the ice crystals in a cloud). One can actually see this warming air and transport of heat in action in the infrared photographs of e.g. hurricanes and storms. Heat is actively picked up from warm, wet oceans, rapidly transported upward, rapidly given up to surrounding cool air, warming the air and cooling the water, which then falls down again.
And the warmer air it leaves behind? It goes up. The troposphere is characterized by warmer on the bottom, cooler at the top, with heat transport dominated by vertical turbulent mixing. At the top — the tropopause — the dominant heat transfer mechanism switches from being convective to being radiative with good lateral mixing (horizontal winds) but with relatively little vertical mixing. Warmer molecules diffuse (relatively slowly) upwards, resulting in a stratified thermal profile with cooler air on the bottom, warmer air on top — this is the stratosphere. Stratospheric cooling is dominated by radiation, and heat trapping there is indeed dominated by greenhouse gases, including water. Stratospheric water is considered to be a wild card in the global greenhouse effect, even by NOAA:
To quote the lead paragraph in this story: “A 10 percent drop in water vapor ten miles above Earth’s surface has had a big impact on global warming, say researchers in a study published online January 28 in the journal Science. The findings might help explain why global surface temperatures have not risen as fast in the last ten years as they did in the 1980s and 1990s.” Or it might help to explain why they may well have fallen.
This is the second way that surface evaporation and subsequent cloud formation impacts global temperature, and it may prove to be a truly significant (and incorrectly modelled) contributor to global temperatures. Stratospheric water vapor is still unstable — a variety of processes remove it and it has to constantly be replenished from below. It is one of the most powerful greenhouse gases — much stronger even as a vapor than CO2, and much stronger (as noted) as clouds, e.g. cumulus (troposphere) or cirrus (tropopause). The 1980′s were, apparently, a high point in stratospheric water vapor, and this water vapor may have been the, or at least a, primary factor in the anomalous warming. NOAA itself recognizes this. Since 2000, water vapor in the upper stratosphere has decreased by around 10%, and I would wager that it is not yet equilibrated because I would bet good money that the state of the Sun is a major factor in the mechanisms that transport water vapor into the stratosphere.
Here it is in a nutshell. Increased cloud formation in the troposphere, wherever it occurs, leads to decreased water vapor in the stratosphere. Once the water is bound up in ice and/or water, it simply isn’t generally available for transfer into the stratosphere. This creates feedback and climate sensitivity with the wrong sign for catastrophic global warming, because increased sea surface temperatures increase evaporative cooling rates of the surface, which in turn causes more rapid transfer of more heat vertically via convection (especially at tropical latitudes where the thermal differential is the greatest). Even without solar modulation of the cloud formation process, this leads to increased net cooling by transferring more heat up above the troposphere where there is a smaller amount of atmosphere left to reflect its heat back down. However, it also carries water vapor with it to the stratosphere, and effectively thickens the stratospheric blanket, slowing the radiative loss of (primarily) upper tropospheric air.
Any modulation of the cloud formation process in the troposphere, however, can reverse the sign of this feedback. Warm SSTs that lead to more tropospheric clouds, especially at tropical latitudes, also lead to less water vapor (on average) at the tropopause and reduced active transport into the stratosphere. This in turn can increase the cooling rate of the tropopause by reducing a primary component of the stratospheric greenhouse. This drops the temperature of the upper troposphere, which favors even more rapid cloud formation, which further reduces the stratospheric greenhouse. As long as the factor modulating increased cloud nucleation is active, there are several feedback loops enabled that all lead to active net cooling until a new equilibrium is reached.
The timescale for this reduction is of great interest. Stratospheric mixing and transport through the layers is slow. E.g. CO2 levels in the stratosphere lag changes in the troposphere by as much as a decade, and the transport mechanisms themselves are strongly modulated by decadal oscillations such as ENSO, PDO, NAO that have a chaotic character and which strongly affect things like where, when and why clouds are formed in the first place. There is further “random” variation in the form of e.g. volcanic aerosols. All of these things produce decadal trends in global temperature (in particular SSTs and the distribution of SST anomalies) that are potentially strongly coupled to stratospheric mixing and water transport. Based strictly on the known timescales of these oscillations (which may not be strictly periodic) one would expect lag times of 10-30 years between the point where e.g. cloud nucleation rates changed and the effect of those changed rates impacts things like stratospheric water vapor content and mean dayside albedo.
This is more than long enough to avoid confounding a solar-cloud modulation hypothesis, and may in small measure explain why observed global temperature variation is as strongly coupled to the lengths of solar cycles as to their strengths. It is very probable that there are resonances in the decadal oscillations, resonances with long, and different, periods. Even small shifts in the period of solar cycles may shift the feedback drivers from net positive, general warming, to net negative, general cooling as it drives cooling mechanisms in resonance and warming mechanisms out of resonance. In resonance problems it isn’t the strength of a driver that is of primary importance — it is how close you are to the resonance!
This is, truly, the evil of the infamous hockey stick. Global temperature is interesting and complex. It goes up, it goes down, by order of ten degrees C on timescales on the order of 1000-10,000 y., by amounts on the order of degrees C on timescales on the order of 100-1000 y, by tenths of a degree C on timescales on the order of 1-10y. This is clearly demonstrated in an overwhelming number of studies over timescales out to millions of years ago. There are 100,000 y quasi-periodicities. Things on geological time scales such as continental drift play an important role. Alleging that average global temperature would be flat if it weren’t for CO2 is absurd; it is contradicted by enormous amounts of long term proxy data and temperature reconstructions.
Alleging this when we have a 33 year baseline of actual reliable global temperatures and a similar baseline (plus perhaps a decade or two) of borderline sound physical science and data addressing the issue is naive. Hence the NOAA article (largely ignored, I’m certain, in the latest IPCC fiasco) which doesn’t seem to have been anticipated in GCMs, and hence is an actual error in the models that has now more or less conclusively been demonstrated. Are there other errors in those models? Of course there are. This is hardly settled science, or NOAA couldn’t have “just” discovered that stratospheric H2O is being strongly modulated by some process that is completely ignored in GCMs and that is at least capable of completely overriding the expected CO2 signal in the short run.
Asserting that the science is settled in an issue like this where a lot of money is on the line is not just naive or absurd, it is criminal. It is also openly inviting Nature (the reality, not the journal) to step in and bitch-slap you upside the head and demonstrate just how wrong “settled” science can be, especially in what is arguably the most complex open thermal system we have ever studied, or maybe the second most complex one — second only to the Sun. And of course the Sun is the other wild card in the global climate equation, isn’t it?
I would never assume that all the people who are doing or have done climate research are either venal, political, biased or stupid. Some may well be, and Climategate has gone a long way to reveal to the world at least one relatively closed subset of people that probably are one or more of these things. Scientists who do indeed have an open mind, and who may well have been convinced by prior arguments for CAGW have the right and privilege to at any time change their mind based on new data and results. Indeed, as negative evidence is accumulated and gradually mounts up, any good scientist should, slowly, decrease their degree of believe in CAGW and increase their degree of belief in alternative hypotheses, especially ones that can also explain the observational data of the past and better explain the present and the future as it is unveiled by the passing of time.
It is not at all implausible that Caldiera is just such a good scientist. It seems that his own work is starting to show some of the cracks in the CAGW hypothesis. The idea that increased surface evaporation leads through GCMs to increased net global cooling all by itself casts doubt on the egregious end of the assertions for climate sensitivity, and I’m quite certain that whether or not he has formally estimated its impact he intuitively recognizes that it almost certainly adds to the already substantial evidence that the higher “catastrophic” values are excluded by existing data and observations. This doesn’t mean that AGW is entirely wrong — he is quite correct that it is dead certain that anthropogenic CO2 has contributed something to the average global temperature. What it may mean is that the “C” is no longer a concern.
What it may also mean — what the NOAA observation means — what the strong correlation between solar state and global temperature over century to millennial time scales means — is that suddenly the magnitude of the CO2-connected AGW has once again become a seriously open question.
The problem that the IPCC faces is very simple. From the very beginning nobody sane has denied that CO2 is a greenhouse gas, or that increased CO2 very likely results in (on average) an increase in global temperature relative to some baseline. What has been doubtful throughout that entire interval and is doubtful today is the climate sensitivity, because it is a single parameter that hides a huge amount of chaotic dynamics in a complex, open system with multiple time scales and reduces it to an enormously (over)simplified term in a differential equation. People attempt to estimate it, people attempt to measure it, people attempt to model climate using it and compare the results to observed temperatures. The uncertainty associated with this process is clear from the fact that even the proponents of CAGW don’t claim to know the sensitivity to within a factor of two or three. They want to establish lower bounds, but those bounds are clearly not well-justified from a statistical point of view given the range of possible variation.
Observation, especially reliable observations made with modern instrumentation over decadal time scales, restricts the permissible range of the climate sensitivity. We’ve seen a number of posts on this issue on WUWT already. Existing observations already make it extremely unlikely that the higher “catastrophic” values for this parameter are correct, as if they were one cannot explain the last decade of stable to cooling temperatures (or, as Bob Tisdale is fond of pointing out in great detail, the fact that abrupt warming followed by stable temperatures, with e.g. ENSO correlations to the abrupt warmings, is rather the rule and very difficult to explain with a CO2 driven model with very high sensitivity).
Caldiera has very likely come to realize this, whether or not he has yet gotten to where he believes it sufficiently to state it in print. His own work, as noted carefully above, makes it likely that the upper reaonable bound will get lowered again over the next two or three years, even if the GCR and other explanations for solar modulation of cloud nucleation are not further supported by observation and experiment as they may well be. At the same time, his own work makes the lower ranges, the ranges where e.g. Roy Spencer asserts that it should be, more likely, again even without solar modulation. Solar modulation would simply be a dagger that instantly closes the topic and results in a threefold decrease in the temperature anomaly we can manage with AG CO2 forcing coupled with a natural variation that is 3 to 10 times more important as a determinant of the actual global mean temperature.
I do think that viewing this as a “rat leaving a sinking ship” does a disservice to the entire debate. Instead of expressing bitterness, it is much more appropriate to recognize that a scientist has the right to an opinion, that the opinion is probably not completely ill-founded (mistaken or not) and that a good scientist has the right to change their mind as new evidence emerges that confounds their earlier beliefs. His resignation might be nothing more than his personal acknowledgement that the “C” in CAGW is finished, and that the actual AGW anomaly is suddenly an open question once again.
Yes, you probably think that it is, or should have been, open all along. I probably agree with you. But let’s give credit to a man who is willing to change his mind as evidence weakens belief, and accord him some respect.
Respect is, most unfortunately, a feature that is sadly lacking in the entire debate.