Water vapor feedback: evaporation

by Judith Curry

New research from Carnegie’s Global Ecology department concludes that evaporated water helps cool the earth as a whole, not just the local area of evaporation, demonstrating that evaporation of water from trees and lakes could have a cooling effect on the entire atmosphere. 
The press release further states:

Evaporative cooling is the process by which a local area is cooled by the energy used in the evaporation process, energy that would have otherwise heated the area’s surface. It is well known that the paving over of urban areas and the clearing of forests can contribute to local warming by decreasing local evaporative cooling, but it was not understood whether this decreased evaporation would also contribute to global warming

The Earth has been getting warmer over at least the past several decades, primarily as a result of the emissions of carbon dioxide from the burning of coal, oil, and gas, as well as the clearing of forests. But because water vapor plays so many roles in the climate system, the global climate effects of changes in evaporation were not well understood.

The researchers even thought it was possible that evaporation could have a warming effect on global climate, because water vapor acts as a greenhouse gas in the atmosphere. Also, the energy taken up in evaporating water is released back into the environment when the water vapor condenses and returns to earth, mostly as rain. Globally, this cycle of evaporation and condensation moves energy around, but cannot create or destroy energy. So, evaporation cannot directly affect the global balance of energy on our planet.

Increased evaporation tends to cause clouds to form low in the atmosphere, which act to reflect the sun’s warming rays back out into space. This has a cooling influence.

Video of Julia Pongratz speaking about the cooling influence of evaporation

Video of Ken Caldeira asking: increased evaporation lead to global cooling?

The press release doesn’t mention a specific publication, but I spotted this one, which appears to be the relevant publication:
Climate forcing and response to idealized changes in surface latent heat and sensible heat
George A. Ban-Weiss, Govindasamy Bala, Long Cao, Julia Pongratz, Ken Caldeira

Abstract. Land use and land cover changes affect the partitioning of latent and sensible heat, which impacts the broader climate system. Increased latent heat flux to the atmosphere has a local cooling influence known as ‘evaporative cooling’, but this energy will be released back to the atmosphere wherever the water condenses. However, the extent to which local evaporative cooling provides a global cooling influence has not been well characterized. Here, we perform a highly idealized set of climate model simulations aimed at understanding the effects that changes in the balance between surface sensible and latent heating have on the global climate system. We find that globally adding a uniform 1 W m−2 source of latent heat flux along with a uniform 1 W m−2 sink of sensible heat leads to a decrease in global mean surface air temperature of 0.54 ± 0.04 K. This occurs largely as a consequence of planetary albedo increases associated with an increase in low elevation cloudiness caused by increased evaporation. Thus, our model results indicate that, on average, when latent heating replaces sensible heating, global, and not merely local, surface temperatures decrease.

Environ. Res. Lett. 6 (2011) 034032 (8pp)   Link to full paper [here].

I find this to be quite an interesting paper.  I tend to be more of an ocean person than a land person, so I find the idea of land use changes resulting in latent heat replacing sensible heating to be interesting (but not at all surprising once you think about it).
Ocean evaporative feedback
The evaporative flux from the ocean surface  depends on surface wind speed (u) and the gradient of specific humidity (q) between the ocean surface and the lowest few meters of the atmosphere.     Surface temperature enters into the evaporative flux through its control on the humidity right at the ocean’s surface, which is determined by the Clausius-Clapeyron equation that relates saturated vapor pressure to temperature (note: to calculate the surface specific humidity, you also need to account for the solution effect of salts in the ocean water).  All other things being equal (which they rarely are), warming the ocean surface would increase the evaporative flux.
In terms of the climatology of ocean evaporative flux, see this global estimate from satellite observations [here].  While generally higher values are seen in the tropics and lower values in the high latitudes, the largest values are seen in the boundary currents such as the Gulf Stream, and some of the lowest values are seen on the equator.  These variations reflect the impacts of atmosphere and ocean dynamics and land/sea temperature and humidity contrasts in influencing the variation of ocean evaporative flux.
I have been thinking of some issues related to the ocean evaporative feedback, which are discussed in Chapter 13 on feedbacks in my book Thermodynamics of Atmospheres and Oceans, see especially section 13.6, pages in a discussion on the tropical ocean warm pool.  Here is a brief summary.
 Using experimental data obtained from research ships (in the tropical Pacific), we determined the derivative of the surface wind speed and humidity gradient with respect to surface temperature (To).

It was shown that dq/dT0 > 0, while du/dT0 < 0. Wind speed dependence of the surface latent heat flux dominates for T0 > 301 K, while the humidity dependence dominates for T0 < 301 K. The decrease in surface latent heat flux at high surface temperatures is hypothesized to arise from the following mechanism: high surface temperature —> increased instability and convection —> increased large-scale low-level convergence —> weaker surface wind — > lower latent heat flux. In interpreting the magnitudes and signs of these derivatives, it should be kept in mind that an apparent empirical relationship between wind speed and humidity with T0 is no guarantee that the primary factor giving rise to changes in wind speed or humidity is T0. Wind speed and T0 may appear to be related because both fields are related to a third and much more dominant factor, such as the large- scale coupled atmosphere-ocean circulation, which is controlled largely by the horizontal gradient in T0 rather than the value of T0 itself.

While googling around, I spotted this paper on Convection-Evaporative Feedback in the Equatorial Pacific, which is also relevant to this discussion.

Surface evaporative feedback as a component of the global water vapor feedback

A quick google search finds discussion of the evaporative feedback on local and short time scales, e.g. in the context of the Madden Julian Oscillation.  I am not finding much discussion of evaporation in the context of global water vapor feedback.

The basic physics giving rise to evaporative feedback is included in climate models, and causes surface cooling.  But there is an evaporative feedback, that is rarely discussed or included in the diagnosis of global thermodynamic feedbacks.  I am having a vague recollection that Will Kinnimonth has been making some of these arguments, but I recall the analysis that I saw had some significant errors, but the broader issues he was raising were interesting.  I can’t find anything right now, I will dig it up.   I know that skeptics have been talking about evaporation, I think I recall Tallbloke discussing how the shallow IR penetration depth couldn’t possibly warm the ocean, he argued that only the surface layer warms, which then increased evaporation.  This is incorrect since turbulence does mix heat in the upper ocean, and the physics of the cool skin layer right at the surface does not preclude heat exchange between the skin layer and the ocean mixed layer.

The broader issue raised in the Carnegie study that is of interest is to me is that the thermodynamic feedbacks – water vapor, cloud, surface albedo, lapse rate, evaporation – are not really separable and additive, they are linked.  Of these thermodynamic feedbacks, evaporation is much less frequently discussed and should arguably be integrated with the water vapor feedback (which is difficult to really separate from the lapse rate feedback and the cloud feedback).

So hopefully the Carnegie study will reinvigorate the discussion of evaporative feedback processes (over both land on ocean) in the context of global climate.

Moderation note:  this is a technical thread, comments will be moderated for relevance.

JC note:  a very busy week for me, I hoped to research this topic more before posting, but this is all I had time for.  Hopefully this will generate some good discussion and others can provide links and arguments.

118 responses to “Water vapor feedback: evaporation

  1. I urge readers to watch Ken Caldiera’s short video presentation.

    The questions here are fundamental to the output of GCMs and to our whole climate debate. The study, while extremely provocative, and done by excellent scientists, is but a start — nowhere near the finish — of what might be the most crucial technical issue about global warming: does increased evaporation lead to negative feedbacks?

    The authors seem to suggest, based upon their modeling results, that increased warming due to increased GHGs will cause increased evaporation, which in turn will cause an increase in low lying, white clouds, which will in turn reflect enough light back to space that the net result will be negative feedback. Meaning, that the amount of warming caused by GHGs before taking into account the complex web of feedbacks will be reduced, once this particular feedback is taken into account.

    So, assuming my summary isn’t too far off the mark, it seems to me that the most important thing for climate scientists to focus on right now is whether the authors of this new paper are largely correct. Yes or no??

    • Evaporation in the Arctic leads to fierce negative feedback.

    • Scientists who are wedded to the AGW theory have already transitioned to catastrophic climate change that is directly related to the hand of man. For some time now AGW True Believers Hand held that even global cooling Is the result of human-caused CO2.

      • Yes but reality is a merciless taskmaster – even of world leaders.

        The danger is that world leaders and leaders of science have aligned themselves with two scientific falsehoods for four decades:

        a.) Anthropogenic climate change, and the
        b.) Bilderberg model of the Sun as a stable H-fusion reactor.

        Having failed to “whitewash” Climategate, their only escape is to:

        c.) Admit deceit and risk retaliation, or
        d.) Ban open discussion of Climategate.

        Statesmanship is needed to resolve the situation and restore:

        e.) Integrity to government science.
        f.) Citizen control over government.

        With kind regards,
        Oliver K. Manuel

    • John

      The authors seem to suggest, based upon their modeling results, that increased warming due to increased GHGs will cause increased evaporation, which in turn will cause an increase in low lying, white clouds, which will in turn reflect enough light back to space that the net result will be negative feedback..

      Isn’t that exactly what Spencer + Braswell 2007 observed?

      Our measured sensitivity of total (SW + LW) cloud radiative forcing to tropospheric temperature is -6.1 W m-2K-1.

      S+B observed an expected increase in SW flux from reflection with warmer temperature, which was at first essentially offset by an increase in LW absorption (resulting in a negative LW flux); this surprisingly switched to a positive LW flux, so that the net feedback from clouds with warmer temperature was strongly negative.

      Variations in the CERES TOA all-sky (cloudy plus clear) SW and LW fluxes (Figure 2c) reveal the expected increase in reflected SW flux associated with clouds produced by the rain systems. But the transition from negative to positive LW flux anomalies during the period of above average rainfall is somewhat surprising.

      It appears to me that these latest model results showing increased low level cloud cover caused by evaporation resulting from warmer temperatures were confirmed by the actual CERES satellite observations of S+B 2007.


      • Manacker, as much as the Climategate emails have poisoned me on the Trenberths and Manns and Jones, etc., of the climate science world, I don’t necessarily buy what the skeptics say just because they are the (quite properly) aggrieved skeptics. Steve McIntyre has criticized the lack of statistical robustness in the results of the recent articles by both Dressler, and by Spencer and Braswell, for instance. I happen to regard some of the skeptics, Christy and Spencer and Lindzen, in particular, in high regard, but what is interesting to me is that the study we are now discussing is by a very centrist, very highly regarded mainstream climate scientist. He can’t be trashed. They are going to have to deal with the science Caldiera and company have come up with, whereas they would continue to try to simply trash S&B, Lindzen and Choi, without engaging.

  2. It’s a wonderful witches brew, which only the addition of a test tube full of Arrhenius’s marvelous gas can make magical.

  3. You talked about wind and evaporation, but you did not include the Arctic. We have been warming since the little ice age. This is because the Arctic Sea Ice was frozen and it did not snow much and ice albedo has been decreasing. Now, the Arctic sea ice is melted and Arctic water is exposed to fierce Arctic winds and the snows have started. We are entering a period during which we will have fierce snows during the fall, winter and spring and a period in which Albedo will increase and earth will cool.

    • Better yet, look at this series of papers.
      Arctic Evaporation Causes Snow that changes the weather in the Northern Hemisphere in Europe, Asia and North America.


    • I agree the Arctic aspect is very important and interesting. If you are following along in Chapter 13 of my book, see section 13.7, and especially Figure 13.9 http://curry.eas.gatech.edu/climate/pdf/chapter13_figs.pdf. An excerpt:

      Consider an equilibrium situation where a positive perturbation to the freshwater flux (e.g., excess precipitation or sea ice melt) is imposed at high latitudes. A decrease in salinity corresponds to a decrease in density, which diminishes the sinking motion and the thermohaline circulation. The weakening of the thermohaline circulation reduces the poleward transport of relatively salty water from lower latitudes, which further decreases the polar salinity (loop 1 in Figure 13.9), which is a positive feedback.
      At the same time, the decreased strength of the thermohaline circulation also reduces the northward heat transport, increasing the high- latitude surface density, which in turn intensifies the high-latitude convection and hence the overturning circulation (loop 2 in Figure 13.9), and hence is a negative feedback. This negative feedback partially compensates the positive feedback associated with salinity, but the compensation is not complete and the positive salinity feedback dominates.

      Diminished overturning leads to lower surface temperatures, which results in reduced evaporation. If all of the evaporated water returns to the ocean locally in the form of precipitation, then there is no net effect and loop 3 is inactive. If precipitation falls outside the region of evaporation, then the feedback in loop 3 is positive. Lower surface temperature also results in increased sea ice formation, which increases the density and hence the thermohaline circulation.

      A further result of decreased high latitude surface temperature is that the meridional atmospheric circulation is enhanced by the stronger meridional surface temperature gradient, resulting in increased northward transport of atmospheric heat and moisture, which increases precipitation and decreases the high latitude surface ocean density (loop 4 in Figure 13.9).


      Also, we have a new paper submitted that related to this issued, unfortunately the paper is under review in an embargoed journal, so I can’t post until this is published

      • Dear Dr. Curry:

        Inherently, the earth is thermodynamically stable, and it has been as such for billions of years. There are no feed backs for a stable system. Surface temperature is not a measurement of earth’s instability-It is the sensible heat exchanged that does, and the later is constant.

        Contrary to the conventional wisdom, observations show that rainfall has been decreasing with surface temperature rise, suggesting that surface evaporation has not been increasing with observed surface temperature rise. Therefore, the suggested surface cooling as a consequence of increased evaporation over land is unlikely to contribute to global cooling. Furthermore, land contribution to climate energy balance is negligible for land has negligible thermal capacity.

      • Consider an equilibrium situation where a positive perturbation to the freshwater flux (e.g., excess precipitation or sea ice melt) is imposed at high latitudes.

        If all of the evaporated water returns to the ocean locally in the form of precipitation, then there is no net effect and loop 3 is inactive.

        What happens in a non-equilibrium situation? doesn’t the process of evaporation, cloud formation, and rain fall speed the transfer of energy away from the Earth surface, compared to a case with no clouds? Isn’t a non-equilibrium analysis more pertinent to a hypothetical small change in Earth mean temperature due to an increase in CO2, especially over a time span less than 100 years?

        The equilibrium situation is a good start to an analysis, but surely it is not a good end?

      • How much we can explain by the non-equlibrium phase can be estimated by the heat capacity of the atmosphere and the size of variation in latent heat content of the atmosphere. Those numbers can then be compared to the size of energy accumulated over one year by some realistic level of radiative forcing.

        The mass of atmosphere is 10 tons/m^2. That means that the heat capacity of dry atmosphere is 10 MJ/m^2/K. For 1% water vapour by mass we would have 100 kg/m^2, which has about 250 MJ/m^2 latent heat. Warming by one degree might correspond to about 6% more vapor or 15 MJ/m^2. On the other hand such warming corresponds to 2-4 W/m^2 in climate forcing, which gives 60-120 MJ/m^2 in one year.

        Of course warming by 1 degree takes much more than one year. Thus we can see that the non-equilibrium effects of the atmosphere cannot be significant.

      • My question was about the way that changes in water vapor and cloud formation change the rate at which energy is transferred from near surface to high altitude. Earth surface and near surface atmosphere are warmed several degrees in a few hours by the sun each day, and then they cool again at night. At no time is the whole atmosphere above a square meter of Earth surface warming equally altogether, much less the whole atmosphere above the Earth surface. Can we really rule out the possibility, in this non-equilibrium condition, that a doubling of CO2 will speed the transfer of energy from near surface to high altitudes, or reduce total daytime surface insolation by increasing afternoon could cover, at enough places to matter?

        If the equilibrium approximation says that a doubling of CO2 will lead to a 2000 – 4000 year time span to re-equilibrate, and the exact model says (or would if it were known) that a doubling of CO2 will lead to a 20,000 – 40,000 time span to approximately re-equillibrate, is that a “significant” error entailed by the equilibrium assumption? that is certainly a very tiny difference in the annual rate of change of mean temperature.

        Consider the Clausius-Clapeyron relationship above a few square miles of Pacific Ocean. The atmosphere warms in the daytime, starts cooling in the late pm, and continues to cool throughout the night; is there any time during this transient that the C-C relationship is not wrong by more than 10% in nearly every sphere of atmosphere to which it is applied? Isn’t the relationship even a worse approximation when water rises to make clouds, rain falls, or the dew settles out at night?

        I think that the error entailed by the equilibrium approximation is large enough that even the sign of the effect of increased CO2 on temperature can not be confidently known.

        Also, I think that the public policy problem is in part a rate problem, not an equilibrium problem: Will the CO2 effect, if any, occur faster than humans can react?

        thank you for your reply. Probably my answer sounds disrespectful (what I call the “electronic effect”). If so, I apologize.

      • Pekka, Would it really correspond to 2-4 W/m^2 in climate forcing? Water vapor scatters and absorbs incoming solar and a large percent of that is radiated to space plus ice crystals scatter more plus there is the heat of fusion loss. Turbulent moisture laden air at altitude with liquid-solid-vapor cycling can pump a good deal of heat.

      • MattStat,

        My purpose was to illustrate using numbers, how small is the capacity of the atmosphere alone to slow down the change (and more specifically of the troposphere, because the slow mixing may allow the stratosphere to be slower in parts of its response). When oceans, biosphere and cryosphere are included, timescales may be totally different and the atmosphere has an important role in that.

        You are perfectly right in noting that the atmosphere is never in equilibrium and that the diurnal variations mean that it’s not even in a stationary state. Diurnal and other fast variations do, however, average rapidly to reach an average determined by slower changing external influences.

        The Clausius-Clayperon relation affects evaporation and condensation. These occur in some parts of the atmosphere and at certain times, but moving the boundaries is moving also the interior. Thus it’s certainly justified to draw many conclusions of general nature from the simplistic arguments. All the details of atmospheric physics belong rather to meteorology than to climate science, but there’s lot of knowledge on, how the averages that form the climate are obtained from the momentary states that form the weather.

      • Dallas,

        You should not read too much from my choice of the range 2-4. I just picked some numbers in the right ballpark. The settings are not defined in sufficient detail to tell, how far feedbacks are taken into account making the precise meaning of the numbers badly defined. You may extend the range as you wish, my main argument is insensitive to that.

      • Pekka wrote: Thus it’s certainly justified to draw many conclusions of general nature from the simplistic arguments.

        No denying that. But what exactly will happen over a 50 year time span if CO2 increases as it is increasing? If CO2 in the atmosphere increases daytime heating rate, increases cloud formation, and increases afternoon rainfall, and decreases subsequent nighttime temperature, then the conclusions of a general nature do not provide the answer.

        The whole atmosphere equilibrium approximation may be accurate enough for some purposes, even as the near surface model is sufficiently inaccurate as to be useless for human planning.

  4. Look at these related links.
    Watch this, “Explanation on Global Warming from mainstream media”, http://www.click2houston.com/video/27156168/index.html
    See this story about accurate forecasts

  5. I have been push wet lands or a long time. The University of Florida has done a great deal of studies from a wet land recovery perspective, but I have not seen studies of local climate impact of storm runoff ecosystems.


    Since retention ponds are mandatory in Florida, some of the designs are functional and very attractive.

  6. The evaporative flux from the ocean surface depends on surface wind speed (u) and the gradient of specific humidity (q) between the ocean surface and the lowest few meters of the atmosphere.

    Isn’t a more nearly accurate summary something like, “Presently used models of evaporative flux, based on correlations of empirical data, generally use the surface wind speed and the gradient of specific humidity as independent variables. And these are used almost exclusively and solely due to their convenience as correlating variables.” The important real-world physical phenomena and processes that determine the evaporative flux are those that dominant at the interface between the atmosphere and the ocean. The most important resistance to the driving potential at the interface governs the flux. And even this summary is very incomplete because the definition of ‘the interface’ is fuzzy, at best, given the compliant, deformed ( frequently as discrete chunks ), and constantly-in-motion nature of the ‘interface’. Absorption of gases from the atmosphere into water in various forms ( rivers, flumes, estuaries, lakes, oceans, etc. ) has been, I think, very comprehensively described using only the characteristics of the turbulence of the motions in the water at the interface ( when that interface is somewhat well defined ).

    Surface temperature enters into the evaporative flux through its control on the humidity right at the ocean’s surface, which is determined by the Clausius-Clapeyron equation that relates saturated vapor pressure to temperature (note: to calculate the surface specific humidity, you also need to account for the solution effect of salts in the ocean water).

    The Clausius-Clapeyron equation has been found to be useful basically due to it being readily convenient. There are models of mass and energy exchange at liquid-gas interfaces that do not rely on the equation. The application of the equation in the particular context here represents a major departure from the theoretical basis of the equation; the liquid and vapor phases of a pure substance at equilibrium ( pressure and temperature ) at a planar interface. I recall that the Gibbs function for the liquid phase and the vapor phase are equal on the co-existence line.

    All other things being equal (which they rarely are), warming the ocean surface would increase the evaporative flux.

    In order to impact weather and climate beyond the local region, the vapor must survive in the atmosphere and that is a function of the thermodynamic state of the atmosphere, not the thermodynamic state of the ocean surface.

  7. ps:
    I forgot that I have posted on this subject.

  8. The Earth has been getting warmer over at least the past several decades, primarily as a result of the emissions of carbon dioxide from the burning of coal, oil, and gas, as well as the clearing of forests.

    Was that necessary? Does that have anything whatsoever to do with the point of the article?

    • I tried to italicize just the part about fossil fuels, but the blockquote italicized everything.

    • The CO2 dogma is still strong. It takes time.

    • It’s an insurance policy, there are countless number of reputable scientists who have had their reputations tarnished by forgetting to mention such a thing!

    • Skeptics refer to it as The Genuflection. It’s part of the plan to get people to buy AGW through repetition by requiring The Genuflection in all Climate Science literature until it is accepted without thinking.


      • Do we walk or do we crawl?
        We are human, and we dance so.

      • Bad Andrew

        I’m skeptical.

        I’d think carrying on the same disambiguating practice in a climate paper as is used throughout the sciences can hardly be faulted by special pleading that because it’s climatology, disambiguation has a hidden agenda.

        If I read a Physics paper that mentions, “because the speed of light is the limit,” or an epidemiology paper that says, “as bodily fluids are the primary vector of transmission of this class of diseases,” in the abstract and introduction, I’d think nothing of it other than, ‘okay, I know where this paper is coming from.’

        This ignorant kneejerk reactionism to a commonplace simply underscores the lack of scientific credibility of some critics.

        If you want to read the paper and understand it, you need to know the baseline assumptions unambiguously that are made in the research.

        If as a skeptic you disbelieve the baseline assumptions, then you’ll want to know that you’re reading something founded on an assertion you do not agree with, don’t you?

        Or would you rather people hide their point of view?

      • Quoting from Tom Lehrer

        First you get down on your knees,
        Fiddle with your rosaries,
        Bow your head with great respect,
        And genuflect, genuflect, genuflect!

        Do whatever steps you want if
        You have cleared them with the Pontiff.
        Everybody say his own Kyrie eleison,
        Doin’ the Vatican Rag.

        Get in line in that processional,
        Step into that small confessional.
        There the guy who’s got religion’ll
        Tell you if your sin’s original.
        If it is, try playin’ it safer,
        Drink the wine and chew the wafer,
        Two, four, six, eight,
        Time to transubstantiate!

        So get down upon your knees,
        Fiddle with your rosaries,
        Bow your head with great respect,
        And genuflect, genuflect, genuflect!

        Make a cross on your abdomen,
        When in Rome do like a Roman;
        Ave Maria,
        Gee, it’s good to see ya.
        Gettin’ ecstatic an’ sorta
        Doin’ the Vatican Rag.

      • ibid

        Oh we will all fry together when we fry.
        We’ll be french fried potatoes by and by.
        There will be no more misery
        When the world is our rotisserie,
        Yes, we will all fry together when we fry.

        Down by the old maelstrom,
        There’ll be a storm before the calm.

        And we will all bake together when we bake.
        There’ll be nobody present at the wake.
        With complete participation
        In that grand incineration,
        Nearly three billion hunks of well-done steak.

        Oh we will all char together when we char.
        And let there be no moaning of the bar.
        And the party will be “come as you are.”

        Oh we will all burn together when we burn.
        There’ll be no need to stand and wait your turn.
        When it’s time for the fallout
        And Saint Peter calls us all out,
        We’ll just drop our agendas and adjourn.

        You will all go directly to your respective Valhallas.
        Go directly, do not pass Go, do not collect two hundred dolla’s.

        And we will all go together when we go.
        Ev’ry Hottenhot and ev’ry Eskimo.
        When the air becomes uranious,
        And we will all go simultaneous.
        Yes we all will go together
        When we all go together,
        Yes we all will go together when we go

    • P.E.

      Have a look at http://judithcurry.com/2011/09/25/trends-in-tropospheric-humidity/ for an example of a portion of the introduction being apparently unnecessary, and seeming to have nothing whatsoever to do with the point.

      Indeed, when I read the introduction, I say to myself, “okay, I know where this paper is coming from.” The author’s disaffection, alienation, and sense of persecution is amply set out. I know to expect sour grapes and spin, and I’m not disappointed.

  9. Dr. Curry,
    Regarding “shallow IR penetration depth couldn’t possibly warm the ocean, he argued that only the surface layer warms, which then increased evaporation”,
    You state ” This is incorrect since turbulence does mix heat in the upper ocean, and the physics of the cool skin layer right at the surface does not preclude heat exchange between the skin layer and the ocean mixed layer.”
    Have there been emprical studies confirming this? Obvious questions to me are how effective the mixing is and if so whether it’s meaningful.
    Thanks for writing a great blog.

  10. I am confused by this article, because evaporation over land should be going down due to deforestation and urbanization, yet their experiment has it increasing. So we should expect warming not cooling. Furthermore as the oceans are warming less quickly than the land, the global relative humidity should be decreasing especially over land, which also points to less clouds, less rainfall and drier conditions favoring warming.

    • What is unclear to me is that not all urbanization leads to less trees. Many prairie or desert cities have more trees than than was on the land before the city was there.

      Places like deserts and barren arctic areas have extreme changes in temp but areas with lots of vegetation tend to be mild.

      And isn’t irrigation (turn semi arid and arid regions green) and increasing crop yields (now feeding 7 billion people) adding vegetation to the planet?

      The land appears to be supporting more life now than in the past.

      • Kermit

        While I don’t disagree that in places some land appears to be supporting some life more now than in the past, most ocean is apparently supporting much less life both overall and of almost every particular type.

        Further, replacing deep-rooting wild species with shallow-rooted lawns and cash crops and altering soil composition by adding fertilizer, plant hormonal agents, and other contaminants that have unknown effects on microbes — by far the greatest fraction of terrestrial biomass — has unpredictable and inobvious but likely detrimental impacts.

    • I would agree that irrigation works in the direction of cooling, and this points to the need to increase irrigation to counter the natural effect towards more dry conditions. The way I see it, urbanization and deforestation work towards less evaporation by channeling precipitation into ground water and rivers more effectively rather than having it re-evaporate, while irrigation diverts river water to areas where it can re-evaporate more. These are the opposing forces.

  11. Seems to me the study misses the point.

    Evaporation has always had some effect.

    With an increasingly warm globe, it’s pretty clear the cooling portion of that net effect is falling relatively compared to other effects.

    Liebig’s law of the minimum means evaporation from plants will approach a peak limit, and open water (& ice) evaporation(&sublimation) will dominate the evaporation portion.

    The warmer the troposphere, the longer in proportion H2O in the atmosphere will spend as warming water vapor gas, compared the cooling water particulate clouds. Clouds themselves approach peak limits on rate of formation and of overall influence due cloud depth.

    I don’t see what the models have to say about these limits and the dynamical nature of the heating/cooling ratio. Did I miss this?

    • I am missing this too. More water vapor due to warming means more GHG, and if more clouds result this is just on the continuum path. In other words the free energy gradient points in a warming direction. I don’t understand how to formulate a problem based on thermodynamic and statistical mechanical arguments whereby this can completely reverse itself. At the very least someone could provide an analogy to a real world example where something like that happens, i.e. a phase change that causes the overall gradient to reverse.

      I just found this paper and will read it and see if it can add some value
      “Understanding Atmospheric Behaviour in Terms of Entropy: A
      Review of Applications of the Second Law of Thermodynamics
      to Meteorology”
      One of those MDPI papers, I know…

      • thanks for the link, i did a quick glance, it is worth a closer looks. I originally had visions of discussing topics like that here, but we never managed to get past the basics of the greenhouse effect. FYI, here is an early paper of mine on entropy and convection.
        Duane, G. and J.A. Curry, 1997: Entropy of a convecting water-air system and the interpretation of cloud morphogenesis. Quart. J. Roy. Meteorol. Soc., 123, 605-629.

      • WHT, This is one of the things I have been trying hard figure out how to explain. There are two unique things about the atmosphere and the oceans that seem related as thermostats, The tropopause and the 4C density of water. The tropopause is a near infinite heat sink with respect to the troposphere. If the local lapse rate changes the minimum temperature of the tropopause changes. There can be 30 C degree changes, decreases, associated with high convection. During day light hours, the impact of H2O’s radiative impact to outgoing long wave is reasonably predictable, but the radiative impact on SW is highly variable. At night, the radiative balance is more predictable. Water vapor has a non-linear impact on climate sensitivity. If the tropopause varied uniformly with warming, H2O would have a linear impact.

        The 4C ocean water temperature boundary may have a similar impact on OH uptake.

        I don’t have the math skills to properly define the mechanism of either, but since warming linearly increases water vapor, more rapid turn over of water vapor in the atmosphere would serve to enhance cooling

      • WHT,
        The gradient is always towards cooling.

    • Bart, are you sure on cause and effect? Does getting hot make it drier, or does getting drier make it hot? I seem to recall a paper that postulated (and presented evidence for) the latter, but I can’t for the life of me think what that paper was titled or who wrote it.

      • Neil Fisher

        Trying to find out where I claimed getting hot makes it drier, or drier makes it hot.. so I’ll see if I can follow your reasoning.

        I assert that the rate of cooling under increased CO2, ceteris paribus, due various mechanisms of evaporation is bound to approach limits for each mechanism, while rate of heating due evaporation from any mechanism does not reach any limit before the cooling effects are largely extinguished.

        Evaporative cooling due plants under increasing CO2, due either the parahormonal influence of CO2 on uptake or due the increasing temperature will have to slow, while increasing heat is at least constant until plants no further contribute to evaporative cooling.

        Likewise, evaporative cooling due clouds varies mainly with cloud cross section to the sun (all other things held to be equal), and cloud cross section (area) increases more slowly than does cloud volume, so is slightly slower overall than heating due increased H2O vapor concentration which increases antilogarithmically.

        Moreover, what I will call back-evaporation, the shift from cloud to vapor, is more rapid in a warmer atmosphere, and I’m unaware of any evidence of the shift from vapor to cloud increasing with warmer atmosphere due any mechanism.

        Do clouds getting hotter make them drier (shift H2O from albedo-contributing cloud to GHG vapor)? I’d think so.

        Do drier clouds (more H2O GHG vapor and less albedo) tend to make the surface hotter? It seems reasonable, during the day.

        After sunset, with increased water vapor under the cooling influence of night, cloud formation increased due the higher prior evaporation and water vapor concentration, tending to retain heat nearer the surface. (And half the heat radiated due condensation radiated downward, too.)

        Is that what you mean?

  12. Wind aids evaporation. Putting my green hat on, since wind turbines block and slow the wind and so they contribute to global warming.

    • Sure windmills block wind but shouldn’t this be compensated by the venturi effect of the wind passing below the vanes being faster and cooler?

  13. Nabil Swedan “There are no feed backs for a stable system.”

    There certainly can be. In fact, many systems are stable only because of their feedback loops. For example, the transmitter and receiver in your cellphone are stable only because of feedback loops.

    • Indeed, even chaos is a form of stability due to nonlinear feedbacks. The system oscillates within the strange attractor but does not venture beyond it. Climate may be such a system.

      One of the leading skeptical hypotheses is that negative feedbacks are neglected by the climate models, including nonlinear ones.

    • Can’t wait to see your complementary paper in a few days, Pere.

    • That is an impressive array of studies, Roger. However, is this line of evidence regarding water vapor feedback all merely about the neglect of other human influences? Or is it also about neglected negative feedbacks in general, natural as well as anthro?

  14. Just off the top of my head, evaporation wouldn’t only lead to cloud formation, but also would act similarly to a heat pipe.


    Only the heat would be carried from a liquid-air or solid-air surface to the upper reaches of the atmosphere.

    • That is a good description, a heat pipe with an infinite heat sink on one end, you are just changing refrigerant flow rate.

    • Yes; in a nutshell, the cycle’s effect is the “lifting” of sensible heat via evaporation-condensation. Heat is grabbed from the surface, and released at the condensation point. In effect/practice, the upper (radiating) layer of a cloud or cloud deck is the relevant relocation point.

      Necessarily, more OLR makes it out of the atmosphere from higher altitudes than from lower or surface radiation sources. Thus, acceleration of water cycling also accelerates cooling.

      Note that this is completely independent of any albedo effects.

      • I know that you probable know this, but the efficiency of the heat pipe changes with the distance between source and sink, so some natural oscillation cool more efficiently than others.

  15. It seems that the only certainty in global climate is that cooling causes warming and vice versa. It never fails.

    • We know what causes warming and it is not cooling. It’s the sun, stupid.

      • Stupid? The sun may cause the warmth, but it need not cause any particular instance of warming. Feedbacks alone can cause warming or cooling, under constant solar input. Moreover, if the system is oscillating under constant solar input due to feedbacks, as it well can, then yes cooling causes warming and then in turn warming causes cooling, etc. It is called nonlinear dynamics.

      • David, in general I agree that feedbacks or any other “natural” variabilities can cause warming or cooling. But in this case it seems the sun does cause almost all instances of warming/cooling since people have been observing the sun and its cycles/oscillations. Take solar cycle length (~11 y) for example (only as indicator of solar influence, there are others) – it correlates very well with temperature, taking into account that there must be other factors (feedbacks or anything else) doing their impact.

      • Edim, you can have your speculation and I will have mine. Just don’t insist that your speculation is known to be true. Leave that to the AGW folks (or fools). We do not know why climate changes. That is the state of the science.

      • Yes, when we are warm, because the sun made us warm, that melts Arctic Sea Ice and that causes snow which cools us.
        When we are cool, the Arctic Sea Ice prevents evaporation and precipitation and it don’t snow much and that lets the sun warm us.
        The sun warms us and the snow cools us.
        We do get cool because we are warm and we do get warm because we are cool.

    • People who look at data can, and some do, make mistakes and maybe, some of them do lie.

      The data don’t lie.

      The data does show that when we are warm, we always get cool.
      The data does show that when we are cool, we always get warm.

      So “a certainty in global climate is that cooling causes warming and vice versa. It never fails.” The Data shows this to be true!

    • We better hustle up and double; Livingston’s and Penn’s findings may well cool us more than a degree Centigrade.

    • Moncton’s reasoning is fascinating but it depends on a strong claim about control theory, which climate scientists have no basis for accepting. I call this a leaping concept. The barrier is high.

      • The applicability of linear control theory to the climate system is highly dubious, I agree.

      • Sorry, but I did not mean to suggest it was dubious, quite the opposite in fact. Leaping concepts can be revolutionary, for example the concept of the quantum leaped from the scientific domains of heat to light to atomic structure, revolutionizing physics. See my little blog note on this at https://www.osti.gov/ostiblog/home/entry/leaping_concepts_and_global_discovery

        The leap from control theory to climate science is even longer, as concept distances in science go. Thus the problem is that the folks in climate science are not likely to understand control theory enough to appreciate the point. That is the cognitive barrier I refer to. We would have to hear from some control theory experts, preferably those without a stake in the climate debate.

      • Not at all. Perturbations to smooth nonlinear systems in near equilibrium can always be linearized about the equilibrium point. Controls engineers use linear controls theory extensively even for highly nonlinear systems in this way. We make sure our systems stabilize large excursions into the nonlinear regime, but once the system settles, we look to linear theory to tell us how robust the system will be and how well it will respond to local disturbances.

      • Furthermore, many very nonlinear, non-smooth systems behave on average like linear systems. Ehrenfest’s theorem for quantum mechanics is a prominent example. Pulse width and pulse frequency modulators used extensively in linear control loop design are others.

        Moreover, many of the climate models use linear approximations to model specific processes. Dessler’s paper assumed a first order linear response to cloud forcing. We know that is an invalid assumption, but only because the relationship is demonstrably not a 1st order response. In any case, you cannot use linear models and then deny that linear analysis is applicable.

    • Judy – In addition to the cherry picking, there were too many serious mistakes in Monckton’s piece for me to find it interesting. Here is one example, in which he derived a low climate sensitivity from the Kiehl/Trenberth energy budget diagram (he used an outdated version, but I don’t think that makes a difference):

      ” The diagram shows surface radiation as 390 W m–2, corresponding to a blackbody emission at 288 K, equivalent to today’s mean surface temperature 15 °C. If the surface radiative flux were indeed the blackbody flux of 390 W m–2, then by differentiation of the fundamental equation of radiative transfer the implicit value of the Planck parameter λo would be ΔT /ΔF = T/4(F+78+24) = 0.15 K W–1 m2 (after including 78 W m–2 for evapo-transpiration and 24 W m–2 for thermal convection)”.

      From the above, I wonder whether he is aware that the λo he cites is not the λo used for no-feedback (Planck only) climate sensitivity estimates, since the latter involves flux changes at the tropopause as reflected in changes in the Earth’s radiating temperature of 255K. The surface flux (with or without the artificial additions of non-radiative heat losses) is not the basis for calculating climate sensitivity, which is determined by the radiative balance at the tropopause. With the appropriate values, the calculation for λo yields a value of 0.27 K W-1 m2.

    • Deriving climate sensitivity by dividing the central IPCC sensitivity estimate by the tropical “hotspot” amplification factor? Uh…what

    • Monckton’s use of linear control theory doesn’t make good sense. That does not mean that control theory may not be useful for describing the amplification of sensible heat removal by moisture laden air with variable vertical mixing. Using the heat pipe analogy, the more rapid temperature variations over land can amplify heat loss compared to over oceans. A hail storm model should be a good example.

  16. I don’t know who is missing the point here. Perhaps it is me.

    But I would have thought that there would have been more discussion of the potential importance of this study by Caldiera et al on model projections of the rate of temperature increases.

    Caldiera is in all likelihood a better scientist that most of the skeptics, possible exceptions being Christy and Lindzen, and he certainly is mainstream, and he seems to be saying that the basic science of whether increasing water vapor leads to a positive or negative feedback needs a good deal more work. Yet all the IPCC GCMs include various forms of positive water vapor feedback. Does the IPCC Emperor have no clothes?

    I can’t say, I don’t think I can unravel this complex ball of yarn from the outside. The only thing I have to go by, to judge whether the IPCC is right or not in its catastrophic temperature increase projections, is how the projections match the realities of current temperature rates of increase, current rates of sea level rise, current rates of Greenland ice melt, and in all cases, the rates are considerably less that IPCC projections.

    But that still leaves someone in my position — someone who studies the science passionately and as carefully as I can, but still is not an expert — with cognitive dissonance.

    Caldiera’s article is the first from a mainstream scientist of considerable repute which provides a possible explanation: the models may be wrong in the way they treat water vapor feedbacks.

    That is why I see this article as potentially of great importance.

    And why I am surprised that so few of the comments here seem to address this absolutely fundamental issue.

    • Perhaps so few comments address the issue because so many of the lukewarmers are aware there is a discrepancy in the models likely due to water vapor.

    • John – The paper is interesting, but it may not say what you think it does. For example, it does not claim that water vapor feedback is negative, and I would say that the net positivity of water vapor feedback is now well established – in any case, it wasn’t addressed here. What it does state is that evaporative cooling locally associated with latent heat transport can have global effects, at least according to the models utilized.

      Latent heat transport is a well known contributor to the so-called “lapse rate feedback”, which is a negative feedback considered to partially offset the positive water vapor feedback. The uncertainties in both the lapse rate and water vapor feedback are actually greater than the uncertainty of their net effect (water vapor minus lapse rate), and so the two are often combined into a “water vapor/lapse rate feedback” quantity, which is positive. (The lapse rate feedback is negative because cooling at the surface is associated with heat released at higher altitudes when the water condenses, and therefore escapes to space with less interference from greenhouse gases).

      Since these general principles are well known, probably the main new conclusion from the paper is that there is also a negative feedback from evaporation due to an increase in low cloud formation from the evaporated water. Low clouds are potent cooling influences because their reflection of sunlight outweighs their greenhouse effects, whereas high clouds tend to have a neutral or warming influence. If we look at actual cloud data over recent decades of temperature increase, it turns out, however, that low clouds have either remained stable or decreased slightly, while the low cloud/high cloud ratio has declined. These are warming phenomena rather than cooling phenomena, and so if the process described in the paper is operating, it appears to have been offset or even outweighed by other processes operating in the warming direction. Whether we can call these positive feedbacks is uncertain, but they are not net negative cloud feedbacks.

      For more on cloud changes, the ISCCP data (see also Part 1) and HIRS data will give you a sense of the observed changes.

      • Fred, “I would say that the net positivity of water vapor feedback is now well established.”

        I would say that once natural variation establishes a mutually agreeable “trend” then the issue of water vapor non-linear feedback can be discussed.

      • For some time I have been trying to point out that the frame of reference you select is important for evaluating feedbacks. The heat pipe analogy is not a surface to tropopause direct heat transfer path. It is a cyclic path between the mid to upper troposphere. Moisture is the refrigerant required to transfer the sensible heat between the mid to upper troposphere and the efficiency of the process varies with the amount of moisture, degree of turbulent flow and the temperature differential between ends of the heat pipe. Natural climate oscillation move heat from one location to another where the efficiency of the heat pipe varies resulting in natural climate change. An unforced variation with varying degrees of efficiency.

        Land use impact that changes available water changes the efficiency since land surface warming provides more turbulent vertical air mixing.

        Perhaps we should have an atmospheric heat pipe thread?

      • Although slightly out of date, this year 2000 paper by Held and Soden analyzes water vapor feedback in some detail. The observational data since then have not greatly altered theconclusions discussed there, including non-linearity from feedback interactions and the effects of regional inhomogeneities. Elsewhere, the authors have provided evidence that within water vapor feedback itself, an increase in water vapor will cause the feedback to increase in a supra-linear fashion, but only at very high concentrations.

      • Fred, I’ll read that. Since my computer is finicky about spread sheets, if you can, compare the mid troposphere to the stratosphere (I used RSS but both UAH and RSS should be compared). What I did was plot the 82 to 0 (NH), mid-trop, strat and mid-trop minus strat to see if there was a trend relationship. I was expecting to see an inverse relationship. It is interesting that when the mid-troposphere cools slowly, the stratosphere cools rapidly. The general trend is warming tropo and cooling strat as expected. The slope of the longer term trends change fairly rapidly around 1994 and the odd direct relationship between mid-trop and strat temperature change becomes more pronounce. That is what started my looking into the atmospheric heat pipe possibility.

      • Fred, I agree with what you say. However, because so many things happen at the same time in the atmosphere, it isn’t quite right to observe that if, as you say, “low clouds have either remained stable or decreased slightly, while the low cloud/high cloud ratio has declined,” that the Caldiera et al process must not be operating.

        From the 1980s through about 2000, for example, world sulfate concentrations declined, mainly because of the fall of the FSU and the huge decline in sulfate when very inefficient, and uncontrolled, factory and electric generation emissions fell sharply when so many inefficient state factories closed around 1990 and beyond. Less sulfate, less low level clouds (since sulfate is potent in cloud formation). So the Caldiera effect might have been operating in the opposite direction, but might have been overwhelmed by the sulfate effect, for example.

        It does still seem to me that what the authors have suggested — in a modeling exercise — is a phenomenon that certainly verges on negative feedback, although the authors understandably don’t use the word. Don’t forget that the IPCC acknowledges that clouds formation and effects are the weakest areas of GCMs. I don’t think we cast this theoretical work aside just because, in a climate world where many facets are changing at the same time, we don’t observe this theory’s manifestation.

      • Water vapor is a positive feedback, no question, if everything else is equal. But if everything else is not equal — if added water vapor causes increased cloudiness, which in turn causes a loss in energy to space due to increased total cloud albedo — that is water vapor still a positive feedback? Isn’t that the question this study raises?

      • John, Water vapor is definitely a positive radiative feedback. During el nino years the mid tropo minus strat is a nearly perfect example, similar amplitude and a slight delay, even with the noisy data. Other times the correlation changes to out of phase. Even with poor quality data I would not expect a phase shift between the two if radiative forcing and feedback were the only consideration. That looks like negative feedback under certain conditions. In general, clouds have a negative radiative feedback, about 20 W/m^2, with low level clouds generally positive and high level clouds slightly negative. So a shift from low level to high level can definitely be a positive feedback. So my question would be if positive cloud feedback is dependent on the rate of change in radiative forcing? With slower change, the feedback looks to be negative. The temperature records do show more rapid rises with slower declines in the north hemisphere. The heat pipe analogy with varying efficiency seems to match that behavior.

      • John – Briefly, I don’t think the positive water vapor feedback is at issue (see for example the Paltridge thread). Caldeira’s group doesn’t dispute this in the paper, but if you have any doubts, you will probably be able to contact him for his answer.

        Cloud feedback is a separate issue from water vapor feedback,even though cloud formation starts with water vapor, and is associated with more uncertainty. I think your point that aerosol reductions may have affected low (and high) cloud amounts is valid. However, the aerosol reductions from the late 1970s to the early 1990s overlap but don’t match the low cloud reductions that continued in the 2000s. Overall, these are more consistent with a positive long term cloud feedback than a negative one. They don’t prove either, but tend to exclude any strong negative feedback long term. Cloud feedback to short term climate fluctuations such as ENSO is probably different in magnitude, and perhaps even in sign, from long term feedback resulting from changes in atmospheric constituents.

      • Fred said, “John – Briefly, I don’t think the positive water vapor feedback is at issue (see for example the Paltridge thread).” and that is correct IMHO. The issue is the overall atmospheric water impact. I don’t think you can separate water vapor from phase changes in the atmosphere to determine the net impact.

    • I have just put up on my blog a paper challenging the universal tendency of IPCC models to report positive water feedback. I propose that they would not generate positive feedback if they did not mishandle vertical shifts in air density.

      In a GCM, the layers are restricted by the hydrostatic approximation from exchanging mass. Air mass cannot move up or down in the models. The most that can happen in terms of changing density is a local expansion of a given layer, in strict proportion with temperature. That is, if the air gets warmer in a layer, density can only go down.

      But in the real world, air mass is subtracted at sea level and winds up being added in the upper third of the troposphere. Warming at sea level leads to a displacement from lower layers to higher ones, over long periods of time, that a hydrostatic model cannot accommodate. There should be a density feedback “signal” propagating from layer to layer, but it is ruled out by the method of computing densities. The hydrostatic relation (rho) * T = constant makes it impossible for any layer to warm AND grow denser at the same time.

      In short, I suggest positive water vapor feedback is an artifact of modeling errors.

      The paper is here: http://declineeffect.com/?page_id=189

      I am not a climatologist, just someone with training in nuclear power systems who has some experience with complex simulation code and the physics of moist air. But the issues here seem to me astonishingly basic. Please take a look at the paper and decide for yourselves.

      • Dean – I realize you’ve put much effort into this paper, but you would have been better off if you had first informed yourself more about basic geophysical principles, including those utilized in climate models. There are a rather large number of errors in your analysis, and it would be a considerable effort to address all of them, but it’s useful to point out that you start with a serious misconception on which you premise much of your subsequent discussion.

        Contrary to your assertions, convective vertical displacement of air as a function of changing temperature (and to a lesser extent humidity) is one of the most extensively addressed features of climate change in models as well as in basic formulations. In many cases, but not all, it involves the upward movement of less dense surface air to a higher altitude, reducing the density at the higher altitude below that which would have persisted without the convection. This is not the only form of convective vertical mixing, but it is one of the common ones. In any case, vertical movement of air masses is well recognized as fundamental to climate physics rather than being ignored.

        Just to mention one other misconception, when the atmosphere is warmed, it expands, and this can reduce density at every level without a change in mass. This is certainly true of the troposphere, whose height (i.e., the altitude of the tropopause) rises with warming. I’m not sure that changes in the stratosphere are as well defined.

        It’s commendable for you to decide that you should examine these phenomena from your own perspective without unthinkingly accepting conventional wisdom. Unfortunately, in a complex field that you admit you are not very familiar with, this is likely to lead you astray, as it has in your case. Climate scientists, like other scientists, are not infallible or omniscient, but they are smart, experienced, and knowledgeable, and the probability that they would have overlooked phenomena as fundamental as the ones you suggest should have been considered by you to be remote. They haven’t.

        I’ll finish by acknowledging that it’s also commendable of you to ask for opinions rather than dogmatically asserting the truth of your conclusions. That spirit should guide you to a better future understanding of climate dynamics.

      • Fred,

        Thank you for the cordial response. I appreciate the positive tone of your post enormously. I have to disagree on several points, however.

        First, I am aware that convective vertical displacement of air is well-studied in climate science. The entrainment / detrainment of air mass in updrafts and downdrafts is modeled in elaborate detail in models such as the CAM3. However, these mass calculations only serve as a basis for estimating energy and humidity transfers. They are, so to speak, “virtual” and are discarded. Dry air mass is not allowed to be transferred vertically between layers in the CAM3, as stated on page 90, Collins et al (2004):

        “Because there is an explicit relationship between the surface pressure and the air mass within each layer we assume that water mass can change within the layer by physical parameterizations but dry air mass cannot.”

        I passed over the issue briefly in the paper. Perhaps it would help to cite documentation and explain the entrainment parameterization specifically. In any case, to my knowledge, all GCM’s resemble the CAM3 in this regard.

        As for density going down in every layer, I expect to hear that response from many climatologists before this debate is over. I urge you to sit down with a spreadsheet and try plotting before-and-after density curves in height coordinates. You will find density must rise somewhere.

      • Dean – As far as I can tell, the CAM3 part you quoted on effects within layers doesn’t contradict the principle that models address convective movement of air masses over large vertical extents, but you didn’t cite an exact reference or a link to one, so I couldn’t review the context. You may have overinterpreted the passage you quoted. It seems to me that convective movement of air masses between layers is a prominent feature of most models, involving complex parametrizations in an attempt to simulate the process accurately – one example is found at convective parametrization. There are many other examples in which models address convective updrafts of air masses, because this is a standard feature of radiative/convective equilibrium. It involves something other than a simple expansion of individual layers, but rather a movement of warmed air parcels in a lower layer up through the atmosphere to a higher altitude while mixing with higher layers in the process. It is this mixing that makes the troposphere what it is, in contrast to the stratosphere, where convection is inhibited, and each layer adjusts to climate change without much mixing. The stratosphere is dominated by the absorption of UV by ozone, and its dynamics are quite different. Most radiative effects involving the warming potency of ghgs involves the troposphere, with some adjustment for stratospheric changes.

        Regarding standard assessments of density in a warming atmosphere, you state “density in the models is going down everywhere, even though this makes no sense physically”. Physically, it does make sense in an expanding atmosphere (the same number of molecules in a larger volume), but it depends on how we define “everywhere”. When all parts of the troposphere are warming, density is going down in all parts in the absence of some extraneous addition of molecules into a volume that is not permitted to expand (or some other localized inhomogeneity). However, I should have stated this more precisely by pointing out that this applies to all air parcels that are warming rather than altitudes, making clear that since these parcels are also rising to higher altitudes, this reduction in density will also involve a change in altitude, because the previous altitude is being replaced by air from below (which is also less dense than before the warming). In essence, though, there need be no transfer of density from one part of the atmospheric mass to another based on simple principles of physics.

        I believe there are quite a few other mistakes in your article, and I cited the ones I did for illustration. I’m not trying to dodge the issue by not addressing every one of them, but it’s simply a matter of available time. I think you have underestimated the ability of climate science to consider all possibilities consistent with atmospheric physics, as well as their ability to measure some of the variables, including temperature, pressure, and humidity at different altitudes and latitudes. There are undoubtedly areas that the science doesn’t yet have quite right, but that’s because the deficiencies are subtle. Obvious possibilities have not been overlooked.

      • Just to illustrate the point about a density reduction in all air masses in a warmed, expanding atmosphere, consider a simplistic example of an atmosphere exactly 10 Km high, with no molecules above that height.

        If we now warm the atmosphere so that it expands to 11 Km, the altitude between 10 and 11 Km will have a positive density rather than the previous density of zero. I don’t think this would be disputed within the climate science community. However, all air masses within the expanded atmosphere will be less dense than before (although sea level density will not decrease as much as in your figure because of convective mixing). In the real atmosphere, the troposphere does expand and become less dense with warming. I’m not sure exactly how this affects density at different altitudes in the stratosphere, but this should have little influence on the basic greenhouse mechanism.

      • I’ve looked at the Collins et al technical paper on CAM3. To my eye, it addresses in detail the large scale vertical convective movement of air masses through multiple layers, with mixing. Emphasis is on moist convection because of its importance both for latent heat transfer and cloud formation, and because tropical air is moist, but dry convection is also mentioned. The section starting on page 75 includes much of this, but there are other mentions throughout the text.

      • Fred,

        First, sorry for not replying directly to your last, the blog logic won’t let me nest my reply there.

        I appreciate your reviewing the CAM3 documentation and making a detailed reply. The language in the documentation takes some getting used to, and it does creates the impression that mass is being moved. On page 76 right after equation 4.6 it says, “Mass carried upward by the plumes is detrained into the environment … ” which sounds like they truly do follow through and alter the mass balances. But they don’t. There is an elaborate simulation of different cloud types rising from different levels of origin, then depositing moisture and energy in higher levels. The simulation requires mass values, and the documentation refers to those mass values as if they really do move, but then at the end, on page 90, there is a clarification which I quoted which confirms that only the energy and moisture gets transferred, not the dry air mass.

        Here’s the sentence preceding the one I quoted. “At the end of the dynamics update to the model state, the surface pressure, specific humidity, and tracer mixing ratios are returned to the model. The physics update then is allowed to update specific humidity and tracer mixing ratios through a sequence of operator splitting updates but
        the surface pressure is not allowed to evolve.”

        In other words, the elaborate mass flux calculations (from the physics update) don’t actually move dry air up or down. The CAM3 outline gives the equation by which the dry air mass is computed in the next time step, right in the same paragraph on page 90, and none of the mass flux terms defined from page 75 onward appear in that equation.

        Now as for density going down everywhere, I notice that your example gives the atmosphere a fixed upper boundary, above which there is zero atmosphere. But you must know that’s not physically correct! The atmosphere thins out asymptotically toward zero, but doesn’t reach zero even at hundreds of kilometers of altitude.

        The monotone decreasing shape of the density distribution is such that if you reduce sea level density, you can only maintain a constant total by increasing density in the upper atmosphere. This is shown step by step in my paper, starting with the International Standard Atmosphere as a basis, then assuming warming at sea level. We have to debate this issue using the actual atmospheric density curve, not an example in which the atmosphere abruptly transitions to absolute vacuum at 10,001 meters altitude.

        Finally, let me say that I’m not underestimating the climatologists. They’re really smart and dedicated guys. But even the smartest scientist can go down the wrong road based on a faulty but widely held assumption. Then an outsider like myself is their best hope to discover their error.

      • Dean – I’m afraid you’ve misinterpreted the models, which all involve the principle of movement of air masses up and down over extensive vertical distances, with turbulent mixing as part of the process – this is one of the most fundamental principles of radiative/convective equilibrium and it involves real movement, not “virtual” movement. If you have doubts, consult a geophysicist and/or climate modeler. There are many other problems with your analysis in my view, but my sincere suggestion is that you begin to acquaint yourself better with what both the models and the basic principles of geophysics actually say. I wish you well, and I appreciate both your intelligence and independent-mindedness, but your conclusions are simply quite wrong (for many reasons), and I think you should become aware that you can’t reinvent a complex field of science until you understand what it says.

      • “I have just put up on my blog a paper challenging the universal tendency of IPCC models to report positive water feedback. I propose that they would not generate positive feedback if they did not mishandle vertical shifts in air density.”
        In regards to feedback I would suggest wattsupwiththat:
        But in general, I don’t think there is much mixing in tropical regions- though of course, this is where hurricanes form and hurricanes are severe examples which is counter to this- but hurricanes are large scale affect which somewhat freakish rather than norm.
        And tropical regions are where most warming occurs on the planet.
        Routine predictable winds occur near ocean and land areas- a daily cycle occurs in such locations most of the time.
        The tropics generally have a high humidity, and there is a limit to humidity which commonly reached in the tropics. And you can’t amplify this affect- you can’t get 101% humidity.
        In the tropics you vast areas of ocean and variation is a cycle water evaporating turning into clouds and raining- in daily cycles. These cycles similar to onshore and offshore winds can be disrupted- but there are the norm.
        Since most of sun energy is going to tropics and since in close to 100% humidity a lot of the time, one can’t get amplification by increase water vapors in these region. So dominate regions in which heat is generated on earth can not have significant increase in the greenhouse affect.

        If you could somehow increase the transport of warm wet air and warm ocean from the Tropics one could increase average temperature in Temperate Zones.
        I think the only way Earth could get significantly warmer is changing oceans and continents are on the Planet. And there now way to get anywhere vaguely close Venus like condition on this planet. {put Venus in Earth orbit- and you will get “runaway affect of cooling” which could result in a ice cover planet a few millions of years. You can’t get much hotter than the tropics and temperate zone can’t get near the average temperate of the tropics- the sun dominate earth’s temperature.

  17. Leonard Weinstein

    The statement “The Earth has been getting warmer over at least the past several decades, primarily as a result of the emissions of carbon dioxide from the burning of coal, oil, and gas, as well as the clearing of forests.” is again used as if this is a given. This misuse is a big part of the problem. We DON’T KNOW the cause of the warming or cause of the present level to cooling trend enough to make such positive statements. I wish the authors had enough logic to understand this. It may be the Sun, it may be long term ocean currents, or a mix of these and general natural variation, or it may in fact be CO2 and deforestation. However, we just don’t know.

    • Good point, Leonard. But do we know who wrote the press release? Maybe it wasn’t the authors of the paper.

  18. Nominally, it is the sun, stupid. That much is a given– like gravity. There is no such thing as climate ‘change.’ There is only climate. We do not have a problem with the language. We have a problem of a dishonest use of language by scientists. The null hypothesis of AGW theory — that all global warming is natural — has never been rejected.

  19. Judy,

    The paper by Ban-Weiss et al. is very interesting, but I wonder the results especially their analysis about low cloud might be model-dependent because it is well-known that the low-cloud feedback in CAM is negative ( see the figure in http://www.atmos.washington.edu/~breth/CPT-clouds.html ) while in GFDL and some other models the low-cloud feedback is positive. It would be interesting if the authors repeat their simulations with CAM4 or CAM5.

    • Jianhua, they did a very specific experiment of increasing latent heat flux and decreasing sensible heat flux by the same amount. This does not mimic any climate projection as they did this to the current climate. So, I think it says nothing about feedback. In a real climate change, both latent and sensible heat flux would increase.

      • Jim,

        Thanks. I agree with you that they are talking about climate projection.

        Ban-Weiss et al’s experiment is meaningful in that the experiments mimic the change in land surface ( an extreme example might be the change from land to aqua-surface; another example might be reforestation). The results in the experiments may be considered as the response to the land use forcing so the feedbacks still work in the experiments though different from the ones in 2xCO2 experiments.

        You said ” In a real climate change, both latent and sensible heat flux would increase.” In fact, in all of climate simulations the latent heat flux increases and sensible heat flux decreases over global ocean. I had a paper on the topic and might of interest to you ( http://journals.ametsoc.org/doi/abs/10.1175/2008JHM1058.1).

      • sorry typo in the last post, it should be:

        they are NOT talking about climate projection.

      • Thanks for your paper. The change of Bowen ratio over oceans is an important factor. I was not aware it should reduce as the ocean gets warmer according to C-C, and that leads to lower sensible heat flux even over warmer water, which is counterintuitive. I always assumed the ratio was quite fixed near 0.1.

  20. In his video clip Caldeira mentions the observable local cooling effect on New York city provided by evaporation from trees and lakes in Central Park, and goes on to suggest that such evaporation may also lead to a global cooling effect through low cloud formation and consequent reflection of solar radiation. This prompts a question which someone more knowledgeable than me may be able to answer: is there also a cooling effect (possibly global in scale) from the reflection – in part directly back into space – by trees (such as those in Central Park) of near infrared (c. 0.7 to 1.0 microns) solar radiation? And if so has such an effect been measured?

    • Coldish,

      The tails of the incoming solar are interesting. The near infrared and blue to ultraviolet are pretty small percentage wise but then doubled CO2 is only about 1 percent. A couple percent change in UVa would make virtually no noticeable impact on deep ocean temperature but since it penetrates to a greater depth, it may have an impact on the overturning rate. The near infrared is pretty stable, but cloud cover/water vapor percent and cloud/water vapor height would change the small impact. Near infrared is not strong enough to make a measurable direct surface temperature impact, but may vary mid-troposphere air currents, clear air turbulence for example, stimulating the heat pipe.

      • Thanks, Dallas, for your prompt reply. But is the near infrared tail really so insignificant? How opaque is the clear sky atmosphere in that range? How do we know that “..Near infrared is not strong enough to make a measurable direct surface temperature impact…”? Just questions!

      • “But is the near infrared tail really so insignificant?” To me no, it may not be insignificant. Small impacts, synchronized, can have a much larger impact. Water vapor has non-linear aspects. A prolonged solar minimum should only have a 0.1 degree or so impact on climate, but the prolonged impact of .25 W/m^2 in the deep ocean is cumulative. Near infrared is similar. In a warming climate, the increased turbulence mixing of the atmosphere it would have no impact. In a more stable climate, it would amplify the mixing higher in the atmosphere where it would increase the cooling rate of the mid-troposphere. For that to happen, water vapor is required and more water vapor would enhance the cooling. So there a two small, but sympathetic impacts generated by a prolonged solar minimum.

  21. I have been reading though the comments but I do not have enough expertise to contribute to the exchange. I would ask those more knowledgeable to reply to a question however.

    The way technology seems to be developing, it seems fairly likely that hydrogen powered cars will be in our future in the not distant future. What will be the potential impact to the climate of millions of cars pumping out addition water vapor into the lower atmosphere?

    I appreciate your thoughts

  22. If you would like to take much from this paragraph then you have to apply these methods to your won blog.

  23. This paragraph offers clear idea in support of
    the new viewers of blogging, that actually how to do
    running a blog.

  24. If some one desires to be updated with hottest technologies then he must be
    go to see this web page and be up to date everyday.

  25. Very quickly this website will be famous among all blog viewers, due to it’s pleasant articles or reviews