Another Climate Change question: Temperatures at the tropics

[wolfstu]

One of the main political parties in Canada is proposing taxing carbon emissions, as a lead-up to a later cap-and-trade scheme. The national public broadcaster hosted a public question-and-answer forum about it on their website, with the questions answered by a university professor who is a climate-change skeptic.

In answering the questions, he claims that global warming probably isn’t happening as imagined, since we should be seeing disproportionately rapid temperature rises in tropical regions, but this is not being observed.

@Dr. McKitrick

My view is that there are some key model predictions about what the climate should be doing if greenhouse gases were the dominant forcing that we are clearly not seeing in the data. For instance the tropical troposphere should be warming about twice as fast as the rest of the atmosphere (see http://ipcc-wg1.ucar.edu/wg1/Report/suppl/Ch10/Ch10_indiv-maps.html Figure 10.7 for an example of what models predict under greenhouse-induced warming. I don’t see this in the satellite data, which tells me that the IPCC has probably jumped to a conclusion to say that greenhouse gases are very likely the cause of observed changes over the past few decades.

So my questions are:

1. Should we see disproportionately rapid temperature rises in the tropics? Why?

2. Has this been observed? Is there a discrepancy between the model and the observations?

3. Is this guy right to criticize the general scientific conclusion based on this?
   I’d put this in GQ, but it’ll probably end up here, I figure. I don’t know enough myself to guess at this specific question of the tropical troposphere.

[jshore]

We’ve been discussing this a lot in a couple of threads recently. Here are a few important points:

(1) The prediction is not that tropical regions will be warming more rapidly than other places (in fact, the polar regions, particularly the arctic is predicted to warm the fastest). It is rather that, in the tropics, as you go up in the atmosphere, the temperature trend is supposed to get stronger. I.e., there is supposed to be a an “amplification” of the warming as you go up in the troposphere relative to the warming at the surface.

(2) Despite what many skeptics like McKitrick tend to argue, this prediction is not a prediction specific to the mechanism that the warming is caused by greenhouse gases. It is in fact a very general feature in the climate models that can be traced back to a basic piece of physics, what is called “moist adiabatic lapse rate theory”. And, it is not just predicted for the long term temperature trends. The same sort of amplification is predicted to occur for temperature fluctuations at shorter time scales.

(3) As for what the observations show: It is correct that there seems to be a disagreement between models and observations for the decadal temperature trends, at least for some of the observational data sets. However, it is important to note first that this disagreement is only for these long term temperature trends. If you look instead at the fluctuations in temperature that occur on timescales of months to a year or so, then the expected amplification as you go up in the atmosphere is seen. So, there are two possible explanations for this disagreement: either that there is a new bit of physics that the models don’t capture that comes in only once you get to the long (decadal) timescales or there are problems with the observational data sets. There are good reasons to believe that the second explanation is more likely to be correct: First of all, noone to my knowledge has been able to come up with good candidate physics that would cause the tropical atmosphere to behave differently on these longer timescales while still preserving the amplification behavior over the shorter timescales. Second of all, while the observational data sets are expected to be accurate for the fluctuations over the shorter timescales (where they agree pretty well with the models), there are good reasons to be worried about how well the they can capture longterm trends (where they disagree with the models, at least in some cases). Basically, the observation data sets are based on two things, weather balloons (radiosondes) and satellites. Both of these have problems that make them difficult for determining long term trends. For the weather balloons, coverage in the tropics is sparse and there have been changes over time that are likely to produce a spurious cooling trend (namely, better shielding of the instruments from the sun). For the satellite data, there are various subleties having to do with orbital decay, splicing the data from the different satellites, etc. These problems are reflected in the fact that different groups analyzing the same satellite data get quite different trends. This is similarly true for the balloon data.

(4) It is also worth noting the history of this field. When the satellite data was first analyzed by Roy Spencer and John Christy in the 1990s (the available satellite data itself goes back to like 1979), it actually showed the atmosphere globally to be cooling. This became a huge skeptic talking-point, i.e., that global warming did not appear to be happening except for surface observations (which could presumably be affected by urban heat island effects and so forth). However, over time, as the time series grew long but, more importantly, as there were various corrections made to the analysis from other scientists pointing out problems such as the failure to account for the effect of the decay of the satellite orbits, the cooling trend has become a stronger and stronger warming trend. In addition, a couple other groups also did independent analyses of the satellite data and got a stronger warming trend. So now, it is generally agreed (as noted in a recent report by the U.S. Climate Science Research Program), there is no longer statistically-significant disagreement between the satellite and surface data on a global scale. It is only in the tropics that there remains any significant disagreement. So, in fact, what the skeptics are desperately holding on to is basically the last piece of disagreement between these data sets where, thus far, the disagreements have basically all been resolved in favor of the models (& surface observations).

See also, this RealClimate post and the links to earlier RealClimate posts therein.

[jshore]

@jshore

First of all, noone to my knowledge has been able to come up with good candidate physics that would cause the tropical atmosphere to behave differently on these longer timescales while still preserving the amplification behavior over the shorter timescales.

I should note that in one of the thread recently where we discussed this, I seem to recall intention throwing out some hypotheses…but I don’t remember what they were. He would have to remind us and explain to us how these could fix up the disagreement at the longer timescales without messing up the agreement on the shorter timescales. (It is not hard to come up with lots of ideas for mechanisms that might make moist adiabatic lapse rate theory too simplistic. What is harder is finding ones that would still cause the data to agree with the theory at the timescales of months to years…which are already much longer than the timescales over which the relevant convective processes operate…but deviate for the temperature trends over the timescale of decades.)

[Alex_Dubinsky]

This is on the topic of has global warming already happened?.

It’s a bullshit debate to get into. I don’t know why everyone keeps piling fucking in. The assertion it has already is in truth tenuous. Yet it doesn’t matter because noone, nobody reasonably doubts that global warming will happen. I just don’t understand why the discussion keeps repeating itself about this more uncertain, and not even all that important, question.

[MrDibble]

@wolfstu

3. Is this guy right to criticize the general scientific conclusion based on this?

jshore’s dealt nicely with the meet of things so let me just address this:
No, he isn’t right to criticise the general conclusion based on 1 paper whose findings are being handily challenged and basically debunked. Global warming theory doesn’t rest on this one thing, and the process of science is self-correcting. The IPCC report is a comprehensive literature review. One blip doesn’t overturn it. Especially a blip that turned out to be more of a damp squib. I mean, this was dealt with in 2007, it’s amusing that it’s still surfacing now.

[MrDibble]

or the “meat” of things, even.

[wolfstu]

jshore, thanks for the explanation. To aid my own understanding, though, I have a further question:

@jshore

(2) Despite what many skeptics like McKitrick tend to argue, this prediction is not a prediction specific to the mechanism that the warming is caused by greenhouse gases. It is in fact a very general feature in the climate models that can be traced back to a basic piece of physics, what is called “moist adiabatic lapse rate theory”. And, it is not just predicted for the long term temperature trends. The same sort of amplification is predicted to occur for temperature fluctuations at shorter time scales.

So why is that, exactly? I understand what lapse rates are, but why should there be amplified warming at altitude in the tropics? Is this just a question of ‘hot air rises’? Is it because the atmospheric warming ends up being confined mostly to the troposphere, and the troposphere is thicker in the tropics than elsewhere?

[jshore]

wolfstu: The answer to your question is as follows. There are two different adiabatic lapse rates, the lapse rate of “dry” (meaning not completely saturated) air and the lapse rate of “moist” (completely saturated) air.

The dry adiabatic lapse rate follows from the fact that as air rises and the pressure on it drops, it expands doing work on the air around it and thus drops in temperature. It is at least approximately independent of temperature.

However, the moist adiabatic lapse rate includes the additional factor that, as the air cools, the amount of water vapor it holds decreases and so some of the water vapor has to condense out. This releases latent heat and thus warms the air back up somewhat. Hence, the moist adiabatic lapse rate is smaller than the dry adiabatic lapse rate. However, not only is the amount of water that air can hold an increasing function of temperature, such a graph also has positive (upward) curvature. So, the amount of water that must condense out when the air drops in temperature by a certain amount is larger at higher temperature than at lower temperature. As a result, the moist adiabatic lapse rate is a decreasing function of temperature.

And, there you have the basic reason for tropical amplification: Consider two moist parcels of air rising from the surface of the earth that are at two different temperatures. The one at higher temperatures will have a lower lapse rate and hence cools more slowly as it rises, so that when you look at these two parcels of air higher up in the atmosphere, the temperature difference between them will be amplified relative to what it was at the surface.

I’m not completely sure why this effect is more important in the tropics than elsewhere but my guesses are that it is because convection plays such a large role in the tropics. And, the tropics may be the only places where dewpoints are consistently high enough that the difference in the moist adiabatic lapse rates is significant.

[wolfstu]

jshore: Gotcha. Thanks for the explanation. I suppose just the fact that the tropics are quite a bit hotter (and consistently so) than the temperate or polar regions means the lapse rate will be lower and so the high temperatures will persist farther up in altitude. The tropics are also pretty moist, I guess.

[ralph124c]

As the tropical oceans heat up, it is reasonable to assume that the clouds resulting will be denser and larger. As such, will the sunlight falling on tropical regions be reduced? also, as the clouds condense into rain, a tremeondous amount of latent heat is released-where does this heat end up? is it heating up the upper atmosphere?

[MrDibble]

@ralph124c

As the tropical oceans heat up, it is reasonable to assume that the clouds resulting will be denser and larger. As such, will the sunlight falling on tropical regions be reduced? also, as the clouds condense into rain, a tremeondous amount of latent heat is released-where does this heat end up? is it heating up the upper atmosphere?

Clouds are … tricky. They could have a net positive or negative feedback effect, depending on their altitude, composition, etc…
Here’s what the IPCC TAR had to say.

[jshore]

@ralph124c

As the tropical oceans heat up, it is reasonable to assume that the clouds resulting will be denser and larger. As such, will the sunlight falling on tropical regions be reduced?

What MrDibble said. Clouds are the source of the largest uncertainty in the models. Note that it is not even trivial to predict whether cloud cover will decrease or increase with warming (as warming leads to more evaporation but also warmer air can hold more water vapor before condensing) and what the effect of a change in cloudcover will be (as clouds can both reflect sunlight and reflect infrared radiation). High clouds tend to have a net cooling effect and low clouds a net warming effect…or maybe its the other way around; I always forget!

What you are proposing is basically akin to Richard Lindzen’s “iris effect” that he proposed to produce a negative feedback. However, that hypothesis has not gotten much observational support. There was a paper very recently by Spencer and Christy (famous…or infamous…for their analysis of the satellite temperature record) that claimed to find observational support for something akin to it on short timescales but there are lots of reasons to be skeptical of their results, which at any rate are too new to have received much evaluation from their fellow scientists. Also, it becomes difficult to explain things like the ice age / interglacial cycles with our current understanding if one assumes that a strong stabilizing feedback of this sort exists.

also, as the clouds condense into rain, a tremeondous amount of latent heat is released-where does this heat end up? is it heating up the upper atmosphere?

Please read what I wrote in this thread in regards to the moist adiabatic lapse rate theory. The answer is that, yes indeed, such latent heat release is an important source of heating in the atmosphere and is basically responsible for the amplification of temperature fluctuations as you go up in the tropical atmosphere that is predicted by the models (and is also observed…except for the case of the longterm trend at the decadal timescales, depending on which data sets you believe).