Guest Post by Willis Eschenbach
I wrote before of my investigations into the surface air temperature records of the TAO/TRITON buoys in the Pacific Ocean. To refresh your memory, here are the locations of the TAO/TRITON buoys.
Figure 1. Locations of the TAO/TRITON buoys (pink squares). Each buoy is equipped with a sensor array measuring air and sea temperatures and other meteorological variables.
I have hypothesized that there is a thermostatic mechanism involving clouds and thunderstorms that maintains tropical temperature within a certain range. To investigate this mechanism, I decided to look at what happens at a given buoy on days when dawn temperatures are warmer than average, versus what happens at the same buoy on days when dawn temperatures are cooler than average.
My speculation was that when it was warmer at dawn, there would be more cloud and thunderstorm activity during that day. This would tend to drive the temperature down. On the other hand, when it was cooler at dawn, there would be less or no clouds or thunderstorms during that day. As a result, this would tend to drive the temperature upwards. And while I did find this, I was still surprised by the exact patterns.
To begin with, I compared the overall average of all days for each station with the overall average of the warmer days for each station, and the overall average of the cooler days for each station. Here are those results:
Figure 2. Average of all buoy records (heavy black line), as well as averages of the same data divided into days when dawn is warmer than average (heavy red line), and days when dawn is cooler than average (heavy blue line) for each buoy. Light lines show the difference between the previous and the following 1:00 AM temperatures.
First, the black line, showing the average day’s cycle. The onset of cumulus is complete by about 10:00. The afternoon is warmer than the morning. As you would expect with an average, the 1 AM temperatures are equal (thin black line).
The days when the dawn is warmer (red line) show a different pattern. There is less cooling from 1AM to dawn. Cumulus development is stronger when it occurs, driving the temperature down further than on average. In addition, afternoon thunderstorms not only keep the afternoon temperatures down, they also drive evening and night cooling. As a result, when the day is warmer at dawn, the following morning is cooler.
In general, the reverse occurs on the cooler days (blue line). Cooling from 1 AM until dawn is strong. Warming is equally strong. Morning cumulus formation is weak, as is the afternoon thunderstorm foundation. As a result, when the dawn is cooler, temperatures continue to climb during the day, and the following 1AM is warmer than the preceding 1 AM.
So this is the result that we would expect with a thermostat operating on a daily basis. If the dawn is warm, clouds and thunderstorms ensure that the following day starts out cooler. And when the dawn is cool, extra sun and few clouds and thunderstorms warm the day up, with the warmth lasting into the night.
Now … is this just a statistical oddity? One way to determine if we’re looking at a real phenomenon is the “dosage effect”. That is to say, the response should be proportional to the dosage. In this case, the “dosage” is the overall average temperature for that particular buoy. My hypothesis says that the effect seen above in Figure 2 should be greater in those buoys where the average air temperature is warmer, and less in those buoys where the air temperature is lower. And indeed, that proved to be the case, as is shown in Figure 3. This shows the buoys divided into four quarters (quartiles) on the basis of annual average temperature.
Figure 3. Differences between warm days (red line) and cool days (blue line) for the TAO/TRITON buoys divided into quartiles by temperature. Black line is average for all days.
Note that the response systematically grows larger and more exaggerated as we go from the first quartile (the coolest quarter of the buoys) sequentially to the fourth, warmest quarter of the buoys.
I hold these results out as strong support for my hypothesis that the temperature of the tropics is regulated by the combined action of clouds and thunderstorms. The difference in the temperature response of the warm and cool days shows the homeostatic mechanism in action, with warm mornings having cooler afternoons, and vice versa. All of this shows the clouds and thunderstorms at work.
I will ask again that if you disagree with something I’ve said, please quote it so that we both know what we’re discussing.
All the best,
w.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
This looks to me like just regression to the mean. If a day is unusually hot then it is more likely that it will cool down to a more common normal temp day, and if a day is unusually cool then it is more likely that it will warm up to a more normal temp day. Hot days are more likely to cool, and cool days are more likely to warm. I see no need for any thermostatic principle to explain it other than just settling to the average after disturbances.
It could even be partly a simple result of the progress of the seasons. If you start out at 1AM on the average hottest day of the year, then on average, 1AM on the next day is going to be cooler. This will continue until the average coolest day of the year, where the next day will on average be warmer.
This would be more obvious if the graph were plotted so that instead of how they’ve been adjusted to coincide at the beginning of the day, they were instead adjusted to coincide at the end of the day.
MindBuilder says:
August 25, 2011 at 9:57 am
I’m not sure what you are calling “regression to the mean”, or how you distinguish it from a homeostatic process.
w.
I appreciate the comments pointing out that near the equator there is virtually no variation in time of sunrise and sunset. With the resolution of data Willis is working with, then, my suggestion is not really necessary. I would however like to know what the picture looks like for the other tropical oceans besides the Pacific. Likely they will be similar, but it would still be interesting.
Any or all of this may be wrong so feel free to correct it.
The mainstream view is that temperature is controlled by radiative forcing which can be approximated to a linear relationship. But that doesn’t mean that temperature slavishly follows RF just that it averages out over a long time.There are other processes that can lead to temperature moving away from this linear relationship on different time scales (such as ENSO). So you have temperature ‘trying’ to be in equilibrium with radiative forcing but never quite getting there. Either being pushed higher (or lower) by multiple mechanisms.
In this scenario any higher than average temperature induced by non-radiative mechanisms would more likely be followed by lower temps as the system “tries to return to equilibrium’. And the reverse if the temperature is lower. This is what you have found on a daily basis. So I don’t necessarily see that your result distinguishes between your mechanism and the mainstream position.
Some have commented above about the need to adjust for sunrise/sunset times. There are a couple of other corrections that are of the same magnitude or greater.
The array in the eastern pacific is spaced at 15 degrees of longitude, so using the zone time for each works well. For the western half of the array the longitudes are not on the same grid and local mean noon will have +/-30 minute offsets from the mean noon at longitude of the center of the time zone. So one easy correction to get things to correspond better would be to use a fixed correction for the difference between the center longitude of the time zone and the longitude of the buoy.
The other correction is the Equation of Time; the difference between actual sun longitude and the longitude of the mean sun. +14 minutes to -16 minutes over a year period. It can be expressed neatly as the sum of two sine functions. One has a period of one year, the other a period of 1/2 year.
——————————
This is a very interesting, intriguing, though-provoking series of posts.
I guess the wider question is whether Willis mechanism can induce long term trends. Whether he has identified a processes that maintains temperature around a sort of equilibrium or optimal temperature on a daily basis. But that equilibrium or optimal temperature is constantly being changed by the change in radiative forcing.
By regression to the mean I just mean that days that are warmer than normal tend to cool towards normal and days that are colder than normal tend to warm towards normal.
A homeostatic process implies something in the system that resists external forcings other than mere thermal inertia or settling to the average resulting from the long term balance of incoming and outgoing radiation. This tropical thunderstorm/cloud cooling effect is sort of homeostatic, but the graph you’ve presented today which shows that warm days cool and cool days warm, is kind of a separate issue. It doesn’t establish the homeostatic nature of the thunderstorms.
I doubt that the thunderstorms are really homeostatic anyway. It looks to me more like they are reacting to differences in temperature rather than stabilizing towards any particular temperature. Perhaps a difference between the temperature of the water and the air, or a difference between the air temp at low and high altitudes. In one of the previous analysis it was shown that this thunderstorm cooling happens at a wide range of buoys that have considerably different average temperatures. So I suspect that if the entire globe warmed up one degree, for example from a steady long term increase in solar output, these thunderstorms may resist the change at first, but after a while the new average temperature around a buoy would settle in, and then these thunderstorms would resist changes away from the new average, even if the change was back toward the old average.
Willis said in comments:
“the balance is restored by the fact that there are more warmer-than-average days than colder-than-average days”
So that says that the magnitude of the colder than average days was greater than the magnitude of the warmer than average days, which is evident by the sharp point on the bottom of the morning temperatures on the cool days. If, in stead of dividing the days according to deviation from the mean, you divide them so that there are an even quantity of warm and cool days, does that change the results of the analysis? Do you still see the increasing trend when you break the data up into quartiles?
Another question:
How many days follow this pattern, and how many days stray from this pattern? What is the probability distribution of any given day fitting the pattern you suggest? Does that distribution say anything informative about the strength of the pattern you suggest? How statistically significant is it?
this is win, willis. 🙂 wb.
The more I hear about the governor/thermostat approach, the more I like it. It seems a lot more intuitive than the “cycle” approach in regard to what drives the weather and climate.
With the governor approach, the weather/climate makes its own cycle in the sense that it gets out of whack, and through its own in built mechanisms, it corrects itself, and because it does this over time, it looks like a cycle.
It also shows that the Earth is a very resilient (robust) planet, in that it can accommodate a wide range of extremes, and still have a mechanism to bring itself back to/through equilibrium.
It’s probably just me, but since Willis elucidated the idea, I’m seeing thermostats everywhere. It’s govern(ors/esses) all the way down to the bottom.
For instance, it seems that open water at the poles dissipates a lot more energy than it can possibly get at those latitudes, so counterintuitively, low sea ice extent ends up leading to a cooler Earth (ergo is a cooling mechanism).
Stephen Wildes’ “iris” effect would be another one that springs to mind.
I’ll thank the alarmists for one thing (and one thing only), and that is that in their disasturbating, they’ve made me open my eyes and realises what an amazing place this planet is.
/p
Imagine how different Earth would be if all the land mass in the northern hemisphere was centred on the equator.
Your analysis is really about how cloud formation responds to the local temperature to moderate variations. Since everyone is talking about the CERN cloud experiment, have you looked at different periods of time such as near or just after a solar maximum (about 1999-2001) or near or just after solar minimum (such as 2007-2009). Do clouds form more easily and are they more persistent just after solar minimums?
HR says:
August 25, 2011 at 10:19 am
The mainstream position is that there is no “equilibrium”. Instead, their position is that everything is linearly ruled by forcing.
But the solar forcing is increasing and decreasing throughout the day, and despite that the temperature response varies depending on both the temperature and the time of day. The net effect is that the system does not passively “return to equilibrium”, it actively restores the “preferred” temperatures.
In addition, if (as I show) the response is based on temperature, then there is no climate sensitivity as it is currently understood.
w.
To R Gates – where are you when we need the voice of sanity?
Sean says:
August 25, 2011 at 1:43 pm
Interesting, Sean. Hadn’t thought about that, but there might be something to be seen there. The TAO/TRITON buoys return a host of hourly information on things like wind, relative humidity, sea temperatures at various depths, and the like. I’ve barely scratched the surface of the data.
paulhan says:
August 25, 2011 at 1:42 pm
First, thanks, paul. I’m glad someone out there sees the beauty of the way nature does it. I agree with you that there are a variety of homeostatic mechanisms, at a variety of temporal and spatial scales, that work in unison to modulate and control the temperature of the planet. Discovering what those mechanisms are and how they interact and intersect should be the job (and the joy) of the climate scientist, not looking for fancied linear relationships.
w.
Gary Swift says:
August 25, 2011 at 12:31 pm
The quartiles, as the name suggests, are broken up the way that you suggest—equal amounts above and below the median.
Gary Swift says:
August 25, 2011 at 12:35 pm
And good questions all. I have no answers. It’s early days yet, and there’s a monumental pile of unexplored TAO/TRITON data out there. I report things as and when I find them. To answer even a simple question (what is the difference when dawn is warmer versus when it is cooler than average) takes hours. New code has to be written and debugged, and I don’t speak R as fluently as I might wish. Then plot the results. Then slice it in quartiles. Plot each result with a different title. It’s a slow process.
So, I just send back periodic reports from the scientific battlefront as time and the tides allow.
All the best,
w.
I agree with Mindbender on this. If a day begins hotter than a mean, then there is a smaller probability that it will end hotter still, than there is it will end cooler. There may be many reasons for this regression toward the mean, one of which could be the hypothesized thermostat.
sorry, Mindbuilder in the previous post, not Mindbender!
Willis Eschenbach says:
August 25, 2011 at 4:59 pm
And thank You, Willis. You’re among the top five people that have really opened my eyes. I always look forward to your postings
/p
Hi Willis this is most interesting.
From your write up and from subsequent comments, I can see two possibilities:
(1) you have found a true homeostatic phenomenum.
(2) your results ae just due to random fluctiations.
I note your comments about the practical difficulties of doing such analysis.
However, I would like to see the data combined into sets of say ten days duration.
If there really is a daily homeostatic reaction at work, then it should persist – if it starts high then it will end low which creates a low start for the next day and so forth.
If not, then the trend will quickly evaporate.
I tend to suspect the latter – that what you have got is just a random selection of chaotic data.
If I’m wrong and you’re right, then you are on to something big which is not at all recognised in the IPCC style modelling.
So I feel that it would be worth the effort to extend your analysis as suggested.
I’d do it myself, but my skill at R is far less than yours (close to zero in fact Pr 0.000001 in reality).
Fascinating stuff anyway and thanks for sharing it with us.
AusieDan says:
August 25, 2011 at 7:35 pm
AusieDan, thanks for the thoughts. If it were a “random selection of chaotic data”, the tendencies would not vary by quartile. But they do vary by quartile (Figure 3), and in the expected direction—the warmer the average temperature at the buoy, the more separation and difference there is between the warm dawn and cool dawn days.
w.
Kevin Kilty says:
August 25, 2011 at 5:29 pm
Kevin, thanks for the thoughts. If this is just a regression to a mean, then how do you explain the difference between the shape of the temperature rises on the warm dawn and cool dawn days? (See lower right panel, Fig. 3)
Obviously, on the two types of days there are very different factors shaping the evolution of the daily temperature. When it is warm, cumulus reverses the temperature rise of the morning, and then thunderstorms knock down the expected afternoon peak temperatures to the point where the peak temperature actually occurs before noon.
And on cold days, there is little interference from clouds or thunderstorms, so the temperature rises throughout the day, with the afternoon much warmer than the morning.
Which is why my hypothesized mechanism is a better explanation than some relaxation towards a mean …
w.
PS–I got to thinking about this claim:
Curiously, there’s no reason that should be true.
Consider a hypothetical temperature record where every day has a 50/50 chance of being either warmer or cooler than the previous day. This kind of a path is sometimes called a “drunkard’s walk”, meaning there’s equal change the drunk temperature will stagger warmer or cooler.
So we might end up with a daily record like 10°, 10.2°, 10°, 9.7°, 8.5°, 9.2°, and so on. These might have a mean of say 9.8° or something. Now here’s the question.
Is it true that if a day is hotter than the mean, it is more probable that the following day will be cooler?
Well … absolutely not. Remember that despite the fact that it has a mean, it is a drunkards walk. At every point, the odds of warming and cooling are equal, 50/50.
So the existence of a mean does not imply a tendency of a “reversion to the mean”. The days don’t know what the average temperature is.
w.
W.;
yes, “reversion to the mean” is just a matter of any current deviation from the true average/mean being swamped and reduced to insignificance by the piling up of future instances which will, by definition, average out to the mean. Since that’s what mean means.
😉