Guest Post by Willis Eschenbach
The TAO buoy array is an array of moored buoys in the equatorial tropical Pacific ocean. Here’s a map showing their locations along with the average sea surface temperature.
Figure 1. Locations of all active and historical TAO buoy sites.
And here is what a typical buoy setup looks like:
Figure 2. Details of the TAO moored buoy and sensor arrays.
I like the TAO buoy data because we can be sure that it is free of urban heat islands, changes in location, instrumentation changes, and many of the other problems that plague land-based stations. It is also measured very frequently, typically every ten minutes. This lets us explore the daily cycles of air and sea temperature, solar radiation, longwave radiation, humidity, and the like.
The eight buoys located on the Equator between 165°E and 95°W have the most data covering the longest time, so I’ve looked most at those locations.
The coldest of these eight buoys is the one at 95° W longitude, near the Americas at the far right of the map above. Average sea surface temperature (SST) there is 23.8°C (75°F). The warmest is the buoy at the other side of the Pacific at 165° East longitude. Average SST there, near the warmest part of the Pacific, is 29.2°C (85°F).
Now, something that I like to look at and consider are the differences between the buoys. The buoys at the colder eastern edge of the Pacific are often different from the buoys at the warmer western edge of thePacific. Here are the daily air temperatures from each of the eight buoys:
Figure 3. Daily cycle of air temperatures at eight equatorial TAO buoys.
Next, here are the same eight buoy air temperature records, expressed as anomalies about their respective means.
Figure 4. Daily cycle of air temperature anomalies at eight equatorial TAO buoys.
Note that the daily air temperature cycles at the warmest buoys (reds) have a very different shape and amplitude than do the coolest buoys (blue), particularly during daylight hours.
I’ve mentioned before that what I like best about science is getting surprised by what I find. Here’s my surprise for today. I got to thinking about what is called “delta T”, usually written as “∆T”. The “delta” means “difference”, and the “T” means temperature. For example, winds on the ocean are often driven by temperature differences.
One of the important ∆T’s in the climate system is the difference between the surface temperature and the air temperature. Over the ocean, the air is generally cooler than the sea surface. When the difference between the surface and the air temperature (∆T) gets large enough, when the ocean gets significantly warmer than the air or the air cools significantly below the sea temperature, we start to see things like cumulus clouds and thunderstorms.
So let me start with the absolute values of the differences between sea surface temperature and air temperature at the eight TAO buoys. I’ve used the same color coding as above. Light blue is the coolest buoy at 95°W, and red is the warmest buoy at 165°E.
Figure 5. Daily cycle of differences between sea temperatures and air temperatures at eight equatorial TAO buoys.
This was the first surprise. The overall difference between the sea temperatures and the air temperatures was not in any order by temperature. The coolest and the warmest buoys had the widest differences between sea and air temperatures … odd.
But the real surprise came when I plotted the delta T values as anomalies around their respective means, as I’d done in Figure 4 above …
Figure 6. Daily cycle of anomalies of differences between sea temperatures and air temperatures at eight equatorial TAO buoys.
How interesting. All along the Equator across the Pacific, from the cold edge to the warm edge, the sea-to-air temperature difference anomaly is just the same in every location—lowest at eight AM, peak at one PM, trough at five PM, peak at six PM, trough at nine PM, highest point at about three in the morning.
Not only that, but the temperature swings have the same amplitudes, to within a few hundredths of a degree. Given that these are eight totally different results from sixteen independent temperature datasets (eight air temperature, eight sea temperature), this is an astounding degree of agreement.
I must say that I do not have any coherent explanation for the afternoon and evening peaks and troughs. All I can conclude from this is that all across the Equatorial Pacific, the daily temperature swings are strongly constrained by some unknown combination of natural phenomena, such that the average daily swings are identical in both timing and amplitude regardless of the average temperature of the location …
Go figure … the joys of settled science.
w.
AS ALWAYS: Let me politely request that when you comment, you QUOTE THE WORDS YOU ARE DISCUSSING so that everyone can understand just what you are referring to. I can defend my own words. I can’t defend your vague restatement of something you think I said. Quote what you are talking about, it’s the only way to refute what someone says.
FURTHER READINGS: Here are some of my previous posts about the TAO buoy data:
The Tao That Can Be Spoken … 2011-08-14
As I mentioned in an earlier post, I’ve started to look at the data from the TAO/TRITON buoy array in the Pacific Ocean. These are an array of moored buoys which collect hourly information on a variety of environmental variables. The results are quite interesting, because they relate directly to…
Pinatubo and the Albedo Thermostat 2011-08-21
I got to thinking that the eruption of Mount Pinatubo should provide a good test case for my theory that changes in albedo help regulate the temperature and keep it within a narrow range. When a big volcano erupts, it throws both black and reflective particles and aerosols high into…
TAO/TRITON TAKE TWO 2011-08-25
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…
Cloud Radiation Forcing in the TAO Dataset 2011-09-15
This is the third in a series ( Part 1, Part 2 ) of occasional posts regarding my somewhat peripatetic analysis of the data from the TAO moored buoys in the Western Pacific. I’m doing construction work these days, and so in between pounding nails into the frame of a building I continue to…
TAO Buoys Go Hot And Cold 2015-06-16
I got to thinking about how I could gain more understanding of the daily air temperature cycles in the tropics. I decided to look at what happens when the early morning (midnight to 5:00 AM) of a given day is cooler than usual, versus what happens when the early morning…
Mr Eschenbach, one thing I have noticed on the Air/Sea Difference Graph is the the lowest one actually goes Negative 4 times, ie the Air is Warmer than the Sea?
Is that correct?
If it is correct it must mean that the Air is not moving the heat away from the surface as quickly as you would expect.
I assume that the Time of day for the readings is the local time of day?
Are all the buoys in very deep water?
As the depth can have an affect.
@Willis
“The eight buoys located on the Equator between 165°E and 95°W have the most data covering the longest time, so I’ve looked most at those locations.”
But map in Fig. 1 shows ten buoys distributed in the Pacific along the equator. Could you mark the buoys that you used? (Yes, I also looked at your ‘TaoTriton Take Two’ reference posting. But it also shows 10 buoys on the equator.)
Questions:
1. Why did you pick two consecutive days? Why not a larger, random sample, to help eliminate any possible bias?
2. Did you investigate any buoy data at other locations? For example, the equatorial buoys in the Atlantic (“PIRATA”) and Indian (“RAMA”) oceans?
https://www.pmel.noaa.gov/tao/drupal/disdel/
Comment:
Looking at your “first surprise” (Fig. 5), the coherence of the dT’s at widely separated locations does indeed suggest that these instruments are somehow in “lock-step”. For example, between 4-6pm, all of the plots seem to do a similar zig-zag dance: down, then quickly up, then quickly down again.
I’m skeptical that we’re looking some global law of nature. Looks more like some local instrumental process or data processing phenomenon.
Looking at more data should reveal what’s really going on.
😐
… one more question, with respect to this statement: “One of the important ∆T’s in the climate system is the difference between the surface temperature and the air temperature.”
3. How did you obtain/compute the “surface temperature”? Looking at Fig. 2, “surface temperature” appears to be measured at a depth of 1 meter.
So, since most IR is absorbed at the surface “skin”, it would appear that readings at 1 meter would be significantly affected by wave action and other flux mixing processes (Stokes Law etc). How could this generalize into the ‘lock-step’ results you observed?
I meant to say ‘Navier-Stokes Law’, of course, which generalizes and characterizes flux and motion within viscous liquids. :-]
Hi mr Eschenbach could it have something to do with the air humidity, at a certain point energy will be lost in evaporating mowe water to keep the humidity at a certain level, before the temp can rise again ?
Yes Jan I agree could the peaks and troughs be due to water vapour. early in the morning the temperature at the sea surface would warm up uninhibited until there was sufficient heat to increase evaporation and humidity then as the sun moves higher the temperature would then increase at the same time as evaporation. When both of these peak there would be the rapid drop off as intensity of the sun decreases and precipitation and cloud cover increase. In the later part of the afternoon the suns intensity reduces, the clouds disperse and there is a period where there is increased surface warming.
Such a constant probably does have a simple answer (applying Occams Razor).
I hasten to add my opinion here is pure conjecture from a lurker but mostly based on what Willis has said before. However does this temperature data coincide with humidity/cloud cover?
The first thing that I noted was the delta T slope at sunrise (which I take is at hour six for each buoy) illustrated on Fig 3. It seems steeper at the warm end of the pacific. Is this explained (expected) due to salinity differences? Sunrise also seems graphically the max temp slope at all buoys, almost matched by the cooling slope at the end of the day. Second, it seems the cooling slope sets in before sunset (I take to be hour 18)? Is this a misinterpretation on my part of the sunrise / sunset time at each buoy location?
Willis
What is the radiative budget of the buoys located at the equator at 165°E and the radiative budget of the buoys located at 95°W?
If it is the radiative budget that drives temperature, why is there such a big difference in temperature between the buoys on the same equatorial latitude but different longitudinal location?
What temperature would the ocean reach at around 165° East, if evaporation did not act to cap temperature?
If anything the eastern buoys receive more solar radiation (less clouds). The temperature rise is due to the influx of cold water (Humboldt current) in the east which gradually warms up as it flows west.
trade winds—> transport of heat from east to west at the equator with ongoing warming—> warmer SST in the west tropical pacific (IPWP); in the east: upwelling (colder waters from the deep) due to this wind driven transport—> colder SST in the east tropical pacific.
w ==> In your last graph — how many days of data are in the mix? Is this a long-term, consistent pattern? (is the data deltas of monthly averages by hour?) If you are using what you used in Fig 4, then yes — longer-term, repeating pattern.
It is really quite interesting ….
If you dig into it any more, let us know if there is some quirk in the data or if the pattern holds in different seasons etc.
After reading the recent
Willis Eschenbach’s index of articles,
which was so long, that it exhausted me,
there were so many subjects covered,
I was wondering what was left
for him to write about.
And then he found another subject,
and hit another home run today.
The quality of
the science comments
following this article
is another indication
of the quality
of the article.
At this website
a byline that says
“Willis Eschenbach”
means “must read” to me.
Keep up the good work.
This website would not be
the same without all
the Eschenbach articles!
http://www.elOnionBloggle.Blogspot.com
Very interesting. Cyclical input from sun with rapid transfer of heat to surface of ocean via direct e-m radiation, similar cyclical input to atmosphere but with a delay before the heat reaches the surface, outgoing radiation also on both direct and delayed transfers. Add up the 4 sine (approx) waves and solve for steady state but not static equilibrium. Could produce curves like that seen, probably worth investigating, mash up an electrical analogy perhaps with capacitors and resistors and a delay?
It would be most interesting to see the same data for the buoys at 5 degrees north that are right under the ITCZ.
Willis wrote: “This was the first surprise. The overall difference between the sea temperatures and the air temperatures was not in any order by temperature. The coolest and the warmest buoys had the widest differences between sea and air temperatures … odd.
There could be a calibration problem between thermometers measuring SST and air temperature. If these devises were thermometers, they could be calibrated by taking readings in both water and air at different temperatures. The device measuring air temperature also measures relative humidity. So it is possible that these instruments haven’t been calibrated to produce the same reading when air and water have the same temperature. This hypothesis would explain why they record similar daily changes in temperature, but inconsistent absolute differences. Hypothesizing that calibration could be off by almost 1 degC isn’t a very attractive hypothesis. However, these devices have been exposed to the marine environment for more than a decade (?) and may have deposits on their surfaces.
The winds, currents and irradiate all are parallel along the equator. The starting temps vary because of cold water input on the east and warm water pileup on the west. The extremes lie at each end where the temp differences are most “out of phase” with the other inputs. Does this help define what effect temps have on the process? One might expect a balance to be reached halfway across the ocean. The 5C gradient across 11,000km is not much. It may, however, play into defining ENSO “regions”. It would be interesting to see similar data under el nino, la nina and madoku(?) situations. Similar data at 40L would be very interesting because the irradiance would remain the same along the latitude, the winds would be westerlies and more variable, the water temps would be the same (warm west-cool east), but the water would be moving longitudinally rather than latitudinally. Just might say something about the PDO, the development, strength and longevity of semi-permanent pressure systems and related jets etc. Fun article – and great discussion.
Willis wrote: “I must say that I do not have any coherent explanation for the afternoon and evening peaks and troughs.”
Nor do I. However, I’ll note that the buoys are measuring SST are 1 m below the surface of the ocean and air temperature is measured some distance above the ocean. During night time, only the top 10 um of the ocean is losing heat by evaporation and exchanging LWR photons with the atmosphere. Nighttime surface cooling doesn’t reach 1 m below the surface by conduction. It could reach a depth of 1 m by wind-driven turbulent mixing. Or it could begin when the surface cools enough to be become cooler and more dense than the water below. The rapid changes between 6:00 and 7:30 and between 16:30 and 18:30 could be associated with the end and beginning of density-driven convection in the ocean.
Likewise, the effective heat capacities of the ocean and air could be quite different. Again, a lot depends on how fast vertical mixing is occurring in the ocean. During daytime, the air near the surface probably isn’t absorbing an appreciable amount of heat from SWR – all of the infrared wavelengths that water vapor absorbs have been depleted by absorption before SWR reaches the surface. So the air is only warmed by LWR emitted by the top 10 um of the surface of the ocean. Early morning sunlight might raise the temperature of those top 10 um of the surface (and the air immediately above) long before it raises the temperature 1 m below the surface (where some direct heating by absorption of SWR occurs when the sunlight is near vertical).
“Likewise, the effective heat capacities of the ocean and air could be quite different.”
They are. Enormously different as a matter of fact.
“I must say that I do not have any coherent explanation for the afternoon and evening peaks and troughs. All I can conclude from this is that all across the Equatorial Pacific, the daily temperature swings are strongly constrained by some unknown combination of natural phenomena, such that the average daily swings are identical in both timing and amplitude regardless of the average temperature of the location …”
That’s easy. It’s the afternoon wind.
And exactly why would you expect “afternoon wind” out in the open ocean? Diurnal winds depend on local temperature contrasts which are largely absent at sea.
“Diurnal winds … are largely absent at sea.”
Indeed, but not they are not totally absent at sea. Large temperature swings occur at coastlines with diurnal winds. But looking at this TAO buoy data temperature swings are largely absent. But they are not totally absent. It’s only .5 or .6 C.
I think I have some ideas on this, Willis.
It is called the di-urnal pattern. It has been studied for years and there is still no clear description as to what causes it. So let me try.
There are two main layers of our atmosphere, the upper and the lower. The upper which extends into space allows the light from the sun to wrap around the inner layer. Sunlight shines through this layer, which is outside the magnetic field, becomes heated by the utra-violet portion of the light. At around 4am It begins to puff up the upper atmosphere and puts pressure on the lower atmosphere which also increases its pressure. That causes the temperature to drop.
As the day continues this UV heating of the upper atmosphere becomes less once the magnetic field of the earth becomes the blocking force of the UV rays. The normal daily heating begins around 10:30am. And again the temperature begins to rise until 4:pm. After the sun sets, the heating all ends and the temperature drops to a low point to recycle.
This di-urnal cycle does not require the atmosphere to move. The expanding character of the air provides all the action.
And I must say the UV light is invisible. I have noticed that the roosters (cocks) provide a nice detector for UV first light. They start crowing between 3:45 and 4 each morning along with other wild birds.
Lee
Figure 6 has a horizontal line at 0.0 and a vertical at 12AM that intersect at the center of the figure.
I would like to see 2 days of data running… first day standard…second day with a shade that blocks
direct and reflected sunlight from reaching the buoy.
On Figure 4, I would expect to see a parabolic shape between 6hrs and 18hrs through the daylight cycle if there were no dampening due to cloud cover etc between 9hrs and 15hrs. The brief uplift from 18hrs and also between ~3hrs and 6hrs are fascinating, both stronger in the west. A very interesting post, thanks.