Guest Post by Willis Eschenbach
I have put forth the idea for some time now that one of the main climate thermoregulatory mechanisms is a temperature-controlled sharp increase in albedo in the tropical regions. I have explained that this occurs in a stepwise fashion when cumulus clouds first emerge, and that the albedo is further increased when some of the cumulus clouds evolve into thunderstorms.
I’ve demonstrated this with actual observations in a couple of ways. I first showed it by means of average photographs of the “view from the sun” here. I’ve also shown this occurring on a daily basis in the TAO data. So I thought, I should look in the CERES data for evidence of this putative phenomenon that I claim occurs, whereby the albedo is actively controlling the thermal input to the climate system.
Mostly, this thermoregulation appears to be happening over the ocean. And I generally dislike averages, I avoid them when I can. So … I had the idea of making a scatterplot of the total amount of reflected solar energy, versus the sea surface temperature, on a gridcell-by-gridcell basis. No averaging required. I thought well, if I’m correct, I should see the increased reflection of solar energy required by my hypothesis in the scatterplots. Figure 1 shows those results for four individual months in one meteorological year. (The year-to-year variations are surprisingly small, so these months are quite representative.)
Figure 1. Scatterplots showing the relationship between sea surface temperature (horizontal axis, in °C) and total energy reflected by each gridcell (in terawatts). I have used this measurement in preference to watts/square metre because each point on the scatterplot represents a different area. This approach effectively area-averages the data. Colors indicate latitude of the gridcell. Light gray is south pole, shading to black at the equator. Blue is north pole, shading to red at the equator. Click to enlarge
So … what are we looking at here, and what does it mean?
This analysis uses a one-degree by one-degree gridcell size. So each month of data contains 180 rows (latitude) by 360 rows (longitude) of data. Each point in each graph above is one gridcell.That’s 64,800 data points in each of the graphs. Each point is located on the horizontal axis by its temperature, and on the vertical axis by the total energy reflected from that gridcell.
The main feature I want to highlight is what happens when the ocean gets warm. From about 20°C to maybe 26°C, the amount of solar energy reflected by the system is generally dropping. You can see it most clearly in Figure 1’s March and September panels. But from about 26° up to the general oceanic maximum of just above 30°C, the amount of solar energy that is reflected goes through the roof. Reflected energy more than doubles in that short interval.
Note that as the ocean warms, the total energy being reflected first drops, and then reverses direction and increases. This will tend to keep ocean temperatures constant—decreasing reflections allow more energy in. But only up to a certain temperature. Above that temperature, the system rapidly increases the amount reflected to cut down any further warming.
Overall, I’d say that this is some of the strongest evidence that my proposed thermoregulatory system exists. Not only does it exist, but it appears to be a main mechanism governing the total amount of energy that enters the climate system.
It’s very late … my best regards to everyone, hasta luego …
w.
[UPDATE] A commenter asked that I show the northern and southern hemispheres separately. Here is the Southern Hemisphere
And the Northern. The vertical lines are at 30.75°C, nothing magical about that number, I wanted to see the temperature shift over the year and that worked.
@ur momisugly Willis Eschenbach.
Thank you for this well reasoned and polite response.
You claim to do “science”, When people make scientific criticisms of your work, which is completely reasonable as you are publishing your ideas on the web, You never address their questions. .
As a case in point, I an another professional signal processer have questioned your work on a basis that seems to us, based on our training and experience, to be sound.
If you had a genuine scientific interest in the problems that you try to tackle, you would take our remarks much more seriously and try to understand them.
I have reviewed some of your “mathematical” posts. The effects of filtering on correlation, the periodicity transform, your first oder model of heat transfer. The intellectual standard is lamentable. You really haven’t got a clue.
As regards your feedback hypothesis, do you really think that other people haven’t explored this? It is after all, one of the burning topics in climatology and has proved very difficult. You seem to think that you’ve cracked it in about a day. Well, forgive me, I’m sceptical. Any scientist, would trest this claim with caution and the first step is criticism. The demonstration of negative feedback, particularly with non-ideal data, is extremely difficult and I don’t think that you have demonstrated this. I would remind you that if you make these claims the onus is on you to defend them. The purpose of my “model” is to demonstrate that the processing methods involved makes it almost impossible to distinguish between a feedback and a non-feedback system. If you were to attempt this, you would discover this and you might learn something. If you can’t do it, I would suggest that your knowledge is too limited to perform reliable processing and analysis or even to understand the issues that I have raised.
WUWT is worth visiting because there are people who post on it who are genuine scientists and what they write is sensible. You do not fall into that category.
Willis: “Next, I’ve asked Richard to provide evidence that aliasing is an issue in this dataset. I know it’s an issue in theory, and can easily be a problem in certain kinds of analysis … but is it an issue in this dataset for this particular analysis?”
So it doesn’t count unless it Richard that says it , is that the game?
I have pointed to artefacts in the rad variable that are at least 30% across large areas of the data. In fact if you look closely there are many more such artefacts.
I have also suggested a possible cause, that they are time of day changes in cloud cover being aliased into the rad/temp data because of longitudinal scanning of the instrument.
This may be exactly the evidence you are looking for to test your hypothesis. Instead of looking at what I pointed out you seem more interested in continuing your pissing contest with professional engineers , presumably in the hope that you can prove untrained you is smarter than they are.
Your ill-mannered snarky comments do not flatter you, How about you put your ego in the fridge for the evening and get back to doing some science.
@Willis Eschenbach 12:45 am
Here’s the full quote (emphasis mine):
Had it said, every one day it would be technically true, if deceptive because it would include highly oblique coverage grid cells of almost half the planet each day.
But “one hour”? That can’t be done from any one satellite.
What is the minimum number of satellites to give “complete global coverage every one hour?”
It can easily be done with five: 3 geosynchronous 120 degrees apart and 2 low polar orbiters with orbital periods in the 90-120 minute range.
It might be done with four: 2 Lagrangian L1 and L2 satellites and two low polar orbiters, and you fudge the “one hour” by letting the earth’s grazing limb rotate into view during the hour.
But you cannot do it with one, no matter how close or far away.
On what satellite(s) does CERES ride? It is on Terra. Is it on any others?
From My 10/8 11:18 pm
The Terra satellite was launched on December 18, 1999 and began collecting data on February 24, 2000. It operates in a polar sun-synchronous orbit at 705 km above the Earth’s surface, crossing the equator on descending passes at 10:30 AM, when daily cloud cover is typically at a minimum over land. Because of this morning equatorial crossing time, “Terra” (a mythical name for “Mother Earth”) was originally named EOS-AM-1. Terra has a repeat cycle of 16 days, meaning every 16 days it crosses the same spot on the Earth. It is roughly the size of a small school bus. Follow-on missions are planned to continue key measurements made by the five instruments aboard Terra: ASTER, CERES, MISR, MODIS, and MOPITT.
Back on 10/8 1:33 pm I ran some orbit numbers on Terra. That where I got the distance on the ground between consecutive passes: 2737 km or 25.83 degrees longitude at the equator. The distance on the ground from the overhead point to the satellite’s horizon 2767 km, almost exactly the spacing of the ground track. So from the satellite’s horizon to horizon it is 5334 km or 51.66 deg of longitude. Once per day. Won’t be back until 24 hours and 42 minutes on a track 1163 km further west.
So. “complete global coverage every one ” day. Provided you will accept readings 1400 km off track from an altitude of 700 km which means the satellite is only about 15 degrees above the horizon at that grid cell midway between the day’s passes. I don’t know how precise those measurements will be.
CERES is not giving you 720 samples per grid cell per month. At the tropics, Terra/CERES is giving you 30 daylight passes per month confined to between solar 9:40 am and 11:20 am, and 30 night time passes at 9:40 pm to 11:20 pm. with half of these measurements at an oblique 15-30 degrees above the horizon.
All,
Well obviously some of you didn’t read my post or review the link I posted. There is of course more than one satellite that has CERES instruments on board. The Satellites that have CERES instruments on board have 3 each. Currently there are 3 Satellites in orbit which allow for the hourly sample rate. As far as signal processing goes the CERES data is double checked with land based instruments to ensure there is no Temporal (alias type) errors. This type of data collection, while there are obviously electronic circuits in the instruments, is not signal processing! It is measurement of physical process which change relatively slowly.
By Changing slowly I mean that on average it takes around 24 hours to achieve an entire cycle of temperature change and while cloud cover is somewhat chaotic its more about moving than creating and clouds tend to persist for periods of time greater than 12 hours. A sample rate of 24 times a day is more than enough to capture those changes. The data is then averaged by the CERES folks for the monthly mean. Once again the base data is sound.
So Willis has done nothing to the data, other than plot it. I refuse to think that direct plotting data creates errors in and of itself. I find that Willis logic and thoughts on this are sound. Arm waving and calling names are not helpful. If you think there are true errors with the data or the plotting, put your math where you mouth is and explain it mathematically or graphically or explain it so a non electronics technician can understand. Oh, and do a little better research on the CERES project, quite a few EDU sites have information on it, even though the official site is down.
v/r,
David Riser
@David Riser 1:26 pm
Burried in your 10/7 1:59 pm linked reference
– To improve scene identification, the CERES data products make full use of coincident 1-km imager data (e.g., MODIS on Terra and Aqua) to provide detailed information about cloud and aerosol properties within the larger 20-km CERES footprints.
Oh, so there is another satellite that carries CERES. Nice to know. Any others?
BTW, this is the first use of the word “Aqua” on this thread.
Willis
I didn’t mean to devalue anything you’ve done. It’s easy to nit-pick with hindsight.
the “Nyquist frequency” is half the sampling rate
OK yes – fair point. Talking too loosely.
But then being pedantic…
The Nyquist rate is approximately twice the highest frequency of interest.
Should be…”the sampling rate should be at least twice the highest frequency component.”
That means that the Nyquist rate is about two samples per month …
But the variable of interest changes at a much higher rate than this ~20mins. You’re sampling way below this. Perhaps I’ve missed something.
Both he and you have returned with nothing but theory and handwaving
I think an acknowledgement that there might be an issue is all that is needed. I’m not as dogmatic as others who seem to be suggesting that unless you have the perfect experimental setup everything that follows has no value. Yes there does look to be artifacts, but the sheer weight of data would suggest you have something.
IN THE REAL WORLD IN THIS PARTICULAR ANALYSIS
I guess I’m just suggesting that from what you have described, it would appear that aliasing is an issue. But again, even if RC is right, I’m not suggesting that this falsifies anything you’ve stated. I think RC is being too dismissive and you’re being a bit to obstinate.
Willis oops:
”the sampling rate should be at least twice the highest frequency component.”
Should’ve been:
“The sampling frequency should be at least twice the highest frequency contained in the signal”
CERES is apparently on Terra, Aura, and Aqua All are in sun-synchronous polar orbits, with very similar periods, 98.8, min, 98.82 minutes and 98.4 min. (differences are probably due to Wike edits.) So they have the same coverage parameters. They each measure one tropical grid cell once in the day, once at night.
Wikipedia says Aura crosses Equator about 1:30 pm. But Aqua and Aura are in the “A-train” a formation of 5 satellites. So Aqua and Aura are only 8 minutes apart at the equator on the same track. Puzzling. Instead of three tracks, it is two with duplication/redundancy.
So, we are up to 2 daylight measurements per grid cell per day. Where is CERES getting it’s every hour coverage? Is it from the GOES geostationary satellites? I cannot find a reference.
Greg says:
October 9, 2013 at 11:41 am
Oh, please. I’ve asked Richard because he is the one who always promises and never delivers. I see you want to join him. OK.
So you, like RC, think there might be a problem. Sure, there might be … but the sampling rate is way above the corresponding sampling rate for e.g. temperature data, and monthly averages of temperature are used around the world. You claim that there is aliasing from a monthly average of ~ 720 samples per month … sorry, bub, but you’ve gotta demonstrate that, not just claim it. Because I don’t believe it.
But instead of demonstrating it, what you’ve done is point at some flyspecks or other and pompously declaimed that they might be showing signs of aliasing …
… just what part of my request to “provide evidence that aliasing is an issue in this dataset” has escaped you? You, like RC, just issue claim after claim. I don’t care in the slightest that you are worried that there’s a chance that some aspect of my graph might possibly be showing the effects of aliasing. That is an OPINION. What you need to support your OPINION is some EVIDENCE.
And that’s why I didn’t bother answering you, Greg. You provided nothing more than RC, just another opinion unsupported by any evidence.
w.
RC Saumarez says:
October 9, 2013 at 10:59 am
You’re welcome.
Yeah, that’s the ticket. I’ve “never” addressed anyone’s questions … RC, that’s total hogwash. I answer more questions than just about any climate blogger you can name. That claim is a joke.
Do I answer dumb questions? Often not. Do I answer spiteful or “gotcha” questions? I try not to. Do I answer questions designed to deflect or to troll rather than to actually find something out? Not when I see them first.
So if your question didn’t get answered, RC, you might consider all of the above. At that point, what you need to do is what I do: point out the question, and ask it again. Because you see … no one here but you has any idea what question or questions you are referring to.
This is because, once again, your rant is without a single example, link, citation, or number. You wave your hands and say I don’t answer questions … but you don’t identify the questions.
Near as I can tell, your whole post a valiant but futile effort to avoid answering my question, viz: what evidence do you have that aliasing is affecting my analysis?
You made the claim that aliasing was an issue. When I asked you to back it up, you erupted with fantasies about my never answering questions, and a host of other equally fanciful objections … but the question is still there, RC.
Where’s your evidence that aliasing is affecting this analysis?
w.
.
Ok. That gives the hourly coverage in grid cells outside the 10:30am and 1:30 pm CERES passes.
In the same paragraph….
It seems to me that the paragraph was badly worded to confuse CERES and ERBE processes.
David Riser, thanks for your contribution. You say:
There’s a good explanation of the whole sample rate question on page 863 here.
It says in part:
See the original document for more details.
Given that there is “excellent temporal coverage of the diurnal radiative cycle for most of the globe”, I’m not seeing how that would cause aliasing. Even if it’s only four samples per day, that’s plenty to give us accurate monthly data.
w.
RE: Oct 9, 3:21 pm
That gives the “CERES dataset” the hourly coverage in grid cells outside the 10:30am and 1:30 pm CERES passes. But calling it CERES data isn’t exactly “truth in lending” since 85% of it is GOES imager data “calibrated” to look like CERES.
I am quite curious what the GOES data looks like before and after the CERES calibration.
How much does the calibration change day to day? Month to Month?
How much does it change from 10:30am to 1:30pm? from 40N to 20N to 40S?
You see, if it is a rock steady calibration, and we can use GOES reliably for 8:30 am, 3:30 pm and 5:30 pm estimates of cloud emissions…. why do we need the CERES instrumentation at all?
Well, one answer is resolution. CERES is capable of 1 km x 1km studies of clouds at two snapshots at 10:30am, 1:30pm and 1:38pm each day (time approx at equator. an Equal number on the night side).
So, in principle, you take 400 1km x 1km CERES measurements and match them to one 20 km x 20 km reading from GOES. Calibrate GOES from that. Repeat for every 20 km x 20 km GOES cell that you have simultaneous CERES pass. Cross plot GOES vs. CERES. There is your calibration. Take 5 to 25 GOES cells, convert by the calibration cross plot into a 1×1 degree grid cell. Presto! A 1×1 degree CERES dataset.
Would it not be lovely to see those CERES vs GOES cross plots by time, by latitude, by season? I’d settle for knowing the error bars. If we calibrate on the10:30 data, how far are we off on the 1:30 data? And vice versa. Only then would we know how good the GOES in-fill in the CERES dataset.
BTW, a cross plot of Aqua and Aura CERES data would be interesting too. How wide is that cloud of data points measured 8-15 minutes apart?