The Size of Icy Reflections

Guest Post by Willis Eschenbach

In my continuing wanderings through the regions cryospherical, I find more side roads than main highways. In my last two posts here and here, I discussed the curious inverse relationship between temperature and ice accumulation rates in Greenland and Antarctica.

Wanting to understand the changes in the polar oceans that occur when the sea ice forms, I got to wondering about the albedo changes between sea ice and water. Obviously, ice reflects more sunshine than water does … but how much more does it reflect? It is important because the more the ice melts, the less solar energy is reflected, the warmer the ocean becomes, and this melts even more ice, and so on. This is a positive feedback called the “ice-albedo” feedback, which will tend to increase a given warming. Naturally, the size of this ice-albedo feedback is of interest.

To start with, I looked at the relationship between the “clear sky” albedo and the ice coverage. The clear sky albedo is the albedo measured by the CERES satellite at the top of the atmosphere when there are no clouds in the sky, so it is an estimate of the albedo of the surface itself. Figure 1 shows the relationship between the two variables, for all 1° latitude by 1° longitude gridcells which have ice during some part of the year.

albedo versus sea ice clear skyFigure 1. Ice coverage as a percentage of gridcell area (horizontal axis) versus clear sky albedo (vertical axis). The data is composed of the 12 monthly averages for each gridcell. There are 11,646 gridcells (1°x1°) which contain sea ice at some point during the year, meaning that the total number of data points N is 139,752.

It is clear that as the ice coverage increases, so does the albedo. And there is a fairly steep relationship, going from a polar ocean albedo of about 25% with no ice to an albedo of about 55% with complete ice coverage. This is an albedo change of about 30%.

However, that’s just the surface albedo. Of more interest is the “all sky” albedo, which includes the clouds. In Figure 2, I have added the all sky data in blue to the clear sky data shown in Figure 1.

albedo versus sea ice all skyFigure 2. Ice coverage as a percentage of gridcell (horizontal axis) versus both clear (red) and all sky (blue) albedo (vertical axis). The data is composed of the 12 monthly averages for each gridcell. There are 11,646 gridcells (1°x1°) which contain sea ice at some point during the year, meaning that the total number of data points N is 139,752.

The most obvious change is that the slope of the all-sky data (blue) is much less than that of the clear-sky data (red). Rather than a 30% albedo change from no ice to full ice, in the real world there is only about an 18% albedo change from no ice to full ice.

I was surprised to find that the clouds are brighter (greater albedo) than the ice itself. At all different amounts of ice coverage, including 100%, the albedo with clouds is greater than the surface albedo of just the ice itself. (I haven’t thought through all of the ramifications of this finding, I’m just pointing it out.)

However, this still doesn’t tell us just how much extra energy is reflected by the ice. The problem is that in each hemisphere the ice is at its largest extent when there is the least sunlight and vice versa. So what I did was to actually calculate the amount reflected based on the relationship given by the black line in Figure 2, which shows that the change in the albedo is equal to 0.18 times the change in the ice coverage. I calculated for each gridcell just how much difference that ice-based albedo change makes given the variations in the incoming sunlight. This will not be exactly accurate, but is certainly close enough for a first-cut analysis, and is shown in Figure 3 below.

total solar reflections global and sea iceFigure 3. Monthly averages of the total solar reflection from the entire globe (black) and the amount of the reflection that results from the existence of sea ice.

Here we finally have what I started out to find. This shows that on average, sea ice is only responsible for 1.1% of the total solar reflection. This is the result of what I mentioned above, that when there is a lot of ice there is little sun, and vice versa.

Finally, remember that the blue line is the full effect of the existence of sea ice. Let us assume that we get say a 10% reduction in sea ice. This will have 1/10 the effect of the full change, or about a tenth of a percent of the total reflections.

As a result, I’ve gotta say that on a global level at least, even a 10% change in the amount of sea ice makes very little difference to the total reflections. It only makes the total global reflections vary by a tenth of a percent. Now conveniently, total global reflections are about 100 W/m2, so that means that averaged over the planet, if all the sea ice disappeared it would only make a difference of 1 W/m2 in the global reflections … and this means that a 10% change in sea ice amounts to a globally averaged change of 0.1 W/m2

And this, of course, means that the effect of the ice-albedo feedback is vanishingly small globally. It is certainly possible that it makes some larger difference in the immediate neighborhood of the ice, but in terms of a global effect, it is what I call a third-order variable.

Ranking the variables is my own system for trying to understand what is important in a system. I divide variables in a system into first, second, and third order variables. A first-order variable can change the output measurement by greater than 10%. If for example we’re talking about solar reflections, the clouds are obviously a first-order variable.

A second-order variable can change the output by between 1% and 10%. Regarding solar reflections, an example of a second-order variable is snow cover.

Finally, we have third-order variables, which are those that make a change of less than 1% in the output measurement. That is why I said that variations in sea ice reflections are a third order variable. And typically, third-order variables can be ignored in all but the most accurate analyses … and generally we can’t do analyses anywhere near that accurate in climate science.

Anyhow, that’s what I found out about the size of the ice-albedo feedback … it is a third-order variable, so small that it disappears in the noise.

Jupiter burning in the midnight sky, ah, dear friends, another springtime is upon us here, it is good to still be on the upper side of the grass.

My best wishes to all,


My Usual Request: Misunderstandings destroy communication. If you disagree with me or anyone, please quote the exact words you disagree with, so we can all understand the precise nature of your objection. I can defend my own words. I cannot defend someone else’s interpretation of some unidentified words of mine.

My Other Request: If you think that e.g. I’m using the wrong method or the wrong dataset, please educate me and others by demonstrating the proper use of the right method or the right dataset. Simply claiming I’m wrong doesn’t advance the discussion.

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
M Courtney
April 2, 2016 12:25 am

Is the effect of sea ice greater for ice attached to land at lower latitudes?
I’m speculating that Greenland may be more significant than the Arctic for that reason – even though the area is smaller.

Peter Sable
April 2, 2016 12:26 am

Oh dear. I wonder how many papers this debunks. Nice analysis.

Reply to  Peter Sable
April 2, 2016 11:51 am

“Debunks” is probably not the right word. “Reduces to triviality” is more the case. Still, I’d sure like to know how the climate models handle albedo.

Reply to  Peter Sable
April 2, 2016 3:22 pm

If this had any debunking power then what is it doing here ?.
Submit it to a proper scientific journal for peer review. Oh, I forgot, everything is corrupted and psychologists know more than climate scientists.

Ed Bo
Reply to  WTF
April 2, 2016 5:59 pm

Willis has published several papers in peer-reviewed journals. I believe all of them started out as posts on WUWT and benefited greatly from comments, suggestions, and (substantive) criticisms received here.
Are you going to help (even pointing out fundamental errors that discredit his whole idea is helping) or just continue with your useless sniping?

Reply to  WTF
April 2, 2016 7:31 pm

Ed Bo,
Can’t find anything climate related that he has published, maybe Psychology related only ?, fair enough.
Please correct me.

Jean Parisot
Reply to  WTF
April 3, 2016 6:24 am

If the trend continues, being published in a climate science journal is going to look like a police report on your resume.

April 2, 2016 1:50 am

It is notable that there is a much higher effect if you include snow cover over land.
The global planetary albedo varies annually by as much as 2.5%, from 0.293 to 0.318
For Planetary albedo see This plot

Reply to  Jan Kjetil Andersen
April 2, 2016 5:15 am

Very useful link Jan.
I was initially struck by significant 6mo signal. Even splitting off NH , SH it is still there. SH seems to have a 6mo component larger than the annual change! Clear-sky, all-sky seems similar.
Then I looked at SW and bang, we’re back to a fairly clean 12mo cycle. So the big signal in LW.
SH , LW all-sky seems to be the wild card.

Reply to  Greg
April 2, 2016 6:11 am

Thank you Greg
The site is actually mine. Nice that you find it useful

Reply to  Jan Kjetil Andersen
April 2, 2016 10:13 am

The biggest effect of ice/no ice is how clouds respond. Below 0C when the surface is covered by ice and snow, incremental clouds warm the surface by slowing down surface emissions into space while above 0C, incremental clouds reflect more energy then they trap at the surface and the net effect of incremental clouds is to cool, rather than warm. This is evident when the fraction of the surface covered by clouds is plotted against the surface temperature.
The LTE behavior of clouds relative to the surface temperature is clearly affected by whether or not surface ice is present. Anyone with practical experience with real (rather than imagined) feedback control systems should recognize from this that clouds are the control variable that drives the surface towards some otherwise required steady state average temperature and that cloud feedback is positive 0C. Unfortunately, nobody on either side of the climate science debate understands how real feedback systems actually respond and this has lead to the claim of massive positive feedback that defies the constraints of the math and physics.

Reply to  co2isnotevil
April 2, 2016 10:57 am

Below 0C when the surface is covered by ice and snow, incremental clouds warm the surface by slowing down surface emissions into space…

Do you mean surface cooling slows down vs actual warming of the surface?

Reply to  co2isnotevil
April 2, 2016 11:47 am

Incremental clouds act like a broad band GHG and returns some fraction of captured surface emissions (about half) back to the surface as emissions from the cloud. Like GHG’s, this makes the surface warmer than it would be otherwise by slowing down the rate at which the surface cools, in effect, feeding back some of the energy emitted by the surface back to the surface preventing it from escaping. This is the actual physical manifestation of feedback, relative to the energy balance of the planet, but becomes obfuscated when the units of gain and feedback are expressed as temperature changes per W/m^2. since feedback linearly affects the energy/power, but non linearly affects temperature (i.e. Stefan-Boltzmann power is proportional to T^4). Meanwhile, the IPCC asserts linearity over a narrow range only to silently extend the assumption of linearity across all T.
Clouds and GHG’s do not actually warm the surface but only slow down its cooling. Only the Sun warms it and only when the surface has a lower reflectivity than the clouds (i.e. no ice and snow) does increasing clouds arising from increasing water vapor as the temperature rises cause the surface to cool by reflectiing away more solar input than can be replaced by cloud emissions returning to the surface. Can you see this as the strong negative feedback it really is?

Bill Treuren
Reply to  co2isnotevil
April 2, 2016 4:14 pm

Further if there really is a control knob for the climate it is the cloud at or near the equator. Willis has discussed this in the past.
The radiation increase at the poles as a result of the open oceans would dominate any reflective loss due to ice loss.
And if that does nothing to satisfy the CAGW crowd explain how the world is not boiling because it has been warmer before and came back to refreeze itself without solar panels and wind turbines.

Reply to  Jan Kjetil Andersen
April 2, 2016 12:01 pm

One of the finer contributions I’ve seen here, Jan. I read the “About” page. Most impressive. I liked the interactive plot, too.

Reply to  jorgekafkazar
April 3, 2016 8:33 am

Thank you for the kind words Jorge.
I have tried to make an intuitive and user friendly interface to objective facts related to climate science.
Most of the graphs are based on monthly data series which I automatically copy to the homepage.

Reply to  Jan Kjetil Andersen
April 3, 2016 10:11 am

This annual change is not due exclusively to snow/ice, even tnough that is presumably the dominant factor. The state of vegetation is also important. Fresh green vegetation and leafed trees in summer have a markedly lower albedo than dead/dry vegetation and leafless forests. After all absorbing sunlight is what photosynthesis is all about.

Reply to  tty
April 4, 2016 12:52 am

That is a good point tty

April 2, 2016 1:51 am

Willis, very interesting and thought provoking stuff, as usual.

The most obvious change is that the slope of the all-sky data (blue) is much less than that of the clear-sky data (red). Rather than a 30% albedo change from no ice to full ice, in the real world there is only about an 18% albedo change from no ice to full ice.

The first thing that struck me about figure one was the that the single linear fit did not really represent what was seen in the scatter plot. I saw a lower line from left to right with a much lower slope. Then a higher group of albedo points which had a similar, low slope line going left.
I would want to see what these two groups represent. Is it geographic position, latitude, fast coastal vs sea ice? It seems that there two groups of points both with a lower slope and fiting a single slope to the whole dataset may be misleading.
When I saw figure 2, it turns out that the all sky data has a very similar slope to the line I’d noticed in the lower group on the clear sky data, though shifted up to higher albedo.
Like yourself, I’m not interpreting what this means, just observing.
In the same way that all sky has a lower slope but higher offset, I think this slope also applies to all the data with a certain and significant amount of the clear sky ice data showing similar higher overall albedo and the same slope.
This represents a considerably lower magnitude +ve f/b but of course the one the simplistic alarmists focus on to predict tipping points and death spirals that are not born out by the data.
The other side of the coin is negative feedbacks in the IR. Water is nearly “black” : zero albedo in the infra-red compared to quite high as ice. This means open water as well a increasing evaporation radiates a lot more heat.
This is clearly the dominant effect otherwise we would be seeing “tipping point” behaviour, not the massive recovery in ice volume that happened following the 2012 minimum, nor the general slowing of melting.
Slowing down is not compatible with the concept of a system dominated by a +ve f/b. Once a vase starts tipping it can not slow down a bit half way over.comment image

Reply to  Greg
April 2, 2016 1:56 am

Up until 2007 this data could be said to roughly compatible with a parabolic curve, or accelerated melting. That was a reasonable *possible* interpretation at that point.
It is not longer tenable. The bedwetters need to change the sheets and find something else to worry about.

Reply to  Greg
April 2, 2016 2:15 am

The NH sea ice extent flows the AMO cycle very closely.
The Arctic sea ice extent is almost exactly where it should be, about 1sd below the 1980-2000 mean.

Reply to  Greg
April 2, 2016 2:21 am

Arce is STILL anomalously HI compared to the first 3/4 of the Holocene.
Biodata clearly shows that there was quite often ZERO summer sea during the Holocene OPTIMUM.
Arctic sea ice is still recovering from the dangerously vast amounts of the LIA.
Let us all pray that the LIA is never repeated in our life times.. or our children’s or grandchildren’s for that matter.

Reply to  Greg
April 2, 2016 2:22 am

first word should read…… Arctic sea ice

Reply to  Greg
April 2, 2016 2:23 am

? typing letters, and they are not displaying ????

Reply to  Greg
April 2, 2016 2:24 am

Yes, that’s interesting Andy. The trouble is if you put enough spaghetti on one plate you can’t see what’s going on beyond a general trend which have been arbitrarily scaled and offset to roughly fit.
Also AMO is detrended so I don’t see any reason to use detrended temp data. Straight SST would seem more relevant.
I think the may be an agreement with the slowing of warming in SST and the slower melting but you graph masking the lack of correlation on the decadal scale. ( Or at least as well as I can tell trying to read between the spaghetti ).
In short, you may be right but not showing particularly clearly what is going on.

Reply to  Greg
April 2, 2016 7:40 am

Greg, you are right. We have to use AMO values without detrending.

Reply to  Greg
April 2, 2016 5:19 am

“Water is nearly “black” : zero albedo in the infra-red compared to quite high as ice.”
Point of order – there is no such thing as “IR Albedo”. Albedo is the fraction of solar energy (shortwave radiation [visible]) reflected from the Earth back into space. It is a measure of the reflectivity of the earth’s surface. Please don’t change or create the definition of things to suit your perspective. Also, water may be nearly “black” (in visible) ..except at shallow angles & at near zenith (that is what sun glint is). Then it can get quite high.
“This means open water … radiates a lot more heat.”
Of course, because it is warmer than ice…isn’t that kind of obvious?

Leonard Weinstein
Reply to  JKrob
April 2, 2016 6:06 am

Open water can be nearly the same temperature as ice. Sea water typically freezes a bit below the freezing point of pure water due to the salt content. Open water radiates a lot more due to the higher emissivity, not necessarily to a difference in temperature.

Reply to  JKrob
April 2, 2016 11:17 am

@Leonard Weinstein –
“Open water radiates a lot more due to the higher emissivity, not necessarily to a difference in temperature.”
With what Willis says:
emissivity coefficients:
Water 0.95 – 0.963
Ice smooth 0.966
Ice rough 0.985
Sorry Leonard, I don’t get your point…especially since ice can be *much* colder than liquid water (salt or fresh). You won’t get below zero F temps over water but you sure can over ice.

John Harmsworth
Reply to  JKrob
April 2, 2016 2:41 pm

It is absolutely worthwhile to examine how different wavelengths of light are absorbed or reflected by water, ice, clouds, trees or nudists if it helps to understand what the ultimate fate or effect is of the energy that we receive from the sun.

Ed Bo
Reply to  JKrob
April 2, 2016 6:07 pm

JKrob: You say “Point of order – there is no such thing as “IR Albedo”.”
If you’re going to play pedant, you darn well better get your facts right. “Albedo” is a general and flexible term, applicable to many different spectra. There is a reason that the reflectivity to solar radiation uses the specific term “Bond albedo”; that distinguishes it from albedos to other spectra.
If you really knew your stuff, you would have pointed out, as others have, that ice/snow do not have high albedo in the (thermal) infrared.
So I’m afraid you’re batting 0 for 2: one error of commission and another of omission.

Reply to  JKrob
April 3, 2016 12:14 pm

This is why I really dislike the term “albedo” . If anything , If anything , it should be cited as “albedo with respect to ( some specific source spectrum )” .
But even that is crude compared to thinking in terms of the dot product between the source spectrum and the object absorption=emission spectrum .

george e. smith
Reply to  JKrob
April 6, 2016 3:34 pm

I’m with Willis. Real ice, and open water are nearly BB radiators; at least for the 5-80 micron range of wavelengths that you would expect from a zero deg. C black body spectrum.
And the open water sans stormy weather can be even higher reflectance than rough ice, because of the Brewster angle polarized reflection effect.

bit chilly
Reply to  Greg
April 2, 2016 2:48 pm

just to pick up on the “water is nearly black term”. in a post last year caleb highlighted that due to the low angle of incidence of sunlight in arctic summer even open calm water reflects a lot of sunlight. he had a very good image highlighting it.

Bill Treuren
Reply to  bit chilly
April 2, 2016 4:40 pm

the real issue is that the surface of the ice is very low especially when the sun is very low or gone thus the radiation of the arctic region is hindered by the insulating ice.
The surface of the arctic ocean (unfrozen) is almost always above -1.8C the surface of the frozen ice will dip to -40C in moments after sunset thus releasing little energy from planet earth.

george e. smith
Reply to  bit chilly
April 6, 2016 3:44 pm

If Bill’s Temperatures are correct (no dispute) then I agree with his assertion, the -40 deg. C ice won’t radiate as much as the -1.8 deg. C water.
The polar regions do not cool the planet. It’s the tropical dry deserts that are doing all the radiating, as much as 12 times the rate for the Antarctic highlands.
My Automobile radiator is not running at either -1.8 or -40 deg. C ; far too slow a cooling rate.

Reply to  Greg
April 3, 2016 10:24 am

“I would want to see what these two groups represent. Is it geographic position, latitude, fast coastal vs sea ice? ”
At a guess ice that is snow-covered and ice that is not snow-covered. This would fit with the higher albedo line being very weakly defined at low ice-densities. Ice that is melting is unlikely to be covered by snow. Snow always has higher albedo than ice.

April 2, 2016 2:08 am

Jupiter burning in the midnight sky

Oh what, Jupiter’s burning? Does that mean they have a CO2 problem too?

Reply to  Willis Eschenbach
April 2, 2016 4:49 am

By Jove ! It’s worse than we thought.

Reply to  Willis Eschenbach
April 2, 2016 10:36 am

Above grass joviality, long may it persist.

Reply to  Greg
April 3, 2016 8:55 am

Has Willis Metis match? Io know.

Reply to  Greg
April 5, 2016 4:21 pm

Actually Jupiter radiates pretty strongly for a planet, about twice as much energy as it receives from the sun IIRC. There is some mild debate about the source. The simplest suggestion is that the heat radiated by Jupiter is left heat from the formation of the planet.

April 2, 2016 2:11 am

And to further counteract the extra open water absorbing more sunlight, open water radiates more energy than ice. So I’m not worried about sea ice being a positive feedback for warmer temperatures as I think any major change needs a significant external change.

Reply to  Rod
April 2, 2016 2:41 am

What makes you think water radiates more energy than ice?

Reply to  MikeB
April 2, 2016 3:29 am

Well if you compare salt water at -1 and sea ice in -15 C, which of them radiates more? Which evaporates more?
‘All being equal’ is not a realistic assumption.

Reply to  MikeB
April 2, 2016 4:30 am

Mike, see my comment above about this. Water has a low IR albedo : the corollary of that is an emissivity near 100%. Ice IR emissivity is, from memory, about 30%.
That is a huge increase in IR emissions 24/7 for a as long as there is open water. ( Don’t forget: ice “extent” can be as little as 15% actual ice. ). The >40% increase in ice volume from 2012 to 2013 is a pretty solid indication of whether the +/ve or -/ve feedbacks dominate here. As is the graph showing slowing of melting that I posted above.

Mike M the original
Reply to  MikeB
April 2, 2016 5:42 am

Ice is not a very good insulator but snow is and it accumulates on top of ice not water. Snow covered ice is trapping the heat in the water. Keep in mind that we hit minimum ice about the time of equinox or low solar incidence angle so much solar is reflecting off the water.
Polar sea ice is a strong negative feedback mechanism IMO.

Leonard Weinstein
Reply to  MikeB
April 2, 2016 6:30 am

Greg, you are confusing albedo with emissivity when you say IR albedo. There is no such thing as IR albedo.

Reply to  MikeB
April 2, 2016 7:16 am

OK, let’s use standard physics terms, not this “albedo”.
Water has a very low IR reflectivity and a high IR emissivity. The point stands as I explained.

Reply to  MikeB
April 3, 2016 2:42 am

Kirchhoff’s law of thermal radiation states that emissivity is equal to the absorptivity.
For an opaque body absorptivity is the complement of reflectivity.
From the graph Alex provided below, we see that reflectance of snow drops off notably as we move in to IR ( longer wavelengths on the right. )
That would give an absorptivity and emissivity of about 30-40% for snow ( so long as it is thick enough to be considered opaque.)
That was probably the number I was recalling. Ice without snow may be different and Eskimos would not be pleased about us only having one word for snow.

george e. smith
Reply to  MikeB
April 6, 2016 3:50 pm

Well albedo is solar spectrum NOT LWIR. And at near zero deg. C the LWIR spectrum is 5.0 to 80 microns for 98% of the energy radiated, but 2-3% for solar spectrum reflectance.
Albedo is one number for the entire earth. It is NOT a local variable.

Reply to  Rod
April 2, 2016 11:31 am

What “water has low reflectivity and high emissivity” means is that absorbs and radiates strongly in the IR.comment image
This can be seen the broad area of strong absorption between wave numbers 10 and 10,000. Water absorbs poorly/reflects a lot in the region of the most intense surface solar radiation centered on the visible bands. This is why albedo from snow covered LAND is so important.
Ironically, water and CO2 share an extreme fondness for photons in the 15 micron/667 wn range such that the water gobbles them up in about the thickness of a human hair and atmospheric CO2 within a few meters. What gets gobbled up gets spit out, albeit at much lower proportion. This is the basis for the famous photon food fight.

richard verney
Reply to  gymnosperm
April 3, 2016 12:23 am

the water gobbles them up in about the thickness of a human hair

In fact, it is just a fraction of a human hair. If I recall correctly, human hair varies from about 20 to 180 microns, say with a width of about 50 microns being typical.
Due to the omnidirectional nature of DWLWIR, about 80% of all DWLWIR is fully absorbed in about 5 microns of the ocean. I do not think that most people appreciate just what a small volume of water is fully absorbing DWLWIR, and your example of a human hair helps puts this in perspective but even though a hair width is small, it is actually an order of magnitude too large!!!
See generally (which shows the vertical penetration of LWIR in water, but DWLWIR is omnidirectional with much of it having a grazing angle of less than 35degs):comment image

Bill Treuren
Reply to  Rod
April 2, 2016 4:52 pm

I do wonder what the earths warming would be by the calculation that is used, for better or worse, if the regions that were ice and now water are taken out.
My suspicion is that much of the Arctic’s reported warming is an artifact of warmer open ocean rather than ice cover. Just look at the fact that the north is blow torch and the south freezing and somehow this is cause and effect in most peoples mind when in fact this may just be a symptom of a sea current or prevailing wind change through a largely natural cycle.
We may yet see this as the oceans refreeze due to natural variation and I do wonder whether the Hiatus was more an artifact of Ice cover loss stopping during that period.
This is all way past my pay scale!

April 2, 2016 3:13 am

If ever there was someone deserving of a climate research grant, it is you Willis. Of course if you did get one, you’d probably need to get political and your original ideas would head South.

Reply to  Tony
April 2, 2016 7:04 am

Actually I think Willis would take the money and HEAD south.

Reply to  Willis Eschenbach
April 2, 2016 12:36 pm

I suggest that wuwt readers that think as much of Willis Eschenbach as I do write to their congressman and direct to the National Science Foundation and suggest they give him $100,000 grant, unrestricted other than the work be related to climate science. They’d get results equal to any $10,000,000 they ever spent.

Reply to  Willis Eschenbach
April 2, 2016 1:03 pm

…except to my friend who gets me all the paywalled scientific studies I can’t afford…
Look! Willis is funded by Big Oil/Coal/Nuclear!

Reply to  Willis Eschenbach
April 2, 2016 4:07 pm

Piper Paul! Maybe Willis is funded by a mole. Maybe he’s a spy on the sky! 😉

Reply to  Willis Eschenbach
April 2, 2016 4:14 pm

In all seriousness, your comment should be immortalized within the scientific community.

Reply to  Willis Eschenbach
April 3, 2016 7:11 pm

Thanks again. You keep contributing to human knowledge. Glad you fell in love with numbers.

Bloke down the pub
April 2, 2016 3:27 am

As, at low angles, water reflects more light than ice, does your data give any insight into how important it is as to how far north the open water is? For example, will there be a greater change if ice is missing from around the tip of Greenland than from the Barents Sea?

Reply to  Bloke down the pub
April 2, 2016 4:40 am

Yes incident angle and surface reflection is something that gets totally overlooked. One observational study was done but from memory it was relatively low latitude, around Halifax, I think. Wind is quite an important variable since there is less low angles to do reflection when it’s choppy.
We are all fami1iar with strength of the reflections of sunlight off a pool or sunset on a calm sea. It is always “sunset” in the Arctic. More so the higher you get.
Also shallow melt pools will have a fairly smooth surface.
I did bring the question up with some researcher who was doing spectral measurements in situ in the Arctic but had failed to even look at it. He said he was “not aware” of the effect being significant. Not surprising he was not aware if he was not looking while being paid to do so.
Rather than address the problem he walked away saying he was “not familiar with my work”, when this sort of thing dates back to 19th science.
More interested in “saving the planet” that doing valid research it seems.

Reply to  Greg
April 2, 2016 5:40 am

I don’t see how angle has anything to do with it. Maybe a 4th order problem. If you look at a full moon the average reflect light is very even across the “disc”. What your suggesting is the edges should be darker due to less reflection at low angles. Same with a picture of earth from space. Very even intensity for the entire image. The low angle is correct for by the larger area represented.

Reply to  Greg
April 2, 2016 5:58 am

The moon is not made of water !
check out surface reflection and Brewster angle and you get the idea.

Owen in GA
Reply to  Greg
April 2, 2016 6:12 am

The moon is made of a granular powder on the surface. I would suspect that each grain reflects rather well in any direction. NASA recently published all the lunar surface pictures taken by the Apollo astronauts on the surface, and the material looks very much like it would not respond like a sheet of glass the way a calm stretch of water would, but like a trillion individual tiny mirrors. If the moon were covered in water or some other such medium with those optical properties, you would be correct.

Reply to  Greg
April 3, 2016 6:48 am

With a full moon the greatest concentration of light is from the centre. The photon flux diminishes to the edges following the cosine rule. Your eye may not detect it but it is true. The moon is a sphere and not a disc. Same goes for the sun and other heavenly bodies.

george e. smith
Reply to  Greg
April 6, 2016 3:57 pm

Windy weather on the horizon gives a black line on the horizon. Mexican Baja fisherfolk use that to know there could be some bad wind coming.
Ergo the grazing angle reflectance (of water) is much lower with wind.

Reply to  Bloke down the pub
April 2, 2016 4:47 am

Have often wondered if this factor is properly accounted for also. The higher the latitude the greater is the degree of specular reflection so then in the high Arctic you would expect open water to exhibit higher albedo than ice and obviously enough vary with the seasons.

Reply to  cephus0
April 2, 2016 4:55 am

A couple of years ago there was new flap about surface melt water causing more solar to be absorbed and thus accelerating melting…. +ve f/b …. tipping points OMG , etc. Models were modified to include more melting due to this. At the time models were under-estimating the historical rate of melting in the Arctic.
None of this helped them anticipate the massive recovery of 2013. They clearly have very little understanding of the primary forces determining Arctic ice coverage.

Reply to  Bloke down the pub
April 2, 2016 6:23 am

comment image

Reply to  Alex
April 2, 2016 6:25 am

above graph is for water

Reply to  Alex
April 2, 2016 7:18 am

Thanks Alex, a picture and 1000 words etc. very nice.

Reply to  Alex
April 2, 2016 7:30 am

And “No”.
And “That (classic) plot of water’s albedo with respect to incident angle is (almost) irrelevant to the world’s open ocean seas.”
Perfectly true – for a laboratory environment of pure water in a sterile environment with no wind, no water, and “pure” incident light: No diffuse radiation, no atmospheric re-reflections, no multiple-reflections.

Reply to  Alex
April 2, 2016 7:39 am
Air mass should also be considered. Considerable reduction of energy due to atmospheric thickness

Reply to  Alex
April 2, 2016 7:53 am

Totally agree. Performed in a laboratory. Wave action will increase and also decrease the angle of incidence so it kinda averages out. You also have foam on waves (happens to be white). The effect is still there and shouldn’t be dismissed out of hand.
Apparently the planet is going to go into meltdown because of an alleged imbalance of 5-6 watts /metre squared. And who knows whether everything has been factored into the equation

Reply to  Bloke down the pub
April 2, 2016 7:15 am

Calculations of critical angles of reflection have been around for a long time:

Reply to  rbabcock
April 2, 2016 3:11 pm


Calculations of critical angles of reflection have been around for a long time:

True. In theory for two theoretically pure substances with perfectly flat planar surfaces, each composed of perfectly theoretical lab-measured values in still air at lab pressures and temperatures.
Like the Fresnel approximations (er, equations) you CANNOT simple read the textbook(s) and then assume the theoretical values are actual Arctic or Antarctic values. You MUST read the papers that acvtually GO there and MEASURE it. And, even then – or especially even then, you have to look at time-of-day, day-of-year, pressure, temeprature, cloud cover, and the total environment.
In one case, a lab measured open ocean albedo while transiting from Brazil to Africa, then measure it again on the return trip. For the same sea state and wind conditions, they got two different albedos at low elevation angles at different times of the day. Turns out the device was properly calibrated for elevation angle and ship rolls and pitch motions, but the instrument was catching reflected energy from the hull AND the sea surface, but only at certain times of the day and certain solar elevation angles.

Reply to  Bloke down the pub
April 3, 2016 11:09 am

At an angle above 70-80 degrees, the reflection off water, increases albedo to being nearly the same as ice. And this isn’t just important for latitude, early and late sun would have the same high angle.
So for maybe 6 hours a day, some polar water would absorb excess energy, while the rest of the day it’s a very effective radiator to space, if it’s clear out.

April 2, 2016 4:29 am

data point: as a photographer brought up in the manual camera days full cloud was a 75% reduction in light; some of tat may be absorbed by the cloud,but its a massive difference

April 2, 2016 4:41 am

Hi Willis,
The clear sky plot shows a lot of very high values (80-100%) for albedo at high ice coverage. This looks suspicious to me, because the highest albedo surface (fresh clean snow) has a maximum of ~85%, and it is usually much less. So there may be a quality control problem with the data. Another tip-off there is a problem is that you see lots of very high clear sky values even when the sea ice coverage is well under 100%, even though open water has a relatively low albedo. Something seems messed up.

Reply to  stevefitzpatrick
April 2, 2016 9:20 am


Hi Willis,
The clear sky plot shows a lot of very high values (80-100%) for albedo at high ice coverage. This looks suspicious to me, because the highest albedo surface (fresh clean snow) has a maximum of ~85%, and it is usually much less. So there may be a quality control problem with the data. Another tip-off there is a problem is that you see lots of very high clear sky values even when the sea ice coverage is well under 100%, even though open water has a relatively low albedo. Something seems messed up.

Not true. Well, true, but actually irrelevant to the Arctic and Antarctic waters.
1. You are probably mentally picturing NASA’s (Trenberth’s) almost-deliberately-misleading famous diagram of sunlight and water reflectivity.
That mental image is dead wrong. Well, it is correct – but only for sea ice floating at latitudes closer to Florida than the north coast of Siberia or around Antarctica. For solar elevation angles above 33 degrees, it is a (barely) acceptable approximation.
But never under any other circumstances.
There are several factors that control the Arctic and Antarctic regional albedo:
A. Day-of-Year (Affects darkness of the sea ice, average dirtiness of the surface snow and ice over the sea ice, type of snow and ice, aount of surface melt ponds, the average latitude of the sea ice, the elevation of the sun each hour of the day,
B. Hemisphere. (Arctic and Antarctic behave differently, melt differently, re-freeze differently, and have different albedo values from each other.
C. Cloud Cover, Percent of Surface Water (on top of the sea ice), and Percent of Sea Ice (in the area).
D. Solar Elevation Angle of the incident sunlight. (Affects direct-to-diffuse light ratios, atmosphere absorption (air mass the light must penetrate to get down to the surface to either be reflected or absorbed), atmosphere scatter (direct and diffuse ratios), and (most important!) the albedo of direct radiation from open water.)
E. Percent of Direct Radiation and Diffuse Radiation hitting the Arctic or Antarctic surface.
F. Wind speed. (Controls wave height, which affects albedo)
G. Local “Weather”: Air temperature (dry bulb, 2 meter), wet bulb temperature (which is how you calculate relative humidity, which controls local evaporation, which controls low wave radiaiton losses), wind speed ( which also controls evaporation and convection and conduction heat losses), length and width the ocean leads between ice floes (which also controls film coefficient calculations.)
Willis talks about only one of these above: Percent of Snow and Ice-Covered Surfaces in the local area of the sea ice.
But, in total, all three (Percent of Cloud Coverage, Percent of Open Water on the Sea Ice, Percent of Snow and Ice-Covered Surfaces in the local area of the measurement) work WITH each other to affect the Arctic/Antarctic regional albedo.
They all combine with each other by affecting reflections, re-reflections and re-re-re-reflections between the clouds above the local region (which reflect light back down to the local surfaces) and the local area (which reflect light back up to the clouds to be reflected back down.) Curry measured this effect during the 1997-1998 SHEBA ice expedition to the Arctic. Clear skies – at EVERY time of year! – have a lower sea ice area albedo than cloudy skies. Sea ice albedos measured near open water (ANY open water: leads in the sea ice, ponded water on the sea ice surface, AND sea ice albedo at the ocean edge of an area covered by sea ice) are generally 5% darker than sea ice albedoes measured on 100% ice-covered regions – regardless of time-of-year. Down south, the higher albedo of Antarctic sea ice at comparable times of the year are higher than Arctic sea ice albedos because there is almost no surface ponding on Antarctic sea ice (it melts from below, not above) and the relative absence of open water leads in the sea ice.
Solar Elevation Angle.
The albedo of ice and snow surfaces, of bare sea ice, and of snow-covered sea ice do not very significantly with the angle of the sunlight onto the surface.
The albedo of open ocean water hit by diffuse solar radiation does NOT vary with Solar Elevation Angle, and IS CONSTANT at the “classic” (Wikipedi-verified!) albedo = 0.067.
However, the AMOUNT of diffuse solar energy hitting the open ocean surface depends on SEA and on cloud coverage: So, on average, diffuse radiation is only 5-10% of the direct radiation present (clear days) up to as much as 30% of theoretical direct radiation on cloudy days. Diffuse radiation will NEVER be greater than theoretical (clear day) direct radiation.
The albedo of water hit by direct radiation varies strongly with Solar Elevation Angle, but NOT according to the simplistic Fresenal plot above.
The albedo of open ocean water is almost immeasurable at SEA angles below 5 degrees above the horizon. (Willis should get a year-long grant from the NOAA to measure this at some convenient point: Say looking west towards the sun every evening from the pier at Key West, recording wind speed, wave height, solar elevation angle, cloud cover and position of the cloud cover, direct and indirect albedo, and relative elevation of every beer glass above the table.
The albedo of direct solar radiation of open ocean water is constant at all wind speeds and all wave heights at Solar Elevation Angle above 33 degrees at the “classic” value of 0.067 (At very high winds (storm conditions) spray and spume DO increase reflectivity of light, but realistically, the amount of light energy getting through the dark storm clouds that are associated with spray and wind-driven froth makes the change neglibile. ( Except for the guys who got their grant and their research money and their published paper writing about the effect of spray and wind froth on ocean albedo.))
The albedo of direct solar radiation from open ocean waters is a straightforward calcuation for all Solar Elevation Angles between 5 degrees and 33 degrees. Pegau and Paulsen extended the original formula derived from measurements from light towers in open waters to account for both SEA and wind speeds. (Yes, Virginia, I believe in ocean waves, and the ocean’s waves and winds ARE included in the albedo measurements reported. ) Do NOT let any body use the simplistic ideal Fresnel equations! They are valid ONLY for theoretical laborartory albedoes under very specific lab conditions.
Real world? You MUST use the measured values – NEVER a simplistic approximation of a “constant” albedo for water either.
Curry’s measured Arctic sea ice albedoes vary according to day-of-year: Starting at TOA maximum solar radiation on 5 January at 0.83, Arctic sea ice albedo remains at 0.83 through mid-April, then decreases as melt-water ponds up and Siberian and Canadian dust and Chinese poluttion falls through the spring and summer. It reaches a yearly minimum of 0.43 (averaged value) on 5 July, then increases after 12 August back towards its wintertime maximum of 0.83 in early October. Given a date (day-of-year) you can calcuate a reliable Arctic sea ice albedo for any day of the year.

Reply to  RACookPE1978
April 2, 2016 6:58 pm

After all that, it appears you offer no explanation for albedo approaching 100%, even with 10% to 20% open water. Can you explain that very high albedo?

Mike M the original
April 2, 2016 5:04 am

What about angle of incidence comparing reflection to open water?

Philip Mulholland
April 2, 2016 5:17 am

Nice one Willis.
Data overrules theory every time.
From the master:-

It doesn’t matter how beautiful your theory is, it doesn’t matter how smart you are. If it doesn’t agree with experiment, it’s wrong.

Richard P. Feynman

Reply to  Philip Mulholland
April 2, 2016 4:05 pm

Sometimes data and theory disagree and they are both correct.
Sometimes data is correct
Sometimes theory is correct.
‘Data overrules theory every time.” Nope: This is a theory, a principle. it’s empirically wrong.
And Feynman himself knew it to be wrong He knew from his own experience that sometimes data is wrong.
he also knew that sometimes the conflict between data and theory cant be decided.
“I was a young person and very enthusiastic and full of calculations that we’d done about the sun, and I argued with Ray. But Ray said “John, I know the director of Brookhaven. He often says that no astrophysicist can calculate anything with sufficient precision to be of interest to any particle physicist.” He said, “Trust me, forget about your models of the sun. Our only chance of selling the experiment is talking about your nuclear physics, and I’ll talk about how to do the experiment.”
So I deferred to him. I talked about the nuclear physics, [and] the director was very enthusiastic about that. Eventually his wife did an experiment to test those ideas, and Ray described how he could do the experiment. About three weeks later the director approved the experiment with no proposal ever having been written, and very shortly afterwards Ray began to look for a mine where he could do the experiment.”
“Then, of course, when his experimental results came out and they were in conflict with our calculations, he and I would be invited to give theory and experimental talks everywhere. ”
And yet there was a nagging discrepancy between your results and his, right?
Well, right from the beginning it was apparent that Ray was measuring fewer neutrinos events than I had predicted. He came to Caltech in early 1968 to spend a week with me while he and I wrote our papers up describing for me a refined calculation, for him the first measurement of the rate in his tank. It was clear that the rate that he was getting was a factor of three smaller than I was predicting, and that was a very serious problem.
There was a famous meeting at Caltech, just a few physicists—Dick Feynman, Murray Gell-Mann, Willie Fowler, Bob Christie, and a couple of others—in a small meeting room, where Ray presented his results and I presented my calculations of what he should have measured. There was some discussion of it afterwards, and it was pretty inconclusive. There was a discrepancy; it looked like one of us was wrong.
“I was very visibly depressed, I guess, and Dick Feynman asked me after the meeting if I would like to go for a walk. We just went for a walk, and he talked to me about inconsequential things, personal things, which was very unusual for him, to spend his time in quite idle conversation; it never happened to me in the many years that I knew him that he did that before or afterwards. And only toward the end of the walk, which lasted over an hour, he told me, “Look, I saw that after this talk you were depressed, and I just wanted to tell you that I don’t think you have any reason to be depressed. We’ve heard what you did, and nobody’s found anything wrong with your calculations. I don’t know why Davis’s result doesn’t agree with your calculations, but you shouldn’t be discouraged, because maybe you’ve done something important, we don’t know. I don’t know what the explanation is, but you shouldn’t feel discouraged.”
For me I think of all of the walks or conversations I have had in my professional life, that was the most important, because I was a young man without tenure, and [while] I’d done many calculations by that time, this was the one that was most visible and people had paid the most attention to, and it looked like it was wrong. I really was feeling very, very, very discouraged. And for a person whom I so enormously admired, Dick Feynman, to tell me “You haven’t done anything that’s visibly wrong, maybe you’ve done something important”—for me that was a huge boost.”””
“But there were plenty of scientists who did think there was something wrong with your model of the sun.
Well, initially very few people paid any attention to this discrepancy, but the discrepancy persisted. … And every year for 30 years I had to look at different processes that people would imaginatively suggest that might play a role in the sun, and it didn’t matter how convinced I was that they were wrong. I had to demonstrate scientifically that these processes were not important in order to convince people [that] yes, the expectation from the sun was robust and therefore you should take the discrepancy seriously. It took I would guess three and a half decades before I convinced everybody.”

Reply to  Steven Mosher
April 2, 2016 10:03 pm

” ‘Data overrules theory every time.’ Nope: This is a theory, a principle. it’s empirically wrong.”
Is there a human out there who doesn’t interpret Feyman’s words as (essentially) “Correct data overrules theory”? Obviously, if the data is bad, the theory cannot be said to be thereby refuted, can it? Otherwise, what Feyman must have been saying was, “Any and all data, regardless of how good or bad it is, trumps any and all theories every single time if the two are not in agreement.”
Who interprets his words as such?
No one.
The argument, therefore, is not over whether good and proper data trumps theory, but over whether the data is in fact both good and proper. And those are two very different arguments.

richard verney
Reply to  Steven Mosher
April 3, 2016 12:58 am

Is there a human out there who doesn’t interpret Feyman’s words as (essentially) “Correct data overrules theory”?

Bang on!
The sort of comment posted by Steven does him no favours.
Bad data proves nothing, and herein lies the problem that most besets climate science. Almost all the data is bad data, and not fit for the scientific purpose to which it is being put.
The land thermometer record is a classic example. The data is bad because it was never collected under scientific conditions with laboratory standards, and then poor data is so hopelessly bastardised that it has become completely worthless.
A better approach would be to collect a series of data sets using data only from extant stations recording data throughout the entirety of the period in question, and then assess appropriate error bars for those stations which error margins will no doubt vary over time, ie., if one wanted to look at temperature between 1880 to date, one would look at the stations that existed in 1880, identify these and use data from only such stations that still exist today returning data. That might only be a handful of stations, and that in itself will have an impact on spatial/global coverage and that issue also needs to be carefully identified and explained .
If one wanted to look at data from 1900 to date, one would identify the stations that were returning data in 1900 and look at only those that are still extant today returning uninterrupted data and use only and only those stations to create a time series.
The current data sets are meaningless since at no time is one ever comparing apples with apples, the input data and its siting constantly changes, some go, some come, but we are certainly not comparing temperatures today with those measured back in the 1990s, or 1900s, or 1920s, etc.
The data needs completely reworking, and with sensible error margins being clearly stated, and unfortunately B EST missed a golden opportunity to get a proper handle on what is obviously poor raw data .

Philip Mulholland
Reply to  Steven Mosher
April 3, 2016 2:45 am

Your reply is a masterclass in obfuscation.
Thank you. I shall treasure it.

Science or Fiction
Reply to  Steven Mosher
April 3, 2016 3:17 pm

This is the third time I have seen this piece from you, and I have commented on it twice before. I might as well copy my reply from the first time:
What Feynman said here was:
“We’ve heard what you did, and nobody’s found anything wrong with your calculations. I don’t know why Davis’s result doesn’t agree with your calculations, but you shouldn’t be discouraged, because maybe you’ve done something important, we don’t know. I don’t know what the explanation is, but you shouldn’t feel discouraged.”
If the theory is neither corroborated nor falsified, you should keep your theory, but you should also suspend judgement about it. Which is, I think, exactly what Feynman seems to have told John Bacall, the man who was quoted in this interview.

This time I would like to add quote you might like, or maybe not:
“I don´t know, therefore I suspend judgement. I dont believe or doubt. I have no informed opinion. That’s skepticism.”
– Steven Mosher

April 2, 2016 5:20 am

But if the ice stays all summer, as it does during a growing ice age, the albedo factor is very significant as your fig 1 demonstrates (very little cloud on the ice sheets during the ice age).
I am still surprised by the 60% albedo figure, during the 100% ice coverage period. Warren and Svensson in their surface experimentation found fresh ice sheets had more like a 90% albedo. I wonder if this is because the incident sunlight in the polar winter (what little there is) is so oblique, and it is not being measured properly by the satellite.

Reply to  ralfellis
April 2, 2016 10:15 am

But if the ice stays all summer, as it does during a growing ice age, the albedo factor is very significant as your fig 1 demonstrates (very little cloud on the ice sheets during the ice age).
As well as that combo of extensive land ice sheet, sea ice and snow cover, the cloud would also migrate closer to the tropics and equatorial zone relative to where it is now, i.e. the high albedo clouds move to where more of the input energy could have actually been absorbed, but would then be reflected off also.
Despite these feedbacks, this post by Willis confirms what must be true, that the feedback mechanisms of Earth’s weather and longer term climate are smaller than theories suppose, but they can flip ‘rapidly’, but even then the catch themselves again just as quickly, to prevent run-away, or one-way permanent irreversible changes (yes, overprinting cycle-mania does matter).
This has not always happened though for if you look at the period of the Cretaceous to present, the steady global temp drop makes it look like Earth was doomed to become an ice block. And it still could, as much of that cooling has happened recently, geologically speaking, 5 to 15 mya.
So long-term it can seem to not be as stable or buffering, but the closer we get the human scale, the more common the oscillations become (but predating humans by millions of years, so it’s not us doing it), the more stable the feedbacks in those oscillation’s extremes also are.
Fortunately the natural feedbacks the have come into play, and to climate prominence over the past ~6 million years, have served to moderate the extremes of the accelerating planetary cooling-induced oscillations of the Quaternary. If these feedbacks hadn’t been self-moderating and buffering, as opposed to amplifiers only, we wouldn’t be here to talk about the weather, let alone climate.
If we ignore the anthropomorphism involved in that, this fact, that we are here, in a basically 11 k year ‘stable phase’, illustrates that the theories that the earth’s climate is about to spin out of control, is largely bunk. it won’t.
It seems to only become actually extreme (i.e. extreme 10k years of cold) when the orbital arrangements can override/over-power these natural feedback mechanism’s ability to moderate the orbit-induced cooling inertia.

Reply to  ralfellis
April 3, 2016 10:43 am

“I am still surprised by the 60% albedo figure, during the 100% ice coverage period”
Not me. Arctic sea-ice isn’t at all like winter ice on a lake. It is broken up, churned, piled into compression ridges, has lots of cracks and small leads even when classed as “100 %”. There is a lot of shadowed nooks and crannies. Reflection is anything but simple. Have a look at this for example:

April 2, 2016 5:29 am

Concerning albedo of ice/snow, it is important to remember, hard pack ice without snow cover is not bright white at all but shades toward blue which will reduce it’s albedo. Loose ice (fresh snow cover) will have higher albedo.
As for cloud albedo, it is a direct reflection (ha!) of the density (thickness) of the cloud. The thicker, denser the cloud, the higher the albedo. Deep cumulonimbus will have the brightest albedo & thin cirrus will have the lowest with infinite variability in between. It is not a ‘yes/no’ thing like sea ice/water.

Reply to  JKrob
April 2, 2016 6:44 pm

Which means the blue end of visible light, or the higher photon-energy end of the spectrum is reflected. If it’s being reflected, more than red or green, then the reflected albedo is reflecting the shorter wave (ice and water-warming end) visible spectrum, and absorbing the least energetic end of visible light.
However … if ozone blocks UV out … but ozone concentration falls at high latitudes for some reason … will the ice and water absorb more of the incident UV rather than reflect it, and thus actually make a more significant difference than ‘third order’?

Ron Clutz
April 2, 2016 5:31 am

Arctic ice in March was not as low as reported.

Don K
April 2, 2016 5:32 am

Willis, I don’t think you’re wrong, but two points.
First, are you sure that CERES albedo readings in high latitudes are accurate enough to support what you are trying to do? It would seem that a significant portion of the time, the satellites would be looking at or near the sun. That’s unlikely to work very well, so I imagine that the sky area near the sun is masked somehow before you see the data.
Second, I’m pretty sure I’ve seen charts that show the reflectivity of water varies dramatically with sun elevation angle. Sea ice coverage should tend to be negatively correlated with sun elevation angle. Which might mean that part of any apparent increase in albedo from sea ice might actually be enhanced water reflection when the sun is low in the sky.

Reply to  Don K
April 2, 2016 3:31 pm

( … and the water is generally relatively flat, I wish to add to your question, Don)

Mickey Reno
April 2, 2016 6:15 am

Interesting as usual, Willis.
Sea ice is a true blanket that inhibits infrared energy flowing from ocean water to open air. I suspect that the effect of extra sea ice is most influential on “climate” in that it reduces this transfer of infrared energy. Ironically, sea ice warms the oceans, and creates another sort of “greenhouse” effect-one much closer to the mechanics of an actual greenhouse than are so-called greenhouse gasses in the troposphere.
But of course, relatively warmer ocean water underneath extra sea ice will begin to melt more ice from below, at the margins, creating an additional feedback.

bit chilly
Reply to  Mickey Reno
April 2, 2016 3:05 pm

it would be interesting to calculate the volume of ice that has melted since the peak of the late 70’s/early 80’s and work out how much it cooled the north atlantic by . once it cools to a certain point, the sea ice volume begins to increase again and the cycle starts all over again . amo ?

John Harmsworth
Reply to  Mickey Reno
April 2, 2016 7:03 pm

In fact, the heat transfer coefficient for ice is .5 while that of water is 1. At high latitudes the air temp is lower than the water temp through most of the winter. The ice does slow down the heat transfer but not by much. The ice can thicken by a foot or more overnight across a huge area. That is a massive amount of heat transfer.

April 2, 2016 6:38 am

Here’s an issue that I have been waiting to bring up and this post offers the opportunity.
First question: Does the sensing device on the satellite “read” the reflected light from an angle normal (90 degree) to the surface and what would be the “field of view” of that device?
Think of the spotlight from a boat striking the water and being reflected up into the trees on the shoreline. The result is clearly visible. Relate that to the reflection of the sun off water or ice/snow. Much light is reflected at a similar angle as that which it strikes the reflecting surface. Now the point that I am making/or asking is this: Is the satellite picking this up while directly overhead?
The satellite may be picking up more reflected light from clouds due to the fact that clouds are not as flat on top as the land/water surface and therefore offer more of a normal surface angle with regards to the location of the satellite. Even then there will be a dispersion that is not in the direction of the satellite.
Just how accurate are the measurements of albedo? Are we getting an accurate value in the data? Is this where the “missing heat” has gone?
I hope someone will respond as this has been on my mind for years.

Reply to  eyesonu
April 2, 2016 7:25 am

Hey, do your own leg work:
I think CERES is part of the A train of satellites, in near polar orbit. The downward viewing instruments are usually protected by closing a trap when in new dawn and dusk situations but will be picking up direct reflection before hitting this geometry.
I think albedo will be assessed from wide field of view ( WFOV ) sensors.

Reply to  Greg
April 2, 2016 9:14 am

Thank you for the link. I will spend more time with it later.
What I got from an initial look is it was released in June 1997 (19 years ago). There must be a more recent link so that I can do my own leg work. But then another viewing this thread may know and comment with answers that others would find interesting and add to the overall value of the thread.
Anyway, I’ll continue to do my own legwork as you say.
By the way, what is the value of the “wide field of view ( WFOV ) sensors you think they use”. I would prefer an actual number over just a thought or guess.
Again thanks for the link.

Reply to  eyesonu
April 2, 2016 7:31 am

The idea of missing heat is nothing (little) to do with energy budget at the poles.
I comes from having incorrectly diagnosed climate sensitivity to both volcanic aerosols and CO2. Amongst other things because they cannot a linear regression correctly and screw up the fitting and get spurious results which they then base their models on.
It is not missing heat in the real world but spurious retained heat in models.
It looks like we have about another 20y to go before they admit that possibility.

Reply to  Greg
April 2, 2016 8:51 am

The idea of missing heat has everything to do with reflected solar energy that may not be accounted for at the poles (as well elsewhere) that is not identified/captured by the satellites on a global scale. Any missed energy reflected away and not accounted for adds to the modeled global energy input that Trenberth can’t seem to find. I would suggest he look here if he hasn’t done so.

Reply to  eyesonu
April 2, 2016 9:25 am

I can see that there are numerous comments added above that also bring up some of the same questions that I have in my comment posted @ 6:38 am. This may well be a very interesting and informative thread. Willis has a way of causing/creating that. 🙂

John Harmsworth
Reply to  eyesonu
April 2, 2016 7:24 pm

You have hit upon a point that I find curious. Why oh why can’t we measure energy input to the earth (I know we do), measure energy radiated from the earth and compare these two numbers? If the most recent measurement of energy flow from Earth to space is 1997 then I would suggest they don’t want to know. With 7B being spent annually on flaky climate studies, there must be money to measure reality instead of creating more models that don’t work.

Reply to  eyesonu
April 3, 2016 6:59 am

@ John Harmsworth
The link offered by Greg above was some of the details regarding the CERES satellite and the report was released in 1997. Without going over the report’s various pdf contents again I’ll just note that there were references to future satellites that were not yet launched at the time. I stopped reviewing the various pdf attachments after a cursory look and was hoping that others on this thread may be more informed and would comment. I don’t recall seeing anything with regards to actual data being collected. I believe the link was more of a conceptual planning report. It is very interesting to read if you have the time. I’ve been busy this weekend and this thread also has my attention. This is a very interesting topic being discussed here. Please verify/question what I have written in this comment as it is from a limited peek at the info provided.

April 2, 2016 6:49 am

As a rough analysis regarding global influence, this is interesting.
My question/concern is regarding how the ice concentration is established for the data points in the scatter plots. My understanding was that the ice concentration is estimated based on assumptions regarding the albedo of the pixel being evaluated. If that’s so, then this is a bit of circular analysis, isn’t it? The ice v water albedo assumption is cooked into the data to get the ice concentration parameter.

April 2, 2016 7:53 am

Science obviously already knows this. The biggest contributor to planetary albedo is by far atmospheric albedo. Surface albedo has a small contribution from Antarctica which is icy at all seasons, and an even smaller contribution from Greenland. Arctic sea ice contribution is negligible as you say.
Figure from:
Donohoe, Aaron, and David S. Battisti. “Atmospheric and surface contributions to planetary albedo.” Journal of Climate 24.16 (2011): 4402-4418.

Reply to  Javier
April 2, 2016 2:41 pm

“Science obviously already knows this.”
True, but it’s bingo night in the Church of On-line Rediscovery.

Reply to  Willis Eschenbach
April 2, 2016 5:48 pm


Naturally, the size of this ice-albedo feedback is of interest.

And this, of course, means that the effect of the ice-albedo feedback is vanishingly small globally.

Science obviously already knows this. This is not surprising to anybody with some knowledge of planetary albedo, even from perusing the bibliography, like me.

I fear your citation says nothing about the subject of this post, which is the actual measurements of the relative size of sea ice reflections. Not one single word. Instead, it is a study of some CMIP5 model results.

The figure posted shows observations as a light blue line. The Arctic surface contribution to planetary albedo is very small (figure 6c right side). Arctic sea ice contribution is only a small part of that as its surface is very much reduced during the summer and there is water over part of that ice then.

Reply to  Javier
April 3, 2016 11:12 am

““Science obviously already knows this”
Since the paper You refer to is 100% based on climate modelling with no actual data being used this seems a somewhat optimistic claim (OK, there is an Appendix where they claim that errors in the measured data in no way affects their results, which is a rather amazing idea in itself).

Reply to  tty
April 4, 2016 3:12 pm

You are incorrect, tty,
The paper I refer compares the model results to observations from ERBE and CERES satellites. Observed values are reflected in the figure above as observations in “light blue.” Hint: Do a search for “observations” in the paper.
Methods to calculate surface albedo from planetary albedo have been developed since the late 70’s. It is therefore common knowledge that surface albedo is a small part of planetary albedo, and that Arctic sea ice albedo is a small part of surface albedo. It is obvious then that Arctic sea ice albedo is a very small part of planetary albedo. Willis and you are free to think this is new and noteworthy, but it is not.
What scientists are discussing about is if the decrease in albedo due to Arctic sea ice melting is strong enough to increase the rate of Arctic sea ice melting, in other words how strong a positive feedback it constitutes.

Bill Illis
April 2, 2016 8:45 am

There is a large variance depending on the time of the year and whether the solar angle is low (and/or non-existent as in the winter north of the Arctic Circle).
Here is the Zonal Mean Annual Albedos and then how much it varies across that latitude (as in Greenland and Antarctic glaciers which are very high – with and without clouds).
From Earth’s Thermal Environments (which is used to control satellites).

Dr. S. Jeevananda Reddy
Reply to  Bill Illis
April 3, 2016 5:57 pm

With overcast sky, these will be modified as the angle of incident of rays will be reduced — angle of incidense reduced — and because of this the blue gradient is less than red [on ice].
Dr. S. Jeevananda Reddy

Pamela Gray
April 2, 2016 8:57 am

On the other side of this coin, (warning: mind experiment) when Arctic sea ice is extensive during the winter, the necessary polar region ocean heat release is diminished. By how much I don’t know but that ocean water that traveled there to get cooled off will eventually circulate back out of the Arctic into global currents again, less cooled. This change, over a very long span of time is of interest. The reverse of that is when this “too warm” ocean water begins to change weather patterns and starts to eat away at the very cap that kept things from cooling off. Under this scenario, the necessary polar region ocean heat release is enhanced. By how much I don’t know but that ocean water that traveled there to get cooled off will eventually circulate back out of the Arctic into global currents again, much cooler than when it arrived.
The winter premise for this mind experiment: The most important evaporative Arctic period is Winter, because the air is colder than the water. Less ice, more surface area for heat release. More ice, less surface area for heat release. During the summer, any changes in extent will not necessarily lead to effective changes in evaporation.
Here is the mind experiment.
Initially: Extensive Arctic sea ice in Winter leads to less ocean cooling which leads to global increase in total stored ocean heat over long periods of time. But because of diminished ocean heat release, we have to bundle up because our heater isn’t sending us much heat.
Subsequently: Extensive Arctic sea ice in winter and less ocean cooling leads to polar vortex staying at home and keeping snow and cold at its breast, uncovering New York so that it can build a city and beckon one and all to come to its shores. At the same time, while the oceans are gobbling up heat and storing it, some of us are living in cold, dry conditions.
Initially: Diminished Arctic sea ice in winter leads to more ocean cooling which leads to global decrease in total stored ocean heat over long periods of time. But it keeps land surfaces basking in released ocean warmth and water vapor.
Subsequently: Diminished Arctic sea ice in winter and more ocean cooling leads to polar vortex looping out of control and sending snow, cold, and land ice advance out to get us which eventually covers New York with a mile of ice, sending one and all to live in Cancun. At the same time, while the oceans are releasing heat, some of us stay warmer and wetter. But not those that live under the loop. They are getting hammered with cold and ice.
Finally: Millennial swings in this imbalance are demonstrated as stadial cold and interstadial warm periods. Will ponder ENSO connections later.
Which leads me to say, I think Columbus should have discovered the West Coast first. Here’s a thought: move the lady in the harbor to just off shore in California. Might as well. They let anybody in there anyway.

Pamela Gray
Reply to  Pamela Gray
April 3, 2016 8:05 am

For those interested in the process of evaporation related to Arctic and subarctic conditions–
Changes in Arctic albedo due to Arctic ice is, as Willis has demonstrated, a fairly small thing in terms of global climate change. However, changes in evaporation rates in the Arctic and subarctic zones may indeed drive some aspects of global climate change. In this case, the transport of collected surfaced warm water to these areas (think the Blob), the less ice there is, the greater LOSS of heat, which is anathema to the mantra of global warming. I doubt we will see any attention paid to this until the current crop of AGW climate researchers move on to a nursing home.

April 2, 2016 9:02 am

Looking at the visible range I took reflective intensity from earth photos. They do not vary much between center and edge of image. Here are some averages over typical and very large areas. I have shown them against dark and white backgrounds for comparison. Its amazing how dirty snow really is. colors are true as well so you can see how the “blue planet” is not really very blue at all.comment image?dl=0

Reply to  Kirkc
April 2, 2016 10:00 am

Wow, just imagine all the billions wasted on putting carefully calibrated spectrometers into space when all we needed to do was scan a few photos. Sish.

Reply to  Greg
April 2, 2016 8:20 pm

Wow. That was a dick comment. And by dick I mean condescending. They ARE scans from a highly calibrated satellite a million miles away. That was my point. You will notice how closely they match the OP numbers. If you can show something is wrong with the results that’s fine.

Reply to  Greg
April 2, 2016 10:25 pm

Gotta agree – that was a super dick comment. Dang, dude!?

Reply to  Greg
April 3, 2016 12:51 am

You were the dick who presented that as “earth photos”, without the slightest indication of where they came from. It would be Apollo astronauts shooting through the LM window !
If you want to be taken seriously, you need to say what you data represents and where it came from. It may ( or may not ) then have some relevance.
Calling people dicks because you failed to say what you were presenting is both offensive and uncalled for.
We still don’t know what these fuzzy grey-blue dots are. Pretty meaningless.

Reply to  Greg
April 3, 2016 12:54 am

How can anyone say what may be wrong with your “data” is we do not know what it is?

April 2, 2016 9:20 am

Thank you Willis for your analysis of the effect of ice on albedo, essentially nada due to the high latitudes. An even more important consideration is the energy flow. This concept gets lost in all of the high amplitude noise floating around the world about climate. You re-stated in your article the oft-exalted mantra: “It is important because the more the ice melts, the less solar energy is reflected, the warmer the ocean becomes, and this melts even more ice, and so on.” You then proceeded to show that the stated mantra is inconsequential. It is not inconsequential. It is flat out wrong. Consider the Stephan-Boltzman Law at the interface with the water. For a given temperature, there is a certain amount of photonic energy that must be absorbed by the material to maintain that temperature. If more energy comes in, the temperature increases. If less energy comes in, the material temperature drops. So little light comes in a such a low angle to the horizon even in the summer months at the poles that the water temperature is far above that equilibrium point. The ice is an excellent insulator so as soon as it is gone, evaporation and direct radiation into the atmosphere begins. The mount of photons coming from the sun is far below that necessary to maintain the existing water temperature much less warm it. The water is at least 100C above the equilibrium temperature according to radiation balance. The water radiates into the local atmosphere. What doesn’t escape directly into space is eventually radiated into space by the atmosphere. Since the warmth of the sea water originated at the equator and flowed to the poles, there are three rules that define the energy flow at the poles: 1) Sea ice maintains a higher temperature balance across the world by impeding the loss of stored energy to space. 2) The loss of all sea ice at the North Pole dramatically increases the rate of energy loss to space from the whole of the Earth’s surface environment. Hold on to your hat because that means the coming ice age is accelerating. 3) The feeble number of photons hitting the surface of open water in the poles is smaller than inconsequential, a point which you proved in your article. A first corollary to #2 is that higher atmospheric temperatures at the North Pole means higher energy loss to space from Earth’s ecosystem. A second corollary to #2 is that the North Pole controls Earth’s environment and the onset of ice ages because water can flow all the way to the pole. That mechanism cannot happen at the South Pole because it is land (or at least an ice mountain on submerged land) that blocks water flow. That makes Antarctica a cold storage point on Earth’s surface with super long time constants.
Think of this simple paradox to help solidify an understanding of the true mechanism occurring at the North Pole. This paradox helped me sort out all the controlling factors for the flow of energy in the Earth’s ecosystem. A) As you descend into a mine, the temperature rises due to heat rising to the surface from the Earth’s core. B) As you descend into the ocean everywhere, the temperature decreases even though it is in line with the same heat flow from the Earth’s core. Why?

Reply to  Joe
April 2, 2016 10:11 am

You then proceeded to show that the stated mantra is inconsequential. It is not inconsequential. It is flat out wrong.

I’m not sure anyone was suggesting it was a consequential part of the global energy budget, so maybe the result is not that important. The usual argument is that it is ” the canary in the coal mine” . That generally means something that is very sensitive to small change. No one says that it is a huge loss to the functioning of a mine if they loose a canary. The point is that it pre-warns change , not that it is a major part of the change.
The point that it is flat out wrong is probably more pertinent, in that it shows how little real scientific knowledge there is ie. capable of *predicting* events ( especially future type ones ! )
The point of having a canary is that you know why if falls over and what it means. If it suddenly gets better and you shut down the mine for nothing, it’s less than useless.
Obama and the EPA are busy shutting down the mines. Quite literally.

Reply to  Greg
April 3, 2016 12:45 am

The Arctic is an epileptic canary, not a dead parrot.

April 2, 2016 9:35 am

Willis: The “consensus” agrees with you. Ice loss leads to more clouds, which negates much or all of albedo loss.
While there are reservations in the paper (using models), measured data shows a negative feedback with a reduction in sea ice.
“”Using satellite data of the 1982–98 period in the area north of 60°N, Wang and Key (2003) found a significant negative trend in the surface albedo in the Arctic during the spring and summer. The authors claim that the expected enhancement of the surface net radiation imbalance was reduced or even cancelled out by a concurrent increase in cloud amount, as well as more frequent occurrence of liquid phase clouds.””
Liquid phase clouds are important, as these types have a higher albedo than ice crystal clouds.
From the conclusion:
Although based on a limited time series, these results suggest that changes in Arctic sea ice are compensated by changes in cloud cover, perhaps, as a result of enhanced evaporation from the sea surface, therefore, leaving the TOA energy budget unchanged. The implications are that any ice-albedo feedback could be dampened because of increased cloud cover and such responses should be sought in climate simulations.
These authors also show a that net increase in shortwave radiation at the surface, is reduced by increasing cloud cover.
Wang, X., and J. Key (2005b), Arctic surface, cloud and radiation properties
based on the AVHRR polar pathfinder dataset: Part II. Recent trend,
J. Clim., 15, 2575– 259
Over the measured period, there was a statistically insignificant decrease in reflected SW radiation (2.0 ± 2.0 Wm2), which matched well with the calculated changes (2.2 Wm2)” Reduced ice causes more cloud, and the resulting albedo change mitigates warming in the arctic.
The more the ice melts, the cloudier it gets…
Loss of ice increases cloud cover.
The arctic sea extent is highly correlated to early summer cloud cover. Also, this is not in the models.
This intimate delayed ASR-SIC relationship is not represented in most of current climate models. Rather, the models tend to over-emphasize internal sea ice processes in summer.

Scott Scarborough
April 2, 2016 9:57 am

I don’t see anything in this analysis that considers the angle of the sun at the poles. The albedo of the sea surface should depend on the angle of the incident sunlight – not just the watts/Meter2 intensity of the sunlight. Was that considered and I just missed it?

Reply to  Willis Eschenbach
April 3, 2016 12:29 am

Willis, none of this data processing is “assuredly” anything ( apart from wrong ). Sometimes there are some pretty stupid assumptions, sometimes less stupid but always lots of assumptions and simplifications.
When I looked at ERBE in detail I came out screaming. You don’t see this in the nice simple time series they put out for general consumption:comment image
That massive hump is an alias of the daily cycle, introduced by them assuming constant cloud cover throughout the day in the tropics in doing the SW extraction !!
I’m sure you know how realistic that is.
I have no idea why it does the 3-2-3-2 pattern on top of that but I would not mind betting is just another alias. They *mask* this by taking 36d averages.
As you know there is also a massive TOA imbalance in CERES. I’m not very ‘assured’ by any of this.

Reply to  Willis Eschenbach
April 3, 2016 12:39 am

The problem is there will be quite a bit of low incident reflection that is not captured by measurements assuming an even, diffuse reflection.
How that is handled I don’t know, but I would guess that quite a lot of out-going reflected solar in not getting counted in clear sky situations.
If they do start getting some stronger readings when the satellite is nearing a position where it would pick it up, I would not mind betting that the readings get excluded by some kind of automatic QA algo that eliminates anything > 2 std dev away from the mean.
Again in looking at ERBE data around Mt P event, there was some data getting thrown that looked like it may have been legitimate deviations.

Reply to  Willis Eschenbach
April 5, 2016 12:17 pm

Thanks for the link to (CERES_EBAF_Ed2.8_DQS-1.pdf) as well as the “google search” suggestion. That along with the link provided by Greg has been very interesting reading. I would likely be out of my league to respond but there my be some issues with the certainty of the values in the data (at least from my point of view). This is not meant as a takeaway from your using data in lieu of modeled “data”.
Please keep in mind that I’m just a serious ‘lurker’. I hope there will be other discussions with regards to the CERES data.
Just sayin’.

April 2, 2016 9:57 am

Willis: Very interesting analysis. You wrote: “This is the result of what I mentioned above, that when there is a lot of ice there is little sun, and vice versa.” This statement appears to be incorrect for sea ice. The maximum and minimum extent of sea ice occur near the spring and fall equinoxes. This is when an AVERAGE amount of radiation is being received. So solar radiation and sea ice coverage are 90 deg out of phase with each other. Lots of sea ice remains near the summer equinox when incoming SWR is strongest. (This doesn’t mean that your calculations are quantitatively wrong. The generalization you use to explain your results is misleading.)
Seasonal snow cover responds more quickly to changes in irradiation. However, as best I can tell, your post deals only with the global impact of areas covered by sea ice, not seasonal snow cover.
(One interest fact about ice-albedo near the poles is that almost all of the solar radiation is indirect, not direct. Indirect radiation is due to scattering by particles in the atmosphere and comes from all directions. There are no strong shadows at the poles, just like there aren’t near sunrise and sunset at lower latitudes when sunlight must pass through a lot more atmosphere. This is important because the albedo of ice and water varies with angle and effects those who calculate the expected change in albedo. You sensibly are relying on observations, not theoretical calculations.)

Reply to  Willis Eschenbach
April 3, 2016 12:04 am

Willis: This graph (from the WUWT Sea Ice Page) shows that the minimum and maximum sea ice for the Arctic are in mid-September and mid-March, very near the equinoxes. Note the large amount of sea ice in May and June. Coverage doesn’t drop below average (about 10 million km2) until July when solar insolation has begun to drop. As I said above, almost 90 deg out-of-phase.comment image

Reply to  Willis Eschenbach
April 3, 2016 12:16 am

Arctic sea ice minimum is usually 2nd to 3rd week of Sept, not August. I would say his estimation of 90 degree phase lag is not too far off. Not surprising when ice area is simplistically at least an integration of the solar input.
However, I don’t thinks his assumption that the equinox are “average” points in the cycle is correct. It is not nice even sinusoid. Solar input is one bump of a cosine with a long flat section. The cut off dependant up on latitude.

Reply to  Willis Eschenbach
April 3, 2016 12:18 am

Thanks Frank you beat me to draw as I was typing.

April 2, 2016 10:24 am

Albedo? Thanks Willis for this post – I actually had to look up the word albedo. Just some basic questions:
At what angle of the sun to ocean water is the reflection equal to the absorption? Is incoming light and incoming solar radiation (heat) equal to each other as far as reflectivity?
I assume that the sun at 90 degrees above the ocean reflects about 10% of the light (and I assume solar energy ie. heat is the same) When the sunlight comes in at say 5 degrees above the horizon, how much light is reflected and how much is absorbed? Is it the same for solar heat energy at that angle? Also is it the same if the ocean is rough (with whitecaps) vs smooth ocean water?
I never studied this, but am just fishing for information. My first reaction is that the nearer the ocean water is to the North (or South) Pole the less the angle of solar light/radiation to the water and therefore the lesser the water’s absorption of incoming radiation.
I know a lot of this has already been discussed, but these were my thoughts as I was reading the post.

April 2, 2016 10:33 am

Willis: Many readers may confused by the units you use in Figure 3 (PW 1 E15 W). Usually we discuss heat fluxes in terms of W/m2. Given the surface area of the planet (510,072,000 km2 or 5.1 E8 km2 or 5.1 E14 m2), means that 1 PW is 2 W/m2 and that Solar Reflectance (the black line) is about 100 W/m2 – which is the normal value. So the sea ice reflectance is 1.1% of that value or about 1 W/m2. That number is much easier to for me to think about.
The IPCC’s best estimate for ice-albedo feedback is 0.3 W/m2/K. If ALL of this feedback came from loss of sea ice, 3 degK of global warming would produce 1 W/m2 more reflected SWR. However, your 1 W/m2 value is a combination of surface albedo and cloud albedo feedback over areas covered by sea ice. Increasing clouds cover neutralizes about half of the change in surface (30% vx 18%), so your surface albedo change due to sea ice is about 1.6 W/m2. In that case, if ALL of this feedback came from loss of sea ice, 5 degK of global warming would produce 1 W/m2 more reflected SWR. If half of ice-albedo feedback come from loss seasonal snow cover and ice caps on land, then 10 K of warming would be needed would produce 1.5 W/m2 less reflected SWR from sea ice.
So, do you agree that nothing in your results that suggests that the IPCC’s value for ice-albedo feedback is unreasonable.

Reply to  Willis Eschenbach
April 3, 2016 12:25 am

Willis: The IPCC chooses to divide “albedo feedback” into cloud feedback and ice-albedo (surface) feedback. You have chosen to combine both in polar regions covered by sea ice. There is nothing “wrong” with either approach, but by combining them you lose the ability to compare with the IPCC’s values. The IPCC’s values come from climate models, which are underestimating loss in the Arctic and failing to predict gain in the Antarctic. Your numbers come from observations and therefore are interesting to compare with the IPCC’s. I’m looking forward to the land analysis and combining them.
Given the vast differences between the Antarctica and Arctic, it might be interesting to see if they are behaving in the same manner.

April 2, 2016 10:38 am

Willis: I have a frustration in my personal life, regarding “housing” right now. It is NOT a crisis. BUT I am an “experimentalism” person at heart. (Despite the extensive engineering background.) And something just dawned on me in the last few weeks..which needs a “constructed device” which demands several tens of hours of work…but could be very revealing. I present it to you to either do it, or find someone who has. This would be similar to the 1910 Robert Woods “greenhouse” experiment, but a minor twist. To build a “Solar Collector” with a PTFE (I believe that is the RIGHT longwave transparent material to use) top. A sealed device, festooned with thermocouples.
First to run it a a set angle and a set time. The angle might have to vary slightly, day to day to make
an “even” solar input.
It would have AIR with a measured humidity in it (30% RH at RT to start?) Then, to run it in a parallel condition, with CO2 and 30% RH in it (or the same amount of water total.)
My guess is, the temperature profile will be virtually IDENTICAL. This will validate the concept that
97% of IR trapping is H2O and also H2O is the primary upflux agent.
This simple experiment would show that CO2…at levels we may POSSIBLY experience has a complete and tolerable limit with regard to the IR balance shift (or rather that it will be miniscule)

Reply to  Max Hugoson
April 2, 2016 4:08 pm

the only problem is this experiment doesnt test the Green house effect.

Reply to  Steven Mosher
April 2, 2016 6:53 pm

Yes it does. AND the problem with your comment Mr. Steve is that it it ISN’T the “greenhouse effect”. I refer you to pages 232 and 233 of Fleagle and Businger (Introduction to Atmospheric Physics) as found here: Get your terminology right, and then we’ll discuss particulars!

charles nelson
Reply to  Max Hugoson
April 2, 2016 7:17 pm

I call these people Water Vapour Convection Cooling ‘deniers’!
And I advise them to look at the behaviour of clouds over the equatorial oceans.
The idea that raising the levels of (the poorly named) Greenhouse gas CO2 to 1/25th part of ONE percent of the earth’s atmosphere could impact on this almost unimaginably powerful planetary heat transfer system is laughable. Like homeopathy is to real medicine!

Reply to  charles nelson
April 4, 2016 2:27 pm

Indeed, radiation worshipers fill the pews in “climate science, while the dominant mechanism in cooling the surface–moist convection–is barely understood by self-ordained high priests. Max Hugoson is spot on in pointing to the passage in Fleagle & Businger that warns: “However, it is not commonly recognized that,
whereas the absorbing effect of the atmosphere results in temperatures well above what they would be without an atmosphere, the high temperatures in a greenhouse are not to be attributed to absorption of long-wave radiation by the glass.” Real-world heat transfer is by no means that simple!

April 2, 2016 10:44 am

Very good, Willis, as usual. I’ll give a couple of my thoughts on positive feedback. I started modeling when the only computer available to me was analog (an EAI PACE 231R, for those of you who are also older than dirt). These operated with vacuum tubes, which often failed. Tube failures were announced by a loud horn which warned the user and repair technician (and every one else in the room) that all subsequent results were to be ignored. An operating simulation that drifted to its control limit (100v.) would also turn on the horn.
Any integration loop with positive feedback would also QUICKLY and LOUDLY sound the horn. The only question then became “did the operator screw up, or did a vacuum tube fail”. Operator screw-ups were typically an integration loop with positive feedback. A beginning programmer would quickly, loudly, and often, announce his incompetence, to everyone in the room.
The point is that positive feedback will usually quickly be driven to its limit of control. Think of a thermostat that turns the furnace off, instead of on – it’s all over, once you feel cool. Granted, in a complex control system with a multitude of variables, other factors can possibly compensate. But positive feedback also, can quickly overcome a whole lot of subtle interactions. In a climatic situation with several billion years to stabilize, it’s difficult for me to believe that anything but negative feedback dominates the control buttons.

Reply to  Tom
April 2, 2016 2:45 pm

“In a climatic situation with several billion years to stabilize, it’s difficult for me to believe that anything but negative feedback dominates the control buttons.”
Bingo in the universe of reality!

April 2, 2016 11:43 am

Great work, Willis. Sacred cows are braying.

Michael Carter
April 2, 2016 12:24 pm

Willis –
What influence will angle of incidence of incoming and reflected energy have? Some of this gets reflected out to lower latitudes during a good part of the year. It will fall outside your window right?
Nature never makes it easy 🙂

April 2, 2016 12:33 pm

I suggest that wuwt readers that think as much of Willis Eschenbach as I do write to their congressman and direct to the National Science Foundation and suggest they give him $100,000 grant, unrestricted other than the work be related to climate science. They’d get results equal to any $10,000,000 they ever spent.

Reply to  Jimtech
April 4, 2016 7:32 am


April 2, 2016 1:02 pm

I think this means I can quit worrying about ice melt due to carbon soot causing global warming or sea level rise.

charles nelson
April 2, 2016 1:33 pm

Maurice Ewing was of the opinion that reduced (winter) Arctic sea ice cover enhanced heat transfer between the Ocean and Space.

April 2, 2016 2:16 pm

Willis wrote: “And this, of course, means that the effect of the ice-albedo feedback is vanishingly small globally. It is certainly possible that it makes some larger difference in the immediate neighborhood of the ice, but in terms of a global effect, it is what I call a third-order variable.”
A closer look shows this statement is grossly wrong.
How much has global sea ice changed? The WUWT sea ice page says that global sea ice has declined perhaps 2 million km2 since 1979 or about 10%. However, the signal is very noisy. In 2014 and early 2015, sea ice coverage was about the same as in the 1980s. Most of the time, however, it has been lower. . Although a value between 5% and 10% decline might be more appropriate, I’ll stick with 10% for simplicity and readers can easily adjust my calculations to suit their views. During the same period, GMST has risen about 0.5 K. So we can say that global sea ice coverage changes about -10%/0.5 K or -20%/K. This predicts that all sea ice will be gone when GMST has risen 5 K – a reasonable estimate when considered in terms of GMST 30 million years ago when there were no polar ice caps or sea ice. Using the information in this post, this reduction is about -0.3 W/m2/K without including cloud feedback above the sea ice or -0.2 W/m2/K if you include the cloud feedback over sea ice (as Willis did).
-0.3 W/m2 is the IPCC’s best estimate for ice-albedo feedback, so Willis and the IPCC are roughly in agreement here. Ice-albedo feedback includes reduction in seasonal snow coverage and ice caps, so the information in this post actually suggests an ice-albedo feedback that could be MORE negative than -0.3 W/m2. So why is Willis telling us that ice-albedo feedback is unimportant, when the IPCC takes it very seriously?
Here Willis is getting approaching “Big Lie” territory, because AGW is all about climate sensitivity – the change in GMST with a change in forcing (W/m2). Ice albedo feedback contributes to climate sensitivity. For clarity, let’s first discuss climate sensitivity in terms of K/(W/m2) rather than the familiar K/doubling. An ECS of 3.7 K/doubling (roughly the IPCC consensus) is 1 K/(W/m2). For those who think energy balance models provide a more realistic estimate (Otto et al, Lewis and Curry), ESC is about half this value or 0.5 K/(W/m2). Those who don’t believe any feedbacks exist are saying ECS is no worse than 1 K/doubling or about 0.3 K/(W/m2). These numbers make more sense physically if we consider their reciprocal, called the climate feedback parameter, which is about 1, 2 or 3+ W/m2/K (for the consensus, lukewarmers, and non-believers.) The climate feedback parameter tells us how much more NET radiation escapes to space for each 1 degK rise in surface temperature. The S-B eqn tells us that a blackbody near 255 K has a climate feedback parameter of +3.2 W/m2/K. To obtain the earth’s climate feedback parameter, we need to add the other feedbacks to Planck feedback (with negative signs for feedbacks that decrease OLR or increase absorbed SWR. Surface albedo decreases absorbed SWR.) Then we take the reciprocal to get climate sensitivity in K/(W/m2) and multiply by 3.7 W/m2/doubling.
So we need to consider how important an ice-albedo feedback of -0.3 W/m2/K (or more negative) is on a planet where the climate feedback parameter is 1 or 2 or at least 3 W/m2/K. We are talking about a 30%, 15% or up to 10% contribution to the planet’s climate feedback parameter. By Willis’s definition, ice-albedo feedback is a “first-order variable” from the consensus, lukewarmer, or non-believer perspective.
So how can Willis fool us so easily? GMST is 288 K. A 1% change in GMST is huge. GMST is the result of 240 W/m2 of post-albedo SWR. From the S-B equation for a blackbody, we can derive the relationship dW/W = 4*(dT/T). In other words, a 4% change in radiation produces a 1% change in temperature. The earth doesn’t behave like a blackbody. If climate sensitivity were 4 K for a doubling of CO2 – 4X the “climate sensitivity” of a blackbody, then the relationship becomes dW/W = dT/T. A 1% change in radiation produces a 1% change in GMST. If you are a lukewarmer, a 2% change in radiation produces a 1% change in GMST. If one keeps these simple relationships in mind, it is easier to not be led astray.
The other trick is that Willis presented sea ice change without discussing the warming that produced it – as a forcing (W/m2) rather than a feedback (W/m2/K). If you want to understand the importance of feedbacks, you need to think in terms of the climate feedback parameter – roughly 1, 2, or 3+ W/m2/K depending on your perspective – rather than climate sensitivity.
Of course, the situation is complicated. Relatively little warming is occurring in Antartica (except for the Peninsula) and sea ice is generally increasing there. Much more warming is occurring in the Arctic, where sea ice is generally decreasing. Natural variability is huge in both locations. Fortunately, Willis was brilliant enough to study at the relationship between reflected SWR and sea ice coverage throughout all of this natural variability. If done correctly, all my calculation would have large confidence intervals. But confidence intervals won’t make ice-albedo feedback is a “third-order variable”.

Steve Fitzpatrick
Reply to  Frank
April 4, 2016 3:01 pm

Putting aside the rather rude tone of your comment (which I think completely inappropriate), you should re-think your calculation, which is simply mistaken. A change in albedo of 1.1% of the total current albedo (assuming a 100% loss of sea ice) is identical to a net change in solar intensity of 1.1% of 30% of the current solar flux, or about 0.011 * 0.3 * 1365/4 = 1.12 watt/M^2 averaged over the earth. We can compare this to the estimated forcing from a doubling of CO2…. about 3.7 Watts/M^2. So complete loss of sea ice would generate roughly 1.12/3.7 = 30% of the forcing from a doubling of CO2. If you want, you can look at it in an even simpler fashion: Based on climate sensitivity of 1C per watt/M^2 (the average of GCM’s) an increase in solar intensity of 1.12 Watt/M^2 (equal to the complete loss of sea ice) would yield an equilibrium warming of about 1.12C. If empirical estimates of sensitivity are correct (sensitivity a bit under 0.5C per watt/M^2), then the expected rise from a 1 watt/M^2 increase in solar intensity would be ~0.56C. More realistic losses of sea ice, say 25%, would of course lead to proportionally lower warming effects. Willis is correct in his analysis: modest changes in sea ice are not that important globally. There could be more important local effects, since the local change in solar flux is much higher than the average.

April 2, 2016 2:22 pm

Hi Willis, trying to understand something you said.
“I was surprised to find that the clouds are brighter (greater albedo) than the ice itself.”
My problem is I don’t see data about cloud albedo. “All sky albedo” includes the clouds.
In the explaining text with the figures it is stated: “The data is composed of the 12 monthly averages for each gridcell. There are 11,646 gridcells (1°x1°) which contain sea ice at some point during the year”.
If the ice coverage would be 100% in the gridcells I think yes you may conclude “he clouds are brighter (greater albedo) than the ice itself”.
As I read the explaining text that means on average there is no 100% ice coverage in the grid cells, as there only has to be sea ice at some point of the year. That would mean the “all sky albedo” could be higher for no other reason than extra albedo because of cloud cover over average areas withing the grid cells without ice.
Well, clumsy explanation I guess I am not well into this stuff, guess I am missing something here.

Reply to  Willis Eschenbach
April 3, 2016 11:34 am

Thanks for the reply Willis, I’m reacting a bit late, being outside today enjoying some spring weather at last in Holland.
Yes, I missed that “at all ice concentrations”. It makes sense to me now.

April 2, 2016 2:38 pm

WOW, there are some guys with really big brains throwing around a lot of ideas and working on this. I think I’ll have another drink (VO & Seven) and ponder all this. Thanks.

April 2, 2016 2:49 pm

..Oh boy !! 30 cm of snow coming to Southern Ontario by Sunday !! …..Dang Glo.Bull Warrming !! AAAARRRRrrrrrrrrggggggg…………

April 2, 2016 3:15 pm

I was surprised to find that the clouds are brighter (greater albedo) than the ice itself. At all different amounts of ice coverage, including 100%, the albedo with clouds is greater than the surface albedo of just the ice itself.
that is surprising, would like to see someone knowledgeable expand on the ramifications.

Reply to  dmacleo
April 2, 2016 4:22 pm

Ahh.. that answers part of my question to Willis, I missed that “at all different amounts of ice coverage”. Should be interesting to see the difference in albedo compared to the difference in ice coverage.

April 2, 2016 3:32 pm

Willis would you like to have some insight into what the effects a ice age weather set up would have on the Arctic ?.
Well next weekend it looks like we will get the chance. Because if the jet stream forecast is correct, Then there is a “weapons grade” ice age weather pattern formation about to turn up. l have waited 3 years for the weather to “play ball” to show about my ideas on ice age formation, and now it looks like the insight will be better then a dared hoped.

April 2, 2016 4:12 pm

Willis writes:

This shows that on average, sea ice is only responsible for 1.1% of the total solar reflection.

I think you’re forgetting that climate changes *are* small. A 1C change in global average temperature is equivalent to a 0.34% change in temperature. Your calculation of sea ice effect on solar reflection is thus equivalent to a > 3C change in global temperature measured in percent.
It’s common in discussing climate for people to dismiss small effects not realizing that a 3C change in global temperature is ‘merely’ a 1 % change. Yet 3C would take us back to an era when sea levels were dozens of feet higher than today.

Reply to  oneillsinwisconsin
April 2, 2016 4:38 pm

..” 3C increase is 1% ” ?? Is that Common Core Mat !!

Reply to  Marcus
April 2, 2016 7:18 pm


Reply to  Marcus
April 2, 2016 8:20 pm

Marcus – you can’t use Celcius and do a percentage calculation. You have to first convert to Kelvin. Using 16C as the average global temperature that converts to 289K. 1% of 289 is 2.89. Using the percentage Willis gave of 1.1% would be 3.18. Take your pick.

Reply to  oneillsinwisconsin
April 2, 2016 4:55 pm

..1% of 58C = 0.58C

Reply to  oneillsinwisconsin
April 2, 2016 5:02 pm

58C + 1% = 58.58C, NOT 61C !

Reply to  Marcus
April 2, 2016 5:23 pm

58C = 331K
1.01 time 331K = 334.31K
334.31K = 61.31C
58C + 1% = 58.58C, NOT 61C !

Reply to  Marcus
April 2, 2016 5:28 pm

..LOL..1% of K is not the same as 1% of C…

Reply to  Marcus
April 2, 2016 6:22 pm

Next your going to tell me that 1% of 100 kilograms is the same a 1% of 100 pounds !! LOL…If you torture the data long enough, it will confess to anything !

Reply to  Marcus
April 3, 2016 12:03 am

rather bad analogy there Marcus. Pounds and kilograms have the same zero point. A 1% difference in the weight of an object measured in kg is the same as a 1% difference measured in pounds.
The reason it does not work with K and C is that they do not have the same zero point, despite the having exactly the same magnitude for each unit.

April 2, 2016 4:14 pm

best do a literature search first.
I used to be a dive into the data kinda guy.
Then I got a schooling from some Nobel types

Reply to  Steven Mosher
April 2, 2016 5:09 pm

..Nobel types ?? You mean like Al Gore and friends ?? LOL

Reply to  Marcus
April 3, 2016 1:00 am

best do a literature search first.
I used to be a dive into the data kinda guy.
Then I got a schooling from some Nobel types

Get with the orthodoxy you mean?
Going to the literature first is good way bias your work with preconceptions before you start. A literature search after a dive into the data is a very good idea.

Reply to  Steven Mosher
April 2, 2016 6:51 pm

Willis Eisenbach already knows that paper, as he is a signing author of a letter to the authors questioning their findings:
Arctic albedo changes are small compared with changes in cloud cover in the tropics
Reply to Legates et al.: Negligible role of arctic cloud albedo changes in observed darkening
But regarding Pistone et al. 2014. “Observational determination of albedo decrease caused by vanishing Arctic sea ice”, that you cite,
the Arctic albedo changes found by Riihelä et al. 2013: Observed changes in the albedo of the Arctic sea-ice zone for the period 1982–2009,
appear to be much smaller. I find it very curious that Pistone et al. will not cite Riihelä et al. work that was published 6 months before on exactly the same subject in Nature Climate. Very curious indeed.
It is obvious that Arctic Ice albedo is decreasing, but since surface albedo is a small part of planetary albedo, and Arctic sea ice albedo is a small part of surface albedo, it is clear and obvious that on a planetary scale the changes in Arctic sea ice albedo are negligible and likely to be overwhelmed by any change in cloud cover. Do you think this is not the case?

Reply to  Javier
April 2, 2016 11:58 pm

When the authors describe sea ice as “vanishing” “in the title of the paper you can see instantly that this is not an objective scientific study. I won’t waste my time even reading the abstract.

Reply to  Javier
April 3, 2016 12:00 am

I don’t see any Nobel prize winners amongst the author list either. No idea what Mosh’ means by that.

Reply to  Javier
April 3, 2016 1:17 am

May be he is confusing a Nobel “Peace” prize for political correctness with a Nobel prize for a science subject.

Johann Wundersamer
April 2, 2016 7:21 pm

Willis, You say
The most obvious change is that the slope of the all-sky data (blue) is much less than that of the clear-sky data (red).
Understanding right :
The more detailed, the less positive feedbacks in AGW?
Regards – Hans

Clyde Spencer
April 2, 2016 7:25 pm

The reason that overhead satellites are able to approximate the Earth albedo is because most things have a diffuse reflectance. In the case of snow, it is close to Lambertian, albeit with a strong forward scattering. (note that, minus the strong forward scattering, light is equally reflected by snow in a downward direction as well as upward or sideways) Water reflects primarily by specular reflection, as quantified by Fresnel’s equation. An overhead satellite will see no reflection (albedo) from water unless it is in the special position of being in the plane of illumination, (even though a minimum of at least 4% is going off into space) opposite from the sun and at the same angle as the incident sunlight has. At the limbs of the Earth, the reflection from water can approach 100%, although nothing will be seen or recorded unless the satellite is looking in precisely the right direction. Downward is not the right direction! Albedo is probably most useful for estimating the relative reflectance of water-less bodies in the solar system where the only viewing platform we have is Earth. To truly estimate the reflectance from the surface of Earth, we need to characterize the land cover, and develop a bi-directional reflectance distribution function (BRDF) for each and every diffuse reflector. Water can be treated as a specular reflector except when it is very rough, and especially if it has abundant white caps.

charles nelson
April 2, 2016 7:30 pm

Ahem…Dear Mods, I used the d word in a post…as you might see if you check the contents it does not contravene any of your eminently sensible rules. Please consider putting it up. Thanks.

April 2, 2016 9:50 pm

“The upper side of the grass” is muuuch more comfortable. Let’s enjoy it while it is still time.

Roy M.
April 2, 2016 10:53 pm

However much albedo there is due to water ice or snow, it’s surface energy reduction due to green house gases; that’s after there are 20 + % surface energy reduction due to the very same green house gases.
Green house gases represent additional radiating mass, at identical temperature to the surface. This parallel route for emission constitutes cooling; not warming.
This is so simple, one pseudo-science doesn’t allow real physics to be discussed regarding it’s cult belief: that the atmosphere heats the planet.
The atmosphere – specifically the green house gases – reduce energy to the surface many percent, then create additional mass from which emission may take place.
There’s no such thing as reducing surface energy density, then emitting that reduced energy from an overall larger, colder total mass, and those thermodynamic processes being energy gain: warming.

Reply to  Willis Eschenbach
April 3, 2016 10:00 am

Willis: I’m very sorry you took my characterization of your conclusions about the importance of sea ice albedo feedback personally. My words were directed solely at the validity of your conclusion, not at you personally. Nevertheless, upon re-reading my remarks, it clear that my deepest apologies are warranted.
The mathematics of climate sensitivity and feedbacks is non-linear. Feedbacks are summed in the denominator. If a feedback of -1 W/m2/K is added to planet with an ECS of 3.7 K/doubling, a run-away greenhouse effect (ECS is infinity) will exist. On the same planet, an addition -0.33 W/m2/K feedback increases warming by almost 50%. On a planet with an ECS of 2 K/doubing, however, an addition -0.33 W/m2/K feedback increases warming by only 20%. Small changes in non-linear systems can have large impacts.
Sincerely, Reformed Pond Scum

Reply to  Willis Eschenbach
April 5, 2016 10:24 am

Willis: Thanks for your kind reply.
If possible, I’d like to return to the problem that your data is consistent with a feedback large enough to be a first-order variation. Modest changes in non-linear systems (like feedback and climate sensitivity) can have large effects.

H. Gutierrez
April 3, 2016 12:37 am

Look – he believes in green house gas theory.
He thinks Angry Bird is a real mathematician
He thinks Rowboat Hansen is a real astrophysicist and programmer
He thinks Phil Jones’ admission he fabricated data with others for over a decade means nothing.
It’s these contrarians who pretend starting argument is scientific discourse when what it is, is trolling. Turning scientific discourse into a game of ”who can act most obtuse and arrogantly ignorant.”
I remember the day the man who released the ClimateGate emails left them and told Mosher stuff it. I looked back at what Mosher had been saying and thought to myself – these fake statisticians are everywhere among these science cons. He’s no different than any of the rest of them.

April 2, 2016 at 10:03 pm
” ‘Data overrules theory every time.’ Nope: This is a theory, a principle. it’s empirically wrong.”
Is there a human out there who doesn’t interpret Feyman’s words as (essentially) “Correct data overrules theory”? Obviously, if the data is bad, the theory cannot be said to be thereby refuted, can it? Otherwise, what Feyman must have been saying was, “Any and all data, regardless of how good or bad it is, trumps any and all theories every single time if the two are not in agreement.”
Who interprets his words as such?
No one.
The argument, therefore, is not over whether good and proper data trumps theory, but over whether the data is in fact both good and proper. And those are two very different arguments.

Toby Bronson
Reply to  Willis Eschenbach
April 4, 2016 12:00 am

Confessing you’re too stupid to follow a conversation,
then launching engaged, mental illness level, ranting skreed about it,
is something no scientific mind would do.
You don’t have the right to tell people how to communicate. No matter who the f**** you wish you were/hope to become before you die.
It’s particularly ignorant of you to try it when the entire history of mankind has been people fighting, to have other people, not tell them how they have permission, to talk.
You’re something out of another era when because you’re ” __________ ” you can tell people what and how they are to communicate to other people. Whoever you think you are, you need to get it through your head that – the typical real scientific mind doesn’t pay attention to wheedling grammar grannies.
They say what they say, and whoever they say it to, can understand it or not. I wouldn’t ask you for permission to say anything. Who are you? Precisely who are you to tell a stranger you never met that – his previous methodology of speaking isn’t good enough any more, that person answers to you. ?
You’re some kook is who you are, and you’re going to find out if you keep up that kind of behavior, how little people think of that sort of arrogance. We invented the net to STOP your kind of speech control.

Willis Eschenbach
April 3, 2016 at 11:11 am
Look – H. Gutierrez thinks he is above polite requests to quote what he is talking about, and as a result, we have no idea who the “He” is that he is abusing.

P. Bergeron
Reply to  Willis Eschenbach
April 4, 2016 12:35 am

Only the insecure join the free exchange of ideas on the modern internet, and believe they have the right to describe how others are to communicate. It’s the mark of the terminal level control freak type intellect.
Adults don’t try to take possession of other adults’ speech that way; the sociologically disordered do.

April 3, 2016 1:23 am

what seems the most interesting to me in this data is the two clear groups in figure 1. It would be informative to know why there are two groups with notably difference albedo. One of which seems to match the all-sky albedo.

April 3, 2016 3:10 am

Willis – I’ll ask again, you write that the change in reflection is only 1.1%, but how much would a 3C change in global temperature equate to in percent?
The answer, as I wrote above, is 1%. You should acknowledge that only small changes are required to have significant impacts.

Reply to  oneillsinwisconsin
April 3, 2016 4:25 am

…A 3C increase to 16C would be a 19% increase, NOT 1% !!

Reply to  Marcus
April 3, 2016 5:52 am

Temperature is a measure of the energy in the system and you can only express a % of it by using the absolute Temperature scale, therefore a 3ºC increase is a 1% increase (in T and energy).
Relevant to the Earth’s energy balance is that the loss depends on T^4, so a 3ºC increase represents a 4% increase in radiation heat transfer.

Reply to  Willis Eschenbach
April 4, 2016 4:14 pm

Willis – your post does not discuss a 1% change in precipitation. It does discuss a 1% change in albedo. Albedo is intimately entwined with energy absorption and emission. Now, why I would be asked to admit to some proposition that I never made, but you are (still) unwilling to admit the simple truth that a 3C change in global temperature is a 1% change – a smaller percentage than the 1.1% you ascribe to sea ice reflection – is an interesting subject on its own. It’s part of why I typically avoid this place – rarely a straight answer or correction to errors or ill-conceived ideas. For the record: small changes in unrelated variables or those variables with low sensitivity coefficients have little to no impact on systemic changes. Albedo is not – in the context of the earth’s energy budget – unrelated, nor does it have a small sensitivity coefficient.
You might also wish to inform Marco that you can’t use Celcius to calculate percent changes in temperature. Obviously he’s not willing to listen or learn from me.

Svend Ferdinandsen
April 3, 2016 6:10 am

Thanks Willis
Allways nice to look at the figures and relations. When searching for anything climate related you get a lot words and very few facts and figures. Like “warming of the globe will melt the ice”.
No one tells that in most of Greenland and Antarctica you need at least 10K warming to start melting any ice at all.

April 3, 2016 6:13 am

I see this study about energy reflection talking about Greenland and Antarctica.
Greenland being lower in latitude gets more sunlight than Antarctica. There are places in Antarctica that get no sunlight at all or a good couple months so you can’t account for reflection of ice or snow where no light is present. How many W/M^2 of energy is lost in the total darkness of the poles during each pole’s winter months? Over the span of time in this study, what angle of sunlight is being calculated and how much loss of energy in sunlight base on the thickness of atmosphere the light energy has to travel through to get to the clouds and sea ice? Then there is rough water compared to calm that is going to affect the rate of freezing or the rate of overall water temperature drop.
I just don’t see any environmental study like this to be simple at all. There are just so many variables that are in constant flux like the sun angle and spinning earth. I know this study is just one element of light energy and maybe there isn’t space to write a book on all the variables your dealing with to keep this understandable to a layman.
I’m a product engineer that does get involved in testing and have to take in account all variables that could cause failures and get accurate results.

April 4, 2016 7:41 am

First Willis thank you for your work, and thanks to the reviewers posing proactive comments. My bottom line on all of this is that the science is far from settled and instead of trading carbon offsets we should be investing more in climate research. My very first impulse in response to CAGW many years ago was how of the multitude of factors that influence climate can one molecule be used to define the whole system.

James at 48
April 5, 2016 9:42 am

Heaping on to the ice vs snow commentary. Sea ice I’ve seen has been all over the map, ranging from nearly blue in color to white. Once there is a snow pack on top it feeds itself. With a large pack, wind scouring becomes less of a factor, additional accumulation becomes easier, and, the snow compresses at the snow-ice interface, adding to the ice thickness. Ah, NOTHING is ever simple.