Science > Physics > The sky isn't blue - it just looks blue because we don't see purple as well as blue
| Topic: |
Science > Physics |
| User: |
"penseur" |
| Date: |
21 Feb 2006 05:32:08 PM |
| Object: |
The sky isn't blue - it just looks blue because we don't see purple as well as blue |
Scattering by particles much smaller than a wavelength should result in
increasing scattering for increasing frequency, in the range of visible
light. So I assume that the power spectrum does not in fact peak at
blue light, but rather gets brighter in violet and in the ultaviolet.
If that is true, a spectrometer would characterise the sky more as
violet than as blue, whereas our eyes repond to blue light far better
than they do to violet light.
So the blue sky is only our perception. If our eyes had a constant
frequency response up to the edge of ultraviolet, the sky would look
purple/violet.
Perhaps bees, which can see some way into the ultraviolet, would view
the colour of the sky as something like violet or ultraviolet.
True? Or does the power spectrum of skylight actually peak in the
region of blue light?
.
|
|
| User: "Edward Green" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
22 Feb 2006 09:56:46 PM |
|
|
penseur wrote:
Scattering by particles much smaller than a wavelength should result in
increasing scattering for increasing frequency, in the range of visible
light. So I assume that the power spectrum does not in fact peak at
blue light, but rather gets brighter in violet and in the ultaviolet.
If that is true, a spectrometer would characterise the sky more as
violet than as blue, whereas our eyes repond to blue light far better
than they do to violet light.
So the blue sky is only our perception.
Well... that's true! Blue _is_ a perception. I suppose we might say
that something isn't "really" blue which appeared blue if we had reason
to believe our perception had been skewed relative to the norm; say by
looking through a red filter for 30 minutes and then taking it off.
But in a case like this, where the majority of the population would
agree the sky is blue, it _is_ "blue" by definition.
If our eyes had a constant
frequency response up to the edge of ultraviolet, the sky would look
purple/violet.
Perhaps bees, which can see some way into the ultraviolet, would view
the colour of the sky as something like violet or ultraviolet.
There are two questions at work here: how do we characterize the
physical frequency response of the apian eye, and what do bees
perceive. The former could be measured. Some people might doubt we
could say anything useful about the latter.
True? Or does the power spectrum of skylight actually peak in the
region of blue light?
Ah... now you are asking a different and more operational question.
As Mark Twain once put it "I don't know".
.
|
|
|
| User: "penseur" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
24 Feb 2006 04:55:31 AM |
|
|
Edward - you are absolutely right to separate the question into "what
is perceived by human eyes" (does apian refer to related species like
"apes"?) and what is measured by a machine. I agree that virtually by
definition, the sky is blue since humans perceive it that way.
But when I look at something I wonder whether I am being fooled by
optical illusions or quirks of perception.
For example, a spectrometer woud look at an old fashioned cathode ray
tube color TV and report that there are never any yellow, orange or
purple colors, and it might complain that "brown" doesn't really exist
as a spectral color. The power spectrum of such a TV would have three
peaks in the visible (and I suppose a general IR and X-ray glow as
well, not to mention low freq stuff like the VHF, UHF radio signals,
15kHz retrace and 3.5 MHz crystal oscillator, 60Hz hum, etc).
WRT bee vision, I have read that they can see abit into the
ultraviolet, and what we think of as "white" flowers may have wild
patterns splashed all over them in the UV!
But I guess my question actually boils down to "what does a
spectrograph of a "blue" sky look like? Does power peak in the blue or
the violet?"
.
|
|
|
| User: "Edward Green" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
24 Feb 2006 11:00:25 PM |
|
|
penseur wrote:
Edward - you are absolutely right to separate the question into "what
is perceived by human eyes" (does apian refer to related species like
"apes"?)
I was hoping it meant "of or pertaining to bees", but since you need a
subscription to the unabridged Miriam Webster site to look up a fancy
word like that, I couldn't be sure. ;-)
<bestirs self to subscription free paper dictionary>
Yup: "of or relating to bees".
and what is measured by a machine. I agree that virtually by
definition, the sky is blue since humans perceive it that way.
But when I look at something I wonder whether I am being fooled by
optical illusions or quirks of perception.
For example, a spectrometer woud look at an old fashioned cathode ray
tube color TV and report that there are never any yellow, orange or
purple colors, and it might complain that "brown" doesn't really exist
as a spectral color. The power spectrum of such a TV would have three
peaks in the visible (and I suppose a general IR and X-ray glow as
well, not to mention low freq stuff like the VHF, UHF radio signals,
15kHz retrace and 3.5 MHz crystal oscillator, 60Hz hum, etc).
We perceive color as the relative excitation of three types of
receptors with peak frequency sensitivities. Any spectrum whatsoever
which excites the three receptors to the same extent will be perceived
as the same color. Then, we confuse things by referring to certain
monochromatic frequencies or ranges of frequencies as light of the
perceived color.
WRT bee vision, I have read that they can see abit into the
ultraviolet, and what we think of as "white" flowers may have wild
patterns splashed all over them in the UV!
But I guess my question actually boils down to "what does a
spectrograph of a "blue" sky look like? Does power peak in the blue or
the violet?"
Well, if you think about it, you might conclude that if power peaked
"in the violet", i.e., in the range of frequencies which the human
eye perceives as "violet", we would most likely see violet. I say
"most likely" because there is no guarantee a broad peak with a
certain center frequency will be perceived as the same color as a
narrow peak. The rub is, the actual spectrum lives in an infinite
dimensional space, whereas our color perception lives in a three or
maybe four dimensional space. We perceive a feeble projection of the
actual information in the spectrum.
.
|
|
|
| User: "" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
25 Feb 2006 08:39:06 PM |
|
|
Scattering by particles much smaller than a wavelength should result in
increasing scattering for increasing frequency, in the range of visible
light. So I assume that the power spectrum does not in fact peak at
blue light, but rather gets brighter in violet and in the ultaviolet
.......
The violets are not scattered as much being higher energy. The lower
frequencies are absorbed more after they are scattered. The reds are
scattered much more easily, which is why at sunset the displaced image
of the sun is red. The sun radiates most strongly in the green. But the
blues that we see are the frequencies that are scattered and survive in
the scattering the most.
Longer than one micron wavelengths are scattered to such a degree that
they do not pass through the atmosphere. Analyzing the sun in these
frequencies, one mainly detects the re-emission of these photons from
the molecules of the atmosphere. Around 10 microns is the intensity
maximum for the temperature of the earth. The sun's intensity in these
wavelengths is much much greater than earth's radiation. The proponents
of the greenhouse gas theory seem to be unaware of the fact that the
sun is the perfect blackbody because it has no reflected light in it's
spectrum. It is easy therefore to calculate the energy of the sun's
radiation it these wavelengths. So where is their consideration for
this radiant energy. The sun heating the ocean gives the earth it's
mean temperature. CO2 cannot trap radiation to warm the ocean without
blocking radiation from the sun to the ocean. Any detected change in
earth's temperature is not caused by CO2 and there is no scientific
evidence to support this conclusion of cause and effect. Such
proponents cannot even establish a scientific analyses of what
frequencies are 'trapped', much less lay out a valid quantification for
the heat capacity of the air and the earth and the increase of radiated
energy for 1 degree which is 1.4% of energy radiated per square cm at
287K.
Kent Deatherage
.
|
|
|
|
| User: "penseur" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
26 Feb 2006 03:52:26 PM |
|
|
Spurred on by the stimulating responses I got here, I have done a bit
more digging, and with the help of the people who run the Usenet
physics FAQ, came upon a paper published in the American Journal of
Physics in July of 2005 by Glenn Smith of the Georgia Institute of
Technology. In the paper is reproduced exactly the graph I was
looking for: the output from a spectrometer which was pointed at a blue
clear sky. Interestingly for colors of longer wavelength than blue,
the result is exactly what you would get from classic blackbody
radiation, multiplied by a Rayleigh scattering factor of
1/wavelength^4. However interestingly, from about the area of the
bluish part of the spectrum up until the edge if human vision below UV,
the spectrum is sort of wiggly and flat. In the UV it starts to
actually start to go down. So one could easily approximate the output
of the spectrometer as half violet and half blue! Which is not how we
see it.
Very interestingly, the claim of the author is that the reaction of the
human eye to a mixture of half violet and half blue is the same as the
reaction of the human eye to a mxture of monochromatic blue light and
pure white light. The claim is that this is the explanation of why we
see the sky as a sort of desaturated blue color. A mixture of red and
green light looks yellow to us due to the fact that a linear
combination of the right amounts of red and green result in the same
stimulation of the "red and green cones" as does yellow light. So the
claim in this paper is essentially (unless I got it wrong) that the
continuous spectrum of skylight, which looks like it is baasically half
violet and half blue, is perceived by our eyes as a mixture of pure
white light and pure blue light, for the same reason we perceive
green+red=yellow.
Thinking about this, I emailed the author and suggested that if one had
a sort of "optical notch filter" which could pass all light apart from
blue, then viewed through this filter, the sky should appear violet!
.
|
|
|
| User: "penseur" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
27 Mar 2006 05:58:18 AM |
|
|
Edward: in response to your posting:
"Well, if you think about it, you might conclude that if power peaked
"in the violet", i.e., in the range of frequencies which the human
eye perceives as "violet", we would most likely see violet. I say
"most likely" because there is no guarantee a broad peak with a
certain center frequency will be perceived as the same color as a
narrow peak. The rub is, the actual spectrum lives in an infinite
dimensional space, whereas our color perception lives in a three or
maybe four dimensional space. We perceive a feeble projection of the
actual information in the spectrum."
I would just comment that it may well be that the power spectrum is
roughly
flat between the blue and violet areas. But if the response function
of our
eyes is lower at violet than at blue, we would not perceive violet as
the sky color.
From all that I have read here, I still have to conclude that the sky
is not "really" blue.
It is the continuous spectrum of blackbody radiation that fits the
solar spectrum,
and which is then attentuated by atmospheric absorption. The fact is,
our eyes
perceive this as a sort of washed-out blue, i.e., blue plus white
mixed. I think that
if our eyes had a flat spectral response the sky would look quite
different
(maybe a blue-violet shade). Probably a lot of other things, like
famous art, would
look quite different as well! TV and LCD screens would probably look
quite odd if
our eyes worked like spectrometers. The whole RGB thing would not work,
and we
would have to use some kind of technology that has a tunable spectrum,
maybe
akin to the technology of tunable dye lasers...
.
|
|
|
|
|
|
|
|
| User: "Andy Resnick" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't seepurple as well as blue |
22 Feb 2006 08:09:04 AM |
|
|
penseur wrote:
Scattering by particles much smaller than a wavelength should result in
increasing scattering for increasing frequency, in the range of visible
light. So I assume that the power spectrum does not in fact peak at
blue light, but rather gets brighter in violet and in the ultaviolet.
If that is true, a spectrometer would characterise the sky more as
violet than as blue, whereas our eyes repond to blue light far better
than they do to violet light.
So the blue sky is only our perception. If our eyes had a constant
frequency response up to the edge of ultraviolet, the sky would look
purple/violet.
Perhaps bees, which can see some way into the ultraviolet, would view
the colour of the sky as something like violet or ultraviolet.
True? Or does the power spectrum of skylight actually peak in the
region of blue light?
You are neglecting absorption.
http://www.biogeorecon.com/tutorial.htm
--
Andrew Resnick, Ph.D.
Department of Physiology and Biophysics
Case Western Reserve University
.
|
|
|
| User: "penseur" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
24 Feb 2006 02:59:50 AM |
|
|
Anderw: If I understand correctly, absorption by the atmosphere
strongly attentuates UV and X-rays. I wonder what the relative
absorption of blue and violet light is? This is of course a vague
question as you would have to agree upon what is defined as blue and
what is defined as violet. But if you could show that a spectrometer
pointed at the sky (and it probably varies with inclination to the
horizon) unambiguously finds that blue intensity is higher than violet,
I will agree that the sky "is" blue. The "is" bit in this case refers
to a non-creature measurement of spectral intensity.
.
|
|
|
|
|
| User: "Ben Rudiak-Gould" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't seepurple as well as blue |
22 Feb 2006 03:46:59 PM |
|
|
penseur wrote:
Scattering by particles much smaller than a wavelength should result in
increasing scattering for increasing frequency, in the range of visible
light. So I assume that the power spectrum does not in fact peak at
blue light, but rather gets brighter in violet and in the ultaviolet.
If that is true, a spectrometer would characterise the sky more as
violet than as blue, whereas our eyes repond to blue light far better
than they do to violet light.
Something like that is true, but the details seem to be pretty complicated. See
http://math.ucr.edu/home/baez/physics/General/BlueSky/blue_sky.html
-- Ben
.
|
|
|
| User: "penseur" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
24 Feb 2006 04:40:22 AM |
|
|
Ben - superb website - thanks! Especially interesting was that of the
blue/red/green receptors in our eyes, red cones respond more than green
do to the indigo/violet part of the visible spectrum! But I still want
to see what a spectrometer "sees"....
.
|
|
|
| User: "tadchem" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
26 Feb 2006 06:08:30 PM |
|
|
An important consideration is that sunlight is significantly more
intense at 500 nm than at 400 nm.
Even the scattering can't make up for the fact that the violet light is
just very weak.
http://www-spof.gsfc.nasa.gov/stargaze/Q4.htm
has a graph showing the comparative spectra ("power curve" if you
prefer) of sunlight outside the astmosphere and at the earth's surface.
Google "solar spectrum"
Tom Davidson
Richmond, VA
.
|
|
|
| User: "penseur" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
06 Mar 2006 02:06:12 PM |
|
|
looking at that graph i found it tricky to tell, but it almost looked
as if 400 nm radiation was more intense! I would like to see a graph
of this, zoomed in betwen 400 and 500nm (is 500 really blue?) and with
the colors we would perceive.
.
|
|
|
| User: "tadchem" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
07 Mar 2006 02:29:52 AM |
|
|
Try this one:
http://ceos.cnes.fr:8100/cdrom-98/ceos1/science/dg/dg1.htm#anchor5699875
The peak is at about 500 nm (0.5 micrometers) and falls off *sharply*
at shorter wavelengths.
Tom Davidson
Richmond, VA
.
|
|
|
| User: "penseur" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't see purple as well as blue |
27 Mar 2006 05:43:46 AM |
|
|
Link seems to not work - is that graph anywhere else?
.
|
|
|
|
|
| User: "Sam Wormley" |
|
| Title: Re: The sky isn't blue - it just looks blue because we don't seepurple as well as blue |
06 Mar 2006 02:12:38 PM |
|
|
penseur wrote:
looking at that graph i found it tricky to tell, but it almost looked
as if 400 nm radiation was more intense! I would like to see a graph
of this, zoomed in betwen 400 and 500nm (is 500 really blue?) and with
the colors we would perceive.
If you put up a continuous spectrum from
green to blue and ask folks to make the boundary...
there will be a lot of marks in different places.
A large sampling should have a normal distribution.
.
|
|
|
|
|
|
|
|
|
Related Articles |
|
|
