Green Flash Musings

I've been enjoying the University shutdown between the Christmas and New Year's holidays.  Melinda has been unable to work after her recovery from oral surgery has been slower than expected, and her continued use of pain meds.  So we used the chance to continue chasing the sunset from the Mount Lemmon Highway from a different location than normal this last Friday, 28 December.  While we missed an exact Observatory alignment (though collecting another data point for future use!), we were able to enjoy an excellent, clear sunset, with a touch of a "Green Flash" at the upper edge of the sun.  What causes this green upper edge?  Well, let's talk about it!

I've noted some of this atmospheric effect several times on this blog.  For instance nearly 3 years ago I imaged Mercury very low in the sky and it showed the characteristic colored spectrum in its image, shown here at left.  And about 2 years ago, I imaged the bright star Archernar, which only gets a couple degrees above the horizon from Tucson shown at the right.  Both of these, taken with the Celestron 14" with 4,000mm effective focal length as well as the sun image above with a small telescope of about 1,000mm focal length, show the characteristic blue-green fringe atop the image, with the reddish part below.

 

There is a simple explanation for this - in all cases we're looking very near the horizon through long paths of the Earth's atmosphere.  Because the earth is spherical in shape, as is its air surrounding it, long pathlengths show the prismatic effect of the curved atmosphere.  One of the world's authorities on atmospheric optics, Les Crowley, who runs the Optics Picture of the Day site, has numerous examples and discussions about the Green Flash, including links to commentary, more examples, simulations and a reading list by Andrew T. Young deferring to him as the ultimate authority.

 

I had e-mailed Les, taking exception to some of his remarks regarding the requirement that mirage must be present for the Green Flash to be seen.  His response, as well as excellent examples of his discussions follow.  Be sure to click on the links to see those examples!  My initial comments to him preceded by a >:

 

Hi Dean,

> In my recent readings to catch up, I read your comment that "A green
> flash needs a mirage, ordinary atmospheric dispersion is too weak to separate colours
> sufficiently." I'm not sure I totally agree.

I was being deliberately provocative because many explanations cite normal
atmospheric dispersion or 'prisms' as the cause and therefore mislead rather
badly. I've campaigned for some time that flashes are rather more than that
and I think that the tide is turning. But we still get vague and misleading
statements like this:

http://epod.usra.edu/blog/2011/10/green-flash-on-the-tyrrhenian-sea.html

But not as bad as this howler on corona formation:
http://ww2010.atmos.uiuc.edu/%28Gh%29/guides/mtr/opt/mch/diff.rxml

The green rim (http://www.atoptics.co.uk/fz231.htm) is another matter.

> As an astronomer-type, I've
> spent many years observing atmospheric dispersion, and while landlocked in
> Southern Arizona, where the green flash is rare (perhaps proving your point),
> atmospheric dispersion is rampant. Note the following blog post of mine:
>
> <http://theketelsens.blogspot.com/2011/01/more-fun-with-colors.html>
>
> Which noted the strong prismatic effects of the atmosphere at low elevations.

I had intended to compute the angular dispersion for a normal atrmosphere at
zero sun elevation. However, Rayleigh (Proc. R. Soc. Lond. A 1930 126,
311-318) did the calculation a long time ago and that should suffice. He
gets, for a refraction at the horizon of 35', a dispersion between H-alpha
and H-beta (red-green) of 22 seconds of arc. That is less than the eye's
resolving power. Rayleigh argues via some experiments involving masks and
slits that the dispersion is sufficient for a green flash at an horizon.
However, he talks only of inferior mirage type flashes
(http://www.atoptics.co.uk/atoptics/gf2.htm) and not mock-mirage flashes where
the green is at the top of the sun's disk and above the horizon
(http://www.atoptics.co.uk/atoptics/gf3.htm). Rayleigh's experiments used
comparatively short path lengths and might therefore represent flash visibilty
under ideal conditions rarely experienced outdoors. In terms of duration,
the red to green a mask occultation would take 1.5s upwards depending on
latitude and solar declination and that is about right (but not supportive of
Rayleigh) for an I-Mir flash duration.

There are some example of Rayleigh mask type flashes
(http://www.atoptics.co.uk/atoptics/gfim7.htm) but even here, more extreme
refraction (not mirage) effects are likely. I agree with you that some
extreme but non miraging refraction could give inferior-mirage like
flashes but consider that most flashes at or near the horizon are mirage amplified.

Notwithstanding Rayleigh, modern flash understanding is that extreme
refraction normally leading to miraging better accounts for observations. I guess you
are familiar with Andy Young's extensive site and modelling on this
(http://mintaka.sdsu.edu/GF/explain/explain.html). Distortion and
multi-imaging of the sun prior to the flash(s) and apparently choppy horizons
are visible indications of mirage conditions.

> We rarely get the thermal layering in the desert you say is needed for the
> green flash, but we do occasionally see it as the last second of sunlight
> disappears behind distant mountains, that I suggest is caused by pure
> atmospheric dispersion.

Agreed - also the layering over hills can be accentuated by the airflow.

The layers over flat terrain can be quite thin and this one
(http://www.atoptics.co.uk/atoptics/gfim12.htm) in the Libyan Desert
comes to mind.

> Mostly I observe from Kitt Peak National Observatory, home of a 1.6
> meter solar  telescope that produces an 80cm diameter of the sun. On occasion I've been
> able to watch the sunset with that instrument and have, in fact, seen
> the "blue  flash" as the upper clearly blue-violet edge of the dispersed sun slips below
> the horizon (certainly not visible to the unaided eye, though).

On that solar image scale I would think that normal refraction would be
sufficient to show a green or blue rim which would normally be
unnoticed by the unaided eye. If at the last minute it appears to spread horizontally along
the horizon then some abnormal refraction is present.

>Anyway, very much enjoy your postings, perhaps I can contribute someday

I hope so, and thanks for provoking me in return on green flashes!

Best wishes,
Les

Les Cowley
Atmospheric Optics - www.atoptics.co.uk
Optics Picture of the Day - www.atoptics.co.uk/opod.htm

I think the gist of his discussions is that the Green Flashes we are seeing from our mountaintop and desert locations are not the traditional Green Flash, but rather masked Rayleigh dispersion.  The inferior mirage effect that occurs over water (but not over desert terrain), amplifies the Rayleigh dispersion effect for seaside observers.  I'd be glad to hear of opinions of others!