The Green Flash

M. M. Dworetsky, Department of Physics and Astronomy, University College London

(Originally published in Gnomon, Newsletter of the Association for Astronomy Education, December 1995)


Permission to link to, copy, or reproduce this article by electronic means is hereby granted for non-profit educational purposes or for personal use by individuals, provided that it is unaltered and the title, author, and this notice are copied with the article. Quotations from this article are permitted provided the source of the information is acknowledged. Those wishing to use it for other purposes should contact the author for permission. (c) 1995, 1998 M. M. Dworetsky.


Some of you must be wondering why I am writing about a comic-book hero in an astronomy newsletter, but please be assured that no such hero ever existed! The green flash is an astronomical, or rather, an atmospheric, event. When conditions are right, at the last moment of the setting Sun, its upper edge or limb blazes with an emerald green color for a few seconds before disappearing below the horizon. Few people have seen the green flash, yet it is one of the most startling and colourful of sunset or sunrise phenomena, requiring patience and good luck to be seen. Until recently, I had personally never been fortunate enough to observe this remarkable sight, and many of my astronomical colleagues say they have never seen it despite repeated attempts. So, what is it, and why is it so elusive? I explain all below, and at the end of this article I'll give some hints about how you can join an exclusive club--the green flash observers.

To understand the green flash requires some background knowledge. Even the most cursory everyday observation reveals that the Sun fades and appears to turn reddish-orange as sunset approaches. (All effects described in this article are seen in reverse order at sunrise, of course.) This reddening is caused by Rayleigh scattering of light by molecules in the atmosphere. These molecules are very small compared to the wavelengths of visible light, which has the consequence that the scattering is proportional to the inverse fourth power of the wavelength. Thus scattering of violet light (400 nm) is 7 times more effective than that of red light (650 nm). The result is that the Sun looks red because so much blue light is removed from the line of sight.

Astronomers use the term "air mass" (symbol A) to describe the Earth's atmospheric thickness in the line of sight. An object overhead (altitude h = 90°) is seen through one airmass, at altitude 30° degrees through two air masses, and so forth. If the Earth were flat, the formula A = sec (90° - h) would apply. But the Earth is spherical, so at sunset we look not through an infinite amount of air but (as detailed calculations show) about 38 air masses. In a clear atmosphere the setting Sun is about 1/1000 its mid-day brightness, but even a moderate amount of dust or haze can reduce this figure by a further factor of 10 to 100 or more. (It occurred to me, as I wrote this paragraph, that the fact that the setting Sun does not disappear but is only dimmed is yet another proof that the Earth is not flat!)

A more detailed examination of the setting Sun shows that it also appears somewhat flattened and may show horizontal structure and banding. (Note: viewing precautions are needed as the setting Sun can be too bright for safe viewing without using dark filters, although quick glances, once parts of the disk are below the horizon, should not cause more than dazzling and afterimages.) The flattening or ellipticity is due to atmospheric refraction, which raises the Sun's lower limb by about 35 minutes of arc ( '), while the upper limb is raised by only about 29', when the lower limb is tangent to the horizon. The banding effects are due to layers of differing temperature and density in the atmosphere.

The setting Sun's disk is made up of light of all colours. Green and blue light are refracted by air slightly more than red light, so the disk actually consists of a flattened red disk, with a yellow disk slightly above it, a green disk above that, and blue and violet disks at the top. This phenomenon is called atmospheric dispersion and is easily visible to anyone who looks through a telescope with high magnification at a bright star or planet low in the sky. The vertical separation of red and green varies with conditions, but is typically about 1' at the horizon.

In the case of the setting Sun, the upper limb is green (or sometimes blue). If the horizon is provided by the sea or by very distant flat land, and the air is very clear for a hundred kilometres or more towards the sunset, then it is possible to see the green flash. As the upper rim of the setting disk approaches the horizon, it begins to spread into a thin bar of light, then runs through the spectrum from orange to yellow, then pale green, and finally reaches a deep emerald colour for two or three seconds. Under favourable conditions a brief blue blob of light may be seen after this, but exceptional clarity of the air is needed.

During the summer of 1995 I had to travel to California and Hawaii for guest investigator observing runs at Lick Observatory near San Jose and at the Mauna Kea Observatory's Canada-France-Hawaii Telescope. Between these runs I also visited relatives in Los Angeles with my family. Thus my travels took me to several places with clear air and unobstructed ocean horizons, ideal for seeking a view of the green flash. I was not disappointed.

My first sight of the green flash was memorable. The Lick astronomers on Mt Hamilton (height 1280 metres) would gather on the catwalk of the 3-metre telescope dome each evening, to watch the sunset (and check for signs of approaching clouds). On 23rd July 1995, at 18:29 PST, Dr Elizabeth Griffin and I observed an excellent 2-3 second green flash, which rapidly faded into a distinctly blue bar of light for about half a second before disappearing. Although we had a run of twelve clear nights, this was our only sighting of the green flash. Conditions that evening were exceptionally clear as we could see the Pacific Ocean beyond the San Francisco peninsula. A blue flash is a rare sight indeed.

A month later my family and I were staying in a delightful oceanfront apartment in Redondo Beach (altitude 10 metres--we were on the top floor) with panoramic views of Santa Monica Bay. The first week of our stay was rather misty, but each night during the second week we would stand on our balcony and watch the setting Sun, hoping to see the green flash. Only once, on 27th August, were conditions just right to see a classic display of emerald green. On other occasions when no flash was seen, we could see indications of very distant clouds on, or just over, the horizon.

Although the astronomical work I did at Mauna Kea (altitude 4200 metres) in the following week was extremely successful, no green flashes were observed because one rarely sees the true horizon from the peak--it is almost always cloudy over the surrounding sea. But I was rewarded with one final green flash as I was waiting on the evening of the 10th of September for a local flight from Kona airport, which is built on the west coast of the Big Island (height 4 metres). The waiting area is outdoors, with a clear view over both the runway and the Pacific. A jet aeroplane taxied out of the line of sight moments before the upper limb of the Sun reached the horizon, and I witnessed one more green flash.

From reading I have done, it appears that the correct explanation of the green flash was surprisingly slow in emerging during the 19th century. There are apparently no known written descriptions of it from previous centuries. Some scientists speculated that, because most reports were from observers at sea, it was due to a filtering of light by the crests of distant waves. Others suggested that it was an optical illusion caused by afterimages of a red setting Sun which merely appeared green. The green flash even had a place in romantic Victorian literature, such as Jules Verne's 1882 novel Le Rayon Vert (The Green Ray), one of the first recorded mentions of it: ``If there is a green in Paradise, it cannot be but of this shade, which most surely is the true green of Hope.'' (It has been a long time since I read it, but if I recall correctly, Verne's characters spend the entire novel travelling to remote places whence reports of the green flash have emanated, but something always happens to prevent them from seeing it. Meanwhile they philosophized a lot and had adventures.)

Right. If you got this far, I hope that I have intrigued you into trying to see the green flash yourself. Or, perhaps you will encourage your students to look for it. First, it is regrettably more difficult, though far from impossible, to see the green flash from Britain. The best time to try is on a very clear day from a sea-side location with a clear view of the setting Sun, any time of the year. It is more likely that you might see a green flash during a seaside holiday abroad, although you would need to make sure your chosen location has open sea to the northwest (assuming it is a summer holiday), or in the appropriate sunset direction at other times of year. Next, never ever miss an opportunity to watch the last moments of the sunset. Do not blink or glance away. One of my colleagues once saw a brief but intensely green flash from a mountain in Oman. He excitedly asked his companion, ``Did you see that?'', to which the reply was, ``See what?''

It is, of course, possible to observe a green flash at sunrise. But a good deal of preparation is needed to note as accurately as possible beforehand where the Sun will first appear, preferably using fixed direction markers overlooking open sea. And, unless you are a very early riser the discipline needed to check every sunrise probably won't exist. Of course, you may have been up partying all night, but chances are that you would not then be in suitable condition to observe anything. . . .

Mike Dworetsky
University College London

e-mail: mmd@star.ucl.ac.uk