Explaining Green Flashes

Green Flashes and Mirages

Geometrical optics of flashes

Green flashes are by-products of the large variations in astronomical refraction near the horizon. Although there are several kinds of green flash, almost every kind is a by-product of a corresponding mirage. All the common forms are magnified images of the green rim produced by dispersion in the bulk of the atmosphere. The rim is too narrow to be seen without magnification, but sufficient magnification can be produced by mirages of various kinds.

That means a separate explanation is needed for each kind; and you have to understand the mirages before you can understand the flashes. (The detailed explanations of the different green-flash forms are on the simulation pages; and there are also mirage simulations, if you are interested.)

There is, however, one important difference between astronomical mirage phenomena (distorted sunsets, green flashes, etc.) and terrestrial mirages. Because astronomical objects are practically at “infinity” (optically speaking), astronomical mirages depend only on the directions of rays at the top of the atmosphere. But terrestrial objects are relatively near at hand, so ordinary mirages depend on the height where each line of sight meets the miraged object; and this height depends on distance from the observer. The distance dependence complicates the explanation of terrestrial mirages, but not astronomical ones.

I attempt to explain these phenomena here. I have provided an introduction to mirages, as well as some simulations of a few common green-flash forms; consult that simulation page for detailed explanations of the individual types. Because green flashes and the related distortions of the low Sun are all part of astronomical refraction, I've also added some general information about astronomical refraction.

It's easy to get lost in the details of these complex phenomena. I have a page devoted to basic principles that may help you see how the most important principles in atmospheric refraction fit together.

Colors of green flashes

Green flashes are not always green; they can be yellow, green, blue, or even violet — see the color examples. But green is the hue seen most often.

The basic cause of the color is atmospheric dispersion: refraction by air is larger at shorter wavelengths. So, at sunset, the refractive delay of the sunset is usually a second or two longer for blue and violet than for red. In general, then, the red image of the Sun (or of some miraged part of it) sets or disappears first, followed by yellow, green, blue and violet.

So why isn't violet the last color to be seen at sunset? There is another effect at work: atmospheric extinction. Both air molecules and aerosol particles scatter the shortest wavelengths most strongly (which is why the sky is blue: the strongly-scattered blue light goes in all directions, so we see it when we look anywhere in the sky). At the horizon, the path length through the air is very long, and the shortest wavelengths are almost completely removed.

Scattering by molecules alone is not quite enough to make the shortest wavelengths invisible; so if the air is very clear, violet is the last color seen. But usually there is enough haze in the air that violet, and even blue, is completely removed, so that green is the last color seen at sunset, or the first at sunrise.

These processes (dispersion and extinction) are fairly well treated in many of the explanations on the Web. But I have added another page here that goes into the colors of green flashes in more detail. There are also simulations that show the interplay of aerosols and the colors of the flashes.

And at sunset, there are additional effects, due to visual adaptation, that make even the (nominally) yellow stage of a sunset flash appear green.

Visual effects

Green flashes involve physiological optics as well as physical optics. At sunset, the eye is almost invariably affected by looking at the bright red Sun, even for a few seconds; the major effect is bleaching of the red-sensitive photopigment from the retina. This bleaching distorts color perception, usually without the observer being aware of it: loss of red sensitivity means that the yellow stage of the sunset flash that precedes the green one is usually perceived as green, not yellow. A technical paper on this subject is now available.

However, green flashes are not afterimages! A great deal of effort has been wasted by proponents of the afterimage theory, which was directly refuted in the 1920s by Lucien Rudaux's pioneering photography of a sunset flash (not to mention the argument that had been made for decades previously, that the existence of sunrise flashes was evidence enough against the afterimage idea.) Although the effects of retinal bleaching can be avoided by observing green flashes at sunrise, few people are ambitious enough to get up that early; and it's difficult to know where to look, before the Sun is visible. So, most observations are made at sunset, and are affected by bleaching.


You will find many other explanations on other people's Web pages. I have tried to select a few that contain a grain of truth. Obviously, I don't think the other explanations offered are adequate; that's why I am providing my own. Nevertheless, you may find it useful to see what other people have to say about this complex and fascinating subject.


For longer accounts, with more details, see the reading list. You might also have a look at some simulations I have prepared for the commoner types of flash.


There are still many loose ends to the Green Flash business. I have a list of unsolved problems you might be able to help with. Take a look and see if you know something I don't.

© 1999 – 2006, 2010, 2012, 2013 Andrew T. Young


Didn't find what you were looking for? Try the tables of contents or the glossary or the alphabetical index.


Back to the GF home page