Green Flash
For the most part the following is in agreement with the URL found by wofahulicodoc. But it extends the reasoning and elaborates on some additional points that might be of general interest.
(From the book Color and Light in Nature by David K, Lynch and William Livingston.) Extracted and paraphrased below...
***Contrary to popular opinion, green flashes are quite common, especially over water. Indeed they happen to some degree every time the sun sets (or rises). The difficulty is in observing them. They last only for a second or two and they happen when the sun's light is rapidly fading. If we look at the sun's disk directly, our eyes are likely to be dazzled and there is a risk of missing the smaller green flash just above, before the solar disk goes dark.
*** Observing green flashes at sunrise is more difficult. It is hard to determine the exact location on the horizon where the sun will rise. The green flashes first, then the sun rises red beneath. So if you see the edge of the rising sun first you've missed the flash. Turn the boat around and head for home.
*** (Contradicting wofahulicodoc's URL) "...binoculars or a small telescope may be helpful. A setting sun presents no danger to the eyes."
*** The duration of the green flash depends in part on the rate at which the sun sets. During summer at high latitudes, the sun approaches the horizon at a grazing angle and passes below the horizon so slowly that the green flash may last for several seconds. During Admiral Bird's expedition to Little America in 1929, the green flash was observed off and on between irregular ice floes for a period of 35 minutes.
*** The green flash occurs because of dispersion in the atmosphere. Since the index of refraction depends on wavelength, all images passing through the atmosphere are dispersed vertically. Near the horizon, the apparent sun is composed of a continuum of such images, each at its own wavelength and location. Refraction is greatest for the shortest wave lengths, so that the blue image of the sun is highest in the sky. As the total amount of dispersion is much less than the diameter of the sun the images overlap except for the extreme upper and lower edges. The lower part of the sun is red for the same reason that the upper part is green: dispersion.
*** Recent work on green flashes shows then to be more complicated than first thought...