Supernumerary rainbow

It's been another rather unpredictable day weatherwise. Alex and I went down to Silsoe to carry out our annual recording of a restoration experiment in a disused sandpit, and dropped Pete at Clophill on the way, to go and record an area of alder woodland. We arrived early to get ahead of the rain showers that were predicted after 3pm - only the first one arrived at about 9.30! Despite this irritation, none of the showers lasted long, and we managed to get all our work completed by 2pm, just as the sun decided to appear...

This evening we had several much heavier showers and for a short while there was low sun and rain, which produced a double rainbow, the lower one of which was a supernumerary rainbow, with extra bands.  Supernumerary bows occur when raindrops responsible for the main rainbow are uniform in size. Slightly different ray paths through a raindrop yield slightly different path lengths and slightly larger exit angle. As a result, there is constructive and destructive interference of each color in the spectrum as a function of ray exit angle, and a set of bows become visible inside the primary rainbow. Usually, there is some variation in size of raindrops, and the supernumeraries are washed out altogether. There is almost always some washing out of colors, and the bows show much green and red in them, and not the other colors in the spectrum.