Ursa   meteoptic-l/summary  

 

meteoptic-l [ät] ursa.fi

message archive

This is meteoptic-l [ät] ursa.fi message archive. Note, your can reply the messages on this page only if your are already subscribed the list.

» To the end of the list/message

 

From: Les Cowley (sun1_at_hidden_email_address.net)
Date: 05/19/2003



I don't think the second bow is a supernumerary. As Harald Edens has already mentioned, a supernumerary (interference) bow so well separated from the primary requires very small droplets and it would have very pale colours. It would be, or almost be, a fogbow. Take a look at http://www.sundog.clara.co.uk/droplets/fogdrpsz.htm for the way supernumerary bows widen and separate as the drop size gets smaller.

The double bows could be produced by mixtures of different water drop *shapes*. Namely, small round drops and large oval or oblate ones. It's
the shapes that matter, not the sizes except that surface tension holds smaller droplets much more tightly spherical. Harald has explained how they might occur in storms. Only geometric optics is needed, no diffraction
effects have to be invoked.

To get distinct double bows requires that there be two distinct shapes of drop. A mixture of many drop shapes would merely produce a whitish mush - i.e. no rainbow! Perhaps that is why doubles are rare.

The attached simulation is for a 50/50 mixture of spherical drops and oblate
ones which departed from sphericity by only 2.5%. Even this small departure
makes well separated bows. The non-sphericity of the photographed bows must have been even smaller.

The simulation traced 15 million rays through the drops using program developed by Michael Schroeder and myself. It's a bit rough and ready because there wasn't time this morning to trace more rays. And of course it doesn't rule out other ways of making a double bow!

Les Cowley