Rainbow Twinning ~ The inner primary rainbow is split into two, "twinned". The outer secondary bow stays single. A rare sight captured by atmospheric optics expert Alexander Haussmann on May 11th '14. This image is enhanced to show better the bows. The original is below.©Alexander Hauβmann, shown with permission
Surface tension forces act to minimise the surface area of small droplets making them spherical. In large drops aerodynamic drag become significant and flattens them. A flattening of 1-2% is sufficient to distort the primary rainbow and sometimes produce twinning.

The original image. The twinning is still visible
Atmospheric
Optics
About - Submit Optics Picture of the Day Galleries Previous Next Today Subscribe to Features on RSS Feed
         
Alexander Hauβmann writes: "Twinned rainbows are a quite rare and exceptional sight. I saw my first one in August 1997 when I was 16 years old. Back then, I already had made contact with the German observers� network and asked Claudia and Wolfgang Hinz for advice (I still have the letters in my files). However, we did not come to a satisfactory explanation [about what are now called twinned bows] due to the lack of further observations.

Since this time a considerable amount of reports and photos of twinned rainbows has been collected, and also the theoretical explanations have made some big steps forward [1, 2]. It now seems certain that the raindrops� slight distortions from spherical shapes while falling are the key to the phenomenon: The larger the drops are, the more their outline is changed towards a �hamburger bun�. This will cause the primary rainbow to shift inwards at its top. If a shower consists of a mixture of two different drop sizes, or two such drop sizes follow each other, the observer may see two primary rainbows on top of each other: The lower resulting from the larger, hence more distorted drops, and the upper from the more spherical smaller drops. Usually the secondary rainbow is not seen split because the deviation of the sunlight along its ray path remains rather resistant against shape distortions.

On May 11th, 2014, I was able to add the 6th observation to my personal list. At my home in Hoerlitz, Lower Lusatia (51� 32� N, 13� 57� E) a very weak shower was passing and slowly gave way to sunshine, allowing me even to see a small piece of a 22� ring simultaneously with the primary and secondary rainbows around 18.40 CEST at a solar elevation of 16.9�. Due to the remaining rainfall it was unfortunately not possible to completely shield the fisheye lens from the drops. The rainbow display was not particularly bright and I did not expect anything unusual. But within the next 2 minutes I noticed the primary getting wider, and shortly afterwards the twinning became visible to the eye while some drizzling still persisted � it can be seen quite nicely in the photo lighting up close to the rainbow. Over the next 2 minutes the split primary and the secondary rainbow vanished quickly from top to bottom. I had not even the time to wipe my lens dry.

It is a strange coincidence that I saw my previous twinned bow exactly two years before � on May 11th, 2012. At that time I had the chance to calibrate the photo and take some measurements, aiming for a subsequent analysis of the raindrop size distribution along the rainbow cone [3, 4]. Both these observations, as well as my earlier ones, have in common that some light rainfall was still going on at my place or the rain was at least still nearby. At all occasions the sun stood higher than 15�, though optics would as well allow twinned bows for a lower sun. Maybe this just reflects that some thermal dynamics from a more intense solar irradiance is needed to generate showers with the appropriate drop size mixture.

(1)   L Cowley    https://atoptics.co.uk/rainbows/twin1.htm
(2)   I. Sadeghi et al., ACM Transactions on Graphics 31, 3 (2012).
(3)   A Haussmann   http://atoptics.wordpress.com/2012/05/28/twinned-rainbow-over-dresden-may-11th-2012/
(4)   A Haussmann     http://www.youtube.com/watch?v=N0BkGFYBN34"