|Fogbows & Glories - Lamp optics imaged in Finland by Jari Luomanen (atmospheric photography) on August 25, 2009. (Earlier OPOD). ©Jari Luomanen, shown with permission.|
|Droplets of a dilute night-time fog have individually scattered the light of an intensely bright HID (high intensity discharge) lamp to form a huge diffraction pattern. The brightest ring is 70-80° in diameter and the whole scene demanded a very wide angle lens.
The lamp was behind the camera which was pointing away from it in the opposite direction.
The diffraction pattern components have individual names but they are all continuous parts of a sphere of light scattered by each droplet.
Directly opposite the lamp is a glory.
Outwards from its centre its rings decrease in colour intensity and eventually become the inner supernumeraries of a fogbow. The supernumerary hues become progressively less colour saturated towards the primary fogbow itself.
The primary fogbow - from a single reflection inside the droplets - is almost colourless. It is often mistakenly called a 'white rainbow' but only the very broadest of fogbows from the smallest of droplets are devoid of colour. As the droplet size increases the fogbow outer edges become increasingly fringed with straws and reds and the inner edges take on blues and violet.
Beyond the primary is yet another ring, the faint secondary fogbow from two internal reflections.
The second image shows a close up of the inner glory taken at a different time of the night. There is a ‘hole’ instead of a bright centre because there was insufficient fog in that direction. Most of the droplet scattering took place relatively close to the camera.
Fine structure with many rings in a diffraction pattern speaks of droplets with similar sizes. The lower image compares the fogbow and glory with an exact Mie scattering calculation (IRIS program) for monosized droplets of 12 micron radius. The primary, the larger supernumeraries and the glory are a reasonable fit. The calculation is for parallel light whereas that from the lamp was diverging and so we should not expect exactitude.