We see only part of the sun’s light. Here, Eva Seidenfaden (Paraselene) has captured something always present when a rainbow graces the sky but is never seen by our eyes – the infrared rainbow.
©Eva Seidenfaden, shown with permission.
The central slice is an image of the visible light bow for comparison. The infrared and visible images were taken within 46 s of each other using the same tripod mounted camera and they are essentially simultaneous views in two wavelength bands.
About the bow:
Our eyes see light of 400 – 700 nm wavelengths. Sunlight contains longer wavelengths – infrared – and a near infrared rainbow is just possible because relatively small raindrops transmit sufficient IR to form a bow. Water absorbs longer IR wavelengths and large raindrops do not even transmit sufficient near IR. Robert Greenler first recognised the possibilities of an infrared rainbow just outside the red bow and in 1971 his pioneer image made the cover page of Science.
The refractive index of water for near IR is less than that for visible light and the IR bow is therefore of slightly larger radius as apparent in Eva’s image. That it is also slightly broader could result from some natural broadening at long wavelengths but also because the limited IR transmissivity of water will tend to give prominence to those bows produced by the smaller droplets in the shower population.
The IR bow has a stronger and more widely spaced supernumerary. Supernumerary spacing is a function of the ratio (drop size/wavelength). Small drops give widely spaced supernumeraries, longer wavelengths give widely spaced supernumeraries. See this simulation for the effect in visible light.
How it was done:
Camera CCDs have a small but significant sensitivity to near IR light in spite of installed filters. A visible light blocking filter over the lens permits IR photography. The exposures are very long because of the low sensitivity to IR.
Eva first imaged the visible bow using an automatic white balance setting. Then a filter was fitted (RG 715 nm Heliopan = Kodak Wratten 88 and Hoya 720) to block all wavelengths shorter than 715nm. The camera was then white balanced again and the IR image captured.
A small aperture was needed with careful focusing because the lens IR focus differed from the visible. The IR exposure was 0.3s compared to 1/90s for the visible. The two images had to be captured close together because Earth’s rotation moves the bow 0.25° each minute.
More IR bows.
Is there a UV rainbow?