Blog Archives

Third and Fourth Order rainbow in Russia

Sergei Antipov observed on June 22, 2013 in the Vladimir region, Russia (100km from Nizhny Novgorod city) beside a primary and secondary rainbow, the rainbow third and fourth order too.

Time: 14:00 (UTC + 4h)

Weather condition:
+20.1ºС, relative humidity 98%
atmospheric pressure 747mmHg (normal at 82m is 752-753mmHg)
Min/Max: +14.0º / +20.8º
Rain during the day: 3 times, thunder-storm and heavy rain.
wind: in the morning northern, in the afternoon and in the evening eastern
Light breeze, 1 – 3 meter per second, gusts were not stronger than 10 meters per second

Photo time with 1st and 2nd order rainbows: 19:37 (+4)
Photo time with 3rd and 4th order rainbows: 19:47 (+4) (1st, 2nd were visible too)
sunset: 21:52 (+4), azimuth 315º
sun azimuth @ 19:47 291º, height 15º

In late afternoon there were black clouds that came from the east (usually cumulonimbus comes from the west). Cumulonimbus covered almost all the sky and although it was not raining, there was a bright primary and a good secondary rainbow. The sun was covered by clouds. You can see that on a roof of the house there is no shadow. But two rainbows were visible and were bright!
10 minutes later there was bright sunshine (you can see a shadow on a roof of the house).
The sun appeared at 19:47. Till this time the sun was hidden).
The rain began at about 19:45. 3rd and 4th rainbows are photographed from under an umbrella.
But the rain was very weak. From the sky rare droplets of water fell.
Even the roof of the house remained dry (but with traces of drops).
At this moment the rare rainbow also was observed.
The heavy rain began much later (>20:00)
The sun became covered by a cloud, and the first rainbow gradually disappeared.


  • Good weather (the last hour)
  • clouds (from the East) and sun (in the west) ~ 19:00
  • dark clouds (sky half) and sun ~ 19:20
  • gray clouds (3/4 of sky) and NO sun, No rain (Or very slight rain that I didn’t feel it) = 19:37
  • Beginning of observation of the first rainbow (without rain and without sunshine) within a few minutes there was a sunshine
  • very dark clouds (more, than 3/4 of sky) and bright sunshine (the sun shone from beneath a cloud border)
  • Slight rain (isolated droplets)
  • photo of observation of 3rd rainbow at 19:47

The panorama is made of two photos with an interval 10 minutes; photos are made from different places (about 10 meters). The lens has a bad distortion towards the edge…

Weather that evening was unusual. Cumulonimbus clouds came from the East (usually they come from the West). Therefore I well remember that evening.

The Quality of the original photo is not really good therefore all colors of a rainbow are visible only on “psuedo-HDR” processing (combination of 15 files from one raw with different parameters of brightness, contrast, an exposition and a saturation (12).

From one file it is difficult to receive such picture: red color smoothly passes in green color without orange, without the yellow (edited photo).

Each method of processing has the merits and demerits. For example, processing in the LAB mode very well showed 4th order, but a bad color rendition of 3rd order rainbow.
Processing with imaginary hdr shows 4th worse, but much better color at 3rd order rainbow.

This sketch show the most interesting moment.

My 3rd and 4th order rainbows are very similar to rainbows of Michael Theusner: strictly at level (at height) the sun, rainbows seem vertical. From below and from above, rainbows sharply are rounded. This effect (I think) is explained by that rainbows have the best brightness at sun height. Very much reminds ice halo: at it too (very often) the brightest piece at the left and to the right of the sun.

Nicolas Lefaudeux invented a search method 3rd order rainbow. His method is outlined here and given in more detail.

I used an other (own) method. It is a Processing scheme to find a rainbow in the photo from one 16bit tiff file from RAW (in LAB mode in Photoshop):
RAW file -> Lightroom3 -> zeroed preset -> 16bit tiff file -> Photoshop -> LABmode

I don’t think that my Processing scheme can be suitable for all photos of other photographers.
But, this method very well shows rainbows in my photo (Frankly speaking, I couldn’t repeat Nikolos’s method – I am the novice user of photoshop 🙂 ).
For faint Rainbows it is necessary to work with layers of A and B (in LAB mode).
You can see a layer “L” on this picture and here the result of work with use of my method.

Author: Sergei Antipov, Russia

Related Post: Natural tertiary rainbow 3rd order

3rd and 4th order rainbows

Last evening (11 June 2011) thunderstorms approached my home town Schiffdorf near Bremerhaven in Northern Germany. I went to a field road by car to take some photos of the storm clouds. Just after I had arrived (about 18:00 UTC), heavy rain started which lasted for nearly 20 minutes. To my disappointment, the rain covered the gust front and most of the interesting features of the storm. So I waited and hoped that the sun would come out soon and produce some nice rainbows. When it did I realized that the dark clouds covered the sky to the right of the Sun – just the situation Michael Großmann had had when he took the the first image of a 3rd order rainbow only four weeks ago. I decided to try this out as well. Instead of one image I took sequences of five to stack them and, thus, increase the signal-to-noise ratio. I hoped this would increase my chances to detect the 3rd order bow. I took the images from my car through the open window to protect my camera (Canon 40D) from the rain. Visually, I did not see a 3rd order rainbow. However, in my back, the 1st and 2nd order bow developed nicely.

Back home I converted the raw images to 16-bit-Tiff and stacked them in Photoshop. Adjusting saturation already showed the 3rd order bow in the image sequences taken between 18:17 and 18:22 UTC (first image).  Applying unsharp masking revealed something unexpected in one of the stacked images (from 18:19 UTC): There seemed to be another rainbow close to the 3rd order bow, but, with reversed colors (second image). I checked Les Cowley’s website and realized that my image likely showed the 4th order rainbow!

After some more sophisticated processing including denoising (Neat Image), unsharp masking and increasing saturation, the 3rd and 4th order rainbows both were clearly visible. Finally, I created a composite using masks to retain the natural look of the foreground while still showing the 3rd and 4th order rainbows (third image).

Author:  Michael Theusner, Bremerhaven, Germany