On the way home to prepare for this evenings crescent moon chase, I looked up frequently to the sky. The sky was so milky with cirrostratus and some altostratus clouds when looking west. Through sunglasses I noticed a rainbow arc above the Sun. I stopped to investigate it further. A quick measurement with spread thumb and pinky held at arm's length confirmed a 22° separation between Sun and arc. A partial arc of a 22° halo was observed from about "10 o'clock to 1 o'clock" of a circle's circumference. At the 3 o'clock position, an isolated, rainbow-colored bright spot, detached from the arc, was observed (marked "a" in illustration). At the time, I didn't know that was a sun dog, a.k.a. mock sun or parhelia.
I observed the halo and the sun dog with and without polarized sunglasses. Had I not been wearing sunglasses I would not have detected them; however, after knowing these atmospheric treats were there I could sense there locations without sunglasses by the brightening of the sky in those areas, as a spot and arc. Without the sunglasses, I was unable to see the color.
Later when I arrived home, I looked out the living room to the west. Haze and clouds could be seen down to the horizon. Shadows weren't very sharp, yet the masked Sun was still too bright to look at directly. The sky was awash with grades of white. Above the Sun, with sunglasses, I could see the arc. It rose from a 9 o'clock position up to 1 o'clock where my building obstructed the view. As bright as it was I suspected that it was strong enough to make it to 3 o'clock to join the sun dog previously observed.
I thought the sun dog was a section of the halo, horizontal to the sun, but the observation showed it differently. During this second observation, a second sun dog was observed on the other side, appearing lower at an 8 ~ 8:30 position. Using a building rooftop and my pencil to create a horizontal reference line with the sun, this sun dog was below it. Lastly, the halo appeared to be inside of the sun dog. Since I wasn't seeing the green or blue portion of the arc at this location I thought the sun dog was part of it but it was clearly too far out; perhaps it is tangent if the entire colored spectrum were seen.
Concerning color, red was interior and yellow was the neighboring exterior color. From down on the street, the first observation, green and blue was not so apparent, except on the sun dog. The second observation, the rainbow effect was pronounced where there was evidence of blue on the outside. This is the same color sequence of a secondary rainbow.
Want to know more, visit Atmospheric Optics where most of the links above lead to for more information.
Illustration based on drawing in The Skywatcher's Handbook, edited by Colin Ronan & others.