I have a series of points on a unit sphere that are given in azimuth, elevation coordinates, where azimuth has a domain of -180 < azimuth <= 180 degrees, and elevation has a domain of -90 <= elevation <= 90 degrees. These points are
(90, 75) (180, 75) (0, 75) (-90, 75)
For the center, intuitively I know that the azimuth does not matter, and the elevation would be 90 degrees. How would I calculate this mathematically? I have other situations like this and often with more than 4 points that would not be symmetrical, and would not have a center at 90 degrees.
Ok, figured it out. Yes, finding the centroid is easier using (x, y, z) coordinates of each point, but you have to account for azimuth wrap-around first. And yes, the center in this case needs to be on the surface of the unit sphere.
The sphere needs to be broken into two hemispheres, with the azimuth domain edge as a side and the azimuth line at 0 degrees as the other half.
I have some pseudocode below: