- #1
arlosaur
- 1
- 0
I am doing a geophysics project where I am analyzing a number of lineations on a globe / sphere. That is, I have a data set with many (70 - 180) sets of data points that are grouped together in lineations. I am trying to determine if the lineations are best described by concentric circles about a specific pole. To do this for every longitude/latitude value on the sphere I rotate the entire data set so that the chosen axis of rotation is now the z-axis. I then fit a z-plane through each lineation, which amounts to averaging the z values for each lineation after rotation. I also calculate the sum of the squares of the residuals (z_i - <z>)^2 for each lineation. I then calculate the 'best fit' pole by choosing the pole which minimizes the sum of the squares of the residuals of each lineation with their respective best-fit plane.
Now I wish to come up with a hypothesis test or a goodness-of-fit test to determine if the lineations are best described by concentric circles about this pole (as opposed to two or three poles or something else altogether).
Any ideas? Thanks.
Now I wish to come up with a hypothesis test or a goodness-of-fit test to determine if the lineations are best described by concentric circles about this pole (as opposed to two or three poles or something else altogether).
Any ideas? Thanks.