While hunting around for a good reference on flattening/unflattening an LSPR solution, I came across this excellent summary, written by none other than our first guest speaker, Dr. M. Faisson, way back in 2010. As you read through it, you’ll see it tracks what we did in lecture last week very closely. You’ll need to pay particular attention on the second page, beginning with “More generally,”
Note that the plate constants for a flattened coordinate system will not match the constants I gave for the original test case, but the final coordinates should still match when all is said and done.
As you implement this, keep track of two things in particular:
- Does the model RMS increase or decrease?
- Does your asteroid’s position become more or less accurate?
It comes down to the interplay between the solution improving because of a (small) correction and the solution deteriorating because of the additional errors added by the extra mathematical steps.