Conceptually I had associated v2_ref, v3_ref with the v2/v3 location corresponding to radec_ref. That's what the fields currently mean in Roman.
However, in Webb, the aberration correction is applied after v2/v3 ref, so that instead an "aberrated v2/v3" reference position is what corresponds to ra/dec_ref. This is close to the original v2/v3 ref but scaled by va_scale.
We should consider whether we want to also provide these aberrated v2/v3 reference positions, which would make it so that we are providing the v2/v3 reference positions that actually correspond to ra/dec reference points. Or alternatively add new ra/dec_ref that correspond to the post-aberration locations of v2/v3_ref.
Conceptually I had associated v2_ref, v3_ref with the v2/v3 location corresponding to radec_ref. That's what the fields currently mean in Roman.
However, in Webb, the aberration correction is applied after v2/v3 ref, so that instead an "aberrated v2/v3" reference position is what corresponds to ra/dec_ref. This is close to the original v2/v3 ref but scaled by va_scale.
We should consider whether we want to also provide these aberrated v2/v3 reference positions, which would make it so that we are providing the v2/v3 reference positions that actually correspond to ra/dec reference points. Or alternatively add new ra/dec_ref that correspond to the post-aberration locations of v2/v3_ref.