Lars coordinates are messed up somehow. The keywords specifying the location on the disk exhibit problems: Coordinates are given via two redundant sets of keywords: 1. `SHG_LAT` and `SHG_LON` give the [Stonyhurst Heliographic Coordinates](http://www.mssl.ucl.ac.uk/grid/iau/extra/local_copy/coordinates.pdf) position in degrees according to the header 2. `HPC_X` and `HPC_Y` give heliographic coordinates in arcsec If these coordinate information sets match up, there should be a steady relationship between them. Plotting reveals some sort of mismatch: