Tempo2 silently produces wrong answers when position and proper motion are in different coordinate systems

Tempo2 will process par files where the position and proper motion are in different coordinate systems, but gives incorrect answers. I tried this for all four combination of ecliptic and equatorial coordinates and the two with matched coordinates work (of course), but the two with mismatched coordinates run but give incorrect answers. See attached 4 residuals plots.

PINT also has this problem so I filed an issue there too: https://github.com/nanograv/PINT/issues/654

TEMPO does the correct thing and fails with a sensible error message. This is what Tempo2 should do as well.

% tempo -f TDB_ECLPOS_EQUPM.par TOAs_NUPPI_1484_F4Tp.tim
 TEMPO v 13.000 Princeton/ATNF Pulsar Collaboration
 Data from TOAs_NUPPI_1484_F4Tp.tim,   Input parameters from TDB_ECLPOS_EQUPM.par                                                                                                                                            
ERROR: cannot mix ecliptic and equatorial coordinates

% tempo -f TDB_EQUPOS_ECLPM.par TOAs_NUPPI_1484_F4Tp.tim
 TEMPO v 13.000 Princeton/ATNF Pulsar Collaboration
 Data from TOAs_NUPPI_1484_F4Tp.tim,   Input parameters from TDB_EQUPOS_ECLPM.par                                                                                                                                            
ERROR: cannot mix ecliptic and equatorial coordinates

‌

Comments (2)