Too big parallax from apply_space_motion

I have been trying to plot the path of a star on the plane of the sky as seen from Earth. I have the following example for a source located toward the Galactic bulge and at 8 kpc. However, the parallax I get after calling apply_space_motion and converting from ICRS to GCRS is far too large. Am I doing something wrong or is there an issue in the conversion?

def testing_astropy_parallax():
    from astropy.coordinates import SkyCoord, GCRS
    from astropy.time import Time
    import astropy.units as u
    import numpy as np
    import pylab as plt
    
    ra = 17.5 * 15.0 * u.deg
    dec = -29 * u.deg
    dist = 8000.0 * u.pc
    sc = SkyCoord(ra, dec,
                  pm_ra_cosdec = 0.0 * u.mas/u.yr,
                  pm_dec = 0.0 * u.mas/u.yr,
                  distance = dist,
                  obstime = Time(57000.0, format='mjd'))
    
    t = np.arange(56000, 58000)
    t_obj = Time(t, format='mjd')
    
    sc_t_icrs = sc.apply_space_motion(new_obstime=t_obj)
    sc_t_gcrs = sc_t_icrs.transform_to('gcrs')

    ra_t   = sc_t_gcrs.ra
    dec_t  = sc_t_gcrs.dec
    cosd_t = np.cos(dec_t.to('radian'))
    
    dra = ((ra_t - ra) * cosd_t).to('arcsec')  # in arcsec
    ddec = (dec_t - dec).to('arcsec')          # in arcsec

    plt.figure(1, figsize=(10, 3))
    plt.subplots_adjust(wspace=0.6)
    
    parallax_pred = 1.0 / dist.value
    print('Predicted parallax from manual calculation:')
    print('    {0:.2f} mas'.format(parallax_pred*1e3))
    
    plt.subplot(1, 3, 1)
    plt.plot(t, dra, color='black')
    plt.xlabel('Time (MJD)')
    plt.ylabel(r'$\Delta \alpha^*$ (")')
    plt.axhline(parallax_pred, color='red', linestyle='--')
    
    plt.subplot(1, 3, 2)
    plt.plot(t, ddec, color='black')
    plt.xlabel('Time (MJD)')
    plt.ylabel(r'$\Delta \delta$ (")')
    plt.title('Star Distance = ' + str(dist.to('pc')))
    
    plt.subplot(1, 3, 3)
    plt.plot(dra, ddec, color='black')
    plt.xlabel(r'$\Delta \alpha^*$ (")')
    plt.ylabel(r'$\Delta \delta$ (")')
    plt.axis('equal')
    
    return

[schlafly]

Just adding a +1 to this… I just also wanted to propagate stars from their Gaia epochs to the current epoch. I started staying in ICRS the whole time, and then nothing moves. I think retrospectively that’s correct, since ICRS is centered at the solar system barycenter, but it wasn’t what I originally expected.

After reading this issue, I’m not sure that this example in the documentation https://docs.astropy.org/en/stable/coordinates/apply_space_motion.html#example-use-velocity-to-compute-sky-position-at-different-epochs is correctly including the parallax motion as well. Maybe I should be reading that example as only including the proper motion component, but the presence of the parallax in the SkyCoord constructor certainly suggested to me that the parallax was being taken into account as well.

Reiterating @jluastro’s statement, what’s desired is how the nearby stars move relative to the distant stars. I guess that’s some hybrid frame that has its origin at the center of the earth, but the velocity of the solar system barycenter?

[jluastro]

I would like to be able to plot the proper motion and parallax of a star as it would be seen on the sky; but without the annual aberration. I honestly don’t even know what kind of reference frame this would be. It is a geocentric observation; but the coordinate aberration should be removed. Is this in the existing coordinates somehow?

Essentially I want ICRS + geocentric motion… stars at infinity are fixed (e.g. ICRS), but see perspective change from Earth.