Think about light travel time/tangent plane approx stuff for fast-moving stars

[sblunt]

Eric says to check out https://ui.adsabs.harvard.edu/abs/2019A%26A...623A..72K/abstract

[semaphoreP]

Mainly just for stellar astrometry. Only matters for nearby stars.

[enielsen3522]

You can do a quick order-of-magnitude to see if it's important, using radial velocity and proper motion of the star. The basic question is "how many light-days closer will the star be over the timeframe of interest, and is the expected proper motion over that number of days significant compared to the astrometric error bars?" To pick a completely random star, eps Indi has an RV of -40 km/s, which means over 25 years (the length of time between Hipparcos and Gaia) it gets about 1.2 light-days closer to us. Because eps Indi is so nearby, it moves 4.7" per year, or about 13 mas/day, meaning if you don't take light travel time into account, and your astrometry error bars are 1 mas or less, you'll get the answer significantly wrong. So much like doing rigorous coordinate transformation (also most important for the closets stars), you'd need proper motion and RV to tell if this is important in the first place, but the correction isn't too computationally difficult if it is needed.

[semaphoreP]

Related to #326 but more about absolute astrometry/spherical geometry.