
Despite the cold and a bright moon I went out last night asteroid hunting. This is a dozen lights at 30 seconds 800 ISO with the William Optics Z61(360mm f/5.9) with corresponding biases, darks, and flats. I used the same technique as with uranus to locate the pixel where Ceres should be but it was a bit off – 14 pixels X and 12 Y: 18 pixels total which would be 43″ which is an awful lot. But in my defense, there’s nothing else that bright near the correct location and there are plenty of stars dimmer than Ceres(Mag 9) that are showing up perfectly well so probably Ceres plus some error in my calculations or understanding. Interestingly, Ceres would have moved against the background stars over the time I was shooting but only a couple of arc-seconds – maybe 1 pixel. That does mean though that if I had waited an hour or so and shot again I might have had a convincing shift.
from astropy import wcs
from astropy.io.fits import getheader
import numpy as np
from astropy.io import fits
from astropy import units as u
from astropy.coordinates import SkyCoord as SC
hdulist = fits.open(r"c:\users\bill\downloads\wcsceres.fits")
w = wcs.WCS(hdulist[0].header)
pixcrd = np.array([[0, 0]], dtype=np.float64)
world = w.wcs_pix2world(pixcrd, 0)
print("RA/DEC for [0,0]: ",world)
wcrd=np.array([[335.688, -23.789]])
pixcrd2=w.wcs_world2pix(wcrd,0)
print("Pixel Coords for center as given by astrometry: ",pixcrd2)
uc=SC('22h22m51.65s', '-23d47m30.7s')
print("Ceres Coordinates(J2000): ",uc)
print("Ceres Pixel Coordinates: ",uc.to_pixel(w))
uc=SC('22h23m31.01s', '-24d45m45.2s')
print("49 AQR Coordinates(J2000): ",uc)
print("49 AQR Pixel Coordinates: ",uc.to_pixel(w))

EDIT: I went back and picked a single subexposure, ran it through astrometry, and checked the pixel location and ended up with the same error so i dunno.
reEDIT: I googled around and apparently the positions of minor planets in stellarium are easily outdated due to perturbations in their orbits. I’ll check the web version vs the standalone version when i do pluto.