Drift maps in J2000 coordinates.


    Drift scan are are normally done in coordinates of date (current coordinates).  A starting ra, dec is used with an ra rate of 15 arc seconds per sidereal second. The telescope then remains fixed as the sky drifts by.   This is fine for small maps, but it can be a problem for larger surveys that extend over many years.

    Since the az, za is remaining fixed, the drift is being done in current not J2000 coordinates. If you were planning to map in J2000, or to repeat a strip 2 years later, this would be a problem. To see how large this change was,   a drift scan in J2000 coordinates (15 asecs/sidereal sec epoch J2000) was computed and then the relative motion of the az, za coordinates were plotted. For each day computed:

  1. The telescope was positioned to az=180,za=[2..18] in 2 deg za steps.
  2. The ra, dec for 0 hours utc was computed using the az, za values.
  3. A rate of 15asec/sec sidereal was applied  to the starting ra over the entire day
  4. The az, za was then recomputed for these new J2000 ra, decs over the day. These are the az, za's that would be tracked if the drift was really in J2000 coordinates.
  5. The az, za motion was then plotted over the day removing the starting az, za (which was at the correct J2000 position for the start of the strip).
The plots show the az/za relative motion (.ps)    (.pdf) for the entire day. The top plot is the azimuth encoder motion. The middle plot is the azimuth motion on the sky. The bottom plot is the za motion (encoder/sky). The amplitude is increasing yearly since the precession is moving the pole away from the J2000 position. The plots show that:
processing: usr/a2010/chkdrift.pr