MAPs

We now have a good idea of the distortion map of the prime focus field, based on three MAP fields with 0.4-0.45 arcsec rms errors. The main contributions to the errors come from the instrinsic uncertainty of the stellar positions (0.25 arcsec), AUTOFIB encoder errors (0.1 arcsec) and guiding errors on a highly comatic image (0.3 arcsec). The linear plate scale component is arcsec/mm and 2nd order plate scale term is , as of May 1996 (you can see the history of these coefficients **here**). However, if AUTOFIB2 has just been put on the telescope, one should do a MAP field in order to check out these figures; usually the Support Astronomer would do this. To do a MAP:

• Transfer the MAP file (e.g. map08_1314.cfg) to the appropriate directory on the VAX (see Section 2.3)
• When you are ready to start a MAP, type afsetup mapfile in ICL to setup the map file. This will only take a few seconds.
• Type afpark, to make sure robot is parked.
• Type afmprob 150000 to move the fixed probe to the edge of the field.
• Move the telescope to the field, remembering to have the proper sky PA (usually MAPs are done with sky PA = 90); it's a good idea to get a printout of the .cfg file so you can check this, and also so you can have a listing of your MAP stars. Make sure guiding can be done on fixed probe and centroiding on mobile probe (fm 3 on TCS).
• You should be able to see the guide star in the fixed probe autoguider window. It will look very comatic of course, since it's on the very edge of the 1 degree field. Start guiding on the fixed probe, and close the loop (TO). Check for a few seconds that it's guiding reasonably well.
• Before you actually start a MAP, it would be good to send the mobile probe around to a few of your MAP stars to make sure they actually are seen in the mobile probe autoguider window. If several don't show up, you will have to tweak the telescope translation and rotation and try again.
• You can now start the MAP. Type afmap on ICL, and it will ask you:
  • How many centroids? (3 should be ample)
  • How much integration time per centroid? (2000-3000ms usually, but check brightness of stars with the mobile probe)
  • How many stars in total? (at least 40)
• The mobile probe will now be sent around to each of the MAP stars, and the probe will centroid n times on each star. 40 stars will take **??** minutes.
• When it is finished, you can look at the output file: astrom.af2_dat on the .af2 directory on the VAX, which can be fed directly into the astrom program, available on the SPARCs. If you get a reasonable fit from the 9 coefficient fit, i.e. rms residuals 0.5 arcsec or better, check to see if the current coefficients in wht_prime.dat (the linear plate scale component, and the 2nd order term of plate scale) are significnatly different from the new MAP results. If they are, edit wht_prime.dat on the VAX .af2 directory.