From high frequency commissioning meeting on February 03, 2011:
Item 1:
John Cannon presented a brief description of his Ku reference pointing test, which
he has documented in ECSV-16, sub-task 6. Discussion in the meeting centered
around the two reasons for which one might wish to learn about using Ku band
reference pointing: 1) Would Ku band work better than X band in determining the
pointing offsets to use at high frequency (K, Ka, Q) ? and 2) Does reference
pointing need to be applied (of course, determined before applying) at Ku band ?
Vivek pointed out that the answer to question 1 is known --- since the beam is
smaller at Ku, the blind pointing of the telescopes can be off by a significant
fraction of the Ku band beam (except perhaps on calm, clear nights), so that it
is possible that Ku band reference pointing will fail a lot of the time. So unless
those conditions are met, we probably do not want to suggest to observers that Ku
band reference pointing be used. Vivek and John will continue further discussion.
Historically the answer to question 2 is that applying reference pointing at Ku
band is a wash. But perhaps there is some room for quantifying what we mean by this;
and also, perhaps that discussion is better done in a mid-frequency commissioning meeting.
Item 2:
There is a need for an "Inner Galaxy" flux density calibrator in the
LST range approximately from 16 to 20 hours. This is because in this
range 3C48 and 3C286 are not available, especially for scheduling blocks
(SBs) which are rather short (2 hours), and if an observer wants his/her
flux density calibrator observed above 20 degrees elevation. This problem
sometimes leads observers to submit two blocks --- an early one to observe
3C286, and a late one to observe 3C48, and then it is required that someone
or something keep track of which one is observed. If a reasonably bright
source could be found in this range and demonstrated to either be non-variable,
or the flux density could be tracked, it could be offered to users who need
to observe in this time range. Some criteria for a flux density calibration
source are: 1. Source transits on the south side of the zenith; 2. Right Ascension
approximately 16 - 20 hours. 3. Reasonably strong (what does this mean ?)
4. A point source, or structure than can be reasonably modeled. 5. The flux
density needs to be tied to 3C286, or to be monitored. 3C454.3 (J2253+1608)
is a possible candidate for this flux density calibrator; Lorant has been using
the TCAL0007 project to monitor the flux density of 3C454.3 at several frequencies.
The data he has looked at so far, at high frequencies, is affected by weather,
and so is not convincing that the flux density can be monitored well enough to
use this source. More work is needed. The discussion in the meeting focused mainly
on if there were other possible sources, and how strong such a calibrator really
needed to be. Some suggestions for other sources were 1) thermal sources which are
non-variable, but likely have a lot of structure, especially at high frequencies
and extended configurations; 2) CSOs, which are likely the best candidates, if
a reasonably strong one could be found; and 3) IDV rejects. Suggestions for
finding these sources were the CORNISH survey catalog (for thermal sources)
and asking Greg Taylor about CSOs (probably from the VIPS survey). Lorant said
that he would investigate these other possibilities.
(Please see the comment on ECSV-148 on this topic by Rick Perley.)
Item 3
Claire Chandler and Hendrik Linz presented a report on using some data taken
in September 2011 which was a test delivered in support of Claire's project
AC982. I think Claire will be putting together more details which will go into
this JIRA ticket. The data was, i believe 8 subbands, each 128 MHz and dual
polarization. These observations were made in Ophiuchus, as that is where the
observers of AC982 were having trouble finding a nearby calibrator. Bandpass
calibration was done, and short baselines were used (the EVLA was in the D
configuration). In order to average across subbands, a procedure was needed
to determine the phase offsets among the subbands. There was some non-standard
manipulation that was needed in CASA, which I hope Claire will report on more
carefully. After this procedure was applied, averaging the subbands was done,
and it appears that sources with >~ 30 mJy/beam could be used as phase calibration
sources (total effective bandwidth 2 GHz). A source in this experiment that was
~ 10 mJy/beam was too weak (phases too noisy) for following the phase, and thus
for use in phase calibration.
Some discussion after Claire/Hendrik's presentation yielded the following: 1) In
principle one should not need to do offsets after the bandpass calibration (Michael,
I think); 2) Need a strong source for amplitude calibration every ~20 minutes;
3) Need to see if this will work (or at least to what level) for longer baselines,
i.e. spread-out configurations; and 4) Just when do we need to find these calibration
sources --- which goes back to questions we have discussed before about the
calibration model (e.g. ALMA), calibration source surveys, etc.
The message in the end that for this somewhat limited test that one can get the
expected sensitivity. Data which were taken specifically to use this phase calibration
method for real observing was taken the weekend of the end of January; Claire and
Hendrik will be looking at that data soon.
Item 4
Vivek Dhawan has been doing tests of time averaging of reference pointing. He was
not quite ready to discuss this at this meeting (data having been taken only the
evening before). He did give a report at the ECSV meeting on February 8, 2011 which
will migrate to ECSV-16, sub-task 7.
--
JosephMcMullin - 2011-02-14