- S. Wheelock:
presented the results of her analysis
of the completeness and reliability (C&R) for point sources in the FS16
scan of 5-3-95. This scan represents a best case analysis as it had
very good seeing. Sherry presented statistics on the n/m confirmation
rates in each band, (m=6 in this case), from which she derived the
baseline C&R numbers to J=15.8, H=15.2 and Ks=14.2. The differential
completeness numbers are very good; essentially above 99% in all
magnitude bins. The formal reliability numbers are a little lower,
dropping to 98.5% in some bins. Sherry then described the results of
her detailed checking of all Ks band sources with less than 6 confirmed
sightings, from which it became clear that the true astrophysical
reliability is considerably higher than the formally derived value,
because more than half of the lost sightings were due to double star
confusion, not false detections. Double stars fall outside the
completeness and reliability requirements insofar as counting the two
stars as separate entities near the resolution limit. The group
discussion led to the conclusion that C&R could probably be improved
somewhat by tuning PROPHOT's bright star exclusion radius as a function
of seeing, in order to have a more robust rejection of false sources on
the edge of bright stars. It may also be advantageous to undertake an
analysis of a known set of double stars with the simulator to determine
a measure of PROPHOT's ability to detect/resolve double/multiple stars
under various seeing conditions.
- S. Terebey and T. Jarrett:
described their
progress in tuning GALWORKS to reliably detect true extended sources in
galactic nebulous regions, whilst flagging stars superposed on
nebulosity as detected but not likely to be true extended sources, and
rejecting noise spikes superposed on nebulosity. They are
investigating several promising parameters, the most interesting two
being based on measures of the local background compared to the global
coadd background. Tom compared the results in the Rho Oph region to
Coma, in order to have a test of the performance of the new parameters
on true galaxies in non-nebulous regions. The results are promising;
the nebulous objects have much larger values of the local background
parameters than do Coma galaxies.
- G. Laughlin:
continued the FOCAS saga, reporting
that the idea of subtracting a uniform background failed (FOCAS merely
extracted one enormous source over the entire coadd). She and R. Cutri
did succeed in obtaining an extracted galaxy list from the
unbackground-subtracted image by zeroing in on more reasonable choices
for the SNR thresholds. About 80 galaxies per coadd were extracted.
During this process, Gaylin discovered an error in one of the FOCAS
scripts; when she reported this to the creator of FOCAS, Frank Valdes,
he realized that profile fitting must have been inoperable since the
last release of FOCAS. Gaylin will now do a detailed comparison of the
FOCAS-extracted source list to GALWORKS.
- R. Cutri:
has applied the newly derived
photometric zero points to all the 1995 calibration fields and showed
the resulting total photometric uncertainties for several of these
fields. FS18 is the best case as it contains the largest numbers of
zero point calibrators; it shows a total 3-4% photometric uncertainty
for the brightest stars. More realistic cases are P221 where the sigma
rises to about 7.5% at J, and 3-4% at H and Ks, and FS28 which shows
7-8% uncertainites at H and Ks. Roc pointed out that these fields were
observed early in the run, when there were notes in the observing log
that imply conditions may not have been 100% photometric. Indeed, the
FS23 field observed towards the end of the run, when the photometric
conditions appeared to be better, has much better photometric
accuracy. The overall implication, then, is that the poorer photometry
corresponds to true transparency variations. This points again to how
imperitive is it to have adequate standard star observations. Roc
feels that the baseline standard star observing sequence for the survey
will most likely do the job OK; the May/June protocamera run did not
meet this baseline level. After group discussion it was decided that
these photometric calibrations are now ready to be applied to the
entire 1995 protocamera data (perhaps after a little more tweaking by
Roc), which will then be ready to install in sybase and XCATSCAN.
- C. Beichman and T. Evans:
announced that the new
version of XCATSCAN, including the 92-94 protocamera data, is ready.
It is being tested this week, and if no problems are found will be
installed in the operational version of XCATSCAN, with 2MASS password
protection, on Thursday.