Date: Fri, 5 Feb 1999 18:24:26 -0800 (PST)
To: 2mass-staff
Subject: IPAC 2MASS WG Meeting #179 Minutes
Cc: chas, stiening@cannon.phast.umass.edu, bgreen,
skrutski@north.phast.umass.edu
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-MD5: +ydEdS80Q2KWq5yF2t3U+w==
IPAC 2MASS Working Group Meeting #179 Minutes
2/2/99
Attendees: R. Cutri, S. Van Dyk, D. Kirkpatrick, T. Evans, R. Tam, H. McCallon,
D. Engler, W. Wheaton, R. Beck, G. Kopan, R. Hurt, S. Wheelock,
T. Jarrett, J. Rho, J. White (briefly)
AGENDA
1.) Administrative Details
2.) Project Update
3.) Spring Release
4.) New and Standing Action Items
DISCUSSION
[J. White was briefly at the meeting, but was dragged away by M. Moshir,
undoubtedly for another SIRTF meeting.]
1.) Administrative Details
R. Cutri will be away from Feb. 6 through 18. In his absence, G. Kopan
will be the contact point for software delivery and other software issues. D.
Kirkpatrick will be the contact for all data processing issues. J. Mazzarella
will be the contact for 2MASS-related IRSA issues and will monitor the database
reload.
2.) Project Update
T. Evans reported on the database reload, which is currently about
one-third of the way through. The reload of the Spring release data should be
done within the next 24 hours.
[Note: as of Thursday, Feb 4, the database reload was approximately 60%
complete.]
W. Wheaton gave an update on the next round of PSFs. PSFMake has been run,
and the 600-odd candidates must be put into the table for the select routine to
deal with them. He is about a day from having the table filled. R. Cutri asked
if he needed some help. W. Wheaton reported that the actual process is much
more efficient now; he has be "smarter" in selecting useful runs, although
sometimes only about 20% of the candidate fields had adequate numbers of stars.
R. Cutri emphasized that we within about three weeks of needing this set of
PSFs, for the period after the summer shutdown. This represents a hardware
break, with need for new flats and darks. After the collimation nights that
followed the shutdown, 980919n looks like the first good night for which this
new set of canonicals will be appropriate.
On the QA front, D. Kirkpatrick reported that last Friday, the QA was
almost caught up with processing. Now, it is back to being seven to eight
nights behind.
G. Kopan reported on the revised noise estimator for the point source
detection threshold, especially for crowded fields. The problem is that in H
the completeness, especially for these fields, is underestimated. The goal is
to be able to possibly go deeper in the presence of OH airglow. The point
source filter estimator is still sensitive to confusion noise. Higher noise is
obviously found in confused regions. The trick is to get the estimator to find
real point sources in the presence of airglow. This depends on the steepness of
the background gradient.
T. Jarrett pointed out that extracting more sources in high airglow
regions, particularly at H-band, is not a good thing, because GALWORKS uses the
point sources as its seed list. In particular, if the detection step finds
more spurious sources because of the airglow structure, it will result in more
spurious extended source identifications. This has the potential of
invalidating much of the analysis of extended source reliability in the presence
of airglow structure. R. Cutri indicated that it will be necessary to explore
the ramifications of the new source detection thresholds on GALWORKS and other
downstream systems before final implementation. J. Carpenter is also exploring
the completeness limits in high source density regions, and Gene Kopan was
waiting to hear more from Carpenter about this.
3.) Spring Release
R. Cutri outlined the essential tasks necessary to prepare for the Spring
data release:
1) Define the catalog criteria, including a) scan selection (based on
quality, etc.) and b) source selection (based on SNR, etc.). This
is about 90% of the way there.
2) Extract the release database, which is all sources in 1a) above.
3) DBMapcor version, to identify unmarked artifacts and perform flag
updates.
[Included in between 3) and 4) are the positional uncertainties update.]
4) Duplicate source rectification; H. McCallon and S. Wheelock are working
on this. A flag will indicate that a source is a duplicate, and another
flag will indicate whether or not to use that apparition for the source.
This step will also include positional uncertainty updating.
5) Galaxy classification (This can begin after step 2)
6) Catalog/source selection, based on 1b) above.
7) Source Catalog formatting. A scan information table will also be
included with the release catalogs.
T. Evans reported that she is within one week of having the inner portion
of DBMapcor running, with a first generation of new parameters. She will try,
with R. Tam's assistance, to get all artifacts in the release parameterized.
R. Cutri emphasized that a first-cut toy catalog is necessary as soon as
possible with which to play, hopefully by the end of the month. Detailed
analyses need to be started. Yet, the SNR threshold (see 1b above) still has to
be decided upon.
H. McCallon reported on the positional uncertainties update work. He is
working with L. Fullmer's scan overlap data. He sees some differences with the
Sampler, mostly in that he is now using trimmed averages now, rather than
straight averages, as was done for the Sampler release. R. Cutri asked, if
H. McCallon were to get database files by tomorrow, how long would it take to
have a start on the code to update the positional uncertainties. H. McCallon
answered about one to two weeks. S. Wheelock can have the database files with
overlaps to H. McCallon by possibly the end of the week. The need for "up-down"
overlap was stressed, as opposed to just scan-to-scan overlap, to perform
rigorous testing to develop the routine. For the Spring release, the goal is
to update the uncertainties, not the positions themselves. No threshold on
astrometric accuracy is in place for this release. All will be released, even
ACT-challenged fields, since the opinion is that 2MASS positions are already
far superior to pre-existing source positions, in many cases. So, no image
header updating will be done.
T. Jarrett reported that the galaxy classification routine is ready to go,
and is based on the Sampler procedure.
4.) New and Standing Action Items
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% %
% ACTION ITEMS %
% %
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
(Standing Action Items:)
a) Team members should continue to analyze data for Spring data release.
Members will likely be called upon to assist in steps toward catalog generation
for the release, as discussed above.