IPAC 2MASS Science and Analysis Working Group Meeting #13 Minutes
IPAC 2MASS Science and Analysis Working Group Meeting #13
Minutes,11/29/95
Note - there will be no meeting on December 6th due to the User's
Commitee Meeting
Attendees: C. Beichman, T. Chester, T. Evans, L. Fullmer, D. Kirkpatrick,
G. Laughlin, B. Light, C. Lonsdale, S. Terebey, S. Wheelock
Erratum: Meeting #12 Minutes
B. Light reported on R2-R1 vs. R1 KAMPHOT postional offsets. The minutes
incorrectly stated that there were no systematic offsets seen in plots of
dy versus R2-R1 magnitude, where dx = x(R2-R1) - x(R1-KAMPHOT), and
dy = y(R2-R1) - y(R1-KAMPHOT). Further details will be presented in a
forthcoming memo.
Meeting #13
- S. Wheelock:
presented C&R results in a series of
plots for almost all of the 1995 protocamera data, nights 9 to 22. From
this large compilation she showed the dependence of the point source
completeness and reliability on pfrac; the quality of the data deteriorates
below pfrac of about 0.4 to 0.45 (seeing greater than about 2.5 arcseconds),
though there are few data points in this range of pfrac. Some scans show an
anomalously low C&R for their pfrac, and the reason for this will be
investigated. The next step is to show C&R as a function of source density.
- G. Laughlin:
showed the first detailed comparison
between FOCAS and GALWORKS results for sources which both processors
classify as galaxies in a Coma scan. The agreement between the two processors
is very good in general; there are a handful of objects which FOCAS
detected as galaxies and which GALWORKS did not, and vice versa. These
will be studied in detail to ascertain the reasons the processors behaved
the way they did.
- C. Beichman:
has been investigating the nature of
sources from the '92-'94 database which were found by D. Kirkpatrick to
lack optical counterparts in the APM database. Davy was searching for low
mass stars and high proper motion stars, but his selection criteria also
select very well for spurious sources. Chas has found that a large fraction
of these non-optically identified sources are due to residual persistence
problems. This was done by deriving the density of sources at the predicted
location of up to 5 persistence images trailing bright stars, and comparing
that to random sources in the same pattern of positions but offset 15
arcseconds from the bright stars. His analysis shows that the persistence
rejection filters set when the '92-'94 database was built were too liberal.
Residual persistence is the biggest source of unreliability in the database.
These filters can be tightened, but at the expense of completeness since some
true sources will be found at peristence positions, as shown by the
above-described offset test. The best solution is probably to impose very
strict persistence rejection criteria, but to flag the likely persistence
images rather than reject them, so that a user can recover true sources
at these locations if they wish. The '95 database has much stricter
persistence filters than the '92-'94 database so this is expected to be
less of a problem for it.
- S. Terebey:
presented some 2D simulations of the effects of the 2 arcsecond
pixels and of coadd blurring on the PSF, using two different
assumptions for the seeing model: a Moffett function and turbulence
theory. Starting with a point source having a seeing FWHM of 1.8"
the single frame PSF was found to be 2.5" and the dithered coadd had a PSF of
3.2". The simulations illustrate vividly the effects of sampling and
pixelization on the shape of the PSF. The coadd images also shows a small
positional offset of 0.17". Both seeing models were found to fit the
profiles of real 2MASS bright star profiles well.