Date: Wed, 12 Aug 1998 14:00:14 -0700 (PDT)
To: 2mass
Subject: IPAC 2MASS WG Meeting #160 Minutes
Cc: chas, stiening, bgreen
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-MD5: ZfT5gGAUnhhVvDRxz53LJg==

           IPAC 2MASS Working Group Meeting #160 Minutes

Attendees: R. Cutri, S. Van Dyk, L. Fullmer, R. Beck, T. Evans, R. Hurt, 
           J. Fowler, G. Kopan, H. McCallon, J. White, S. Wheelock,
           R. Tam


1.) R. Tam joins 2MASS
2.) 2MASS Mosaic of the Galactic Center
3.) Project News


1.) R. Tam joins 2MASS

     R. Tam, who recently graduated with a BS from UCLA, joins the IPAC 2MASS
team.  (He is yet another Bruin to continue the siege on IPAC.)  He will assist
with data analysis toward final product generation.  

2.) 2MASS Mosaic of the Galactic Center

     R. Cutri showed one and all an image that he had printed out, to test the 
eyes and astronomical knowledge of all team members.  The image, of course, was
a small mosaic of the region covering the Galactic Center.  The source, 
Sagittarius A, is a bright near-IR source and clearly visible on the 2MASS 
composite JHKs image.  A larger, more complete mosaic will be made and added to
the Image Gallery.

3.) Project News

     The latest news is just a continuation of old news, with more problem
shake-out for 2MAPPS 2.1.  Northern observing should resume from the summer 
shutdown on or around September 1.  A second northern telescope operator is 
being hired.  When northern operations start again in September, R. Steining 
will work on the telescope collimation and polar axis alignment.

     The summer shutdown for northern ops will represent a new break between 
hardware periods.  J. Fowler discussed separate "hardware periods," for which 
distinct data sets must be maintained, and the steps involved in adding new 
hardware periods.  Each hemisphere has a "history" directory; histn is the 
directory for the north, and hists is the directory for the south. Each exists 
in a directory tree structure set up by OPS with pathnames that may change as 
needed for operational reasons.  Each hemisphere's history directory contains a 
subdirectory named darkhist and a subdirectory named pfprep. These are used by 
the DARKS and POSMAN subsystems. Each of these contains subdirectories named a,
b, c, ....  The single-character-name subdirectories correspond to "hardware 
periods", i.e., intervals of time within which the observational hardware is 
modeled as discontinuously different from other intervals of time. Currently,
there are two data items kept separate by hardware period: mask images and
model parameters, used by the PFPREP module stored in NAMELIST files.
     In the darkhist directory underneath histn and hists, a file named maskdat
associates the a, b, c, etc., directories with start and stop dates that define
the hardware period for each directory.  All that matters is that ranges of 
nights are segregated into subdirectories.  In the south currently only one 
hardware period exists, so maskdat in hists/darkhist consists of only one

     OPS uses a program named getmask to automate retrieval of data that depend
on the hardware period; this program reads the maskdat file, finds the
period in which the night being processed falls, and echoes the one-character
directory to stdout, from where it is used to access the corresponding data. 

     When a new hardware period is defined, the following updates must be made:
a) A new one-character directory must be made under darkhist and pfprep for the
hemisphere involved; b) a new set of masks for the period must be placed in the
darkhist subdirectory; c) a new nl.pfprep for the period must be placed in the
pfprep subdirectory; and, d) the maskdat file in darkhist must be edited.

     R. Cutri asked R. Beck to work with J. Fowler on new subdirectory 
generation for mask images and NAMELIST changes, as well as for startup darks.

     R. Cutri stressed that more southern data needs processing, and therefore,
processed data online needs to be moved off-disk to tape, to make disk space.

     The highest priority still is the PSFs for 2MAPPS 2.1 startup.  Some 
progress on this front has been made.  W. Wheaton had reported that the PSF ID
numbers and lookup tables correspond to reality (see WG minutes #159).  A
breakthrough has occurred in the production of the southern PSFs, and the 
resulting chi^2 distributions look better (more like the north).  R. Cutri asked
T. Evans to populate the lookup tables, especially for poor seeing conditions in
the south.

     G. Kopan gave an update on the magnitude-crossscan position bias problem.
Some progress has been made, with analyses for two of the southern RTB nights.
No obvious correlation has been found with telescope temperature.  (In all tests
the default aperture magnitude is selected.)  The magnitude differential as a
function of crossscan position appears to be, and will undoubtedly continue to
be, a variable relation both from night-to-night and even scan-to-scan 
throughout a single night.  L. Fullmer was asked to look at the statistics for
the scan-to-scan overlap regions.  The parameters at work for this problem need
to be isolated.

     R. Cutri hopes to have correction software by the second week of September
for the start of southern OPS processing.  He would like at least a zeroth-order
correction, to reduce the magnitude differential from about 9% to, maybe, 2 or 
3%.  The goal is to get the southern OPS going soon.

     R. Cutri and S. Wheelock discussed the J-band zeropoint drift, which is
quite systematic and not unexpected.  It comes about since the J-band used can
be easily affected by water vapor, which effectively changes the bandpass.
S. Wheelock has a prototype working for a constrained linear fit to the trend,
which amounts to a 0.1 to 0.2 magnitude drift.  (CALMON originally had a linear
variable zeropoint, which was removed by suggestion of the Science Team before
2MAPPS OPS began.)  The effect at H and Ks exists, but is small.  The night
980112n has been tested with the prototype code.  A retroactive application is
necessary for, especially, the most severely-affected cases.

     Finally, a new Enterprise 450 ("Son of Barney") has been ordered, as well
as a new 450 for the database effort (see "Christmas in August" from WG minutes