Date: Thu, 3 Sep 1998 15:49:32 -0700 (PDT)
To: 2mass
Subject: IPAC 2MASS WG Meeting #163 Minutes
Cc: chas, stiening, bgreen
Mime-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-MD5: L/tHeHkMF8DJ9FOl+8vwUA==

           IPAC 2MASS Working Group Meeting #163 Minutes

Attendees: R. Cutri, S. Van Dyk, R. Beck, T. Evans, R. Hurt, J. White,
           J. Fowler, G. Kopan, H. McCallon, S. Wheelock, R. Tam, 
           D. Kirkpatrick, D. Engler, T. Jarrett


1.) Project Update -- Southern OPS Processing to Get Underway
2.) Last Week's Action Items
3.) 2MASS Sampler
4.) S. Schneider Visiting


1.) Project Update -- Southern OPS Processing to Get Underway

     Northern OPS at Mt Hopkins should resume next week.  R. Stiening will 
arrive there on Sept 8, with a restart of the telescope on Sept 9 or 10.
Attempts will be made to improve the image quality.  IPAC 2MASS volunteers
will be needed to provide the analysis of the test imaging.  Full OPS should
resume thereafter.  Pipeline modifications (e.g., PSFs) will be necessary.

     The dome is fixed at the southern facility.  The southern pipeline should
begin operations this Friday.  W. Wheaton is within several days of having all 
production PSFs ready for the southern system.  Preliminary cross-scan 
photometric bias corrections will be derived and applied using code written by 
G. Kopan and S. Wheelock.

     J. Fowler inquired about GETCAN for southern OPS.  Before the main DARKS 
program runs, GETCAN is executed. It gets the obs date on the command line, and
with that information, it searches through the darkhist directory for the most 
recent (relative to the obs date) five sets of acceptable responsivity images. 
The search is limited to the "hardware period" containing the obs date, since 
we cannot use hardware models in other periods of time during which the hardware
behaved differently (e.g., prior to a dewar thermal cycle). There's a file in 
the darkhist directory named maskdat that contains a list of hardware-period 
subdirectory names and the obs dates bracketing the period; GETCAN reads that 
file to find out what period the current obs date is in. The information it 
gathers is passed to DARKS, where it is used to get the right canonical masks, 
darks, and responses.

     Since GETCAN knows how to find hardware-period-dependent data, it was
suggested that it could be adapted to perform the same role for PROPHOT which 
may need to get different selections of production PSFs after hardware 
modifications, such as recollimation.  

     R. Beck is running the first batch of darks to get the canonicals.

     R. Cutri decreed, no deliveries before starting southern OPS.
2.) Last Week's Action Items

     T. Evans was quantifying the scope of the MAPCOR error.  She and R. Tam
have been working mainly with the MAPCOR individual-band (not bandmerged)
output.  R. Tam ran CATCOR to match all sources in the old and new runs of 
971116n within 2 arcsec.  T. Evans then looked at all of the non-matched
sources and found that all of them were either faint (m > 16) or were 
diffraction-spike sources, so the old and new runs matched up very well.  They
found that all of the changes in purge flags between the runs are as follows:
1) 53 H sources and 75 K_s sources are "new glints", i.e., marked as glints in 
the new run, but not the old.  These are the sources that were incorrectly not 
marked because of the namelist problem.  R. Tam found that only 3 of these H 
and K_s sources belong to the same bandmerged source.  The results exactly 
follow what is expected from the incorrect parameters.  2) 19 H sources and 1 
K_s source are "new not-glints" -- marked as glints in the old run, but not the
new.  These are the sourcs that were incorrectly marked because of the namelist
problem -- the reverse of 1) above.  These results also exactly follow what is
expected from the incorrect parameters, and do not indicate any problems in the
reliability of the correct parameters.  3) All other purge flag differences 
between the old and new runs -- 2 J sources, 17 H sources, and 1 K_s source -- 
are all attributable to a) changes in mag making a source toggle between  
diffraction flag = 5 and 6; b) the mess inside bright stars, causing small 
changes in extraction parameters; or, c) in the case of 13 H sources in scan 
102, being artifacts of a bright source that is badly saturated in R1 and not 
extracted properly in any band (except, perhaps K_s; thus, its artifacts are not
properly marked in either the old or new runs).  More information will be 
provided in a webpage to be constructed by T. Evans.

     R. Cutri is maintaining a webpage listing all known anomalies, for which
good bookkeeping is necessary.

     Another action item from last week was the "homework assignment" to test
CatScan and the Survey Visualizer.  J. Fowler's critique of the interface was
directed at its complexity.  H. McCallon commented that legends are needed on
the Visualizer.  B. Nelson commented on the problem of dynamic range in the
Visualizer.  T. Jarrett has suggested a sigma stretch to improve the display.

     Those who have not tried it were directed to do so.  The homework has been
extended one more week.  R. Cutri was not pleased with the delinquents.

     Finally, G. Kopan and S. Wheelock provided an update on what could be 
called DMAGCOR, the routines to correct for the crossscan photometric bias.
The preliminary version of DMAGCOR has been delivered.  This program computes 
the psf-aperture photometric offset vs. cross-scan position.  The program bins 
vs. cross-scan the magnitude offsets for all sources between 9 and 14.5, 8.5 and
14.0, and 8 and 13.5 mags for J, H, & K_s, respectively, which are not confused 
or artifacts.  The mid-average is computed for each bin, as well as the 
estimated error and the number of counts.  In the initial version, all 
calibration and survey scan blocks are processed as units, to increase the 
source counts per bin.  Initial results for 980328s show four calibration blocks
with many bins having less than 20 counts, and corresponding errors of 0.03 mags
or more in the estimated offsets.  Several features in the plots of corrections
vs. time are not well understood at this time.  The offsets should be 
well-determined for survey scans, but high-galactic latitude calibration scans 
are at risk, due to low source densities.  How to diagnose and correct poorly 
determined magnitude offsets is an unsolved problem at this point in time.

     R. Cutri would like to run more southern nights with this correction to
amass statistics, and then reexamine the problem.  S. Wheelock has not yet
checked output results from the calibration correction routine; the results need
a spot check.  The routine runs prior to CALMON in the pipeline, correcting the
.bfpts file.  R. Cutri would not be surprised if the "reset" button does not
need to be hit.  The extreme scenario might be to run the calibration using only
aperture photometry as input.  R. Cutri asked S. Wheelock to expect to have to
test this.

     H. McCallon is about to update the positional uncertainties in the .cal
files for the Sampler night.  He will create new ".calo" files.  The impact
will be to have to reload the night into the database.

     Corrections of this sort for other nights will need to be applied to the
database, which will involve consultation with the archive folks.  Or, a night
will have to be reloaded, with correction then applied to the night, and the
night reloaded into the database.

     R. Cutri asked H. McCallon to draft an implementation plan for deriving and
applying position error corrections using the working databases by the first WG
meeting in October.

3.) 2MASS Sampler

     R. Cutri reported that, from a telecon with UMass the other day, a 
suggestion was made by T. Chester to NOT remove duplicate sources in the scan 
overlap regions as part of the 2MASS Sampler data sets.  The schedule is to make
the Sampler stable by Sept 7, let it ferment in the barrel for a month, with an
intended new release date of Oct 7.

4.) S. Schneider Visiting

     T. Jarrett announced that S. Schneider is visiting IPAC from UMass to work
with T. Jarrett on the "Great Attractor" scans.  There is a Coma-sized cluster
in the direction of the Great Attractor.  2MASS will provide enormous insight
into the nature of the Great Attractor.