IPAC 2MASS Science and Analysis Working Group Meeting #16 Minutes

IPAC 2MASS Science and Analysis Working Group Meeting #16 Minutes, 01/17/96

Attendees: S. Charnoz, T. Chester, T. Evans, T. Jarrett, B. Light C. Lonsdale, S. Terebey, S. Wheelock

Several regular attendees were at the AAS meeting this week.

S. Terebey: has modelled the telescope diffraction pattern to see if it has any measurable effect on the PSF. She finds that the seeing model described in her memo of 12 January gives a good fit to the PSF without any need to invoke the telescope transfer function, therefore she concludes it is of negligable importance to 2MASS image shapes, which are instead dominated by the seeing, the 2 arcsec pixel size, and the relative star-pixel offsets. This results supports the view that we do not need two identical telescopes, since seeing is the more important parameter for the PSF.
Having demonstrated the good fits of her models to the inner 10 arsec or so of bright star images, Sue turned her attention to modelling the outer wings. She achieved a high level of accuracy in fits to Ks=8-9th magnitude stars by doing an angular average with a careful background determination and exclusion of faint stars. A brighter (5.7) star was not well fit by the same model, due to saturation at the center, but the shape of its profile fits when scaled down. She will now study a range of bright stars under various seeing conditions.
T. Evans: has continued to investigate the behavior of R2 vs. R2-R1 magnitudes as a function of pfrac, extending her Ks band analysis of last week to J and H. No dependence on pfrac is apparent in any band, though the scatter within a given scan at a scan-averaged value of pfrac can be quite large and could conceivably mask some effects. The scatter in J was noted to be especially large in the HHE scans; reason currently unknown. It was suggested that it might be due to nebulosity at J in this low latitude field, but Tracey had not noticed any particularly striking nebulosity as she studied these images.
Tracey then showed plots of delta mag versus y (in-scan pixel position) which shows the existence of a shift in delta mag between the ascending and descending scans. The shift is about the same in each band: about 0.02 mag. It is suspected that this is related to the shift seen in pixel positions between R2 and R2-R1, and is due to settling of the secondary mirror affecting the R1 data. Tracey will next look at the behavior of pfrac vs. scan direction and of PSF vs. aperture photometry as a function of pfrac.
T. Jarrett and T. Chester: presented the highlights of the results from their analysis of the effect on galaxy photometry of stellar confusion; details have been released in their memo to S. Schneider and J. Huchra. The main purpose of their work was to determine the merit of a new magnitude derivation requested by Schneider and Huchra. They add artificial stars to real Coma scans, then analyze the accuracy and dispersion in the derived galaxy magnitudes using various methods for handling the stellar contamination.
The current best algorithm of GALWORKS excludes as many stars as possible from galaxy measurements, and subtracts a global background. The new method proposed by Schneider and Huchra involves using an annulus around each galaxy to compute a local background, and either subtracting stars from it in the same way as for the galaxy, or not subtracting stars from either galaxy or annulus. The basic result is that the GALWORKS existing approach produces significantly less scatter in the galaxy photometry than any of the new procedures. A completely 'unbiased' galaxy magnitude that subtracts any stars is not attainable with 2MASS data, because stars on top of galaxies always look different than stars in the annulus and because stars can land on galaxy nuclei where they can never be reliably eliminated. Magnitudes derived without star subtraction have much increased scatter, since one has doubled the number of galaxies affected by stars by the stars in the annulus and pushed those galaxies to large negative values.