[63.24] Improved Crowded-Field Photometry of GOODS IRAC Imaging with the TFIT Package

AAS 207th Meeting, 8-12 January 2006
Session 63 From Here to Eternity: The Spitzer Legacy Programs
Poster, Tuesday, 9:20am-6:30pm, January 10, 2006, Exhibit Hall

Previous | Session 63 | Next | Author Index | Block Schedule

[63.24] Improved Crowded-Field Photometry of GOODS IRAC Imaging with the TFIT Package

N. A. Grogin (JHU), V. G. Laidler (CSC/STScI), H. C. Ferguson (STScI), K. Clubb (U. of Iowa), C. J. Papovich (U. of Arizona), M. E. Dickinson (NOAO), R. Idzi (JHU), E. C. MacDonald (NOAO), M. Ouchi, B. Mobasher (STScI)

The density of sources with z₈₅₀ < 27AB detected throughout the Great Observatories Origins Deep Survey (GOODS) HST/ACS fields is approximately 200 per square arcminute. A large fraction of these ~50000 sources, which are predominantly distant galaxies, has significant mid-IR flux at the extreme sensitivities of the GOODS SST/IRAC observations (1\sigma = 22nJy at 3.6um). Measuring robust mid-IR spectral energy distributions (SEDs) for all these distant galaxies is highly desirable, but also highly challenging because of formidable source crowding at the 2 arcsecond IRAC resolution. At these source densities, reasonably-sized aperture photometry in IRAC is frequently biased by neighbor contamination, and catalog-matching from IRAC-detected source extraction is additionally compromised by the inability to de-blend neighbors. On the other hand, crowded-field, PSF-fitting photometry packages such as DAOphot are unsuitable because many of the extragalactic sources are at least marginally resolved with IRAC.

To address these problems, the GOODS team has developed a crowded-field, resolved-source photometry package called TFIT, which will soon be released to the public through the Spitzer Science Center. In brief, TFIT first constructs lower resolution (e.g., SST/IRAC) template images for sources detected at higher resolution (e.g., HST/ACS) by using the high-resolution position and morphology along with PSF information about both images. These templates are then jointly scaled to optimally match the lower-resolution image. We summarize the inner workings of the TFIT software package, quantify its superiority to SExtractor photometry as a function of source crowding in simulated GOODS-depth imagery, and present preliminary results from our application of TFIT to obtain robust IRAC SEDs of GOODS sources detected from higher resolution imaging at shorter wavelengths.

Previous | Session 63 | Next