[Previous] | [Session 27] | [Next]
A. Dobrzycki, H. Jessop, T.J. Calderwood, D.E. Harris (Harvard-Smithsonian Center for Astrophysics)
{\sc celldetect} is a sliding cell source detection algorithm developed at Chandra X-ray Observatory Center (CXC) for use in both the CXC standard data processing pipeline and the Chandra Interactive Analysis of Observations (CIAO) data analysis software package (Fruscione et al.\ 2000, Chandra Newsletter No.~7).
The tool is based on the programs from the PROS DETECT package, which was based on techniques used in the standard processing of Einstein Observatory IPC and HRI data and in the Level 1 processing of ROSAT HRI data (Harnden et al.\ 1984, SAO Special Report 393).
{\sc celldetect} utilizes variable detect cell size approach necessary since Chandra has a large field of view with PSF dramatically changing with the off-axis angle. That, coupled with effects associated with mirror vignetting, results in non-trivial dependence of source SNR detection threshold versus off-axis angle. Previous efforts aimed at establishing this relation for Chandra were hampered by underestimated background rates (Chandra Detect User Guide, {\tt http://asc.harvard.edu/udocs/docs/swdocs/detect/html/})
We have calibrated the SNR versus off-axis angle relation in {\sc celldetect} by running the tool on sets of simulated Chandra data. We find that the SNR threshold dependence on the off-axis angle can be well approximated with a linear function modified at large off-axis angles by a quadratic component. With no vignetting present, the linear component would suffice, since in such a case SNR would grow as square root of number of counts, and the growth of the PSF -- and of the detect cell area -- is close to quadratic. With vignetting present, the relation had to be modified by multiplicative component, which we empirically found to be parabolic. We test the relation on real Chandra data.
The project was supported by NASA Contract No.\ NAS8-39073 (CXC).