4. Photometry and Basic Measurements
For consistency with the 2MASS XSC and other extragalactic catalogs, the Large Galaxy Atlas includes the same basic galaxy measurements of size, shape, orientation, and photometry. We employ the same algorithms used to produce the XSC, but modified to operate on very large images, some of which are degrees in angular size. Stars are first removed from the image to allow uncontaminated measurements, but we try not to remove sources associated with the target galaxy itself; e.g., unresolved HII regions or globular clusters. This procedure requires visual inspection of the final 3-band combined color mosaics to verify proper star removal, and is part of a larger effort to verify the quality of the mosaic images.
The position of the galaxy is measured from the intensity-weighted centroid of the combined ( J+H+Ks) "super" image. This combined image has the highest signal-to-noise for normal conditions. The absolute astrometry is typically better than 0.3 to 0.5 arcsec (Cutri et al. 2000). Nearly all galaxies have a well-defined nucleus as seen in the NIR, with the exception being low surface brightness late-type galaxies (Sdm and dwarf types). For these extreme cases, we either use the nearest bright source to the nominal optical center, or the centroid measured from a symmetric model of the galaxy (e.g., IC 10).
We use the size and orientation from the RC3 (or UGC or 2MASS XSC as needed) as a first guess at the size and shape of the galaxy, thereafter iteratively solving for the shape/orientation until convergence. To be consistent with the XSC, we solve for the elliptical orientation/shape using the 3-sigma intensity isophote, corresponding to roughly 20.3, 19.4 and 18.8 mag arcsec-2 in the NIR bands (J, H, & Ks, respectively); see Jarrett et al (2000). The elliptical shape parameters, axis ratio (minor to major axis) and position angle, are measured for each band and for the combined "super" image. A reduced CHI2 goodness of fit metric is also reported. These procedures (and photometry, see below) are conducted with the target galaxy "center" given by the intensity-weighted centroid.
The galaxy flux is integrated using 4 different apertures centered on the galaxy: (1) circular, radius = 7 arcsec, (2) elliptical, isophotal, (3) elliptical, Kron, and (4) elliptical, extrapolated "total". Fractional pixels are used for all apertures to avoid deleterious pixelization effects across apertures. The isophotal aperture, r20, corresponds to the 2MASS XSC standard aperture, derived from the Ks-band 20 mag arcsec-2 isophote. The Kron aperture is derived from the intensity-weighted "first moment" (discussed below). The "total" aperture is derived from extrapolation of the radial surface brightness from the standard isophote out to some scale length that covers the deduced extent of the galaxy (discussed further below). The photometric measures are given in mag units (relative to the photometric zero point mag, given in the FITS header) and include their 1-sigma uncertainties, detailed here. The relative calibration of 2MASS is uniformly 2-3% across the sky. The photometric uncertainty includes formal errors owing to the background and target Poisson noise, as well as the estimated uncertainty due to the background removal, large-scale background variations and the fit to the surface brightness profile (relevant to the "total" integrated flux; see below). Although the random components are well modeled (see Jarrett et al 2000), the systematic components (e.g., H-band airglow variations) are not well understood and may induce large errors in the photometry. Verification of the apparent photometry is discussed later.