The ``HIPASS Bright Galaxy Catalog''

B. Koribalski (ATNF) + HIPASS/ZOA Teams

(updated: 3 - August - 2001)

Procedure and selection criteria:

The following steps were carried out to obtain the final catalog.

1. Virginia Kilborn's program gfinder was run on 388 luthered and edge-masked HIPASS cubes, covering the whole southern sky, over all 1024 channels (i.e. -1281 to 12741 km/s), with an HI peak flux cutoff of 60 mJy/beam.
   (March 2000: newg_finder.pl -i cube -c 0.06 -b 5 -d 0 ).
   This procedure generated a list of 245230 source candidates.

   Note that the finder is optimised to find point sources and rejects detections which appear in one channel only. The large number of candidates is mostly due to artifacts generated in areas of strong continuum flux, mostly the Galactic Plane, (causing a baseline ripple), the HI emission in our Galaxy, and lots of interference (1408 MHz birdie, GPS etc.). The candidate list also contains a large number of hydrogen recombination line detections. Many galaxies, especially those with double or multi-horn profiles, also have several entries.

2. From that list a sample with the following parameters was extracted (using BK's fortran program gcutoff2) :
   
   • HI peak flux (*): > 80 mJy/beam
     (*) The peak flux criterion was applied in numerous steps (see Table)
   • systemic velocity: 350 - 8000 km/s
   • velocity width: 30 - 3000 km/s
   • Galactic latitude: > 3 degr

3. Multiple entries were removed (diff(pos) < 6 arcmin and diff(vsys) < 200 km/s; using VK's perl program compare.pl ).

4. For each candidate a cube number and velocity range, both for the HI emission analysis and the baseline fit, was calculated based on the systemic velocity and velocity width given by the finder (BK's program cubeno). All candidate spectra were displayed with the miriad program mbspect and classified as "good", "bad" or "check". The latter positions were inspected in the original cubes and assigned a "good" or "bad" or other appropriate label (e.g. "HVC", "LArm") according to their apparent nature.

5. All sources in the remaining sample were inspected with the miriad program mbspect, based on the gfinder parameter output (RA, DEC, vsys, w50), and non-galaxy entries (recombination lines, interference, etc.) were removed. At the same time the start and end velocities for fitting the source HI spectrum as well as the range of continuum channels to fit a baseline were adjusted.

6. Optical identifications were obtained from NED

   Catalog name	HIPASS HI peak flux	Number of candidates	-multiples	Number of galaxies: all # new + HVCs
   HIP	> 120	2191	1130	618 # 26 + 5
   ADD	100 - 120	1375	896	258 # 25 + 10
   ADE	90 - 100	1342	1060	175 # 10 + 10
   ADF	80 - 90	2478	1980	177 # 23 + 3
   ADG	70 - 80	5094	4091	not done yet
   ADH	60 - 70	9316	7713	not done yet

7. The following sub-samples were searched in addition:

   Catalog name	Selection(*)	Number of candidates	-multiples	Number of galaxies all # new + HVCs
   NAR / HVC	velocity width 10-30 km/s	691	613	20 # 4 + 19
   NEG / HVC	systemic velocity <-350 km/s	307	248	---- + 9

   * only one selection criterion was changed each time, the rest remained as in step (2)

8. The following galaxies were added to this sample:
   • all known galaxies with systemic velocity below 350 km/s (see NED) and HI peak flux above 80 mJy/beam
   • all ZOA shallow survey galaxies (HIZSS; see Henning et al. 2000) with HI peak flux above 80 mJy/beam; there is an overlap region, |b| = 3-5 degr latitude, to check completness.
   • any galaxies missed; e.g. extended galaxies might have been missed by the automatic finder which is optimised to find point sources (none so far)

   Catalog name	Selection	Number of galaxies	HI peak flux > 80 mJy/beam
   NED	systemic velocity <350 km/s	85	24?
   HIZSS	latitude |b| < 5 degr int. HI flux > 4 Jy km/s	110	51? (31 new)

9. Using mbspect preliminary HI parameters were obtained for all objects using the gfinder positions and assuming they are point sources.

10. Accurate galaxy center positions and an estimate of the source diameter were then obtained by using the miriad program imfit on moment-0 maps produced using the recorded start and end velocity information of each source in the sample (by Stuart Ryder; currently complete for HIP, ADD, ADE galaxies).

11. Divide sample into point sources and extended sources (not done yet).

12. mbspect needs to be run again on all extended sources.