R.F. Minchin, PASA, 16 (1), in press.
| Next Section: PICASSO - an automated Title/Abstract Page: Finding the Bivariate Brightness Previous Section: Determining the sample | Contents Page: Volume 16, Number 1 |
Determining the BBD
Once the sample has been selected and the HIcharacteristics determined, it will be necessary to find the optical characteristics of the galaxies in the data set. Optical data will be taken from the ESO-LV catalogue and from existing photometric measurements. Where there is no existing data in the literature, the galaxies will be observed in B and R as part of the HIPASSoptical follow-up programme. The two main parameters to be determined from the optical follow-up are the luminosity and the surface brightness. Inclination and scale-length will also be determined The luminosity and surface brightness will be used to place the galaxy onto the BBD. Here each galaxy will be given a weighting based on the volume density of galaxies of similar HImass, from the HIPASSHIMF (Kilborn, Staveley-Smith & Webster, 1999), and the number of galaxies in that mass bin. These corrections will give a number per unit volume for galaxies similar to the one being analysed.
Once this analysis has been completed, the pixels on the luminosity - surface brightness plane will all have a total number of galaxies per unit volume associated with them. This will give the BBD for the HIselected sample of galaxies. It should then be possible to see if LSB galaxies do make a significant contribution to the numbers in the Universe, and whether this contribution is limited to low-luminosities or spread across the whole range of galaxy sizes.
There are a number of biases that could affect the selection of the sample. The most important of these is that low column-density galaxies will not be found in HIPASS, although if the Disney-Banks scaling is correct then HIPASSwill reach a low enough column density to determine the BBD to a very low surface brightness. However there is the added possibility that the hydrogen in these galaxies will be ionised (eg Corbelli & Salpeter, 1993; Charlton, Salpeter & Linder, 1994; van Gorkom, 1993), which would prevent them being found in HIPASS. It is also likely that galaxies with very high line-widths will be missed as their flux will be spread out over a wide velocity range. Another problem is that LSBG's have different rotation curves to HSBG's generally (de Blok, McGaugh & van der Hulst, 1996), resulting in these galaxies having lower peak fluxes for the same total flux. This problem should be overcome by the automatic galaxy finder, but it does present a difficulty to detections made using the human eye.
Another problem which can occur is that very large galaxies that overfill the Parkes beam (FWHM 15
) may be removed during the on-line bandpass removal or may be measured as sky by the galaxy finder; in either case such galaxies will not be catalogued. As LSBG's are expected to have a longer scale-length for the same luminosity when compared to HSBG's, they may also have larger HIenvelopes. This effect could possibly introduce a bias against finding LSBG's. However, the number of galaxies likely to be affected by the bandpass filter is small, as this will only affect structures greater than 2
across (Barnes et al., 1998) and sky subtraction is much less of a problem for the eye. This has allowed a check to be made on the cube to see if galaxies are being missed due to overfilling the beam. It appears from this that such galaxies do not exist in significant numbers in the velocity range being examined by this project.
| Next Section: PICASSO - an automated Title/Abstract Page: Finding the Bivariate Brightness Previous Section: Determining the sample | Contents Page: Volume 16, Number 1 |
© Copyright Astronomical Society of Australia 1997