M. Zwaan, M. Meyer, R. Webster (University of Melbourne); L. Staveley-Smith, B. Koribalski (ATNF) and the HIPASS team

In 1997 an international team of astronomers set out to survey the entire southern sky for neutral hydrogen (HI) emission using the Parkes radio telescope equipped with the 21-cm multibeam receiver. This instrument is a very efficient survey tool, allowing us to conduct the first large-scale blind neutral hydrogen survey, both in the intermediate neighbourhood of the Milky Way Galaxy, and out to distances of 170 Megaparsecs. The data taking for this survey, known as HIPASS (HI Parkes All Sky Survey), was finished in 2002, and since then much effort has gone into analysing the large dataset. Here, we concentrate on some of the progress that has been made for the extragalactic component of HIPASS.

The HIPASS data consists of 530 cubes, which each cover eight by eight degrees on the sky, and velocities from -1,200 to 12,700 kilometres per second. Together these cubes comprise almost two thirds of the sky, from declination -90 to +25 degrees. So far, most attention has been directed to the 388 southern cubes south of declination 2 degrees. Our first task after creating the data cubes is to identify extragalactic HI 21-cm emission line signals in the noisy background. It was found that this was most efficiently done using two independent automatic finder scripts, the results of which were merged. Galactic high-velocity clouds and detections at frequencies known to be polluted by man-made interference were removed from the list of detections. The remaining 130,000 potential detections were subjected to a series of manual checks for verification by three different individuals. After this process, we were left with approximately 4,300 HI selected galaxies, almost a twenty times larger sample than any existing HI selected sample.

Figure 1 shows the sky distribution of the HIPASS sources. The dashed line at b=0 indicates the plane of the Galaxy. All optical redshift surveys suffer from severe extinction close to this line, which hinders the identification of extragalactic sources over a large region of sky, and hence leads to an incomplete picture of the large-scale structure of galaxies. The figure clearly shows that HIPASS is not affected by extinction and measures the galaxy distribution equally efficiently in all regions of the sky. This illustrates the power of HIPASS to study the large-scale structure in the local Universe.

A more quantitative analysis of the large-scale structure is possible with a tool known as the two-point correlation function. This measures the excess probability of finding another galaxy at a given distance from any galaxy in the sample. For optically selected galaxies, this function has been studied extensively and has been shown to be dependent on galaxy luminosity, morphological type, and star formation activity. Figure 2 shows a two-dimensional form of the correlation function of HIPASS galaxies. The vertical and horizontal axes correspond to separations between galaxies orthogonal and parallel to the line of sight, respectively. Integrating along vertical lines, and applying some mathematics, finally leads to a measurement of the real space correlation function, which is shown in Figure 3. For comparison, the correlation function from the 2dF galaxy redshift survey, at the Anglo-Australian Telescope (AAT), of optically selected galaxies is also shown. This diagram shows that our sample of HI-selected galaxies, representing a population of more slowly evolving galaxies, is less strongly clustered than the optically selected sample. Possibly, two effects are at play here. These gas-rich galaxies might only survive in regions without too many interactions with neighbouring galaxies, which could trigger star formation and burn up the HI. Another explanation is that the HI-rich galaxies are actually formed in lower density regions.

Apart from studying the large-scale structure, HIPASS is useful for measuring exactly how HI is distributed over galaxies of different masses. An analysis of one of the first products of HIPASS, the Bright Galaxy Catalogue containing the 1,000 brightest detections, resulted in the most accurate measurement of the HI mass function to date. This new HI mass function is in good agreement with earlier, much poorer determinations, and allows for a precise evaluation of the neutral gas mass density at the present epoch. Expressed as a fraction of the critical density of the Universe, the neutral gas mass density is only 0.038%. This is approximately five times lower than at the time when the Universe was only 10% its present age, indicating the gradual conversion from neutral gas to stars in the disks of galaxies.

back to highlights page