Australia Telescope National Facility

Summary of the ``Sub-microJansky Radio Sky'' workshop
Andrew Hopkins, Ron Ekers, Carole Jackson, Lawrence Cram, Anne Green, Dick Manchester, Lister Staveley-Smith and Ray Norris, PASA, 16 (2), in press.

Next Section: Galaxy source density -
Title/Abstract Page: Summary of the ``Sub-microJansky
Previous Section: The Square Kilometre Array
Contents Page: Volume 16, Number 2

Subsections

Potential problems

Primordial HI - Lister Staveley-Smith

**Figure 2:** A schematic history of the universe, showing the approximate redshifts at which various events are believed to occur.
$\begin{figure} \centerline{\psfig{figure=universe.ps,angle=0,height=10cm}}\par\end{figure}$

The schematic diagram shown in Figure 2 gives an outline history of the universe as it is currently understood. In typical Cold Dark Matter (CDM) cosmologies the first objects form some time between $z \approx 10$ and 30. This implies there must be sources of ionizing radiation at early epochs so far not observed directly, and these sources must have been sufficiently abundant to reionise the Universe fully by $z \approx 5$ . The temperature of the primordial HI will be raised above the radiation temperature through adiabatic compression of overdense regions well before reionisation, however, and hence the medium surrounding primordial objects, even before the first galaxies have formed, will show emission in the 21cm line. This emission is expected to be directly observable with the Square Kilometre Array. Apart from allowing us to detect objects at unprecedented redshifts, observations of this emission will provide vital information about the processes responsible for the reionisation. Important questions able to be addressed by HI observations include:

When did the first stars form? At some redshift beyond 10 in a CDM Universe objects will start to form which are sufficiently massive to lead to the production of primordial star clusters.
What is the redshift of reionisation? The process of reionisation also probes the
$10^6 - 10^9\,M_{\odot}$ range, allowing discrimination between competing CDM cosmological models which predict fluctuations on such scales.
Where is the peak in star-formation for more recent galaxies (
$z \approx 1 - 3$ )?
When and how do gaseous disks collapse?
What is the relationship between gaseous disks and damped Ly $\alpha$ systems?
How does the Tully-Fisher relationship evolve with redshift?

The primordial universe and the more recent, or `normal,' universe both offer insights into aspects of these questions. However the telescope specifications for making such observations are highly dependent on the redshift of interest, and are shown in Table 1. In some senses, an instrument designed solely to investigate the very high redshift universe would have specifications easier to meet than one designed to investigate more recent redshifts.

**Table 1:** Telescope requirements.
	Primordial	Normal
	$z \approx 10$	$z\approx 3$
Frequency	130MHz	350-1420MHz
Resolution	1kHz	400Hz
Bandwidth	70MHz	300MHz
Channels/Baseline	10⁵	10⁶
Angular resolution	2'	0 $.\!\!^{\prime\prime}$ 3
Longest baseline	4km	600km
Sensitivity	$1\,\mu$ Jy	$0.5\,\mu$ Jy

Potential problems

There are many anticipated problems deriving from various sources of interference, and this subject has been addressed to a greater or lesser extent elsewhere with various proposed strategies for minimisation of the effects. Just to emphasise the importance of the problem, however, $1\,\mu$ Jy is a 100kW VHF TV transmitter at 6000AU. Confusion is also a major problem. At 130MHz, if the confusion level is $\approx$ 1Jy and the rms is

$\approx 1\,\mu$ Jy, then a continuum suppression of 10⁶ is required. Although the Australia Telescope Comact Array (ATCA) is capable of achieving spectral dynamic ranges of almost 10⁴, these are not easily achievable through ordinary techniques. So this area will need a lot of effort if a low resolution array requires this to be bettered by more than two orders of magnitude. Aspects of the confusion problem also imply limitations for the field of view. A wider field of view implies worse confusion. Assuming 4 hours per field, to map the whole sky in ten years only needs a field of view of $\approx$ 1 degree. This in turn implies a maximum element size of $\approx$ 130m. Another recognised problem is the large number (up to 10¹²) of correlators per beam required. (c.f. 3 x 10⁴ for the Parkes multibeam). There may also be problems obtaining the anticipated resolution of the Square Kilometre Array. Even a resolution of 0

$.\!\!^{\prime\prime}$ 3 may be difficult due to problems with scintillation and the ionosphere. Additionally, the receiver temperature, $T_{\rm sys}$ , needs to be cooler than 50K (excluding background).

Next Section: Galaxy source density -
Title/Abstract Page: Summary of the ``Sub-microJansky
Previous Section: The Square Kilometre Array
Contents Page: Volume 16, Number 2

ASKAP

Public