AT Hubble Deep Field South : Project Details

AT Deep Fields Home    |    ATHDFS Home    |    Project Details    |    Data    |    Catalogues    |    Publications    |    ATHDFS Team Pages (password required)  


AT image of the HDFS In 1995 a dark region of Northern sky was observed by the Hubble Space Telescope for 10 consecutive days. This otherwise unremarkable patch of sky in the Big Dipper constellation has since become known as the Hubble Deep Field North. Data was released publicly within weeks and revealed the faint end of the universe to depths of about 30th magnitude. Following the success of the original deep field the Hubble Space Telescope repeated the exercise in 1998, this time in the Southern sky constellation of Tucana and with all three available Hubble Space Telescope instruments. The Hubble Deep Fields are a treasure trove for astronomers studying the history of the universe and evolution of galaxies.

Astronomical sources in the Hubble Deep Field South include normal galaxies, such as our Milky Way, and powerful "monster galaxies" with supermassive black holes in their centre (Active Galactic Nuclei or AGN). These objects give off energy at many wavelengths, such as visible light and radio waves, and it is by comparing images at several different wavelengths that we understand the science.

To this end the Australia Compact Telescope Array observed the Hubble Deep Field South region for more than 5 weeks in total during 1998 - 2001. The ATHDFS survey used four radio bands of the ATCA and reaches about 10 microJy rms. It remains one of the most sensitive radio surveys ever made. The ATHDFS team are using the radio data, coupled with multiwavelength observations, to study the nature of faint radio sources and the evolution of galaxies in the early universe.

Technical Details

The observations consist of single pointing integrations at 1.4, 2.5, 5.2 and 8.7 GHz on the ATCA. An area approximately 60 arcmin in extent around the Hubble Deep Field South has been imaged at 1.4 GHz. There are correspondingly smaller images at 2.5, 5.2 and 8.7 GHz.

A wide variety of ATCA configurations were used to obtain uniform uv coverage. The sixth ATCA antenna was included to obtain the highest resolution possible. The resolution of the final 1.4 GHz image is 6.6 arcsec, while the best resolution of 2.0 arcsec is achieved at 8.7 GHz.

The noise in the ATHDFS images increases as a function of distance from the image centres because of the primary beam corrections. The single-pointing nature of the observations therefore means the area of greatest sensitivity is towards the image centres. The catalogued region has therefore been radially limited. The catalogue limits are 20, 12, 5.5, and 3.5 arcmin at 1.4, 2.5, 5.2 and 8.7 GHz, respectively. See PAPER II for more details.

Summary of ATHDFS Observations
frequency RA Dec Hours Synthesised Beam rms limit full FOV catalogued
(GHz) J2000 J2000 Observed bmaj bmin PA (microJy) (arcmin) sources
1.4 22 33 25.96 -60 38 09.0 345 7.1" 6.2" -5.5° 11.0 34 466
2.5 22 33 25.96 -60 38 09.0 94 5.1" 4.0" 1.7° 10.4 20 84
5.2 22 32 56.22 -60 33 02.7 181 3.0" 2.2" 10.7° 7.8 9.5 24
8.7 22 32 56.22 -60 33 02.7 98 2.3" 1.7" -1.6° 11.0 5.7 6

Scientific Motivation

A variety of projects have been identified by the ATHDFS team. These include:

Optically Faint MicroJansky Sources

Recent deep radio surveys have revealed a population of faint radio sources which are not detected in the faintest optical images. About 20% of faint radio sources (10 microJy
1) luminous dust-enshrouded starbursts at high redshifts of z = 1 - 3,

2) extremely high redshift AGN (z > 6), or

3) luminous obscured AGN at z ≥ 2

Submillimetre observations have shown that at least 30% of these optically faint microJansky sources are dusty starbursts at 1

Our radio survey increases the known number of optically faint microJansky sources. By combining the radio data with available optical imaging and spectroscopy, we will gain further insight into the nature of these sources.

Evolution of Starforming Galaxies

Below a few mJy the radio population starts to be dominated by starforming galaxies. A faint radio sample with near complete redshift information makes it possible to trace the evolution of starforming galaxies from today to the early universe. Photometric and spectroscopic redshifts of our radio sample have been used to trace the evolution of the starforming galaxies to z ≤ 1.

Serendipitous Sources

Just like any blind survey, unique interesting sources maybe found with the data. In the ATHDFS we have identified a few such sources:

  • "Source C" is a radio bright but optically faint galaxy. It is 1-mJy at 1.4 GHz, and has a radio spectral index of about -0.65. It appears to be slightly extended in the 8.7 GHz image, but higher resolution radio imaging is required to see any core-jet structure. Optically it appears to be a faint (V=27.05, I=25.12, J=22.07, H=20.76, K=19.84) red galaxy and it is slightly extended in the WFPC image. Its colour, I - K ≥ 5.0, fits the criterion of Extremely Red Objects (EROs) (see review by McCarthy 2004, ARAA, 42, 477). EROs are thought to be a mix of passively evolving red galaxies at 1 1.

    Depending on its redshift, it could be either an extreme dusty starburst, or an AGN, also dust-enshrouded. Spectra of this source has been impossible to obtain so far because of its faintness. Fernandez-Soto et al. (private communication) have fitted an E/S0 template spectrum to the HST UBVI data and the VLT BVIJHK data (from the EIS survey) and have obtained a redshift of 1.69. At this redshift the radio/optical ratio is roughly 1000 times greater than Arp220, the archetypical local starburst.