The rationale behind making another Deep Field image was to provide observatories in the southern hemisphere with a similarly deep optical image of the distant universe as had been provided to those in the northern hemisphere. The field chosen was in the constellation of Tucana at a right ascension of and declination of. As with the original Hubble Deep Field, the target area was selected to be far from the plane of the Milky Way's galactic disk, which contains a large amount of obscuring matter, and to contain as few galactic stars as possible. However the field is closer to the galactic plane than the HDF-N, meaning that it contains more galactic stars. It also has a nearby bright star, as well as a moderately strong radio source close by, but in both cases it was decided that these wouldn't compromise follow-up observations. As with the HDF-N, the field lies in Hubble's Continuous Viewing Zone, this time in the south, allowing twice the normal observing time per orbit. At specific times of year, the HST can observe this zone continuously, without it being eclipsed by the Earth. Viewing this field, however, has some issues due to passages through the South Atlantic Anomaly and also with scattered earthshine during daylight hours; the latter can be avoided by using instruments with larger sources of noise, for example from the CCD reading process, at those times. The survey again used Director's Discretionary Time. The field was imaged briefly on October 30–31, 1997 to make sure that the guide stars in the field were acceptable; these guide stars would be required to keep the HST accurately pointing on the region during the observations proper.
Observations
The observing strategy for the HDF-S was similar to that of the HDF-N, with the same optical filters used for the WFPC2 images, and similar total exposure times. The observations were made over 10 days in September and October 1998, a total of 150 orbits, and had a total exposure time of over 1.3 million seconds. While the WFPC2 took very deep optical images, the fields were simultaneously imaged by the Space Telescope Imaging Spectrograph and the Near Infrared Camera and Multi-Object Spectrometer. A number of flanking fields were also observed for shorter periods of time. The WFPC2 image is 5.3 square arcminutes, whilst the NICMOS and STIS images are only 0.7 square arcminutes.
Camera
Filter
Wavelength
Total exposure time
Exposures
WFPC2
F300W
300 nm
140,400 s
106
WFPC2
F450W
450 nm
103,500 s
67
WFPC2
F606W
606 nm
99,300 s
53
WFPC2
F814W
814 nm
113,900 s
57
NICMOS NIC3
F110W
1,100 nm
162,600 s
142
NICMOS NIC3
F160W
1,600 nm
171,200 s
150
NICMOS NIC3
F222M
2,220 mm
105,000 s
102
STIS
50CCD
350–950 nm
155,600 s
67
STIS
F28X50LP
550–960 nm
49,800 s
64
STIS
MIRFUV
150–170 nm
52,100 s
25
STIS
MIRNUV
160–320 nm
22,600 s
12
Spectroscopy
G430M
302.2–356.6 nm
57,100 s
61
Spectroscopy
G140L
115–173 nm
18,500 s
8
Spectroscopy
E230M
227.8–312 nm
151,100 s
69
Spectroscopy
G230L
157–318 nm
18,400 s
12
As with the HDF-N, the images were processed using a technique known as 'drizzling', in which the direction the telescope was aimed was changed by a very small amount between exposures, and the resulting images combined using sophisticated techniques to achieve a higher angular resolution than would otherwise be possible. Translational changes were fine during the imaging parts of the observation; however, the telescope had to be rotated by small amounts instead of repointed during the spectroscopic work, such that the centre of the STIS instrument was kept on the central quasar. The HDF-S final image had a pixel scale of 0.0398 arcseconds.
Contents
The cosmological principle states that at the largest scales, the universe is and isotropic, meaning that it should look the same in any direction. The HDF-S would thus be expected to strongly resemble the HDF-N, and this was indeed the case, with large numbers of galaxies visible displaying a similar range of colours and morphologies to those seen in the HDF-N, and very similar numbers of galaxies in each of the fields. One difference with the HDF-N was that the HDF-S included a known quasar with a redshift value of 2.24, J2233-606, discovered during the search for the target field. The quasar provides a probe of the gas along the line of sight where the foreground objects are also observed, allowing an investigation into the association of galaxies with absorption features. Including a quasar in the field of view was originally considered for the HDF-N, but was decided against due to concerns about increased numbers of galaxies associated with the quasar might skew the galaxy number counts, and because there was not a favourably located quasar. For the Southern field, however, such a skewed count wasn't a concern due to the known counts from the HDF-N.
Scientific results
As with the HDF-N, the HDF-S provided rich pickings for cosmologists. Many studies of the HDF-S confirmed results found from the HDF-N, such as star formation rates over the lifetime of the universe. The HDF-S was also extensively used in studies of how galaxies evolve over time, both due to internal processes and encounters with other galaxies.
