A

Figure 2.7. Extracted spectra from the survey by McCarthy et al. (1999) discussed above. The location of the emission lines is marked by an arrow.

Figure 2.7. Extracted spectra from the survey by McCarthy et al. (1999) discussed above. The location of the emission lines is marked by an arrow.

Figure 2.8. The spectrum of the night sky. Marked also are the wavelength ranges of traditional broad-band filters, as well as important emission lines of astronomical sources and of the sky.
Figure 2.9. A synthetic spectrum of the night-sky OH emission in the near-IR wavelength range. From Rousselot et al. (2000). The relative intensity of the lines is proportional to the photon flux.

2.3.4 Planning for a survey Regardless of the chosen methodology, emission-line surveys can be either targeted at sources whose redshifts are known to lie within a relatively small interval around a pre-established value, or blind searches for sources whose redshifts are expected to be distributed over an extended range. The difference between these two cases is important, because information on the sources' redshifts often drives or dictates the design of the experimental configuration, which in turn determines the sensitivity and/or the efficiency of the survey. As always, the design of the observations should be done only after the scientific questions to answer have been defined clearly and the trade-offs between often competing options thoroughly analyzed.

Examples of targeted surveys are searches for Lya emission from damped Lya-absorption systems observed in the spectra of distant quasi-stellar objects (QSOs) or searches for galaxies with strong emission lines, such as Lya, [O ii] or Ha, at the redshift of some known cluster or concentration of galaxies. In the first case a frequently chosen strategy is to obtain deep spectra covering the wavelength of the broad damped absorption feature with the spectrograph's slit centered on the QSO. This technique optimizes sensitivity, since it minimizes the background noise and typically fully covers the redshift interval whence the emission is expected to come. However, it covers only the volume of space subtended by the slit aperture, which includes the portion of the trough that causes the absorption observed along the line of sight to the QSO and the adjacent regions along the slit. Volume coverage can be increased by performing slitless spectroscopy, but the cost is a substantial loss in sensitivity (for a given exposure time). Since the redshift of the source is in a relatively small interval around a known value, narrow-band imaging is also an option. Unfortunately, the quasar prevents one from detecting sources that are closer to it than a few times the size of the point-spread function, depending on the apparent magnitude of the quasar.

Examples of blind surveys include the search for sources supposedly distributed over a large, unknown redshift range, such as distant galaxies at high redshift. Narrow-band imaging is often adopted in these cases, although both slit and slitless spectroscopy have been used as well. Imaging has the advantage of the simultaneous detection of a large number of sources, but covers only the limited redshift range allowed by the narrow-band filter. Spectroscopy generally allows one to probe a significantly larger redshift range but rather limited sky area, unless slitless spectroscopy is adopted (this, however, is penalized by rather low sensitivity).

Was this article helpful?

0 0

Post a comment