Spectrographs on large telescopes are frequently equipped with two-dimensional detectors, able to form independent spectra in a number of contiguous spatial columns along the slit. If the projected column width is sufficiently small that a typical seeing disk illuminates several adjacent columns, a signal-to-noise ratio enhancement for observations of a stellar object is obtained by using appropriate weights in the combination of these individual column spectra, instead of using simple coadding. The enhancement is most significant for faint objects, i.e., those for which statistical noise from the underlying sky (and detector noise if applicable) is comparable to the VN noise due to the object itself. The calculation and use of such column weights is described, covering both photoncounting detectors and CCDs.The combination of spectra from separate successive runs also benefits from the use of weights, since the quality of some runs is often better than others. A scheme for calculation of run weights is given.To illustrate the use of column and run weights, a specific spectroscopic observation of a QSO is discussed. The general criteria for the gains to be expected from use of weighting are given. The gain in signal-to-noise ratio is not dramatic (typically ~ 10% for a faint object) but this would be equivalent to an extra 20% observing time and is obtained merely by following the reduction procedure as outlined. The gain due to weighting is substantially larger if the unweighted reduction includes all columns with detectable signal.Methods for analysis of absorption-line spectra to give wavelength, equivalent width, broadening, and asymmetry of detected absorption lines are given. The wavelengths of lines are derived from the median rather than the centroid because it is less affected by noise and by wings or continuum fluctuations near the edges of the line. These procedures for the extraction of absorption-line parameters can be applied whether or not the weighting scheme given here has been used to produce the spectrum.

A catalogue of 160 extragalactic radio sources stronger than 10 f.u. at 408 MHz has been compiled by selecting sources from the Parkes and 3CR surveys but adopting the most accurate values of flux density that are available. The sky coverage of 10·1 sr omits only the galactic plane and Magellanic Cloud regions. The flux density scale due to Wyllie has been used throughout.

Aust, J. Phys., 1977,30, 241-9 Number-flux density counts of radio sources are given for the first two deep surveys made with the Molonglo Mills Cross at 408 MHz, extending to lower limits of 84 and 88 mJy. Practical techniques are developed for the calculation of corrections to the counts when confusion errors are significant. The resulting corrections due to noise, confusion and other effects are given. Counts are also given for the MC2 and MC3 catalogues in the range 2-10 Jy to reduce the statistical uncertainties there. Results are shown on a composite plot, using relevant Molonglo surveys and an all-sky catalogue of strong sources. The existence of a convergence in the number of weak sources is confirmed.

No abstract

No abstract