The presence of OH lines in the near-infrared spectrum of the night sky currently limits sensitivities out to 2.2 μm. Attempts to rectify this situation have so far only concentrated on elaborate grating spectrographs, such as those described by Iwamuro et al. (1994, PASJ, 46, 515) and Content and Angel (1994, Proc. SPIE, 2198, 757). This is because appreciable gains using filter-based solutions have so far proven to be elusive (Herbst 1994, PASP, 106, 1298). An alternative solution, described in this paper, explores the use of a multiple bandpass filter (MBF) to selectively remove those wavelengths at which the brightest OH lines occur. A numerical evaluation of MBFs as used in the J and H bands is presented, the results from which demonstrate that significant improvements in signal to noise are only possible from such a filter (and indeed, any method of OH line blocking), if more than the brightest ∼50% of the line population is removed. We show here that relative to broadband J and H, gains in signal to noise greater than unity are difficult to achieve for a continuum source. Over narrower bandpasses, gain factors of around 1.8 (J band) and 1.5 (H band) are possible for the same continuum source and can be 2 to 6 for a source with line emission. We conclude that the ordinary rectangular wave profile of an MBF is most suited to blocking over narrow and intermediate bandwidth filters, as might be used for imaging of emission-line sources or searches for such objects in wide-field surveys. Significant gains in continuum imaging will require more complex filter technologies (e.g., "rugate" coating technology), which are still evolving.
|Number of pages||10|
|Journal||Publications of the Astronomical Society of the Pacific|
|Publication status||Published - Oct 1996|