MULTINARROWBAND IMAGING - A NEW TECHNIQUE FOR MULTIOBJECT SPECTROPHOTOMETRY

A technique is described which consists of repeated imaging of a field through a series of narrowband filters. Photometry of each image then provides low-resolution spectrophotometry of all detected objects. For faint objects this method is comparable in efficiency to slitless spectroscopy, and can cover a larger wavelength range. The principal advantages of the method are simplicity, accuracy in crowded fields, accurate sky-subtraction, and the ability to measure many objects simultaneously. In order to investigate the potential of determining galaxy morphological type and redshift from multiband data, computer simulations are performed for a system of 40 bands having central wavelengths equally spaced logarithmically from 400 to 1000 nm. Spectral energy distributions of various galaxy types are redshifted, multiplied by the filter transmission functions, and then degraded by random noise. These simulated data are then compared with model spectral energy distributions, sampled with the filters, for a range of Hubble type and redshift, by means of a two-parameter, least-squares analysis. For a signal-to-noise ratio of 10, the rms redshift error is less than 0.02 for all galaxy types, and the rms error in the morphological type code is less than 0.14. The mean values of these errors are 0.009 and 0.11, respectively. For a signal-to-noise ratio of 3, the corresponding mean errors are 0.06 and 0.6 for galaxies with redshift z < 0.5, and about half these values for galaxies with 0.5 < z < 1.0.