Impact of subsolar metallicities on photometric redshifts

With the advent of deep photometric surveys the use of photometric redshifts, obtained with a variety of techniques, has become more and more widespread. Giving access to galaxies with a wide range of luminosities out to high redshifts, these surveys include many faint galaxies with significantly subsolar metallicities. We use our chemically consistent (CC) galaxy evolutionary synthesis code GALEV to produce a large grid of template spectral energy distributions (SEDs) for galaxies of spectral types E and Sa through Sd - one accounting in a CC way for the increasing initial metallicities of successive stellar generations, the other one for exclusively solar metallicities - for comparison. We use our new photometric redshift code GAZELLE based on the comparison of observed and model SEDs. Comparing the photometric redshifts obtained using solar-metallicity templates when working on a catalogue of artificially created CC SEDs, typical for low-metallicity local late-type galaxies and for intrinsically low-luminosity, and hence low-metallicity, galaxies in the high-redshift universe, we find a significant bias resulting from this metallicity mismatch. This bias consists of a systematic underestimate of the photometric redshift by typically Delta z approximate to 0.1...0.2 until z approximate to 1.2, depending on galaxy type, of distant, faint and low-metallicity galaxies if analysed with solar-metallicity templates.