Survey requirements for accurate and precise photometric redshifts for Type Ia supernovae

In this paper we advance the simple analytic photometric redshift estimator for Type Ia supernovae (SNe Ia) proposed by Wang, and use it to study simulated SN Ia data. We find that better than 0.5 per cent accuracy in z_phot {with σ[(z_phot - z_spec)/(1 + z_spec)] < 0.005} is possible for SNe Ia with well-sampled light curves in three observed passbands (riz) with a signal-to-noise ratio of 25 at peak brightness, if the extinction by dust is negligible. The corresponding bias in z_phot (the mean of z_phot - z_spec) is 5.4 × 10 ^-4. If dust extinction is taken into consideration in the riz observer-frame light curves, the accuracy in z_phot deteriorates to 4.4 per cent, with a bias in z_phot of 8.0 × 10^-3. Adding the g-band light curve improves the accuracy in z_phot to 2.5 per cent, and reduces the bias in z_phot to - 1.5 × 10 ^-3. Our results have significant implications for the design of future photometric surveys of SNe Ia from both ground and space telescopes. Accurate and precise photometric redshifts boost the cosmological utility of such surveys.