The Delay Times of Type Ia Supernova

The delay time distribution of Type Ia supernovae (the time-dependent rate of supernovae resulting from a burst of star formation) has been measured using different techniques and in different environments. Here, we study in detail the distribution for field galaxies, using the SDSS DR7 Stripe 82 supernova sample. We improve a technique we introduced earlier, which is based on galaxy color and luminosity, and is insensitive to details of the star formation history, to include the normalization. Assuming a power-law dependence of the supernova rate with time, ${\rm DTD}(t)=A(t/{\rm 1\,Gyr})^{s}$, we find a power-law index $s= -1.34 ^{+0.19} _{-0.17}$ and a normalization $\rm{log}\ A= -12.15 ^{+0.10} _{-0.13}\, {\rm dex}(M_\odot ^{-1}\, \rm{yr}^{-1})$, corresponding to a number of type Ia supernovae integrated over a Hubble time of $PE = 0.004^{+0.002} _{-0.001}\, M_\odot ^{-1}$. We also implement a method used by Maoz and collaborators, which is based on star formation history reconstruction, and find that this gives a consistent result for the slope, but a lower, marginally inconsistent normalization. With our normalization, the distribution for field galaxies is made consistent with that derived for cluster galaxies. Comparing the inferred distribution with predictions from different evolutionary scenarios for type Ia supernovae, we find that our results are intermediate between the various predictions and do not yet constraint the evolutionary path leading to SNe~Ia.