Delay Time Distributions of Type Ia Supernovae From Galaxy and Cosmic Star Formation Histories

\begin{abstract} We present analytical reconstructions of type Ia supernova (SN~Ia) delay time distributions (DTDs) by way of two independent methods: by a Markov chain Monte Carlo best-fit technique comparing the volumetric SN~Ia rate history to today's compendium cosmic star-formation history, and secondly through a maximum likelihood analysis of the star formation rate histories of individual galaxies in the GOODS/CANDELS field, in comparison to their resultant SN~Ia yields. {We adopt a flexible skew-normal DTD model, which could match a wide rage of physically-motivated DTD forms.} We find a family of solutions that are essentially exponential DTDs, similar in shape to the $\beta\approx-1$ power-law DTDs, but with more delayed events ($>1$ Gyr in age) than prompt events ($<1$ Gyr). {Comparing these solutions to delay time measures separately derived from field galaxies and galaxy clusters, we fined the skew-normal solutions can accommodate both without requiring a different DRTD form in different environments}. These model {fits} are generally inconsistent with results from single-degenerate binary populations synthesis models, and are seemingly supportive of double-degenerate progenitors for most SN~Ia events.