In pursuit of dark energy : improving measurements in Supernova cosmology

Abstract

The Equation of State SuperNova trace Cosmic Expansion (ESSENCE) was a 6 year survey which discovered and followed 228 type Ia supernova (SN Ia). The primary goal of ESSENCE was to characterize the dark energy equation-of-state, w. However, due to limitations in the knowledge of the SN Ia progenitor system and systematic errors related to influences of the host galaxies on SN Ia distances and the photometric calibrations of the survey, the measurements are only accurate to approximately 10%. Combining the 228 SN Ia light-curves with published data, we used a set of 695 SN Ia to search for progenitor signatures in SN Ia light-curves SN Ia. Kasen 2010 predicted that the companion star in a SN Ia would create a hole in the ejecta, allowing heated emission to escape, producing a strong signature in the early phases of a SN Ia light-curve. Using Monte-Carlo simulations, we examine the 695 SN Ia light-curves, searching for the peak flux in the ultra-violet (UV) and blue filters, expected from a red giant companion. Using both the frequency of these progenitor scenarios and viewing angles of the shock break-out, 4-5 detections are predicted. Our analysis yields zero detections, with a 98.2% confidence. We explain that the SN Ia scenario involving accretion from a red giant onto a white dwarf can not make up a significant fraction of the SN Ia population and better models are needed. Also using the ESSENCE survey, we develop a new method for determining the Spectral Energy Distribution (SED) and rest-frame magnitudes of the host galaxies from the UV to the infrared (IR) and use empirical relations to derive galaxy properties. We also calculate the normalized separation of the SN Ia from the center of the host galaxy. We find a substantial amount of UV emission in our passive galaxies, suggesting star-formation in these galaxies. Additionally we find an approximately 4 sigma correlation of the rest-frame Far UV (FUV) - V host galaxy color and distance residuals from best fitting cosmology (Hubble residual) when the sample is divided by host galaxy type. Lastly, we find SN Ia greater than 2 effective radii occur in low-extinction environments, producing a uniform sample of SN Ia for cosmological measurements. SN Ia at these distances have noticeable offsets in SN color and Hubble residual from the entire sample, suggesting improper SN Ia light-curve fitting, potentially biasing SN Ia distance measurements. Finally, we discuss improvements in the photometric calibration of the SN Ia photometry to reduce the systematic errors in distance measurements. We produced a photometrically flat image with less than 1% illumination gradient across the images. We also detail our automated procedure for correcting the flux of the SN Ia in a given aperture. These two improvements, along with a few other implementations, produce SN Ia photometry with approximately 1% accuracy. We compare our SN Ia photometry and light-curves to that of Wood-Vasey et al. 2007 illustrating the full extent of our improvements.