Constraints on dark energy with the LOSS SN Ia sample

We present a cosmological analysis of the Lick Observatory Supernova Search (LOSS) Type Ia supernova (SN Ia) photometry sample introduced by Ganeshalingam et al. (2010). These SNe provide an effective anchor point to estimate cosmological parameters when combined with datasets at higher redshift. The data presented by Ganeshalingam et al. (2010) have been rereduced in the natural system of the KAIT and Nickel telescopes to minimise systematic uncertainties. We have run the light-curve-fitting software SALT2 on our natural-system light curves to measure light-curve parameters for LOSS light curves and available SN Ia datasets in the literature. We present a Hubble diagram of 586 SNe in the redshift range z=0.01-1.4 with a residual scatter of 0.176 mag. Of the 226 low-z objects in our sample, 91 objects are from LOSS, including 45 SNe without previously published distances. Assuming a flat Universe, we find that the best fit for the dark energy equation-of-state parameter w = -0.86^+0.13_-0.16 (stat) +- 0.11 (sys) from SNe alone, consistent with a cosmological constant. Our data prefer a Universe with an accelerating rate of expansion with 99.999% confidence. When looking at Hubble residuals as a function of host-galaxy morphology, we do not see evidence for a significant trend, although we find a somewhat reduced scatter in Hubble residuals from SNe residing within a projected distance < 10 kpc of the host-galaxy nucleus (\sigma = 0.156 mag). We find that Hubble residuals do not correlate with the expansion velocity of Si II \lambda 6355 measured in optical spectra near maximum light. Our data are consistent with no presence of a local "Hubble bubble." Improvements in cosmological analyses within low-z samples can be achieved by better constraining calibration uncertainties in the zero points of photometric systems.