[Previous] | [Session 81] | [Next]
L.-G. Strolger (University of Michigan/CTIO), R. C. Smith (CTIO), A. Clocchiatti (Universidad Catolica de Chile), M. M. Phillips (Las Campanas Observatory), N. B. Suntzeff (CTIO), NGSS Project Team
We compare the {\em B}-band light curves for 11 normal and 7 peculiar 91T/99aa-like Type Ia supernovae, all found from low-redshift searches. Each was chosen based on two criteria: 1) the SN was {\em spectroscopically} identified as either ``normal'' (i.e., resembling SNe 1990N, 1992A, 1992al) or ``peculiar'' (i.e., resembling SN 1991T, or SN 1999aa) either before or near maximum light; 2) the SN light curve was well-sampled, with direct observations periodically obtained from before maximum light to ~80 days after maximum. Each light curve was stretched in the time axis to match a template {\em B} curve made from a few very well sampled, spectroscopically normal Type Ia's. A method of least-squares was used to match the stretched curves to the template curve from slightly before peak to +15 days past maximum. The ``average deviation'', as defined by \Sigma (m(t)-mT(t))/n, between the stretched curve and the template curve was then calculated for each SN. We find that this statistic provides a better measure of the low-order departures from the template curve than does the standard deviation.
Nearly all of the stretched curves of the normal Type Ia SNe have a small average deviation about the template curve, while the peculiars are all significantly brighter outside of the fitted region and have a larger range in average deviations. K-S tests show a significant difference between the two distributions. We compare this result to simple light curve models proposed by Clocchiatti & Wheeler (1997), Pinto & Eastman (2000), and Mazzali et al. (2000). Further analysis on a larger sample of normal and peculiar Type Ia's may eventually reduce the multi-parametered light curve models to only a few parameters, and thus a better understanding of the mechanism(s) of Type Ia supernovae.