It is generally accepted that Type Ia supernovae (SNe Ia) are imperfect standard candles. The available data on well-observed distant () SNe Ia display, without corrections, an rms dispersion in luminosity of about 0.35 magnitudes, which is considerably in excess of measurement and distance uncertainties. In extreme cases, nearby SNe Ia are said to differ by more than 2.5 magnitudes (Phillips, 1993). Thus their usefulness as standard candles in establishing the cosmological expansion parameters has been questioned. However, ways have been found to remove much of the dispersion in selected samples by adjusting the luminosity of each supernova in accordance with its measured decay rate (Phillips, 1993, Hamuy et al., 1995, 1996). Standardized in this way, a collection of 29 distant SNe Ia carefully measured by modern photometric methods and reduced in a uniform and consistent way by the Cal n/Tololo supernova survey (Hamuy et al., 1996), when pruned of three exhibiting excessive reddening, displays a dispersion nearly, but not quite, compatible with the measurement and distance uncertainties alone (Tripp, 1997).
In this paper I show that all 29 of these supernovae, regardless of color, can be standardized by introducing an additional adjustment to the luminosity dependent on the measured color of the supernova. The resulting fit of the corrected data displays no intrinsic dispersion in excess of measurement uncertainties and in fact has a confidence level much higher than expected, suggesting that the quoted measurement errors of the individual SNe may have been overestimated or that the errors may be correlated.
© European Southern Observatory (ESO) 1998
Online publication: March 3, 1998