Previous Article
Next Article
Table of Contents

A Spectroscopic Indicator of Type Ia Supernova...(1Dec93)
Luminosities
(from NOAO Highlights!, NOAO Newsletter No. 36, 1 December 1993)

Due to their extreme luminosities at maximum light, type Ia supernovae
(SNe Ia) have long been considered among the most attractive
cosmological standard candles. Although nearly all work to date has
been devoted to attempts to use these objects to determine the local
rate of expansion of the universe (Ho), SNe Ia also provide one of the
few direct techniques for measuring the deceleration parameter qo.
However, in a recent study of nine well-observed events based largely
on data obtained at CTIO, M. Phillips (CTIO) found clear evidence for a
significant intrinsic dispersion in SNe Ia absolute magnitudes
amounting to ~0.8 mag in B, ~0.7 mag in V, and ~0.5 mag in I. Such a
range in peak luminosity could introduce a substantial Malmquist bias
into searches for distant (z > = 0.3) SNe Ia which, if uncorrected, could
lead to an erroneous value of qo.

Phillips suggested two possible solutions to this problem. For reasons
not yet clearly understood but perhaps hinting of opacity differences
or a range of ejected masses, the absolute magnitudes of SNe Ia at
maximum light appear to correlate strongly with the initial decline
rate of the B light curve, with the slope of the correlation being
steepest in B and growing progressively flatter in V and I. Hence, by
measuring the decline rate during the first 2-3 weeks following
maximum, it should be possible to associate an appropriate absolute
magnitude with any particular event.  Alternatively, Phillips suggested
that measurements be restricted to the I or near- infrared (JHK) bands
where the slope of the peak luminosity-decline rate correlation is
relatively flat. However, both of these solutions are difficult to
apply to distant SNe Ia.  Measuring an accurate initial decline rate
requires catching the supernova at or before maximum, and then sampling
the light curve frequently during the first 15-20 days following
maximum~a difficult feat for relatively nearby SNe Ia, much less for
faint distant ones.  Observing in the near-infrared will also be a
challenge since the contrast between the supernova and its host galaxy
is much less than it is at blue wavelengths. Since it will be necessary
to obtain spectroscopic confirmation of distant SNe Ia candidates in
order to distinguish them from other types of luminous supernovae, a
more practical solution to the Malmquist bias problem would be to find
an accurate spectroscopic indicator of peak luminosity.

                          [Figure not included]

In his original study, Phillips called attention to certain features of
the optical spectra of SNe Ia which appeared to correlate with decline
rate---and, hence, with absolute magnitude.  Using spectroscopic and
photometric observations obtained in the course of the Calan/Tololo
Supernova Survey, M. Phillips, M. Hamuy (CTIO), N.  Suntzeff (CTIO),
and J. Maza (U. of Chile) have now confirmed that the most promising
of these spectroscopic indicators, the intensity ratio of the Si II
l5979 and l6355 absorption features, does indeed correlate tightly with
initial decline rate and with absolute magnitude. If a spectrum can be
obtained within approximately +-5 days of B maximum, this study
indicates that a measurement of the Si II l5979/l6355 intensity ratio
can predict the absolute magnitude of the event to a precision of < =0.3
mag. Further observations of nearby SNe Ia at well-determined distances
are required to confirm these results, but it appears that SNe Ia may
yet provide a reliable means of determining the deceleration rate of
the expanding universe.

Previous Article
Next Article
Table of Contents