Two independent international groups, the High-Z Supernova Search Team led
by Brian Schmidt (Mt. Stromlo and Siding Springs Observatories) and the
Supernova Cosmology Project headed by Saul Perlmutter (Lawrence Berkeley
Laboratories), used high-redshift SN Ia discovered with the CTIO Blanco 4-m
telescope over the past three years to measure the matter density of the
universe M and the cosmological constant,
. Hubble diagrams, which plot
supernova luminosity as a function of redshift, generated by both
groups appear to show that
is significantly different from zero, implying
that the expansion rate of the universe is actually increasing with time.
Type Ia supernovae (SN Ia), which are thought to arise from the thermonuclear explosion of a carbon-oxygen white dwarf, are attractive standard candles because of their extreme luminosity (MB ~ -19.5) and relatively homogeneous properties. Work carried out at CTIO by the Calán/Tololo Supernova project (see papers by Hamuy et al. 1996, AJ, 112) has shown that the SN Ia luminosity correlates tightly with a single parameter, the observed decline rate of the light curve. These corrections, along with reddening corrections based on the multicolor light curves (Riess et al. 1996), yield distances with ~ 8% precision.
Figure 1. Images of three high-redshift SN Ia discovered by the High-Z Supernova Search Team in January 1998 with the CTIO Blanco 4-m telescope. The "Epoch1" and "Epoch2" images were obtained with the Blanco telescope in December 1997 and January 1998 with a separation of ~ 20 days. The "Diff" column shows the subtraction of these two epochs, revealing the presence of a SN in all three cases. The final column shows HST images of the SNe obtained only a few days after discovery.
The supernovae are discovered by observing dozens of high galactic latitude fields with the Blanco 4-m just after a new moon. The instrument used is the wide-field BTC CCD mosaic camera developed by Tony Tyson (Lucent Technologies) and Gary Bernstein (Michigan), and each field contains thousands of galaxies at redshifts between 0.3 and 1.0. Just before the next new moon, the same fields are reobserved and searched for supernovae that have become visible between the two epochs. Rapid data analysis makes it possible to identify the supernovae within hours of the observation so that the follow-up photometry and spectroscopy can begin immediately. Figure 1 shows some high redshift supernovae discovered by the High-Z Supernova Search Team. (A previous report the Supernova Cosmology Project appeared in the NOAO Newsletter No. 47)
Followup observations on ARC, WIYN, HST, Keck, and other telescopes must actually be scheduled well in advance of the discovery of the supernovae; thus the project must guarantee objects to observe. This is why both groups have relied on the CTIO 4m Blanco telescope to find all the supernovae. During the Chilean summer months, the CTIO site has excellent seeing, and the nightly skies are ~ 80% photometric and ~ 90% spectroscopic. To date, over 150 supernovae of all types have been discovered with the CTIO 4m Blanco telescope by both groups.
The High-Z Supernova Search Team and the Supernova Cosmology Project
announced their latest results at the January 1998 meeting of the American
Astronomical Society and a symposium on Dark Matter held at UCLA on 18-20
February. The High-Z Supernova Search Team has since submitted a paper to
the Astronomical Journal (Riess et al. 1998). The
expansion of the Universe actually appears to be accelerating at present,
rather than slowing down as would be expected if matter were acting alone.
Stated in more precise terms, a non-negligible positive cosmological
constant is strongly preferred at the ~ 2.5-5 level depending on how the
analysis is done. For a spatially flat Universe (i.e.,
M +
= 1), the
High-Z Supernova Search Team finds that a best fit to their nine
best-observed, high-redshift SN Ia implies
= 0.76 ± 0.11 and
M = 0.24 ±
0.11; the Supernova Cosmology Project derives essentially identical values
from a larger sample of 40 high-redshift SN Ia. Even if the cosmological
constant is forced to be zero, both groups find at the > 5
level that there
is not enough mass in the Universe to eventually halt the expansion -- we would
live in an eternally-expanding Universe.
Figure 2. SN Ia Hubble diagram. The upper panel shows the Hubble diagram for
the low-redshift and high-redshift SN Ia samples with distances measured via
the D m15 method (Hamuy et al. 1996 AJ). Overplotted are three cosmologies:
"low" and "high" M with
= 0 and the best fit for a flat cosmology,
M =
0.28 and
= 0.72. The bottom panel shows the difference between data and
models from the
M = 0.20,
= 0 prediction. The open symbol is SN1997ck (z
= 0.97), which lacks spectroscopic classification and a color measurement.
Testing these conclusions begins with understanding that the high-redshift SN Ia are observed to be 0.2-0.3 magnitudes dimmer than expected, even in the case of an empty Universe. This can be seen in Figure 2 where we plot the observed Hubble diagram compared to various cosmological models. The lower panel clearly shows that the distant SN are too faint. To explain this effect not using cosmology, there would have to be systematic errors of this size in the data . An assumption implicit is that the range in properties of Type Ia supernovae observed at high redshift are the same as those observed at present epochs. From work carried out at CTIO by Hamuy et al. (1996), we know that the local sample of SN Ia displays a weak correlation between peak luminosity and host galaxy type in the sense that the most luminous SN occur in late-type galaxies. In addition, the SN Ia rate per unit luminosity is almost twice as high in late-type galaxies as in early-type galaxies at the present epoch. Both of these observations suggest that a population of progenitors exist in late-type galaxies which is younger and gives rise to brighter SN Ia than those contained in early-type galaxies, which in turn could indicate an evolution of SN Ia luminosities with progenitor age. Recent theoretical work by Hoflich, et al. (1998) also suggests that there may be observable differences in the light curve shape, luminosity, and spectral characteristics of SN Ia as a function of the initial composition and metallicity of the white dwarf progenitor. However, the range of age and metallicity of SN Ia progenitors in the nearby sample is likely to be larger than the average change in these quantities over the 5 Gyr look-back time to the high-redshift SN (to z ~ 0.5). Thus, to first order, there is reason to expect that the relation between light-curve shape and luminosity that holds for the range of stellar populations encountered in the late-type and early-type host galaxies in the local sample of SN Ia should also be applicable to the range of stellar populations encountered in the hosts of the distant SNe.
The low matter density and positive cosmological constant also may help resolve another cosmological problem: the age of the universe. The dynamical age of the Universe based on the high-z supernova data anchored by the Cepheid distances to nearby host galaxies of Type Ia supernovae as published by Saha, Sandage and collaborators is ~ 14 ± 1.5 Gyrs. This is consistent with the age of the oldest stars in our Galaxy. The same data set, which is dominated by the lower-z supernovae from the Calan/Tololo survey, gives a Hubble constant of 65 ± 1.3 (statistical) ± 3.0 (external) km/s/Mpc.
Nicholas B. Suntzeff, Mark M. Phillips