In 1998, two teams of astronomers, working independently, presented evidence that the expansion of the universe is accelerating (Riess et al. 1998, Perlmutter et al. 1999). This evidence was based primarily on the faintness (by about ~ 0.25 mag) of distant Type Ia supernovae (SNe Ia), compared to their expected brightness in a universe decelerating under its own gravity.
The first suggestions that high-redshift Type Ia supernovae could be used to determine the rate of cosmic deceleration came in the late 1970s (Wagoner 1977, Colgate 1979, Tammann(1979)), but the actual discovery of an even moderate redshift (z = 0.31) SN Ia (SN 1988U) had to wait for a decade (Nørgaard-Nielsen et al. 1989). The samples of high-redshift supernovae started to become sufficiently large to place constraints on cosmological parameters through the efforts of the Supernova Cosmology Project, led by Saul Perlmutter, and the High-z Supernova Search Team, led by Brian Schmidt. By 1998 the two teams gathered sufficient data to be able to show that the universe is characterized by a matter density paremater satisfying M < 1 (i.e., that the universe is not closed by matter; Perlmutter et al. 1998, Garnavich et al. 1998). However, the real shocker was still to come.
8.1. The observations
The two supernova search teams use a similar method to find their supernovae. They take two deep images separated by about a month, subtract the first-epoch image from the second-epoch one, and search for sources above a certain threshold in the difference image. Once a candidate SN is identified, the SN type is determined by its spectrum (if that can be taken; in a few cases one has to rely on the host galaxy type - only SNe Ia were found so far in ellipticals). Supernovae at relatively high redshift are then monitored photometrically to construct their light curves. An image of the host galaxy is obtained at a later time (after a year or more), and subtracted to obtain an accurate measurement of the SN brightness.
The Hubble Space Telescope has proved to be crucial especially for the highest-redshift supernovae. There, the ability to resolve and pinpoint the SN location on the host (including in cases in which the SN was found close to the galactic nucleus) was essential for a correct determination of the SN magnitudes. With samples of a few dozen SNe Ia in hand, the two teams compared their measured distances (derived from the luminosity-distance relation, F = L / 4 DL2; where DL is the distance and L, F are the intrinsic luminosity and observed flux, respectively) with the distances expected for the observed redshifts, for different cosmological models (e.g., Carroll, Press and Turner 1992). The latter is given by
where sinn denotes sinh for M + 1 and sin for M + > 1. The results for the likelihood of the cosmological parameters M and are shown in Fig. (28). As can be seen from the figure, the results favor values of M 0.3, 0.7 and a negative deceleration parameter (a is the scale factor, 0 is the sum of today's energy densities, and w P / characterizes the dark energy "equation of state"; see section VIII D)
corresponding to an accelerating universe.
Figure 28. Joint confidence intervals for (M, ) from SNe Ia. Regions representing specific cosmological scenarios are illustrated.