Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.eso.org/~bleibund/papers/VLTIII.ps Äàòà èçìåíåíèÿ: Sun Jan 7 21:59:01 2001 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 02:16:07 2012 Êîäèðîâêà:

Supernovae at high redshift ?
Bruno Leibundgut
European Southern Observatory, Karl­Schwarzschild­Strasse 2, D­85748 Garching,
Germany
Abstract. Distant supernovae provide a convenient tool for cosmology. As bright
beacons with supposedly very small variation in their peak luminosity Type Ia Su­
pernovae can be used to determine distances and derive cosmological parameters.
Beyond the Hubble constant, which still requires the knowledge of the absolute
luminosity, the cosmic deceleration can be measured from relative distances. Two
groups have undertaken this experiment with the conclusion that the distances are
too large to be accommodated within matter­dominated world models. The dis­
tances can be explained by additional acceleration due to a cosmological constant.
Supernovae also provide direct evidence of the stellar death rate at signific­
ant look­back times. The evolution of the supernova rate as a function of redshift
can give direct evidence of the star formation history throughout the observable
universe. These studies are emerging as a main challenge for large telescopes.
1 Introduction
The observations of distant supernovae has become common place with the
advent of large­aperture telescopes and large­field imaging cameras. It is now
possible to find supernovae, classify them by a suitable spectrum, and follow
their light curves in several filters out to redshifts of about 1. The near future
will see further advances in redshifts and look­back time to probe even earlier
epochs of star formation. Since supernovae are stellar end­products they can
trace star formation more directly than any other method.
Future observational programs will employ all known supernova types and
use their specific characteristics for cosmological purposes. It is of course the
extreme luminosity of Type Ia Supernovae (hereafter SNe Ia) which opened
the distant universe to supernovae first.
To employ any astronomical object for the derivation of cosmological para­
meters requires sufficient physical understanding of its properties in the nearby
? I report results of the High­z Supernova Search Team. The team includes
B. Schmidt (MSSSO), J. Spyromilio and P. Woudt (ESO), M. Phillips (LCO),
N. Suntzeff, R. Schommer and C. Smith (CTIO), A. Clocchiatti (Universidad
Catolica, Santiago), M. Hamuy (Steward Obs.), R. Kirshner, P. Garnavich, S. Jha
and P. Challis (Harvard Univ.), C. Hogan, C. Stubbs, A. Diercks and D. Reiss
(Univ. Washington), A. Filippenko and A. Riess (Univ. California, Berkeley),
R. Gilliland (STScI), and J. Tonry (Hawaii). More information is available at
http://cfa­www.harvard.edu/cfa/oir/Research/supernova/HighZ.html

2 Bruno Leibundgut
universe. Evolutionary effects must then be considered and, if necessary, cor­
rected. In particular, the use of SNe Ia as standard candles demands a suffi­
cient understanding of the explosions and their physical background to reliably
apply them to cosmological problems.
The understanding of the local (z ! 0:1) supernovae forms the basis for
the cosmological results. There are basically four independent indications that
SNe Ia form a rather homogeneous group of objects. These are the Hubble
diagram (log z vs: m), the absolute luminosity at maximum as determined
from direct distance measurements (mostly Cepheid stars), the uniform light
curves and spectral evolution, and the fact that the peak luminosity of SNe Ia
can be 'corrected' with a distance independent quantity, namely the decline
rate after maximum light. We will discuss each of these features in the fol­
lowing.
Several Hubble diagrams of nearby SNe Ia have been published in the past
few years. They all demonstrate the linear expansion to a high degree (e.g.
Tammann & Leibundgut 1990, Hamuy et al. 1996b, Riess et al. 1996, Saha
et al. 1999), although deviations from the linear expansion within a recession
velocity of about 7000 to 10000 km/s have been claimed (Zehavi et al. 1998,
Tammann 1999). The Hubble diagrams show a remarkably small scatter of
the supernovae around the expansion line, which provides ample evidence that
they all seem to reach a very similar absolute luminosity.
Another, more direct, route is the measurement of the absolute luminosities
of SNe Ia. This direct comparison provides a further test of the uniformity
of the standard candle assumption for SNe Ia. Such a program has been
undertaken with the determination of Cepheid distances to now eight nearby
SNe Ia (Saha et al. 1999). It is remarkable that almost all objects reached the
same luminosity to within 20%.
The very uniform appearance of SNe Ia is further evidence for their great
similarity. Although a strong explosion with the complete disruption of a star
gives rise to the supernovae, the temporal evolution of the light emission
follows a rather narrow parameter space. Deviations are observed, but it is
amazing that the chaotic explosion mechanism and the complicated energy
release yield the uniform events we observe.
Several observable parameters of SNe Ia seem to be correlated. The first,
and foremost, such relation was the decline vs. peak luminosity dependence
(Phillips 1993, Hamuy et al. 1996a, Riess et al. 1996, Riess et al. 1998, Phillips
et al. 1999). In the meantime it has been found that the color (Hamuy et al.
1996b), the expansion velocity of the ejecta (Mazzali et al. 1998), the line
strengths (Nugent et al. 1995), and the galaxy type (Hamuy et al. 1996a,
Riess et al. 1996) also correlate with the peak luminosity.
After a correction with these distance independent observables, the scatter
around the Hubble line becomes very small and is of the order of the photo­
metric uncertainties (Hamuy et al. 1996b, Riess et al. 1996, Phillips et al.
1999, Suntzeff et al. 1999, Jha et al. 1999). This remarkable fact, has boosted

Supernovae at high redshift 3
the use of SNe Ia as distance indicators and seems by now accepted by most
people.
Yet, there are a number of unexplained features of SNe Ia exists and an
accepted theory for the explosion and the radiation transport in the ejecta is
still missing. The very relation between the peak luminosity and the decline
rate is not understood (e.g. H¨oflich et al. 1996). The secondary peak in the
RIJHK light curves (Ford et al. 1993, Suntzeff 1996, Vacca & Leibundgut
1996) is likely due to the complicated release of photons in the expanding
fireball. The bolometric light curves show this shoulder as well (Contardo et
al. 1999). The spectrum is not formed in a regular photosphere, but is the
result of complicated scattering of photons out of the ejecta. The input energy
is coming from the radioactive decays of 56 Ni and its daughter product 56 Co,
which is converted into a pseudo­continuum through down scattering in many
lines (e.g. Pinto & Eastman 1999).
The bolometric luminosity of SNe Ia can differ among individual objects
up to a factor of two (Contardo et al. 1999). Extreme cases, like SN 1991bg
(Filippenko et al. 1992, Leibundgut et al. 1993, Turatto et al. 1996) and
SN 1997n (Turatto et al. 1998), can be several times fainter.
Several independent determinations of Ni masses of SN Ia explosions exist
by now (Spyromilio et al. 1992, Bowers et al. 1997, Cappellaro et al. 1997,
Mazzali et al. 1998, Contardo et al. 1999). A fairly large range of Ni masses
is inferred for the individual explosions. The theoretical understanding of this
range in Ni masses is not very complete, although a number of explosion
mechanisms has been proposed and could be responsible for what we observe
(e.g. H¨oflich & Khokhlov 1996).
We have thus a fairly mottled view of SNe Ia as distance indicators. On
the one hand a strong empirical basis for the use of SNe Ia as 'modified
standard candles' has emerged, while the detailed physics of the events are
largely undetermined.
2 Distant Type Ia Supernovae
Supernovae are rare. Although they can outshine their parent galaxies, they
maintain this luminosity only for a brief time, which makes their detection
at significant redshifts (z ? 0:1) very difficult. Only dedicated wide­field
searches which reach a limiting magnitude of at least R ú 23 can hope to
find a sufficient number of SNe Ia out to redshifts of about 0.5 (Perlmutter
et al. 1995, 1997, 1999, Schmidt et al. 1998). More distant objects become
increasingly difficult to find. Searching several square degrees of sky during
a single night and comparing them with observations obtained several weeks
before can provide about a dozen candidates which then are classified through
spectroscopy (Riess et al. 1998, Perlmutter et al. 1999).
The two teams working in this field are now routinely discovering about
20 objects in a search run (Schmidt et al. 1997, Perlmutter et al. 1998b). A

4 Bruno Leibundgut
total of 53 supernovae with z ? 0:15 have by now been published (NÜrgaard­
Nielsen et al. 1989, Riess et al. 1998, Perlmutter et al. 1999).
The High­z Supernova Search Team (Schmidt et al. 1998) has followed a
strategy of trying to characterize the distant supernovae in as many aspects as
possible. We have obtained light curves in at least two filters to detect any dust
absorption in the host galaxy (Garnavich et al. 1998a, Schmidt et al. 1998,
Riess et al. 1998). All objects have been positively identified as SNe Ia from
spectra, and doubtful cases could be investigated in detail (Riess et al. 1998).
The Supernova Cosmology Project (Perlmutter et al. 1995, 1997, 1998a, 1999)
has amassed a sample of 42 distant supernovae. Many of these supernovae
have a classification from spectroscopy and also two filter observations, but
some are lacking this information.
Both teams developed the search and analysis software independently. The
distant supernova samples are completely independent and there exists only
a small overlap in the nearby supernova set, which is drawn partially from
the Calan­Tololo supernova search (Hamuy et al. 1996a). The High­z team
currently enjoys a larger nearby sample (29 supernovae).
It has to be noted that the cosmology depends on the local sample just as
much as it does on the distant one. Since the measurement of the deceleration
is relative, the nearby sample provides the essential comparison set.
The main result is that the distant supernovae (mean redshift for both
groups is around 0.5) are too dim when compared to the local sample (Fig­
ure 1). It is important to note that this result is based entirely on the ob­
servational data and does not depend on any world model. The brightness
difference to a cosmology
with\Omega M =
0:2;\Omega \Lambda = 0 is plotted in the lower
panel. It is clear that the local sample is insensitive to the exact cosmology
and provides the zero­point for the various world model lines. The distant
supernovae are about 0.25 mag fainter (Riess et al. 1998). The exact same
result is found by the Supernova Cosmology Project (Perlmutter et al. 1999).
Various corrections have been applied in the analysis to arrive at Figure 1.
They include the exact photometry of the supernova on the irregular galaxy
background (achieved through subtraction of a galaxy template image), the
K­correction to convert the magnitudes from the observer system into rest­
frame wavelengths, the (1+z) correction for time dilation (Leibundgut et al.
1996, Goldhaber et al. 1997), the light curve fitting, and the correction to the
luminosity for light curve shape. All the details are given in Schmidt et al.
(1998), Riess et al. (1998), Woudt et al. (2000) and Perlmutter et al. (1999).
The faintness of the distant supernovae can be explained in several dif­
ferent ways. Proposed explanations include evolution of the supernovae, dust
with absorption properties different from Galactic dust, and the effect of a
cosmological constant. More exotic explanations have been proposed as well,
but will not be discussed here.
Evolution has been the downfall of every proposed standard candle in the
past. SNe Ia are not fully understood and it is quite conceivable that they have

Supernovae at high redshift 5
Fig. 1. Hubble diagram of the High­z SN Search Team
evolved over half the current age of the universe. However, the basic energy
source of SNe Ia is the thermo­nuclear explosion and explanations have been
sought in abundance differences which lead to slightly different explosion
energetics (H¨oflich et al. 1998, Dominguez et al. 1999, Umeda et al. 1999).
None of these abundance differences are very compelling at the moment, but
can not be excluded. However, the effect is expected to be strongest in the
local supernova sample, which presumably contains progenitors of the largest
age range. The fact, that the luminosity correction works so well in this sample
(Schmidt et al. 1998) gives us confidence, that it also would do so for the
distant supernovae. A systematic sampling difference is largely diminished by
the correction, but is also not seen in the data (Riess et al. 1998, Perlmutter
et al. 1999). No other clear signatures of differences in the light curves or
the spectral evolution have been detected so far, although the quality of the
spectroscopy is not always sufficient to draw stringent conclusions. The High­
z SN Team has now shifted its emphasis to observe higher quality spectra to
explore any evolutionary signatures. A rest­frame I light curve of SN 1999Q
(Riess et al. 2000) also displays the familiar secondary maximum of SNe Ia.

6 Bruno Leibundgut
The first sign of a possible systematic difference between local and distant
supernovae has been reported in their rise times (Riess et al. 1999).
Dust has messed up many astrophysical measurements. We still do not
have a very good handle on dust outside our own Milky Way although in­
dications have been presented that at least in nearby galaxies the optical
properties are not too different from the solar neighborhood. Dust has been
invoked to explain the faintness of the distant supernovae (Aguirre 1999a,
b). The dust properties have to be such that no or very little reddening can
be observed. The precaution taken by the observers to obtain a reddening
estimate has payed off. By extending the wavelength baseline (to include the
rest­frame I­band) it will be possible to test this hypothesis soon (Riess et
al. 2000). It has to be noted that different dust characteristics would have
wide­ranging implications for other cosmological results, such as the inferred
star formation history (e.g. Madau et al. 1996, Madau, this volume) or quasar
luminosities (e.g. Shaver et al. 1997).
A further explanation for the faintness of the distant supernovae is the
existence of a cosmological constant (Riess et al. 1998, Garnavich et al. 1998b,
Perlmutter et al. 1999). Baring any strong evolutionary effects or grey dust,
the distant supernovae do not allow a solution with positive matter density for
any model
with\Omega \Lambda = 0 (Fig. 1). Hence, a cosmological constant (or another
source of energy, e.g. Caldwell et al. 1998) is unavoidable (Caroll et al. 1991,
Turner & Tyson 1999). With the currently measured distances the expansion
must have accelerated during the last ¸5 Gyr.
3 Cosmological implications
The supernova result has some surprising implications for our understanding
of the universe. One very interesting consequence of the accelerated expansion
concerns the estimate of the dynamic age of the universe. Since the expansion
was slower in the past, the inferred age is larger than in an Einstein­de Sitter
or massless model. With a Hubble constant of H 0 = 65 km s \Gamma1 Mpc \Gamma1 an
age of about 14 Gyr is inferred. This for the first time comfortably contains
the age of the oldest known stellar components of the universe (Chaboyer et
al. 1998), thus ending a debate which had started when the expansion of the
universe was discovered by Edwin Hubble (1929).
Recent measurements of the cosmic microwave background have provided
a first hint of the first Doppler peak in the CMB power spectrum (Hancock
et al. 1998, Lineweaver 1998). Since the CMB constraint is nearly orthogonal
to the supernova result (White 1998), the two combined give fairly strict
constraints of the values
of\Omega M
and\Omega \Lambda (Lineweaver 1998, Garnavich et al.
1998b, Perlmutter et al. 1999). Interestingly enough the value found
for\Omega M
corresponds very well with the matter density inferred from clusters of galax­
ies (Turner & Tyson 1999). We thus have a consistent picture of cosmology
at the moment.

Supernovae at high redshift 7
Since the CMB result favors a flat geometry of the universe, inflationary
models are en vogue again. On the other hand, severe problems arise from a
cosmological constant: the value of the vacuum energy has to be tuned after
the inflationary phase and lacks any explanation at the moment.
For a flat universe, the equation of state can be determined (Lineweaver
1998, Garnavich et al. 1998b, Perlmutter et al. 1999). The currently favored
values would exclude cosmic strings or textures as dominating energy sources
and find that the cosmological constant is the best explanation.
0 .5 1 1.5
­1
­.5
0
.5
1
Redshift z
D(m­M)
(0.3,0.7)
(1,0)
(0,1)
(0,0)
Fig. 2. The relative distance modulus for different cosmological models
(\Omega M
;\Omega \Lambda )
The existence of the cosmological constant can be fairly easily checked
with further observations of SNe Ia. At early times the universe was domin­
ated by matter and only later has been turned around by the cosmological
constant. Hence, at sufficiently large lookback times, i.e. redshifts, we should
find a decelerated universe (Fig. 2). With the current best fit parameters we
find that supernovae near z ú 0:8 should be the relatively faintest, while at
redshifts ?1.3 they should appear brighter. This non­monotonic behavior can
be observed and is very unlike any of the other possible explanations. It has
to be expected that evolutionary effects grow as a function of redshift and
also the dust contents of the universe would have to be fine­tuned to mimic
the predicted luminosity behavior for a cosmological constant.
4 Distant core­collapse supernovae as cosmological
probes
After the successful observations of distant SNe Ia it can be expected that
other supernova types may provide important cosmological information as
well. Supernova rates will tell us something about the progenitor systems. In

8 Bruno Leibundgut
particular, the short fuse for core­collapse supernovae provides a simple way
to measure the star formation rate at significant redshifts (Madau et al. 1998,
Dahl'en & Fransson 1999).
There are several ways how to observe distant core­collapse supernovae.
An interesting proposal is to observe their X­ray/UV shock breakout (Chugai
et al. 1999), which essentially depends on the size of the progenitor star.
These are the most luminous moments of a SN II. Time dilation stretches
the approximately day­long flash to several days at significant redshifts. The
flashes have the advantage that they could be observed in the observer's optical
window.
The advent of large telescopes opens up the infrared domain for faint
objects. This is necessary to reach the rest­frame optical wavelengths of these
supernovae as their UV flux is strongly suppressed (Cappellaro & Turatto,
1995, Kirshner et al. 1993) due to very strong Fe line blanketing. Especially
the range from 0:5 ! z ! 4 is interesting as the most intensive phase of star
formation (Madau et al. 1998, Dahl'en & Fransson 1999).
As the brightest signatures of individual stars supernovae are interesting
stellar cosmological probes. Supernovae have an important role to play for
our understanding of the constituents of the universe and their history.
References
1. Aguirre, A. N. 1999a, ApJ, 512, L19`
2. Aguirre, A. N. 1999b, ApJ, in press (astro­ph/9904319)
3. Bowers, E. J. C., et al. 1997, MNRAS, 290, 663
4. Caldwell, R. Dave, R., & Steinhardt, P. J. 1998, Phys. Rev. Lett., 80, 1582
5. Cappellaro, E., Turatto, M., & Fernley, J. 1995, ESA­SP 1189
6. Cappellaro, E., et al. 1997, A&A, 328, 203
7. Carroll, S. M., Press, W. H., & Turner, E. L. 1992, ARA&A, 30, 499
8. Chaboyer, B., Demarque, P., Kernan, P. J., & Krauss, L. M. 1998, ApJ, 494, 96
9. Chugai, N. N., Blinnikov, S. I., & Lundqvist, P. 1999, in Future Directions of
Supernova Research: Progenitors to Remnants, eds. S. Cassisi and P. Mazzali,
Mem. Soc. Astr. It, in press (astro­ph/9901298)
10. Contardo, G., Leibundgut, B., & Vacca, W. D. 1999, A&A, submitted
11. Dahl'en, T., & Fransson, C., 1999, A&A, in press (astro­ph/9905201)
12. Dominguez, I., H¨oflich, P., Straniero, O., & Wheeler, J. C. 1999, in Future
Directions of Supernova Research: Progenitors to Remnants, eds. S. Cassisi and
P. Mazzali, Mem. Soc. Astr. It, in press (astro­ph/9905047)
13. Filippenko, A. V., et al. 1992, AJ, 104, 1543
14. Ford, C. H., et al. 1993, AJ, 106, 1101
15. Garnavich, P. M., et al. 1999a, ApJ, 493, L53
16. Garnavich, P. M., et al. 1999b, ApJ, 509, 74
17. Goldhaber, G., et al. 1997, in Thermonuclear Supernovae, ed. P. Ruiz­Lapuente
et al., (Dordrecht: Kluwer), 777
18. H¨oflich, P., Wheeler, J. C., & Thielemann, F.­K., 1998, ApJ, 495, 617
19. H¨oflich, P., & Khokhlov, A. 1996, ApJ, 457, 500

Supernovae at high redshift 9
20. H¨oflich, P., et al. 1996, ApJ, 472, L81
21. Hamuy, M., et al. 1996a, AJ, 112, 2391
22. Hamuy, M., et al. 1996b, AJ, 112, 2398
23. Hancock, S., Rocha, G., Lasenby, A. N., & Gutierrez, C. M. 1998, MNRAS,
294, L1
24. Hubble, E. 1929, Proc. N. A. S., 15, 168
25. Jha, S., et al. 1999, ApJS, in press (astro­ph/9906220)
26. Kirshner, R. P., et al. 1993, ApJ, 415, 589
27. Leibundgut, B., et al. 1993, AJ, 105, 301
28. Leibundgut, B., et al. 1996, ApJ, 466, L21
29. Lineweaver, C. H. 1998, ApJ, 505, L69
30. Madau, P., et al. 1996, MNRAS, 283, 1388
31. Madau, P., Della Valle, M., & Panagia, N., 1998, MNRAS, 297, L17
32. Mazzali, P., et al. 1998, ApJ, 499, L49
33. NÜrgaard­Nielsen H. U., et al. 1989, Nature, 339, 523
34. Perlmutter, S., et al. 1995, ApJ, 440, L41
35. Perlmutter, S., et al. 1997, ApJ, 483, 565
36. Perlmutter, S., et al. 1998a, Nature, 391, 51
37. Perlmutter, S., et al. 1998b, IAU Circ. 6881
38. Perlmutter, S., et al. 1999, ApJ, 517, 565
39. Phillips, M. M. 1993, ApJ, 413, L105
40. Phillips, M. M., Lira, P., Suntzeff, N. B., Schommer, R. A., Hamuy, M., &
Maza, J. 1999, AJ, in press (astro­ph/9907052)
41. Pinto, P. A., & Eastman, R. G. 1999, ApJ, submitted (astro­ph/9611195)
42. Riess, A. G., Press, W. H., Kirshner, R. P. 1996, 473, 88
43. Riess, A. G., et al. 1998, AJ, 116, 1009
44. Riess, A. G., Filippenko, A. V., Li, W., & Schmidt, B. P. 1999, AJ, submitted
(astro­ph/9907038)
45. Riess, A. G., et al. 2000, in preparation
46. Saha, A., et al. 1999, ApJ, in press (astro­ph/9904389)
47. Schmidt, B. P., et al. 1997, IAU Circ. 6602
48. Schmidt, B. P., et al. 1998, ApJ, 507, 46
49. Shaver, P. A., Wall, J. V., Kellermann, K. I., Jackson, C. A., & Hawkins, M,
R. S. 1996, Nature, 384, 439
50. Spyromilio, J., et al. 1992, MNRAS, 258, 53
51. Suntzeff, N. B. 1996, in IAU Colloquium 145: Supernovae and Supernova Rem­
nants, eds. R. McCray and Z. Wang, (Cambridge: Cambridge University Press),
41
52. Suntzeff, N. B., et al. 1999, AJ, 117, 1175
53. Tammann, A. G. 1999, in General Relativity, ed. T. Piran (Singapore: World
Scientific), in press (astro­ph/9805013)
54. Tammann, G. A., & Leibundgut, B. 1990, A&A, 236, 9
55. Turatto, M., et al. 1996, MNRAS, 283, 1
56. Turatto, M., et al. 1998, AJ, 116, 2431
57. Turner, M. S., & Tyson, J. A. 1999, Rev. Modern Phys., 71, 145
58. Umeda, H., Nomoto, K., Kobayashi, C., Hachisu, I., & Kato, M. 1999, ApJ, in
press (astro­ph/9906192)
59. Vacca, W. D., & Leibundgut, B. 1996, ApJ, 471, L37
60. White, M. 1998, ApJ, 506, 495
61. Woudt, P., et al. 2000, in preparation
62. Zehavi, I., Riess, A., G., Kirshner, R. P., & Dekel, A. 1998, ApJ, 503, 483