String Theory And Cosmology - Proceedings Of The Nobel Symposium 127 9789812701657, 9789812564337

Citation preview

Proceedings of Nobel Symposium 127 Sigtuna, Sweden, August 1 4 - 1 9 , 2003

Editors

Physica Scripta The Royal Swedish Academy of Sciences / World Scientific

String Theory and Cosmology Proceedings of Nobel Symposium 127 Sigtuna, Sweden, August 1 4 - 1 9 , 2003

Editors

U. Danielsson A. Goobar B. Nilsson

Recognized by the European Physical Society Physica Scripta The Royal Swedish Academy of Sciences

^Sra K U N G L y y VETENSKAPSAKADEMIEN W

B

7

THE ROYAL SWEDISH ACADEMY OF SCIENCES

Y|p> World Scientific NEW JERSEY • LONDON • S I N G A P O R E • BEIJING • S H A N G H A I • H O N G K O N G • T A I P E I

Published jointly by Physica Scripta The Royal Swedish Academy of Sciences Box 50005, S-104 05, Stockholm, Sweden and World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.

STRING THEORY AND COSMOLOGY — Nobel Symposium 127 Copyright © 2005 Royal Swedish Academy of Sciences All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.

The contents of this volume were also published as Vol. Tl 17 of Physica Scripta.

ISSN Royal Swedish Academy of Sciences ISBN Royal Swedish Academy of Sciences ISBN World Scientific

0031-8949 (0281 -1847) 91-89621-23-9 981-256-433-0

Printed by Fulsland Offset Printing (S) Pte Ltd, Singapore

Physica S c r i p t a , V o l . T 1 1 7 , 2 0 0 5

Contents Preface

5

List of Participants

7

Programme

9

Supernovae and Dark Energy. J. L Tonry Studying Dark Energy with Supernovae: Now, Soon, and the Not-Too-Distant Future. S. Perlmutter

11 . . .

17

What Have We Learned from Cosmic Microwave Background Fluctuations? D. N. Spergel

29

Quintessential Ideas. P. J. Steinhardt

34

Prospects of Inflation. A. Linde

40

A Non Singular Universe. S. Hawking

49

Unconventional Scenarios and Perturbations Therein. G. Veneziano

51

Holographic Cosmology 3.0. T. Banks and W. Fischler

56

Bouncing Universes in String Theory? J. Polchinski

64

Remarks on Tachyon Driven Cosmology. A. Sen

70

Beyond Inflation: A Cyclic Universe Scenario. N. Turok and P. J. Steinhardt

76

Creating Naked Singularities and Negative Energy. G. T. Horowitz

86

Filtering Gravity: Modification at Large Distances? G. Dvali

92

Cosmological Singularities in String Theory. D. Kutasov

99

Where Do We Stand in Fundamental (String) Theory. D. J. Gross

102

This page is intentionally left blank

Physica Scripta. Vol. T l 17, 5, 2005

Preface The idea of organizing a Nobel Symposium focusing on the recent exciting developments in the overlap between string theory and cosmology was born about two years prior to the meeting. The progress, both observational and theoretical, was sparked off by the spectacular new measurements of the geometry and content of the universe. These included ever more accurate observations of the cosmological background radiation testing our ideas of the very early universe, as well as observations of distant supernovae indicating that the universe is in a phase of accelerated expansion. This development was a fertile ground for cosmologists as well as for the string community where people were looking for contacts with observations by applying string theory at the cosmological scale. We are now witnessing the impact of improved observational data coming in regularly, and on the theory side, the appearence of large amounts of collaborative work between cosmologists and string theorists. In view of these facts, we believe that the timing of the Symposium was close to optimal. As organizers, we felt very pleased with the Symposium, and were very happy to see so many prominent scientists being enthusiastic about the meeting. We direct our most profound thanks for the success of the Symposium to all the participants. A special thanks goes to the speakers and chairmen. With half of the time allocated for discussions in the sessions, each one devoted to a particular issue, it was of paramount importance that the chairmen could get both cosmologists and string theorists to contribute with their views and expertice to penetrate the questions to the limit. The resulting debates in several sessions will linger in the memories of those present for many years to come. The Symposium ended with a question and answer session with the whole audience acting as panel. The participants had been asked to formulate what they thought were the key questions in the field. Although consensus was not always reached, it was clear to everyone that the issues discussed at the Symposium (and documented in this book) are among the most fascinating topics in contemporary Physics. Our most sincere thanks go to professors Lars Brink, David Gross, and Michael Turner for organizing this grand final. The outcome of a Symposium like this depends to a very large extent on the place where it was held and its surroundings. The choice fell on the small historic town of Sigtuna and the Sigtuna Stiftelsen (Sigtuna Foundation), a Mediterranean inspired building complex and conference center from the beginning of the last century, with a rich tradition of hosting the most famous individuals in the cultural life of Sweden. The building itself, the proximity to the lake Malaren, and the picturesque town made the stay a wonderful experience. Our deep gratitude is also directed to the personnel in Sigtuna who with a smile made everything possible and the stay a very pleasant one. Important help was also provided by the scientific secretaries of the Symposium, Martin Olsson and Henric Lars son. We also want to emphasize the crucial role played by the international advisory committee consisting of L. Bergstrom, L. Brink, P. Di Vecchia, D. Gross, R. Kallosh, M. Rees, H. Rubinstein, and M. Turner. We are deeply grateful for their work, especially during the preliminary planning of the meeting. The Klein lecture committee and the Swedish Academy of Sciences are acknowledged for organizing an impressive end to the Symposium in Stockholm. The public lectures by David Gross and Andrei Linde followed by the Klein lecture by Stephen Hawking were attended by an audience of over 1200 people, while several other hundred were unable to enter the already full lecture room. This event was certainly among the scientific highlights of the year in Sweden. Finally, we would like to extend our gratitude to the Nobel foundation for providing the financial support. Uppsala, Stockholm and Goteborg, January 2004 Ulf Danielsson, Ariel Goobar and Bengt E. W. Nilsson

© Physica Scripta 2005

Physica Scripta T117

Physica Scripta. Vol. Tl 17, 6, 2005

From left: A, Goobar, U. Danielsson, L. Bergstriim, F. Ravndal. H. Rubinstein. L. Brink. S. Hannestad. G. Horowitz, B. Greene. T, Damour, L. Page, D. Gross. V. Mukhanov. A. Albrecht, S. Hawking. T. Banks. M. Turner. A. Starobinsky, J. Tonry. R. Kallosh. A. Lindc. C. Bachas. S. Perlmutter. P. Di Vecchia. D. Kiitasov, C. Jarlskog. G. Efstathiou. G. Veneziano, L. Susskind. A. Guth. P. Steinhardt. Larsson (student), A. Sen. Olsson (student). D. Spergel. G. Dvali. B. Nilsson.

Physica Scripta Tl 17

© Physica Scripta 2005

Physica Scripta. Vol. T117, 7, 2005

List of Participants A. Albrecht Andersen University of California Davis USA

B. Greene Columbia University New York USA

S. Perlmutter University of California Berkeley USA

C. Bachas Ecole Normale Superieure Paris France

D. Gross University of California Santa Barbara USA

J. Polchinski University of California Santa Barbara USA

A. Guth MIT Cambridge USA

F. Ravndal Oslo University Norway

S. Hannestad Odense University Denmark

H. Rubinstein Stockholm University Sweden

L. Brink Chalmers and Goteborg University Sweden

S. Hawking DAMTP Cambridge University UK

A. Sen Harish-Chandra Research Institute Allahabad India

P. Carlson Stockholm University Sweden

G. Horowitz University of California Santa Barbara USA

D. Spergel Princeton University USA

C. Jarlskog Lund University Sweden

A. Starobinsky Landau Institute Moscow Russia

T. Banks University of California Santa Cruz and Rutgers University USA L. Bergstrom Stockholm University Sweden

T. Damour IHES Bures-sur-Yvette France U. Danielsson Uppsala University Sweden P. Di Vecchia Nordita Copenhagen Denmark G. Dvali New York University USA G. Efstathiou Cambridge University UK

R. Kallosh Stanford University USA M. Kamionkowski Caltech USA D. Kutasov University of Chicago USA A. Linde Stanford University USA V. Mukhanov Munich University Germany

K. Enqvist Helsinki University Finland

B. Nilsson Chalmers and Goteborg University Sweden

A. Goobar Stockholm University Sweden

L. Page Princeton University USA

© Physica Scripta 2005

P. Steinhardt Princeton University USA L. Susskind Stanford University USA J. Tonry University of Hawaii USA M. Turner University of Chicago USA N. Turok DAMPT Cambridge University UK G. Veneziano CERN Geneva Switzerland

Physica Scripta Till

This page is intentionally left blank

Physica Scripta. Vol. Tl 17, 9, 2005

Programme

August 14 Reception dinner August 15 The dark sectors of the Universe Chair: Page Efstathiou: Large scale structure Tonry: Studying the dark sector with supernovae and lensing Perlmutter: Studying dark energy with supernovae, from now to a future space based mission Chair: Mukhanov Spergel: What have we learned from the CMB1 Kamionkowski: CMB and polarization, expectations Steinhardt: Quintessence, models and predictions August 16 Inflation and other ideas Chair: Albrecht Guth: Birth of inflationary ideas Linde: Prospects of inflation Hawking: Fundamental theory and the universe Chair: Kallosh Starobinsky: De Sitter space-time foam Veneziano: Unconventional scenarios and perturbations therein Greene: Aspects of string cosmology August 17 Holography and de Sitter space Chair: Di Vecchia Susskind: The landscape of string theory Banks: Holographic cosmology Polchinski: Bouncing universes in string theory Boat Excursion on the lake Mdlaren to the castle of Skokloster and Swedish Banquet onboard August 18 New ideas about the universe Chair: Bachas Sen: Dynamics ofD-brane decay Turok: Beyond inflation, a cyclic universe scenario Chair: Damour Horowitz: Comments on time dependence in string theory Dvali: Filtering gravity Kutasov: Cosmological singularities in string theory August 19 Summary and conclusions Chair: Brink Turner: Where do we stand in cosmology! Gross: Where do we stand in fundamental theory! Panel discussion

© Physica Scripta 2005

This page is intentionally left blank

Physica Scripta. Vol. Tl 17, 11-16, 2005

Supernovae and Dark Energy John L. Tonry1 Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822 Received February 23, 2004

Abstract Type la supernovae provide a precise luminosity distance at redshift as high as z 2. At present nearly 200 SN la paint a consistent picture that QA — 1 AQM = 0.35 ± 0.14 (if w = -1). If QT = 1, SN la tell us QM = 0.28 ± 0.05, independent of any other measurements. Adopting a prior based on the 2dF redshift survey constraint on QM and assuming a flat universe, then -1.48 < w < -0.72 at 95% confidence. If we further assume that w > — 1, we obtain w < —0.73 at 95% confidence. Understanding systematic errors is of paramount importance, both because they may affect our present results and they will limit the eventual precision achievable. Possible sources of error include photometric error, dust extinction, evolution of SN la, selection effects, K-corrections, and gravitational lensing, all which could vary with redshift. There are many efforts underway to push to higher redshift and to control possible observational systematic errors; the bottom line is that the acceleration reported earlier appears to be confirmed. The propagation of systematic error in luminosity distance to inferred w or w(z) depends critically on how much is known about QM- For example, at fixed QM a systematic uncertainty in dium of 0.04 mag corresponds to a 10% uncertainty in w, but this becomes much worse if QM is not known. We will have some good constraints on QM, either directly from 2dF or SDSS, or indirectly from WMAP whose constraints in the (QM , w) plane run more or less perpendicular to the SN la (but they can also have systematic errors themselves). Theoretical models for SN la are now able to make multi-wavelength light curves and spectra which are a decent match to observation. The luminositydecline rate correlation depends primarily on the amount of 56 Ni produced in the explosion. There are other, more subtle effects which are also being elucidated, for example, increasing initial metallicity (in the form of neutron-rich 56Fe and 22Fe) diminishes the neutron-poor 56Ni and dims the explosion. Unfortunately we have very little idea at present about what causes a WD to explode. A small fraction of WD explode, and those that do wait a long time between their formation and their explosion. Without initial conditions, there is a lot of room for systematic error at the 0.02 mag level from SN la explosions, and whether they depend in a serious way with z and therefore limit our ability to measure w(z) remains to be seen. It may be that we can measure dium(z) to 1% (0.02mag) by controlling only observational systematics, or it may be that we incur significant systematic errors for which we do not have the requisite information to realize their presence or correct for them.

than SN la, and therefore are much less useful for measuring cosmology. SN la are almost certainly the consumption of a white dwarf (WD) star in a thermonuclear explosion. The history of the use of SN la for cosmology can be loosely broken into three epochs. Zwicky's supernova search from the 1930's through the 1960's culminated in Kowal's Hubble diagram in 1968 which demonstrated the possibility of using supernovae for cosmology. Throughout the 1980's the detailed understanding of core-collapse versus WD explosions was elucidated, and the announcement in 1993 by Phillips of the luminosity-decline rate correlation opened the door for SN la to be used as relatively precise distance estimators. The use of SN la at high redshift started with Norgaard-Neilsen et al. in 1988 and continued throughout the 1990's, mostly enabled by the improvement of wide field imagers and the availability of very large telescopes capable of measuring the redshifts of the supernovae. The startling announcement by Riess et al. in 1998 and Perlmutter et al. in 1999 that the universe appears to be accelerating ushered in the era of dark energy. Since that time there have been several confirmations of these results and several large projects to improve on the precision have been undertaken. 2. SN la and Luminosity Distance SN la display a luminosity function which has a typical dispersion of about 50%, with an tail to the distribution at lower luminosity.

0

20

40

1. Introduction Type la supernovae offer a precise measure of luminosity distance to redshifts as high as z 2. (Higher redshifts may be essentially unobtainable because SN la may be extremely rare at such early times.) This direct measurement of the expansion rate offers the possibility of measuring the content of dark energy (QA), the equation of state of dark energy (w), and time variation in the equation of state (w(z)). SN la therefore can tell us interesting things about particle physics, and the quest for them at high redshift may give us essential clues for understanding the explosions themselves. There are many types of exploding stars, loosely classified into core collapse of massive stars, known as SN II, SN lb, or SN Ic (the "I" versus "II" spectroscopic nomenclature refers to the absence or presence of hydrogen lines), and SN la. The former are a very heterogeneous class and are generally less luminous

[email protected] © Physica Scripta 2005

Fig. 1. SN la lightcurves are at least a one parameter family, but their luminosity can be determined from a variety of distance independent observations. Physica Scripta Till

12

John L. Tonry

.01

.02

.05

.1

.3

redshift Fig. 2. Luminosity distance inferred from SN la at low redshift correlate extremely well with redshift, which allows us to put limits on systematic problems as a function of SN la properties or environment. Redshift z

There is a correlation between luminosity with other features, however, such as decline rate, spectral features, and color which are measurable independent of distance. When this correlation is taken into account, SN la have a luminosity dispersion of about 20%, sometimes reported to be as small as 16 or even 14% in nearby, selected subsamples. This has never been demonstrated in supernova samples at high redshift, and it is likely that the achievable in a sample of N SN la is 20% N~^2, for N < 100. It does appears to be very feasible to measure SN la luminosity distances at redshifts up to z 2 with an accuracy of 10% (i.e. 20% in luminosity). Examination the dependence of luminosity distance on the parameters QM and QA (for w fixed at —1) reveals what SN la observations can tell us. At z 0.5 the luminosity distance relative to an empty universe (a very useful fiducial since no one believes our universe actually has this geometry!) is +0.12 magnitude (roughly fractional decrease in flux) at (QM,QA) = (0.3,0.7), with gradient (dQM,dQA_) = (—0.37,0.34). At redshift 1 these numbers are+0.06 and the slope steepens to (—0.80,0.39); at redshift 1.5 the magnitude offset is brighter at -0.07 with slope (-1.18,0.30); and at redshift 2 the values are -0.21 and (-1.49,0.19). Roughly speaking, at this point our observations tell us that at z = 0.5 dm = +0.13 ±0.05 and at z = 1.0 dm = +0.00 + 0.08. The intersection of these constraints gives us an elongated slice through the (QM, QA) plane which prefers QA — QM ~ 0.4 and has very little power in the QA + QM direction. There are many efforts underway to push to higher redshift (as well as to control possible observational systematic errors), typified by the recent publication of Tonry et al. 2003. The bottom line is that the acceleration reported earlier appears to be confirmed, and we do see SN la brightening at z ~ 1. Assuming a flat universe with w = — 1, Tonry et al. find QM = 0.28 ± 0.05. Additionally adopting a prior based on the 2dF redshift survey constraint on QM, the SN la data indicate that a constant w lies in the range —1.48 < w < —0.72 at 95% confidence. There have been approximately 230 SN la observed with sufficient detail to measure a good distance. Of these 172 have low extinction and are at high enough redshift not to experience a significant peculiar velocity relative to the Hubble flow. Physica Scripta TI17

Fig. 3. The dependence of luminosity distance on redshift becomes sensitive to cosmological parameters at a redshift of about 0.5.

3. Systematics and Future Searches We worry a great deal about systematic errors. Possible sources of error include photometric error, extinction within host galaxies or along the line of sight, evolution of SN la, selection effects, K-corrections (transforming observed bandpasses to fiducial rest frame bandpasses), and gravitational lensing, all which could vary with redshift. With the exception of the last, however, we expect a systematic error with redshift which causes the dimming we see of 0.13 mag at z ~ 0.5 to continue to increase with increasing redshift, whereas the effect of dark energy will cause SN la to brighten relative to an empty universe at z ~ 1. A shortcoming of the existing searches, however, is that they were generally the result of two observational epochs, separated by a month, and then followed with a very heterogeneous campaign to acquire spectra and photometry to delineate the light curve and the host extinction. Ongoing searches improve on this by observing a field continuously over a period of 3-6 months in multiple colors, simultaneously discovering new SN la and following old ones. A good example is the "IFA Deep" survey of Barris et al. 2004. This provides us with 20 new SN la to redshifts beyond 1, but more importantly has better photometric accuracy and less sensitivity to photometric systematics which certainly plague the published compendium of 172. An alternative way to control systematics is to continue in the effort to reach z > 1, even though this is extremely difficult to do from the ground because of the faintness of the SN la and more importantly the redshift of well behaved portion of the rest frame spectrum into the near infrared where the atmosphere becomes very bright and variable. An important campaign in this direction are the "Higher-z" campaign led by by Adam Riess (Riess et al. 2004a, and Riess et al. 2004b), which worked synergistically with the HST GOODS survey and found about a dozen SN la to redshifts of 1.8. The ACS instrument on HST is extremely useful for this purpose, offering as it does a wide field of view, red sensitivity, and an excellent grism spectrograph. The results are not yet published, but we can expect a new point on the Hubble diagram with a mean redshift of about z ~ 1.4 and accuracy of about 0.1 magnitude. The photometric quality of these data is © Physica Scripta 2005

Supernovae and Dark Energy • l < i < I * * . 1 i i T-r-r-pr i I

1.5

_ ; v.

x

•1

1. It is thought that WD which are subject to mass transfer from a companion star are driven to the point of explosion, but there is no convincing model of how this works. It will become more and more imperative to understand the initial conditions which give rise to a SN la explosion, since the Improving explosion calculations can only constrain things to die extent which we understand them. Absent initial conditions, there Is certainly a lot of room for systematic error at the 0.02 mag level (or larger) from the SN la explosion event. Whether systematic errors will depend in a serious way with z and therefore limit our ability to measure w(z) remains to be seen. It may be that we can measure dium(z) to 1% (0.02mag) by controlling only observational systematics. It may be that we ind larger systematics, but by seeking correlations among observed quantities and by reference to the steadily improving models, we can achieve this same limit. Or it may be that we incur significant systematic errors for which we do not have the requisite information to realize their presence or correct for them. For example 22Ne sedimentation could conceivably cause an 0.04 mag systematic difference between z = 0 and z = 1 which is not part of the usual correlations and, although measurable in the 54Fe produced by the explosion, is Impossible to distinguish from low dispersion, low S/N spectra. Physica Scripta 1117

16

John L. Tonry 1 I

i 11

I

|

I

I

I

|

1 i !

1 1

0T = 1 SNIa only SNIa + 2dF

N

I j

_J

^ A \

L^

l a i c iI Ii I i ! i i 1i i1 il m

" 0

1 I

.2

.4

' i i -iJ L i. . I i 1i J I ii ii iI\ M

.6

.8

1

Fig. 14. The systematic errors which occur when luminosity distance is allowed to vary within an envelope of 0.01,0.02, and 0.04 mag is plotted over the CFHTLS expectation for w = — 1.

We have embarked with SN la on an odyssey toward precise cosmology. The first achievements have been to find that dark energy exists with QA ~ 0.7. Our data are certainly corrupt at some level with systematic error, but It is unlikely to affect

Physica Scripta 7117

this conclusion. The next goal is precision cosmology and the determination of the value of WQ If It Is constant. We have results with 25% precision but unknown accuracy, and experiments are underway to significantly Improve the quality and quantity of observations. It is likely that the systematic errors will not be a significant limitation in this effort. The ultimate goal is ultra precision cosmology which would give us w(z) and reveal whether w is a cosmological constant or a dynamical scalar field. There is a very good chance that we will suffer from significant distortion of the truth from systematic error, and we may learn more about the explosions of WD stars than we do about cosmology. This is a worthwhile end, of course, and may In turn tell us how to Improve our models of SN la to the point that they can reliably reveal the nature of the dark energy which dominates our universe.

References 1. 2. 3. 4. 5. 6. 7.

Barns, B. et al.> ApJ 602, 571 (2004). Bildsten, L. and Hall, D., ApJ 549, L219 (2001). Perimutter, S. et al.9 ApJ 517, 565 (1999). Riess, A. G. et at, ApJ 600, 163 (2004a). Riess, A. G. et a/., ApJ 607, 665 (2004b). Timmes, R X., Brown, E. F. and Truran, J. W., ApJ 590, L83 (2003). Tonry, J. et al.9 ApJ 594, 1 (2003).

© Physica Scripta 2005

Physica Scripta. Vol. Tl 17, 17-28, 2005

Studying Dark Energy with Supernovae: Now, Soon, and the Not-Too-Distant Future Saul Perlmutter* Lawrence Berkeley National Laboratory, University of California, 1 Cyclotron Road, Berkeley, CA 94708 Received March 15, 2004

Abstract The nature of the mysterious dark energy is one of the key questions at the intersection of cosmology and string theory - or any fundamental-physics theory. Although dark energy is apparently the dominant constituent of the universe, it is extremely difficult to detect, let alone measure its properties in detail. Our primary route is to measure its effect on the expansion history of the universe; we use Type la supernovae as direct indicators of the relative scale of the universe at each moment in history when such a supernova explosion occurred that we can study. The current generation of supernova-measurement projects now in progress are advancing our knowledge of the time-average behavior of dark energy, but we are faced with a daunting challenge when we seek to make a measurement precise - and accurate - enough to distinguish the different dark energy theories by their predictions for the changes in the dark energy's properties over cosmic time. I here highlight some of the recent results and upcoming advances in our approaches to measuring, analyzing, and identifying details of the supernova explosions - including a new satellite design - that are aimed at putting this tantalizing measurement within our grasp.

1. Introduction: A Challenging Problem At this symposium of distinguished theorists, I do not need to explain the difficult theoretical issues raised by the supernova measurement of an accelerating universe. However I should explain the approach we are taking in pursuing these problems with upcoming instrumentation/observation projects. If we all thought that Einstein's cosmological constant were a nice, simple explanation for the acceleration then we probably would not be developing these ambitious measurement projects. Most scientists, though, are uncomfortable with the tiny value of the required cosmological constant energy density, and its coincidental match (within a factor of two) to the density to which matter has fallen today. This has certainly motivated the whole industry of alternative physics theories that are appearing almost daily in the journals and web archives. Most of these alternative theories - generically called "dark energy," or, more specifically, "quintessence," or "dynamical scalar fields" - can be characterized using the equation-of-state variable w, which is the ratio of the pressure to the density in the acceleration equation. I'll use w in this talk as one of the parameters that we might be after. But stepping back to get the big picture, it is important to note that w is just one way of capturing the measurements that we are making, and, in fact, we don't know whether this p/p ratio is the right way to characterize the explanation for this particular effect in the expansion history of the universe. Nonetheless, taking w as something we might measure, we began in 1998/99 to use the supernova data to obtain constraints [1, 2], and were able to improve these by combining in measurements from large scale structure and CMB (see Figure 1, left panel) [3]. These constraints did not rule out many dark energy models.

Looking at more recent data - Figure 1 (right panel) shows the Supernova Cosmology Project's most recent results [4] - you can see that nowadays one plots w to include the region below — 1, partly to reflect the fact that we don't know if w really is the right parameterization. When we add other data sets like the CMB or the 2dF, we get what appear to be rather tight constraints around a w of— 1.05, with statistical error bars on the order of 0.15 or 0.2, and somewhat smaller systematics. These are rather interesting results, but once again they do not tell us which of these various theories fits best, and most of the current models of dark energy would all fit in these constraints. For that matter these constraints all assume that w is a constant. As soon as you allow w to vary in time, all these constraints become very weak. So we are faced with a very interesting challenge for the future, which is to ask how could we substantially improve these constraints, in particular when faced with the possibility that w could be changing in time. I symbolize this very hard goal with the expression w'(z), the derivative of w with redshift. We want to make a measurement of such precision - and I mean both statistical and systematic precision - that you would actually believe it if you saw a change in the properties of dark energy, a w changing in time. It would be very easy to end up with an experiment where you saw a slight shift in w, and nobody would believe it, because you would say, "well, it's just systematics between one redshift and the next." This is a real challenge; I'll describe both the pessimist and the optimist view today and end up with a sense of why this is a challenge worth pursuing. Along the way, I'll also propose a route to take in trying to make this difficult measurement.

2. The Upcoming Supernova Projects

'Email: [email protected]

Let me begin by showing the Hubble diagram drawn in a rather interesting way [5] (Figure 2), and placing the upcoming supernova projects on it. This version of the Hubble diagram is really just an expansion history of the universe, taking advantage of the fact that the relative brightness of a supernova tells you how far back in time you are looking, and redshift is a very direct measurement of the total expansion of the universe since that time. When we began the high-redshift supernova work in the late-1980s, we had just the points in the "recent," linear regime, and we wanted to look further back in time to see which of these decelerating lines we live on. As you all know, the data set showed us that we are on none of those lines, and that we are on a line of expansion history in which the universe first decelerated, but then began accelerating [6, 7, 8, 9, 1, 4, 10, 11]. So, you can see the supernova data and think of them in terms of the expansion history of the universe. Now I want to lay out for you where the upcoming projects lie on this plot.

© Physica Scripta 2005

Physica Scripta T117

18

Saul Perlmutter Supernova Cosmology Project Knop et al. (ApJ., 2003)

Assuming constant w Perlmutter, Turner, & White (1999) Phys. Rev, Lett, i i i i M

cosmological constant w = -1

0.2

i i i I i i i i I M

i i

0.3 0.4 liM= 1 — wx

Fig. 1. (left) Confidence regions (0.5c to 2a) in the QM~weff plane, from measurements of Type la supemovae (SNe la), Large Scale Structure (LSS), and Cosmic Microwave Background (CMB) (based on [3]). (right) More recent confidence regions (68%, 90%, 95%, and 99%), based on [4]. Both the left and right plots assume QM + QA = 1 ^ d that w is not time-varying.

Expansion History of the Universe Perlniutier. Physics Today (2003)

past - * — today -

• future

1.5 m

m *x m &> m &

c 5

1.0

it AIT, €ik< mi ethers 11 00s j & / |Ne«rt>y SN Fmimy mm

s< 2z *o m -Q

After inflation, the expansion either...

a; ° u c S w m m m m >