Non-Gaussianity as a Probe of the Physics of the Primordial Universe and the Astrophysics of the Low Redshift Universe

Eiichiro Komatsu, Yeinzon Rodríguez, Matías Zaldarriaga, Román Scoccimarro, Daniel Baumann, Eva Silverstein, M. Liguori, Eugene A. Lim, Max Tegmark, S. Matarrese, Steven Weinberg, Neal Dalal, Christian T. Byrnes, Louis Leblond, James E. Lidsey, Paolo Creminelli, E. Sefusatti, Liam McAllister, David Seery, Paul J. Steinhardt, Sarah Shandera, Fuminobu Takahashi, A. Cooray, K. Koyama, J. R. Bond, Patrick McDonald, Richard Easther, B. D. Wandelt, Leonardo Senatore, William H. Kinney, Lev Kofman, Evgeny I. Buchbinder, David Wands, Burt A. Ovrut, Anže Slosar, Andrei Linde, Ding-Yu Chung, David H. Lyth, Andrew J. Tolley, N. Bartolo, E. P. S. Shellard, Hiranya V. Peiris, U. Seljak, Antonio Riotto, Kendrick M. Smith, Amit Yadav, Shinji Mukohyama, Licia Verde, K. M. Górski, Niayesh Afshordi, Juan Maldacena, Alexei A. Starobinsky, Andrei V. Frolov, Justin Khoury, Xingang Chen, Mark Wyman, Jean-Luc Lehners, O. Doré, M. Sasaki, Yadav, A. P. S., Zaldarriaga, M.arXiv 2009 · DOI · arXiv

Abstract

A new and powerful probe of the origin and evolution of structures in the Universe has emerged and been actively developed over the last decade. In the coming decade, non-Gaussianity, i.e., the study of non-Gaussian contributions to the correlations of cosmological fluctuations, will become an important probe of both the early and the late Universe. Specifically, it will play a leading role in furthering our understanding of two fundamental aspects of cosmology and astrophysics: (i) the physics of the very early universe that created the primordial seeds for large-scale structures, and (ii) the subsequent growth of structures via gravitational instability and gas physics at later times. To date, observations of fluctuations in the Cosmic Microwave Background (CMB) and the Large-Scale Structure of the Universe (LSS) have focused largely on the Gaussian contribution as measured by the two-point correlations (or the power spectrum) of density fluctuations. However, an even greater amount of information is contained in non-Gaussianity and a large discovery space therefore still remains to be explored. Many observational probes can be used to measure non-Gaussianity, including CMB, LSS, gravitational lensing, Lyman-alpha forest, 21-cm fluctuations, and the abundance of rare objects such as clusters of galaxies and high-redshift galaxies. Not only does the study of non-Gaussianity maximize the science return from a plethora of present and future cosmological experiments and observations, but it also carries great potential for important discoveries in the coming decade.

Introduction: Beyond the Power Spectrum

For the last few decades, cosmology has been remarkably successful at describing the universe using just a few parameters, such as the age of the universe and the density of dark matter. This "Standard Model" relies heavily on the power spectrum, which measures two-point correlations—essentially, the probability of finding a galaxy at a specific distance from another.

However, the power spectrum suffers from a significant "degeneracy" problem: many wildly different physical theories regarding the inflationary period of the early universe predict nearly identical power spectra. If primordial fluctuations were perfectly Gaussian, the power spectrum would provide a complete statistical description. However, even slight departures from Gaussianity—non-Gaussianity—contain "extra" information hidden in higher-order correlations. Studying these signals is the primary method for "breaking the tie" between competing theories and understanding the specific quantum field interactions that occurred moments after the Big Bang.

Methodology: The Geometry of the Bispectrum

To extract this hidden information, researchers move beyond the standard two-point correlation to the three-point function (the bispectrum) and the four-point function (the trispectrum). The core technique involves analyzing the "shape" of triangles formed by three wave vectors ($k_1, k_2, k_3$) in Fourier space. Different physical processes during inflation leave distinct geometric fingerprints:

  • Squeezed: Characterized by one very long side and two short sides ($k_3 \ll k_1 \approx k_2$). This configuration is a "smoking gun" for models involving multiple fields during inflation.
  • Equilateral: All sides are approximately equal ($k_1 \approx k_2 \approx k_3$). This suggests "non-canonical" kinetic energy or complex field interactions.
  • Flattened/Folded: These shapes typically point to non-vacuum initial conditions.
The three primary bispectrum shapes: (a) Squeezed, (b) Equilateral, and (c) Flattened/Folded.

Additionally, the method of scale-dependent bias allows researchers to measure the non-Gaussianity parameter ($f_{NL}$) by observing how the power spectrum of density extrema (where galaxies form) changes on very large scales, bypassing the need for a full bispectrum calculation.

Observational Probes and Data Synthesis

The study of non-Gaussianity synthesizes results from several major cosmological surveys and looks forward to next-generation experiments:

  • Current Constraints: Data from the WMAP (Wilkinson Microwave Anisotropy Probe) 5-year study and the Sloan Digital Sky Survey (SDSS) provided the first rigorous bounds. WMAP 5-year data constrained $f_{NL}$ to $38 \pm 21$.
  • Future Missions: The Planck satellite, the Large Synoptic Survey Telescope (LSST), and the Square Kilometer Array (SKA) are expected to significantly increase sensitivity.
  • Complementary Probes: Gravitational lensing, the Lyman-$\alpha$ forest, and 21-cm fluctuations offer independent datasets to cross-validate primordial signals.

Key Insights and Scientific Impact

The most profound insight offered by non-Gaussianity is its ability to serve as the ultimate test for inflation. A definitive detection of the "squeezed" configuration would rule out all single-field slow-roll inflation models, fundamentally altering our understanding of the first fraction of a second of time.

Furthermore, non-Gaussianity serves as a dual-purpose probe. While it reveals primordial physics, it also tracks how gravity pulls matter together over billions of years. This makes it a sensitive tool for the low-redshift universe, helping scientists study dark energy and the era of reionization when the first stars ignited. By mining the information that the power spectrum discards, researchers can recover a vast amount of data regarding the universe's evolution.

Challenges and Future Outlook

Despite its potential, the field faces significant hurdles. The primary challenge is astrophysical contamination; non-linear gravity and galaxy formation in the late universe can create non-Gaussian signals that mimic primordial ones. Additionally, new methods like scale-dependent bias require more rigorous study of systematic errors.

The goal for the coming decade is to reach a sensitivity of $\Delta f_{NL} \lesssim 1$. Achieving this level of precision would allow cosmologists to either confirm or rule out entire classes of alternative theories, such as contracting universe models, and establish a clearer picture of our cosmic origins.

Summary updated

Abstract

A new and powerful probe of the origin and evolution of structures in the Universe has emerged and been actively developed over the last decade. In the coming decade, non-Gaussianity, i.e., the study of non-Gaussian contributions to the correlations of cosmological fluctuations, will become an important probe of both the early and the late Universe. Specifically, it will play a leading role in furthering our understanding of two fundamental aspects of cosmology and astrophysics: (i) the physics of the very early universe that created the primordial seeds for large-scale structures, and (ii) the subsequent growth of structures via gravitational instability and gas physics at later times. To date, observations of fluctuations in the Cosmic Microwave Background (CMB) and the Large-Scale Structure of the Universe (LSS) have focused largely on the Gaussian contribution as measured by the two-point correlations (or the power spectrum) of density fluctuations. However, an even greater amount of information is contained in non-Gaussianity and a large discovery space therefore still remains to be explored. Many observational probes can be used to measure non-Gaussianity, including CMB, LSS, gravitational lensing, Lyman-alpha forest, 21-cm fluctuations, and the abundance of rare objects such as clusters of galaxies and high-redshift galaxies. Not only does the study of non-Gaussianity maximize the science return from a plethora of present and future cosmological experiments and observations, but it also carries great potential for important discoveries in the coming decade.
BibTeX
@misc{komatsu2009non,
  author = {Eiichiro Komatsu and Yeinzon Rodríguez and Matías Zaldarriaga and Román Scoccimarro and Daniel Baumann and Eva Silverstein and M. Liguori and Eugene A. Lim and Max Tegmark and S. Matarrese and Steven Weinberg and Neal Dalal and Christian T. Byrnes and Louis Leblond and James E. Lidsey and Paolo Creminelli and E. Sefusatti and Liam McAllister and David Seery and Paul J. Steinhardt and Sarah Shandera and Fuminobu Takahashi and A. Cooray and K. Koyama and J. R. Bond and Patrick McDonald and Richard Easther and B. D. Wandelt and Leonardo Senatore and William H. Kinney and Lev Kofman and Evgeny I. Buchbinder and David Wands and Burt A. Ovrut and Anže Slosar and Andrei Linde and Ding-Yu Chung and David H. Lyth and Andrew J. Tolley and N. Bartolo and E. P. S. Shellard and Hiranya V. Peiris and U. Seljak and Antonio Riotto and Kendrick M. Smith and Amit Yadav and Shinji Mukohyama and Licia Verde and K. M. Górski and Niayesh Afshordi and Juan Maldacena and Alexei A. Starobinsky and Andrei V. Frolov and Justin Khoury and Xingang Chen and Mark Wyman and Jean-Luc Lehners and O. Doré and M. Sasaki and Yadav, A. P. S. and Zaldarriaga, M.},
  title = {{Non-Gaussianity as a Probe of the Physics of the Primordial Universe and the Astrophysics of the Low Redshift Universe}},
  year = {2009},
  eprint = {0902.4759},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/0902.4759},
  howpublished = {arXiv},
}
Figure