Limits on non-Gaussianities from WMAP data

Paolo Creminelli, Alberto Nicolis, Leonardo Senatore, Max Tegmark, Matías ZaldarriagaJournal of Cosmology and Astroparticle Physics 2006 · DOI · arXiv

Abstract

This paper develops a method to constrain the level of non-Gaussianity of density perturbations when the 3-point function is of the "equilateral" type. Such departures from Gaussianity are produced by single-field models like ghost or DBI inflation. The authors show that the induced shape of the 3-point function can be well-approximated by a factorizable form, making analysis practical. Applying this technique to WMAP data, they obtain a constraint on the amplitude $f_{NL}^{equil}$ of $-366 < f_{NL}^{equil} < 238$ at 95% C.L. They also improve constraints on the "local" shape to $-27 < f_{NL}^{local} < 121$ by including a linear piece in the estimator.

Introduction

The standard inflationary paradigm predicts that primordial density fluctuations should be nearly Gaussian. However, exotic models of the early universe—such as those involving higher-order derivative operators or multiple fields—predict subtle non-Gaussian signatures. This 2005 study by Creminelli et al. provides the first rigorous constraints on "equilateral" non-Gaussianity using data from the Wilkinson Microwave Anisotropy Probe (WMAP).

The Challenge of Non-Gaussianity

Measuring non-Gaussianity (NG) involves analyzing the 3-point correlation function, or bispectrum, of the Cosmic Microwave Background (CMB). This task is computationally daunting due to the massive number of pixel combinations in high-resolution maps. Furthermore, real-world data is complicated by sky masks (to remove Galactic interference) and non-uniform noise, which can introduce spurious signals.

Methods: Factorizable Templates and Optimal Estimators

To overcome computational barriers, the authors introduced a "factorizable" mathematical template to approximate the equilateral signal. This allows the complex triple-integrals of the bispectrum to be decomposed into simpler products, reducing computational scaling significantly.

The physical equilateral shape (top) and the negligible difference when using the factorizable template (bottom).

Additionally, the authors derived an optimal estimator that includes a linear term. While standard estimators are purely cubic, the inclusion of a linear piece is necessary to maintain optimality in the presence of anisotropic noise and sky masks. This term effectively subtracts the "fake" non-Gaussianity induced by the mask and noise properties.

Key Insights: The Geometry of Inflation

A central theme of the paper is that the "shape" of non-Gaussianity—how the correlation depends on the relative sizes of the triangle sides in momentum space—reveals the underlying physics of inflation.

  • Local Shape: Predicted by multi-field models (like the curvaton), this signal peaks in the "squeezed" limit, where one momentum scale is much smaller than the others.
  • Equilateral Shape: Predicted by models with higher-derivative operators (like DBI or ghost inflation), this signal peaks when all three momentum scales are comparable.
Comparison between the factorizable equilateral template (top) and the local shape (bottom). Note the different peak locations in the (k2/k1, k3/k1) plane.

Results and Impact

Using the first year of WMAP data, the authors established the following 95% confidence level limits:

  • Equilateral: $-366 < f_{NL}^{equil} < 238$
  • Local: $-27 < f_{NL}^{local} < 121$

The improvement in the local constraint (previously $-58 < f_{NL}^{local} < 134$ by the WMAP team) demonstrated the power of the new linear-term estimator. While the results remained consistent with Gaussianity, they significantly narrowed the parameter space for inflationary models.

Conclusion

This work laid the statistical foundation for modern non-Gaussianity searches. The techniques developed here—factorizable templates and the linear-term estimator—became standard tools for subsequent analyses of WMAP and Planck satellite data, continuing to push the boundaries of our understanding of the early universe.

Summary updated

Abstract

We develop a method to constrain the level of non-Gaussianity of density perturbations when the 3-point function is of the "equilateral" type. Departures from Gaussianity of this form are produced by single field models such as ghost or DBI inflation and in general by the presence of higher order derivative operators in the effective Lagrangian of the inflaton. We show that the induced shape of the 3-point function can be very well approximated by a factorizable form, making the analysis practical. We also show that, unless one has a full sky map with uniform noise, in order to saturate the Cramer-Rao bound for the error on the amplitude of the 3-point function, the estimator must contain a piece that is linear in the data. We apply our technique to the WMAP data obtaining a constraint on the amplitude f_NL^equil of "equilateral" non-Gaussianity: -366 < f_NL^equil < 238 at 95% C.L. We also apply our technique to constrain the so-called "local" shape, which is predicted for example by the curvaton and variable decay width models. We show that the inclusion of the linear piece in the estimator improves the constraint over those obtained by the WMAP team, to -27 < f_NL^local < 121 at 95% C.L.
BibTeX
@article{creminelli2006limits,
  author = {Paolo Creminelli and Alberto Nicolis and Leonardo Senatore and Max Tegmark and Matías Zaldarriaga},
  title = {{Limits on non-Gaussianities from WMAP data}},
  journal = {Journal of Cosmology and Astroparticle Physics},
  year = {2006},
  doi = {10.1088/1475-7516/2006/05/004},
}
Figure