First constraints on the running of non-Gaussianity

TL;DR

This paper uses WMAP data to place the first constraints on the scale dependence of primordial non-Gaussianity, specifically on the running of the parameter fNL, which is crucial for understanding inflationary models.

Contribution

It introduces the first constraints on the scale-dependent parameter fNL(k) in the local non-Gaussianity model using CMB data.

Findings

Constraint on the running parameter n: 0.30 (+1.9, -1.2) at 95% confidence

Demonstrates the dependence of constraints on prior assumptions

Future data from Planck and large-scale structure surveys will tighten these constraints

Abstract

We use data from the WMAP temperature maps to constrain a scale-dependent generalization of the popular 'local' model for primordial non-Gaussianity. In the model where the parameter fNL is allowed to run with scale k, fNL(k) = fNL* (k/k_piv)^n, we constrain the running to be n = 0.30(+1.9)(-1.2) at 95% confidence, marginalized over the amplitude fNL*. The constraints depend somewhat on the prior probabilities assigned to the two parameters. In the near future, constraints from a combination of Planck and large-scale structure surveys are expected to improve this limit by about an order of magnitude and usefully constrain classes of inflationary models.