Evidence of a truncated spectrum in the angular correlation function of the cosmic microwave background

F. Melia, M. López-Corredoira

Research output: Contribution to journalArticlepeer-review


Aims. The lack of large-angle correlations in the fluctuations of the cosmic microwave background (CMB) conflicts with predictions of slow-roll inflation. But though probabilities (≤ 0.24%) for the missing correlations disfavor the conventional picture at & 3σ, factors not associated with the model itself may be contributing to the tension. Here we aim to show that the absence of large-angle correlations is best explained with the introduction of a non-zero minimum wavenumber kminfor the fluctuation power spectrum P(k). Methods. We assume that quantum fluctuations were generated in the early Universe with a well-defined power spectrum P(k), though with a cutoff kmin≠ 0. We then re-calculate the angular correlation function of the CMB and compare it with Planck observations. Results. The Planck 2013 data rule out a zero kminat a confidence level exceeding 8σ. Whereas purely slow-roll inflation would have stretched all fluctuations beyond the horizon, producing a P(k) with kmin= 0-and therefore strong correlations at all angles-a kmin≠ 0 would signal the presence of a maximum wavelength at the time (tdec) of decoupling. This argues against the basic inflationary paradigm-perhaps even suggesting non-inflationary alternatives-for the origin and growth of perturbations in the early Universe. In at least one competing cosmology, the Rh = ct universe, the inferred kmincorresponds to the gravitational radius at tdec.

Original languageEnglish (US)
JournalUnknown Journal
StatePublished - Dec 21 2017


  • Cosmic background radiation
  • Cosmology: theory
  • Early universe
  • Inflation
  • Large-scale structure

ASJC Scopus subject areas

  • General

Fingerprint Dive into the research topics of 'Evidence of a truncated spectrum in the angular correlation function of the cosmic microwave background'. Together they form a unique fingerprint.

Cite this