Supernovae as cosmological probes

TL;DR

This paper analyzes a large supernova dataset to evaluate its evidence for an accelerating universe, revealing only mild support for acceleration under standard assumptions.

Contribution

It provides a rigorous statistical analysis of an expanded supernova dataset, highlighting subtleties in calibration and questioning the strength of evidence for cosmic acceleration.

Findings

Larger supernova dataset shows only mild evidence for acceleration

Calibration subtleties impact the analysis results

Supports the need for careful statistical treatment in cosmological inference

Abstract

The cosmological standard model at present is widely accepted as containing mainly things we do not understand. In particular the appearance of a Cosmological Constant, or dark energy, is puzzling. This was first inferred from the Hubble diagram of a low number of Type Ia supernovae, and later corroborated by complementary cosmological probes. Today, a much larger collection of supernovae is available, and here I perform a rigorous statistical analysis of this dataset. Taking into account how the supernovae are calibrated to be standard candles, we run into some subtleties in the analysis. To our surprise, this new dataset - about an order of bigger than the size of the original dataset - shows, under standard assumptions, only mild evidence of an accelerated universe.