Local gravitational redshifts can bias cosmological measurements

TL;DR

Local gravitational redshifts, caused by our galaxy's gravitational potential, can introduce systematic errors in cosmological measurements, affecting parameters like dark energy properties, especially as observational precision improves.

Contribution

This paper quantifies the bias introduced by local gravitational redshifts in cosmological parameter estimation and highlights its significance for future high-precision surveys.

Findings

Ignoring local gravitational redshift causes ~1% systematic error in density parameters.

Local gravitational potential impacts dark energy equation of state measurements.

The effect is significant for future percent-level accuracy in cosmology.

Abstract

Measurements of cosmological parameters via the distance-redshift relation usually rely on models that assume a homogenous universe. It is commonly presumed that the large-scale structure evident in our Universe has a negligible impact on the measurement if distances probed in observations are sufficiently large (compared to the scale of inhomogeneities) and are averaged over different directions on the sky. This presumption does not hold when considering the effect of the gravitational redshift caused by our local gravitational potential, which alters light coming from all distances and directions in the same way. Despite its small magnitude, this local gravitational redshift gives rise to noticeable effects in cosmological inference using SN Ia data. Assuming conservative prior knowledge of the local potential given by sampling a range of gravitational potentials at locations of Milky-Way-like galaxies identified in cosmological simulations, we show that ignoring the gravitational redshift effect in a standard data analysis leads to an additional systematic error of ~1 per cent in the determination of density parameters and the dark energy equation of state. We conclude that our local gravitational field affects our cosmological inference at a level that is important in future observations aiming to achieve percent-level accuracy.