We discuss the feasibility of using astrophysical observations of white dwarfs as probes of fundamental physics. We quantify the effects of varying fundamental couplings on the white dwarf mass-radius relation in a broad class of unification scenarios, both for the simple case of a polytropic stellar structure model and for more general models. Independent measurements of the mass and radius, together with direct spectroscopic measurements of the fine-structure constant in white dwarf atmospheres lead to constraints on combinations of the two phenomenological parameters describing the underlying unification scenario (one of which is related to the strong sector of the theory while the other is related to the electroweak sector). While currently available measurements do not yet provide stringent constraints, we show that forthcoming improvements, expected for example from the Gaia satellite, can break parameter degeneracies and lead to constraints that ideally complement those obtained from local laboratory tests using atomic clocks.

Mapping the behaviour of dark energy is a pressing task for observational cosmology. Phenomenological classification divides dynamical dark energy models into freezing and thawing, depending on whether the dark energy equation of state is approaching or moving away from w = p/ρ = −1. Moreover, in realistic dynamical dark energy models the dynamical degree of freedom is expected to couple to the electromagnetic sector, leading to variations of the fine-structure constant α. We discuss the feasibility of distinguishing between the freezing and thawing classes of models with current and forthcoming observational facilities and using a parametrisation of the dark energy equation of state, which can have either behaviour, introduced by Mukhanov as fiducial paradigm. We illustrate how freezing and thawing models lead to different redshift dependencies of α, and use a combination of current astrophysical observations and local experiments to constrain this class of models, improving the constraints on the key coupling parameter by more than a factor of two, despite considering a more extended parameter space than the one used in previous studies. We also briefly discuss the improvements expected from future facilities and comment on the practical limitations of this class of parametrisations. In particular, we show that sufficiently sensitive data can distinguish between freezing and thawing models, at least if one assumes that the relevant parameter space does not include phantom dark energy models.