How radiatively inefficient is a radiatively-inefficient accretion flow? Using magnetized neutron stars to test accretion models

A striking diversity of astronomical objects are powered by accretion onto a central object: protostars, X-ray binaries, gamma-ray bursts, supermassive black holes. Some of these objects possess magnetic fields strong enough to disrupt the accretion flow as it approaches the star, giving rise to rotationally-induced pulsations, outflows and other variability. In my talk I will describe how a fast-spinning magnetic field can act as a 'dam', creating a reservoir of gas in the inner parts of the accretion flow that alternately builds up and accretes onto the star.The instability produces strong luminosity variations, similar to variations seen in both accreting neutron stars and protostars. This new picture of accretion onto magnetic fields changes our understanding of accretion at very low or high luminosities, when the flow is expected to become "radiatively-inefficient" (that is, most of the accretion energy is not released as radiation).Understanding the physics of radiatively-inefficient flows is critical for a wide range of accreting objects, from gamma-ray bursts to AGN. I will discuss how neutron stars offer unique and under-exploited opportunities to test models of radiatively-inefficient accretion, and I will describe how some recent observations can put constraints on these models.