Welcome to the 10th International LISA Symposium

A space-based millihertz gravitational wave detector like LISA or eLISA is expected to observe tens to hundreds of inspirals of compact objects into supermassive black holes (EMRIs). Each system can be observed for several years prior to plunge, during which time the inspiraling object completes hundreds of thousands of orbits in the strong-field region of the spacetime. The emitted waves encode the details of these orbits which in turn provide a comprehensive map of the strong-field space-time structure. These maps allow the system parameters to be measured to unprecedented precision while also allowing very small deviations from the predictions of general relativity to be identified in the signals. In this talk I will describe some of the scientific applications of these EMRI observations to astrophysics, fundamental physics and cosmology. These applications include measuring the currently unknown local slope of the black hole mass function, determining the Hubble constant and testing that astrophysical black holes satisfy the no hair property predicted in general relativity.

To be submitted

Modeling extreme-mass-ratio inspirals poses two basic challenges: capturing the essential physics of the smaller object without having to resolve the details of its internal structure; and capturing orbital dynamics on both the short timescale of a single orbit and the long timescale of a complete inspiral. At a formal level, the gravitational self-force formalism, which treats the smaller object as a source of perturbation of the spacetime of the larger, has met these challenges and is now accurate enough to extract system parameters from an observed waveform. However, many tasks remain in the numerical implementation of that formalism. Principal among these are incorporating (i) the spin of the central black hole and (ii) various effects that are second order in the mass ratio. In lieu of such a highly accurate model, simplified ones have been formulated that take into account the dominant self-force effects; while these models are unlikely to be accurate enough for parameter estimation, they may be sufficient for signal detection in many cases.