Limiting alternative theories of gravity using gravitational wave observations across the spectrum

Abstract

The advent of gravitational wave astronomy provides new proving grounds for testing theories of gravity. Recent work has reinvigorated the study of bimetric theories of gravity and massive gravity theories. One of the most interesting predictions of these theories, as well as some string theories, is the subluminal speed of propagating gravitational waves. Multi-messenger astronomy provides a unique opportunity to put limits on the difference (either positive or negative) between the propagation speed of electromagnetic and gravitational waves from these sources. This paper considers limits from multi-messenger cases across the planned measurable spectrum: first, the limits from isolated pulsars based on the current best limits from LIGO on gravitational wave emission, second, the limits from ultra-compact binaries that will be visible to a low-frequency space-based gravitational wave observatory like LISA, and third, limits from super massive black hole binaries using pulsar timing arrays. The required phase comparison between the electromagnetic signal and the gravitational wave signal is derived and, assuming a null result in that comparison, the current bounds on emission are used to place limits on alternative theories that exhibit propagation delays. Observations of the pulsars in the most sensitive range of LIGO could put an upper limit on the graviton mass as low as $10^{-38}\frac{eV}{c^{2}}$ and an upper limit on the fractional difference between the gravitational wave and electromagnetic wave speeds as low as $10^{-9}$. This paper shows results from the initial LIGO limit catalog of known pulsars. The bounds are stronger for binaries. A LISA-like mission bounds $m_{g}<10^{-40}\frac{eV}{c^{2}}$ and $\delta<10^{-12}$. A PTA source gives even better bounds of $m_{g}<10^{-45}\frac{eV}{c^{2}}$ and $\delta<10^{-14}$.