Multi-parameter tests of general relativity using Bayesian parameter estimation with principal component analysis for LISA

Abstract

In the near future, space-borne gravitational wave (GW) detector LISA can open the window of low-frequency band of GW and provide new tools to test gravity theories. In this work, we consider multi-parameter tests of GW generation and propagation where the deformation coefficients are varied simultaneously in parameter estimation and the principal component analysis (PCA) method are used to transform posterior samples into new bases for extracting the most informative components. The dominant components can be more sensitive to potential departures from general relativity (GR). We extend previous works by employing Bayesian parameter estimation and performing both tests with injections of GR and injections of subtle GR-violated signals. We also apply multi-parameter tests with PCA in the phenomenological test of GW propagation. This work complements previous works and further demonstrates the enhancement provided by the PCA method. Considering a supermassive black hole binary system as the GW source, we show that subtle departures will be more obvious in posteriors of PCA parameters. The departures less than 1σ in original parameters can yield significant departures in first 5 dominant PCA parameters.

Keywords

• Gravitational wave
• General relativity
• Bayesian parameter estimation