skip to main content
Caltech

LIGO Seminar

Friday, April 26, 2024
11:00am to 12:00pm
Add to Cal
West Bridge 351 (LIGO Science Conference Room)
Modeling and Inference of Environmental Effects in Extreme-Mass-Ratio Inspirals
Shubham Kejriwal,
  • Internal Event

https://caltech.zoom.us/j/87546916051?pwd=aTByZU5WamhUKzlkRVY2bW05bytDdz09

Speaker: Shubham Kejriwal

Title: Modeling and Inference of Environmental Effects in Extreme-Mass-Ratio Inspirals 

Abstract: Extreme-mass-ratio inspirals (EMRIs) are expected to be measured to sub-percent precision by the upcoming space-based gravitational waves (GW) detector LISA. Their inference will provide unique opportunities to test general relativity (GR) and constrain supermassive black hole environments. The precision of measurement for these "beyond vacuum-GR" effects depends on the availability of their representative theoretical models, and while many attempts have been made to incorporate them in current vacuum-GR EMRI models, their description is often limited by their perturbative long-timescale evolution and poor physical understanding. Consequently, most such effects are introduced phenomenologically in our current models, often as simple power-law-like corrections to the vacuum-GR trajectories. However, this can lead to severe correlations in the parameter space, especially in the joint analysis of multiple beyond vacuum-GR effects. In this talk, I will discuss the results of our latest study, formulating the impact of these correlations in a generic mathematical framework in the context of EMRIs. We find that these correlations - found to be strong for a significant fraction of the analyzed population - can severely impact inference, both, when a modeled beyond vacuum-GR effect is absent in the signal, and an unmodeled effect impacts the evolution of the true source. This result challenges the constraints of previous studies and poses a general problem for modeling beyond vacuum-GR effects in EMRIs. Finally, I will briefly highlight some strategies to circumvent the problem, like looking for electromagentic counterparts of EMRIs, or performing hierarchical inference of the EMRI population. 

For more information, please contact Lucy M Thomas by email at [email protected].