PhenomXO4a: a phenomenological gravitational-wave model for precessing black-hole binaries with higher multipoles and asymmetries

TL;DR

PhenomXO4a is a novel frequency-domain gravitational waveform model that incorporates multipole asymmetries and precession effects, improving accuracy for precessing black-hole binaries with higher multipoles.

Contribution

It introduces the first frequency-domain model including multipole asymmetries and precession angles tuned to numerical relativity, extending previous models with new features.

Findings

Model retains accuracy of PhenomPNR for dominant multipoles.

Current models have comparable average accuracy across parameter space.

Parameter estimation biases are small despite pointwise accuracy variations.

Abstract

In this work we introduce PhenomXO4a, the first phenomenological, frequency-domain gravitational waveform model to incorporate multipole asymmetries and precession angles tuned to numerical relativity. We build upon the modeling work that produced the PhenomPNR model and incorporate our additions into the IMRPhenomX framework, retuning the coprecessing frame model and extending the tuned precession angles to higher signal multipoles. We also include, for the first time in frequency-domain models, a recent model for spin-precession-induced multipolar asymmetry in the coprecessing frame to the dominant gravitational-wave multipoles. The accuracy of the full model and its constituent components is assessed through comparison to numerical relativity and numerical relativity surrogate waveforms by computing mismatches and performing parameter estimation studies. We show that, for the dominant signal multipole, we retain the modeling improvements seen in the PhenomPNR model. We find that the relative accuracy of current full IMR models varies depending on location in parameter space and the comparison metric, and on average they are of comparable accuracy. However, we find that variations in the pointwise accuracy do not necessarily translate into large biases in the parameter estimation recoveries.