Learning Inspiral-Merger-Ringdown Waveforms from a Post-Newtonian Baseline
Abstract
Modeling the full inspiral-merger-ringdown signal requires combining analytically controlled inspiral physics with the nonlinear strong-field information supplied by numerical relativity. We represent the numerical relativity contribution beyond an analytic inspiral waveform as residual amplitude and phase corrections. Retaining the leading-order frequency-domain amplitude and the $3.5$ Post-Newtonian TaylorF2 phase, we use a Kolmogorov-Arnold network to learn these residual corrections from SXS waveforms. After training, the learned corrections are stored as explicit spline functions, so waveform evaluation no longer requires the network itself. On $75$ simulations excluded from training and model selection, the model achieves a median flat-noise mismatch of 2.7×10−5. Our results demonstrate a machine learning driven waveform modeling strategy in which numerical relativity augments, rather than replaces, analytically known waveform structure.
Related
- A new method for inspiral-to-ringdown waveforms of compact binaries: Helmholtz orbits and spherical-Bessel flux
- Semi-agnostic inspiral modeling: the massless scalar case at leading post-Newtonian order
- Distinguishing lensing and precessional modulation in binary black-hole inspiral waveforms
- Gravitational-wave parameter estimation with machine-learning generated surrogate waveforms
- The Mysterious Inspiral of WASP-12b: Why Obliquity Tides Cannot Drive Orbital Decay
BibTeX
@misc{chattopadhyay2026learning,
title = {{Learning Inspiral-Merger-Ringdown Waveforms from a Post-Newtonian Baseline}},
author = {Arghya Chattopadhyay and Shilpa Kastha},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.29115},
eprint = {2609.29115},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.29115},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Learning Inspiral-Merger-Ringdown Waveforms from a Post-Newtonian Baseline
AU - Arghya Chattopadhyay
AU - Shilpa Kastha
PY - 2026
DA - 2026/09/24
PB - arXiv
DO - 10.48550/arxiv.2609.29115
UR - https://arxiv.org/abs/2609.29115
AB - Modeling the full inspiral-merger-ringdown signal requires combining analytically controlled inspiral physics with the nonlinear strong-field information supplied by numerical relativity. We represent the numerical relativity contribution beyond an analytic inspiral waveform as residual amplitude and phase corrections. Retaining the leading-order frequency-domain amplitude and the $3.5$ Post-Newtonian TaylorF2 phase, we use a Kolmogorov-Arnold network to learn these residual corrections from SXS waveforms. After training, the learned corrections are stored as explicit spline functions, so waveform evaluation no longer requires the network itself. On $75$ simulations excluded from training and model selection, the model achieves a median flat-noise mismatch of 2.7×10^−5. Our results demonstrate a machine learning driven waveform modeling strategy in which numerical relativity augments, rather than replaces, analytically known waveform structure.
ER -