Spline Based Search Method For Unmodeled Transient Gravitational Wave Chirps

TL;DR

This paper introduces a spline-based method for detecting and estimating unmodeled transient gravitational wave chirps with smoothly evolving amplitude and frequency, improving detection in noisy data.

Contribution

It presents a novel spline-based approach using penalized least squares for unmodeled transient signals in gravitational wave data.

Findings

Performs well at low signal-to-noise ratios

Extends detection capabilities to a wider class of signals

Effective in highly degraded time-frequency signatures

Abstract

A method is described for the detection and estimation of transient chirp signals that are characterized by smoothly evolving, but otherwise unmodeled, amplitude envelopes and instantaneous frequencies. Such signals are particularly relevant for gravitational wave searches, where they may arise in a wide range of astrophysical scenarios. The method uses splines with continuously adjustable breakpoints to represent the amplitude envelope and instantaneous frequency of a signal, and estimates them from noisy data using penalized least squares and model selection. Simulations based on waveforms spanning a wide morphological range show that the method performs well in a signal-to-noise ratio regime where the time-frequency signature of a signal is highly degraded, thereby extending the coverage of current unmodeled gravitational wave searches to a wider class of signals.