Data analysis of gravitational-wave signals from spinning neutron stars. I. The signal and its detection

TL;DR

This paper develops a comprehensive theoretical framework for detecting gravitational-wave signals from spinning neutron stars, including detailed signal modeling, detection statistics, and analysis methods for Earth-based interferometers.

Contribution

It introduces a detailed signal model with amplitude and frequency modulations, and derives detection procedures requiring multiple linear filters, advancing gravitational-wave data analysis techniques.

Findings

Detection statistics depend on the optimal signal-to-noise ratio.

Four linear filters are needed for amplitude-modulated signals, doubled for the frequency component.

Monte Carlo simulations evaluate detection prospects for current and future interferometers.

Abstract

We present a theoretical background for the data analysis of the gravitational-wave signals from spinning neutron stars for Earth-based laser interferometric detectors. We introduce a detailed model of the signal including both the frequency and the amplitude modulations. We include the effects of the intrinsic frequency changes and the modulation of the frequency at the detector due to the Earth motion. We estimate the effects of the star's proper motion and of relativistic corrections. Moreover we consider a signal consisting of two components corresponding to a frequency $f$ and twice that frequency. From the maximum likelihood principle we derive the detection statistics for the signal and we calculate the probability density function of the statistics. We obtain the data analysis procedure to detect the signal and to estimate its parameters. We show that for optimal detection of the amplitude modulated signal we need four linear filters instead of one linear filter needed for a constant amplitude signal. Searching for the doubled frequency signal increases further the number of linear filters by a factor of two. We indicate how the fast Fourier transform algorithm and resampling methods commonly proposed in the analysis of periodic signals can be used to calculate the detection statistics for our signal. We find that the probability density function of the detection statistics is determined by one parameter: the optimal signal-to-noise ratio. We study the signal-to-noise ratio by means of the Monte Carlo method for all long-arm interferometers that are currently under construction. We show how our analysis can be extended to perform a joint search for periodic signals by a network of detectors and we perform Monte Carlo study of the signal-to-noise ratio for a network of detectors.