Wavelet Denoising of Radio Observations of Rotating Radio Transients (RRATs): Improved Timing Parameters for Eight RRATs

TL;DR

This paper introduces a wavelet-based denoising method for radio observations of RRATs, significantly improving pulse detection and timing accuracy, which aids in understanding their physical nature and relation to other neutron stars.

Contribution

The paper presents a novel wavelet denoising technique tailored for RRAT signals, enhancing pulse detection and timing precision over conventional methods.

Findings

Increased number of detected TOAs by 12% to 69%.

Reduced parameter estimation errors by 16% to 90%.

Wavelet denoising effectively improves RRAT signal analysis.

Abstract

Rotating radio transients (RRATs) are sporadically emitting pulsars detectable only through searches for single pulses. While over 100 RRATs have been detected, only a small fraction (roughly 20\%) have phase-connected timing solutions, which are critical for determining how they relate to other neutron star populations. Detecting more pulses in order to achieve solutions is a key to understanding their physical nature. Astronomical signals collected by radio telescopes contain noise from many sources, making the detection of weak pulses difficult. Applying a denoising method to raw time series prior to performing a single-pulse search typically leads to a more accurate estimation of their times of arrival (TOAs). Taking into account some features of RRAT pulses and noise, we present a denoising method based on wavelet data analysis, an image-processing technique. Assuming that the spin period of an RRAT is known, we estimate the frequency spectrum components contributing to the composition of RRAT pulses. This allows us to suppress the noise, which contributes to other frequencies. We apply the wavelet denoising method including selective wavelet reconstruction and wavelet shrinkage to the de-dispersed time series of eight RRATs with existing timing solutions. The signal-to-noise ratio (S/N) of most pulses are improved after wavelet denoising. Compared to the conventional approach, we measure 12% to 69% more TOAs for the eight RRATs. The new timing solutions for the eight RRATs show 16% to 90% smaller estimation error of most parameters. Thus, we conclude that wavelet analysis is an effective tool for denoising RRATs signal.