High-pass Filter Periodogram: An Improved Power Spectral Density Estimator for Unevenly Sampled Data

TL;DR

This paper introduces a high-pass filter periodogram that improves power spectral density estimation for unevenly sampled data, enhancing accuracy and reliability over traditional methods like Lomb-Scargle.

Contribution

The paper proposes a novel high-pass filter approach to reduce sampling noise in PSD estimation, offering a more precise tool for analyzing unevenly sampled time series.

Findings

HPF periodogram outperforms Lomb-Scargle in accuracy

Improves periodicity detection in noisy, uneven data

Enhances robustness of spectral analysis methods

Abstract

Accurate time series analysis is essential for studying variable astronomical sources, where detecting periodicities and characterizing power spectral density (PSD) are crucial. The Lomb-Scargle periodogram, commonly used in astronomy for analyzing unevenly sampled time series data, often suffers from noise introduced by irregular sampling. This paper presents a new high-pass filter (HPF) periodogram, a novel implementation designed to mitigate this sampling-induced noise. By applying a frequency-dependent high-pass filter before computing the periodogram, the HPF method enhances the precision of PSD estimates and periodicity detection across a wide range of signal characteristics. Simulations and comparisons with the Lomb-Scargle periodogram demonstrate that the HPF periodogram improves accuracy and reliability under challenging sampling conditions, making it a valuable complementary tool for more robust time series analysis in astronomy and other fields dealing with unevenly sampled data.