Complex network view for $\delta$ Scuti stars

TL;DR

This study applies complex network analysis to  Scuti stars' light curves observed by TESS, revealing small-world and non-random properties, and demonstrating the method's effectiveness in preserving stellar pulsation information.

Contribution

It introduces a novel application of complex network theory to analyze  Scuti stars' light curves, uncovering structural properties and efficient data reduction techniques.

Findings

Light curves exhibit small-world and non-random network properties.

Network metrics differentiate HADS and non-HADS star groups.

Reducing window size preserves mode information and amplifies signal over noise.

Abstract

Extraction of characteristics of the complex light curve of pulsating stars is essential for stellar physics. We investigate the complex network (natural and horizontal visibility graphs) properties of the \dsct\ stars light curves observed by \tess. We find that the average shortest path length of \dsct\ light curves is a linear function of the logarithm of the network sizes, indicating the small-world and non-random properties. The small-world property signifies the connection of significant peaks of the light curve with small nearing peaks and other important peaks. The lognormal behavior of nodes' degree is evidence for non-random processes for stellar pulsations. This may be rooted in the different mechanisms of stellar dynamics, such as rotation, excitation of modes, and magnetic activity. The PageRank and nodes' degree distributions of \dsct\ stars collect in two HADS and non-HADS groups. The lower clustering for HADS than non-HADS indicates a more straightforward light curve (containing one or two independent modes) than a more complex light curve (including various oscillation modes) that might be a signature of surface gravity as an indication of stellar evolution. We show that reducing the window size of a light curve to about 5\% of the original one based on the network ranking preserves most of the star modes information. In this case, we also observe that the amplitude of most natural modes amplifies compared to the noise background in the power spectrum. These results indicate the capability of the network approach for interpreting pulsating stars' light curves.