Automated Detection Methods for Solar Activities and an Application for Statistic Analysis of Solar Filament

TL;DR

This paper reviews automated detection methods for solar activities, introduces a new versatile algorithm for solar filaments, and applies it to analyze data over three solar cycles, providing statistical insights.

Contribution

The paper presents a novel automated detection method for solar filaments that can recognize, analyze, and trace their evolution over time.

Findings

Effective detection and feature extraction of solar filaments.

Successful application to three solar cycles of H-alpha data.

Statistical analysis of filament evolution over time.

Abstract

With the rapid development of telescopes, both temporal cadence and the spatial resolution of observations are increasing. This in turn generates vast amount of data, which can be efficiently searched only with automated detections in order to derive the features of interest in the observations. A number of automated detection methods and algorithms have been developed for solar activities, based on the image processing and machine learning techniques. In this paper, after briefly reviewing some automated detection methods, we describe our efficient and versatile automated detection method for solar filaments. It is able not only to recognize filaments, determine the features such as the position, area, spine, and other relevant parameters, but also to trace the daily evolution of the filaments. It is applied to process the full disk H-alpha data observed in nearly three solar cycles, and some statistic results are presented.