Prediction of Solar Flares from a Statistical Analysis of Events during Solar Cycle 23

TL;DR

This paper analyzes solar cycle 23 to identify statistical patterns and active region features that can improve medium- and short-term solar flare predictions.

Contribution

It introduces a statistical framework linking flare periodicity and active region magnetic features to enhance flare forecasting accuracy.

Findings

Identified 13.2 and 26.4-month periodicities in flare occurrence.

Active regions with specific magnetic configurations show high flare productivity.

Sunspot rotation correlates with over 66% probability of energetic flare production.

Abstract

Ways to give medium- and short-term predictions of solar flares are proposed according to the statistical analysis of events during solar cycle 23. On one hand, the time distribution of both C and M class flares shows two main periods of 13.2 and 26.4 months in this cycle by wavelet analysis. On the other hand, active regions of specific magnetic configurations and their evolutions give high productivity of C class flares but relatively low productivity of energetic (M and X class) flares. Furthermore, by considering the measurable kinetic features of active regions, i.e., the rotation of the sunspots, some active regions of specified types are observed to have high energetic flare productivity, above 66%. The periodicity of the activity revealed can be used for medium-term C and M class flare forecasting and the high productivity of active regions forms the basis for short-term prediction of individual energetic flares.