Summary of Advancing Solar Flare Prediction Using Deep Learning with Active Region Patches, by Chetraj Pandey et al.
Advancing Solar Flare Prediction using Deep Learning with Active Region Patches
by Chetraj Pandey, Temitope Adeyeha, Jinsu Hong, Rafal A. Angryk, Berkay Aydin
First submitted to arxiv on: 16 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Instrumentation and Methods for Astrophysics (astro-ph.IM); Solar and Stellar Astrophysics (astro-ph.SR)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper introduces a novel methodology for predicting solar flares using shape-based characteristics of magnetograms from active region (AR) patches across the entire solar disk. The authors develop three deep learning models: ResNet34, MobileNet, and MobileViT to predict ≥M-class flares, assessing their efficacy across various ranges of solar longitude. To address data imbalance, they employ augmentation techniques and undersampling during training, maintaining imbalanced testing partitions for realistic evaluation. A composite skill score (CSS) is used as the evaluation metric, ranking models based on True Skill Score (TSS), Heidke Skill Score (HSS), and CSS values. The primary contributions include a novel capability in solar flare prediction, allowing for predicting flares for each AR throughout the solar disk, with MobileNet achieving a CSS of 0.51 for AR patches within ±30°, ±60°, and ±90° of solar longitude. This advancement opens avenues for more reliable solar flare prediction, contributing to improved forecasting capabilities. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about predicting big events on the sun called solar flares. The scientists developed new ways to use data from the sun’s magnetic field to make predictions. They tested three different models and found that one of them was very good at predicting solar flares across the entire sun. This helps us better understand the sun and makes it possible to predict when big events might happen. |
Keywords
* Artificial intelligence * Deep learning
