Image Details

Choose export citation format:

Short-term Forecasting of Total Solar Irradiance from Solar Disk Images: A Hybrid Unsupervised Deep Learning Approach

  • Authors: Idowu Raji, Rafael Duarte Coelho dos Santos, Luis Eduardo Antunes Vieira, Alejandro C. Frery, Franciele Carlesso, Adriany Rodrigues Barbosa

Idowu Raji et al 2026 The Astrophysical Journal Supplement Series 286 .

  • Provider: AAS Journals

Caption: Figure 3.

SOM pixel-wise clustering of the three-channel solar disk input (continuum intensity, unsigned magnetogram, and μ map) into a cluster-index map. Each pixel is assigned to its best matching unit (BMU) based on the Euclidean distance between its [Ipq,  Bpq,  μpq] vector and the learned SOM weight vectors, yielding an unsupervised segmentation of the disk into 400 photometric–magnetic surface regimes without predefined contrast thresholds.

Other Images in This Article

Show More

Copyright and Terms & Conditions

Additional terms of reuse