Multi-channel coronal hole detection with convolutional neural networks

Abstract

Context. A precise detection of the coronal hole boundary is of primary interest for a better understanding of the physics of coronal holes, their role in the solar cycle evolution, and space weather forecasting. Aims. We develop a reliable, fully automatic method for the detection of coronal holes that provides consistent full-disk segmentation maps over the full solar cycle and can perform in real-time. Methods. We use a convolutional neural network to identify the boundaries of coronal holes from the seven extreme ultraviolet (EUV) channels of the Atmospheric Imaging Assembly (AIA) and from the line-of-sight magnetograms provided by the Helioseismic and Magnetic Imager (HMI) on board the Solar Dynamics Observatory (SDO). For our primary model (Coronal Hole RecOgnition Neural Network Over multi-Spectral-data; CHRONNOS) we use a progressively growing network approach that allows for efficient training, provides detailed segmentation maps, and takes into account relations across the full solar disk. Results. We provide a thorough evaluation for performance, reliability, and consistency by comparing the model results to an independent manually curated test set. Our model shows good agreement to the manual labels with an intersection-over-union (IoU) of 0.63. From the total of 261 coronal holes with an area > 1.5 • 10 10 km 2 identified during the time-period from November 2010 to December 2016, 98.1% were correctly detected by our model. The evaluation over almost the full solar cycle no. 24 shows that our model provides reliable coronal hole detections independent of the level of solar activity. From a direct comparison over short timescales of days to weeks, we find that our model exceeds human performance in terms of consistency and reliability. In addition, we train our model to identify coronal holes from each channel separately and show that the neural network provides the best performance with the combined channel information, but that coronal hole segmentation maps can also be obtained from line-of-sight magnetograms alone. Conclusions. The proposed neural network provides a reliable data set for the study of solar-cycle dependencies and coronal-hole parameters. Given the fast and robust coronal hole segmentation, the algorithm is also highly suitable for real-time space weather applications.

Identification of Coronal Holes on AIA/SDO images using unsupervised Machine Learning

“…The CATCH data in this period is reliable with minimal uncertainties. The total of 237 CATCH CH binary maps consist only contributions from the longitudinal range of [−400, 400] arcseconds in helioprojective coordinates as in this region the CHs can be identified more robustly (Jarolim et al 2021). We also imported CH polygons from the HEK database for the same dates as the CATCH maps, and converted them into binary maps.…”

“…They trained their network using binary maps from Kislovodsk Mountain Astronomical Station. Recently, Jarolim et al (2021) utilized a progressively growing architecture based CNNs using data from all 7 channels of AIA/SDO (94 Å, 131 Å, 171 Å, 193 Å, 211 Å, 304 Å and 335 Å ) as well as line-of-sight magnetograms from Helioseismic and Magnetic Imager (HMI; Scherrer et al 2012) on the SDO. For their network, the authors used binary maps from manually reviewed SPoCA-CH (Delouille et al 2018).…”

SCSS-Net: Solar Corona Structures Segmentation by Deep Learning

“…Now we would like to discuss the comparison of our approach with the previously mentioned CNN-based methods in Illarionov & Tlatov (2018) and Jarolim et al (2021). From the general perspective, we aimed to design SCSS-Net more universally and to investigate if it is able to adapt to different segmentation goals and to segment different types of coronal structures.…”

“…Thanks to their combination, final models can be more flexible and better trained to avoid biases within particular annotation sets. The CHRONOS method, described recently in Jarolim et al (2021), is inspired by the generative approaches and adopts an idea of a progressively growing network. In this case, this idea is applied to convolutional layers combined in similar encoder-decoder architecture as U-Net segmentation networks.…”

“…Illarionov & Tlatov (2018) demonstrated that for the segmentation of coronal holes, CNN provides quantitatively comparable results as the SPoCA and CHIMERA algorithms. Jarolim et al (2021) provided a method called CHRONOS based on CNN with progressively growing network approach for robust segmentation of coronal holes.…”