Discovering Dual-Origin Slow Wind from Solar Orbiter with Self-Supervised Contrastive Learning

Henry Han; Jorge Yero Salazar

astro-ph.SRAug 22, 2026

Discovering Dual-Origin Slow Wind from Solar Orbiter with Self-Supervised Contrastive Learning

Authors: Henry Han, Jorge Yero Salazar

Organizations: Department of Computer Science, Baylor University, One Bear Place, Waco, TX 76798, USA

Abstract

Whether the slow solar wind originates from one coronal source or two distinct channels remains a central open question in heliophysics. Resolving this requires unsupervised separation of two populations that arrive at nearly the same bulk speed and differ mainly in heavy-ion composition. We present Solar-CDC, a self-supervised contrastive deep clustering (CDC) framework that maps plasma observables to a latent space via a Transformer encoder, optimizes a triplet margin loss, and updates pseudo-labels via $k$ -means. Theoretically, we prove that neighborhood-preserving embeddings such as t-SNE and UMAP are fundamentally constrained. Preserving the neighbor graph leaves the cross-cluster cut fraction unchanged, and preserving all but a fraction $\varepsilon$ of its links moves that fraction by at most $\varepsilon$ . Neither bound depends on the target dimension. A margin objective rewrites the graph and drives the cut fraction to zero. Empirically, on 30,602 Solar Orbiter observations, thirty combinations of dimensionality reduction and clustering peak at a silhouette of $0.454$ , whereas Solar-CDC reaches $0.869$ . Escaping the geometric bound alone does not guarantee physical validity: TriMap also optimizes triplets and reaches $0.824$ , yet its clusters score below chance against the published composition taxonomy. Solar-CDC instead recovers clusters with mean charge-state ratios of $0.080$ , $0.160$ , and $0.400$ , placing the intermediate population inside the window associated with coronal-hole boundaries. Even when the defining charge-state ratio is withheld from the inputs entirely, the model still recovers the taxonomy defined on it. Solar-CDC thus connects self-supervised representation learning to coronal source diagnostics. Importantly, a learning loss recovers physical populations only when driven by dynamically updated physically-aware clusters rather than distances.

Figures & tables

Figure 1: Why a threshold on one variable is not enough. (a) The $\mathrm{O}^{7+}/\mathrm{O}^{6+}$ distribution over the 30,602 Solar Orbiter observations, with the coronal-hole ( $0.145$ ) and streamer-belt ( $0.20$ ) reference values marked. (b) Bulk speed against composition: the two slow-wind channels are stacked in composition at essentially the same speed, so a velocity cut cannot separate them.

Figure 2: Solar-CDC. Triplets drawn from the current pseudo-labels pass through the shared encoder; the classification head is trained on those pseudo-labels while the triplet term acts on the latent codes, and both gradients reach the encoder. The lower half of the diagram is the part that makes the method self-supervised: the pseudo-labels are not given, they are produced by $k$ -means, on the observables at epoch 0 and on the latent codes every $p$ epochs thereafter, and they decide in turn which observations are drawn as positives and negatives.

Panel of Figure 3 — Figure 3: The two alternating phases of training. The regular step refines the encoder given fixed pseudo-labels; the reassignment step updates the pseudo-labels given the improved encoder.

Figure 4: Warmup: pseudo-labels are initialized by $k$ -means on the standardized observables, bypassing the randomly initialized encoder.

neighborhood preserving
Space	$\beta_{10}$ , recovered partition	$\beta_{10}$ , threshold partition
Input, seven observables	0.064	0.131
t-SNE [ 12 ] ( $q=2$ )	0.071	0.158
UMAP [ 14 ] ( $q=2$ )	0.084	0.197
PHATE [ 16 ] ( $q=2$ )	0.148	0.228
PCA [ 15 ] ( $q=2$ )	0.174	0.215

Table 1: Cut fraction

\beta_{10}

(

\kappa=10

8{,}000

-point subsample). As Props. 1 – 2 require, neighborhood-preserving methods stay within

\varepsilon

of the input metric, and none improves on it; Solar-CDC reduces it to near zero.

	Latent dimension $m$
Heads	2	4	8	16	32
1	0.869	0.828	0.820	0.755	0.725
2	0.829	0.806	0.773	0.748	0.764
4	0.866	0.814	0.792	0.752	0.728
8	0.840	0.821	0.809	0.756	0.772
16	0.832	0.834	0.779	0.775	0.740

Table 2: Silhouette over the Solar-CDC hyperparameter grid (30,602 observations, seven observables,

k=3

, 200 epochs, pseudo-label period

p=4

, Euclidean triplet distance). All 30 runs converged to three non-empty clusters.

neighborhood preserving
Reduction	K-Means	GMM	Agglomerative
PCA [ 15 ] (2-D)	0.400	0.360	0.414
t-SNE [ 12 ] ( $p{=}30$ )	0.428	0.427	0.388
t-SNE ( $p{=}100$ )	0.438	0.424	0.422
t-SNE ( $p{=}200$ )	0.450	0.440	0.372
UMAP [ 14 ] (15 nn)	0.443	0.452	0.420

Table 3: Silhouette of each dimensionality-reduction baseline inside its own two-dimensional embedding (8,000-point subsample; every method here is superlinear in

N

), against Solar-CDC in its learned representation. The grouping is the one Sec. 5 predicts: the methods that aim to preserve neighborhoods sit together, and the one that optimizes triplet order does not.

$k$	Silhouette	vs. $k=3$
2	$0.856\pm 0.033$	$p=0.42$
3	$0.843\pm 0.035$	—
4	$0.765\pm 0.044$	$p=3.9\times 10^{-4}$
5	$0.688\pm 0.089$	$p=2.6\times 10^{-4}$

Table 4: Cluster-count sweep, ten seeds per

k

(mean

\pm

s.d.), with the silhouette computed exactly on all

30{,}602

observations. Both

k=2

and

k=3

are significantly better than

k=4

and

k=5

; they are not distinguishable from each other.

Method	Inputs	NMI	ARI	Matched acc.
Random labeling	—	0.000	0.000	0.337
Largest class only	—	0.000	0.000	0.501
TriMap [ 1 ] + Agglom.	seven	0.036	$-0.021$	0.465
$k$ -means	seven	0.348	0.284	0.628
Solar-CDC	seven	$\mathbf{0.422\pm 0.079}$	$0.385\pm 0.088$	$\mathbf{0.706\pm 0.046}$
$k$ -means, control	six	0.311	0.259	0.613

Table 5: Agreement with the external

\mathrm{O}^{7+}/\mathrm{O}^{6+}

taxonomy of [ 21 , 5 , 9 ] , defined by published thresholds rather than by archived class labels, and unseen during training. Solar-CDC values are means over five seeds (16 heads,

m=2

). The 6-observable control is trained without

\mathrm{O}^{7+}/\mathrm{O}^{6+}

and scored against the same reference.

Figure 5: The three recovered clusters in learned and physical spaces (colors consistent throughout; $6{,}000$ random observations sampled for legibility). (a) The 2-D latent codes. (b) Bulk speed vs. $\mathrm{O}^{7+}/\mathrm{O}^{6+}$ , with dashed lines marking the published coronal-hole ( $0.145$ ) and streamer-belt ( $0.20$ ) thresholds. The two slow-wind populations share an identical speed range but divide sharply at the compositional boundary. (c) Correlated charge-state ratios. The separation spans the broader composition space rather than relying on a single variable. (d) Bulk speed vs. density. The two slow clusters perfectly interleave. Together, panels (b) and (d) visualize the core premise of our method: the distinct origins of the slow wind are clearly separable by composition, yet completely invisible to classical kinematic properties.

Cluster	$n$	$\bar{v}_{p}$	$\bar{N}_{p}$	$\overline{\mathrm{O}^{7+}\!/\mathrm{O}^{6+}}$	$\overline{\mathrm{C}^{6+}\!/\mathrm{C}^{4+}}$	$\overline{\mathrm{Fe/O}}$	$\overline{Q_{O}}$
0 coronal hole	10,415	526	14.2	0.080	2.14	0.146	6.07
1 boundary type	12,581	374	28.1	0.160	4.37	0.154	6.15
2 streamer belt	0 7,606	379	32.3	0.400	11.00	0.220	6.39

Table 6: Mean plasma properties per cluster, ordered by charge state (16 heads,

m=2

). Cluster 0 isolates the fast wind, while Clusters 1 and 2 partition the slow wind. This compositional profile represents the 11 of 24 runs that successfully resolve the boundary population. For physical context, established

\mathrm{O}^{7+}/\mathrm{O}^{6+}

thresholds are:

<0.145

(coronal-hole interior),

0.145

–

0.20

(boundary), and

>0.20

(streamer belt) [ 9 , 21 , 5 ] . Note: Silhouette scores here are computed across the full dataset rather than the held-out split, precluding direct comparison with Table 2 .

Figure 6: Distribution of each observable per cluster (boxes: quartiles; whiskers:

1.5\times

IQR; outliers omitted). Separation is monotone and largest in the charge-state variables, and smallest in bulk speed, where clusters 1 and 2 overlap almost completely.

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

	seven observables					with time and label columns
Heads	2	4	8	16	32	2	4	8	16	32
1	0.869	0.828	0.820	0.755	0.725	0.857	0.803	0.811	0.746	0.687
2	0.829	0.806	0.773	0.748	0.764	0.842	0.783	0.804	0.753	0.730
4	0.866	0.814	0.792	0.752	0.728	0.853	0.786	0.790	0.721	0.714
8	0.840	0.821	0.809	0.756	0.772	0.850	0.762	0.796	0.750	0.689
16	0.832	0.834	0.779	0.775	0.740	0.846	0.818	0.799	0.746	0.703

Appendix

Table 7: Silhouette over the full grid, with the seven observables and with the two extra columns of Sec. 6.1 appended to the input (30,602 observations,

k=3

, 200 epochs,

p=4

Solar Physics

See all 15 papers

Oct 3, 2026

Cross-Modal Solar Image Synthesis: Adapting the Surya Foundation Model from He I 10830 Å to EUV Translation and Coronal Hole Segmentation

The long observational record of He I 10830 Å offers a means to investigate solar morphology before modern extreme-ultraviolet (EUV) imaging. We adapt the Surya solar foundation model to predict Solar Dynamics Observatory/Atmospheric Imaging Assembly (SDO/AIA) 94, 193, and 304 Å images and a coronal hole (CH) probability map from full-disk helium observations. A convolutional input adapter, low-rank backbone updates, and dedicated output decoders learn from temporally paired, geometrically registered observations, with Spatial Possibilistic Clustering Algorithm (SPOCA) catalog polygons providing CH supervision. On observations held out from downstream fine-tuning, the selected dedicated models achieve disk-restricted correlation coefficients (CCs) of 0.8196, 0.8885, and 0.8398 for the three AIA channels, respectively; the CH model achieves an intersection over union (IoU) of 0.4360. The predictions recover broad solar structure, although local agreement varies substantially by channel. An optional residual refiner addresses spatial detail, and its application on pre-SDO dates improves the correlation of synthetic AIA 304 with Solar and Heliospheric Observatory/Extreme-ultraviolet Imaging Telescope (SOHO/EIT) 304 references from 0.6511 to 0.7085. Together, these results support the feasibility of helium-conditioned EUV morphological proxies and motivate their use in historical reconstruction, within the scope of the downstream test and cross-instrument evaluation.

Marco Marena, Andrés Muñoz Jaramillo, Qin Li +8

New Jersey Institute of Technology, Newark, NJ · SouthWest Research Institute, Boulder, CO · Sam Houston State University, Huntsville, TX +1

Solar Physics Image Generation

Ask PDF

Sep 23, 2026

PISCES: Physics-Informed Solar-wind Convolutional autoEncoder for Space-weather Anomaly Detection and Early Warning

Space weather early warning depends on detecting solar wind transients in in-situ measurements at the first Sun-Earth Lagrange point (L1), before they reach Earth. Fixed thresholds can miss combined magnetic and plasma structure, and many learning methods provide a single anomaly score. We present the Physics-Informed Solar-wind Convolutional autoEncoder for Space-weather (PISCES), a convolutional autoencoder trained without catalog labels on OMNI solar wind measurements under physics constraints. Its loss includes magnetic field consistency, an empirical relation between temperature and velocity, the Parker spiral angle, and penalties on changes between consecutive one-minute samples in derived quantities calculated from the reconstruction. At inference, PISCES separates the anomaly score into magnetic and plasma reconstruction errors, physics relations, and residual corrections, and reports the magnitude of each contribution. Attenuation of the skip connections, selected on validation data, improves average precision for the trained models, while the untrained scores remain nearly the same. The trained models also give a more consistent ordering of these physical contributions. After smoothing with a trailing median, the alarms can precede independently observed sudden commencements, including positive sudden impulses.

Kevin Lee, Alison J. March

NASA Jet Propulsion Laboratory, La Cañada Flintridge, CA, USA · Department of Mechanical and Aerospace Engineering, UCLA, Los Angeles, CA, USA · College of Engineering & Applied Science, University of Colorado, Boulder, CO, USA

Interpretable Anomaly Detection Convolutional Autoencoder Unsupervised Anomaly Detection Physics-Informed ML Anomaly Detection Solar Physics

Ask PDF

Apr 18, 2026

Automated Classification of Plasma Regions at Mars Using Machine Learning

The plasma environment around Mars is highly variable because it is strongly influenced by the solar wind. Accurate identification of plasma regions around Mars is important for the community studying solar wind-Mars interactions, region-specific plasma processes, and atmospheric escape. In this study, we develop a machine-learning-based classifier to automatically identify three key plasma regions--solar wind, magnetosheath, and induced magnetosphere--using only ion omnidirectional energy spectra measured by the MAVEN Solar Wind Ion Analyzer (SWIA). Two neural network architectures are evaluated: a multilayer perceptron (MLP) and a convolutional neural network (CNN) that incorporates short temporal sequences. Our results show that the CNN can reliably distinguish the three plasma regions, whereas the MLP struggles to separate the solar wind and magnetosheath. Therefore, the CNN-based approach provides an efficient and accurate framework for large-scale plasma region identification at Mars and can be readily applied to future planetary missions.

Yilan Qin, Chuanfei Dong, Hongyang Zhou +6

Department of Electrical and Computer Engineering, Boston University, Boston, MA 02215, USA · Center for Space Physics, Boston University, Boston, MA 02215, USA · School of Natural Sciences, Institute for Advanced Study, Princeton, NJ 08540, USA +1

Topic pending

Ask PDF

Abstract

Solar Physics

Explore similar work

Cross-Modal Solar Image Synthesis: Adapting the Surya Foundation Model from He I 10830 Å to EUV Translation and Coronal Hole Segmentation

PISCES: Physics-Informed Solar-wind Convolutional autoEncoder for Space-weather Anomaly Detection and Early Warning

Automated Classification of Plasma Regions at Mars Using Machine Learning