arXiv subjects

John Stefan

Publications and source records attributed to John Stefan.

8 recordsLinked to original sources

arXiv2026

Prediction of Magnetic Flux Evolution During Solar Active Region Emergence using Long Short-Term Memory Networks

Solar active regions (ARs) are the primary drivers of space weather events, making their early prediction crucial for operational forecasting systems. We develop machine learning models capable of predicting the evolution of magnetic flux during AR emergence using 1D time series of the continuum intensity and solar oscillation power maps for 53 active regions and their surrounding quiet-Sun areas. Each observable is sampled over a fixed 30.66{\deg}x30.66{\deg} field of view. These observations capture the temporal evolution of each active region and serve as inputs for training and validation of our MagFluxLSTM and MagFluxEnc-Dec models. The MagFluxLSTM architecture implements a single-stage standard Long-Short Term Memory (LSTM) network. MagFluxEnc-Dec represents an LSTM encoder-decoder with teacher forcing. To test and evaluate the models' performance, we use the continuum intensity and oscillation power maps (calculated for several frequency bands from Doppler velocity) as input to predict the magnetic flux. Among the top 100 hyperparameter configurations ranked by validation derivative RMSE, 98% correspond to MagFluxLSTM, compared to only 2% for MagFluxEnc-Dec. Thus, although the MagFluxEnc-Dec architecture has higher model complexity, it leads to poorer generalization to ARs outside the training set and less stable training than the simpler MagFluxLSTM, which can predict magnetic flux emergence 3-10 hours in advance within a 12-hour prediction window in both experimental and operational-type settings for the 5 testing active regions.

astro-ph.SR↗

arXiv2026

Origin of Coronal Extreme Ultraviolet Shockwaves without a Coronal Mass Ejection Event

A leading theory of sunquake generation involves flare-accelerated particles depositing energy into the photosphere. Simulations of sunquake excitation suggest co-excitation with wavefronts propagating in the corona, similar to large-scale coronal propagating fronts (LCPFs), and also generate Moreton-Ramsey waves in the chromosphere. To investigate observational evidence for the particle-driven mechanism in LCPFs, we compare populations of events associated with and without coronal mass ejections (CMEs). CMEs are known to generate coronal shock waves also observed in EUV emission. We employ visual inspection of flare events that generate LCPFs using Atmospheric Imaging Assembly (AIA) and Large Angle and Spectrometric Coronagraph (LASCO) coronagraph images to find that the large-scale coronal waves associated with CMEs propagate noticeably faster. Then we examine standalone flare events (those that generate coronal waves without CMEs), using soft X-ray (SXR) data from the GOES satellite and focusing on characteristics related to magnetic energy release rate. This reveals that such standalone or confined flares differ from sunquake flares: they are less impulsive and energetic than sunquake flares. However, they are more impulsive but less energetic than LCPF-associated flares with a CME. In particular, coronal waves accompanied by CMEs exhibit significantly higher volume emission measures, suggesting a different generation mechanism.

astro-ph.SR↗

arXiv2026

Forecasting Continuum Intensity for Solar Active Region Emergence Prediction using Transformers

Early and accurate prediction of solar active region (AR) emergence is crucial for space weather forecasting. Building on established Long Short-Term Memory (LSTM) based approaches for forecasting the continuum intensity decrease associated with AR emergence, this work expands the modeling with new architectures and targets. We investigate a sliding-window Transformer architecture to forecast continuum intensity evolution up to 12 hours ahead using data from 46 ARs observed by SDO/HMI. We conduct a systematic ablation study to evaluate two key components: (1) the inclusion of a temporal 1D convolutional (Conv1D) front-end and (2) a novel `Early Detection' architecture featuring attention biases and a timing-aware loss function. Our best-performing model, combining the Early Detection architecture without the Conv1D layer, achieved a Root Mean Square Error (RMSE) of 0.1189 (representing a 10.6% improvement over the LSTM baseline) and an average advance warning time of 4.73 hours (timing difference of -4.73h), even under a stricter emergence criterion than previous studies. While the Transformer demonstrates superior aggregate timing and accuracy, we note that this high-sensitivity detection comes with increased variance compared to smoother baseline models. However, this volatility is a necessary trade-off for operational warning systems: the model's ability to detect micro-changes in precursor signals enables significantly earlier detection, outweighing the cost of increased noise. Our results demonstrate that Transformer architectures modified with early detection biases, when used without temporal smoothing layers, provide a high-sensitivity alternative for forecasting AR emergence that prioritizes advance warning over statistical smoothness.

astro-ph.SR↗

arXiv2025

An Acoustic Model for Sunquakes Unifying the Solar Interior and Atmosphere

One of the leading hypotheses for sunquake generation suggests that flare-accelerated particles originating from the reconnection site in the corona travel down to the chromosphere and photosphere, where they deposit energy through collisions and subsequently drive acoustic oscillations. To properly encompass this top-down excitation mechanism, we extend the domain of a semi-spectral 3D acoustic model of the global Sun up to several 10's of Mm above the photosphere, where the transition region and lower corona are resolved. We then use the radially-dependent heating rates derived from the flare radiative hydrodynamic (RADYN) simulations -- extrapolated to a 3D profile -- to realistically excite sunquakes. In addition to the usual sunquake wavefronts, we also observe waves that propagate through the chromosphere and corona in a similar fashion to Moreton-Ramsey waves and large-scale coronal propagating fronts (LCPFs). We examine the dynamics of these waves and discuss how they may be used to constrain models of sunquake excitation.

astro-ph.SR↗

arXiv2025

Structure and Dynamics of the Sun's Interior Revealed by Helioseismic and Magnetic Imager

High-resolution helioseismology observations with the Helioseismic and Magnetic Imager (HMI) onboard Solar Dynamics Observatory (SDO) provide a unique three-dimensional view of the solar interior structure and dynamics, revealing a tremendous complexity of the physical processes inside the Sun. We present an overview of the results of the HMI helioseismology program and discuss their implications for modern theoretical models and simulations of the solar interior.

astro-ph.SR↗

arXiv2023

Can Proton Beam Heating Flare Models Explain Sunquakes?

SDO/HMI observations reveal a class of solar flares with substantial energy and momentum impacts in the photosphere, concurrent with white-light emission and helioseismic responses, known as sunquakes. Previous radiative hydrodynamic modeling has demonstrated the challenges of explaining sunquakes in the framework of the standard flare model of `electron beam' heating. One of the possibilities to explain the sunquakes and other signatures of the photospheric impact is to consider additional heating mechanisms involved in solar flares, for example, via flare-accelerated protons. In this work, we analyze a set of single-loop Fokker-Planck and radiative hydrodynamics RADYN+FP simulations where the atmosphere is heated by non-thermal power-law-distributed proton beams which can penetrate deeper than the electron beams into the low atmospheric layers. Using the output of the RADYN models, we calculate synthetic Fe I 6173 A line Stokes profiles and from those the line-of-sight (LOS) observables of the SDO/HMI instrument, as well as the 3D helioseismic response and compare them with the corresponding observational characteristics. These initial results show that the models with proton beam heating can produce the enhancement of the HMI continuum observable and explain qualitatively generation of sunquakes. The continuum observable enhancement is evident in all models but is more prominent in ones with $E_{c}\geq$500 keV. In contrast, the models with $E_{c}\leq$100 keV provide a stronger sunquake-like helioseismic impact according to the 3D acoustic modeling, suggesting that low-energy (deka- and hecto-keV) protons have an important role in the generation of sunquakes.

astro-ph.SR↗

arXiv2022

Exploring the Connection between Helioseismic Travel Time Anomalies and the Emergence of Large Active Regions during Solar Cycle 24

We investigate deviations in the mean phase travel time of acoustic waves preceding the emergence of 46 large active regions observed by the Helioseismic and Magnetic Imager (HMI). In our investigation, we consider two different procedures for obtaining the mean phase travel time, by minimizing the difference between cross-correlations and a reference, as well as the Gabor wavelet fitting procedure. We cross-correlate the time series of mean phase travel time deviations with the surface magnetic field and determine the peak correlation time lag. We also compute the perturbation index--the area integrated mean phase travel time deviations exceeding quiet sun thresholds--and compare the time of peak perturbation index with the correlation time lag. We find that the lag times derived from the difference minimization procedure precede the flux emergence for 36 of the 46 active regions, and that this lag time has a noticeable correlation with the maximum flux rate. However, only 28 of the active regions have peak perturbation index times in the range of 24 to 48 hours prior to the flux emergence. Additionally, we examine the relationship between properties of the emerged active regions and the strength of helioseismic signals prior to their emergence.

astro-ph.SR↗

arXiv2020

Machine Learning in Heliophysics and Space Weather Forecasting: A White Paper of Findings and Recommendations

The authors of this white paper met on 16-17 January 2020 at the New Jersey Institute of Technology, Newark, NJ, for a 2-day workshop that brought together a group of heliophysicists, data providers, expert modelers, and computer/data scientists. Their objective was to discuss critical developments and prospects of the application of machine and/or deep learning techniques for data analysis, modeling and forecasting in Heliophysics, and to shape a strategy for further developments in the field. The workshop combined a set of plenary sessions featuring invited introductory talks interleaved with a set of open discussion sessions. The outcome of the discussion is encapsulated in this white paper that also features a top-level list of recommendations agreed by participants.

astro-ph.SR↗