Search

In general, large solar flares are more efficient at accelerating high-energy electrons than microflares. Nonetheless, we sometimes observe microflares that accelerate electrons to high energies. We statistically characterize 39 microflares with strikingly hard spectra in the hard X-ray (HXR) range, which means that they are efficient in accelerating high-energy electrons. We refer to these events as "hard microflares." The statistical analysis is built upon spectral and imaging information from STIX, combined with EUV and magnetic field maps from SDO. The key observational result is that all hard microflares in this dataset have one of the footpoints rooted directly within a sunspot (either in the umbra or the penumbra). This clearly indicates that the underlying magnetic flux densities are large. For the events with the classic two-footpoints morphology, the absolute value of the mean line-of-sight magnetic flux density (and vector magnetic field strength) at the footpoint rooted within the sunspot ranges from 600 to 1800 G (1500 to 2500 G), whereas the outer footpoint measures from 10 to 200 G (100 to 400 G), therefore about 10 times weaker. Approximately 78% of hard microflares, which exhibited two HXR footpoints, have similar or even stronger HXR flux from the footpoint rooted within the sunspot. This contradicts the magnetic mirroring scenario. In addition, about 74% of the events could be approximated by a single loop geometry, demonstrating that hard microflares typically have a relatively simple morphology. We conclude that all hard microflares are rooted in sunspots, which implies that the magnetic field strength plays a key role in efficiently accelerating high-energy electrons, with hard HXR spectra associated with strong fields. This key result will allow us to further constrain our understanding of the electron acceleration mechanisms in flares and space plasmas., Comment: Accepted for publication in A&A. 16 pages, 11 figures

Context. On 2022 January 20, the Energetic Particle Detector (EPD) on board Solar Orbiter measured a solar energetic particle (SEP) event showing unusual first arriving particles from the anti-Sun direction. Near-Earth spacecraft separated 17{\deg} in longitude to the west from Solar Orbiter measured classic antisunward-directed fluxes. STEREO-A and MAVEN, separated 18{\deg} to the east and 143{\deg} to the west from Solar Orbiter respectively, also observed the event, suggesting that particles spread over at least 160{\deg} in the heliosphere. Results. Solar Orbiter was embedded in a MC erupting on 16 January from the same active region as the one related to the SEP event on 20 January. The SEP event is related to a M5.5 flare and a fast CME-driven shock of 1433 km/s, which injected particles within and outside the MC. The hard SEP spectra, the presence of a Type II radio burst, and the co-temporal Type III radio bursts being observed from 80 MHz that seems to emanate from the Type II, points to the shock as the relevant accelerator of the particles. Conclusions. The detailed analysis of the SEP event strongly suggest that the energetic particles are injected mainly by a CME-driven shock into and outside of a previous MC present in the heliosphere at the time of the particle onset. The sunward propagating SEPs measured by Solar Orbiter are produced by the injection of particles along the longer (western) leg of the MC still connected to the Sun at the time of the release of the particles. The determined electron propagation path length inside the MC is around 30% longer than the estimated length of the loop leg of the MC itself (based on the graduated cylindrical shell model) consistent with a low number of field line rotations., Comment: 23 pages, 19 figures

Spatially unresolved observations show that the cooling phase in solar flares can be much longer than theoretical models predict. It has not yet been determined whether this is also the case for different subregions within the flare structure. We aim to investigate whether or not the cooling times, which are observed separately in coronal loops and the supra-arcade fan (SAF), are in accordance with the existing cooling models, and whether the temperature and emission measure of supra-arcade downflows (SADs) are different from their surroundings. We analysed the M5.6 limb flare on 13 January 2015 using SDO/AIA observations. We applied a differential emission measure (DEM) reconstruction code to derive spatially resolved temperature and emission measure maps, and used the output to investigate the thermal evolution of coronal loops, the SAF, and the SADs. In the event of 13 January 2015,the observed cooling times of the loop arcade and the SAF are significantly longer than predicted by the Cargill model, even with suppressed plasma heat conduction. The observed SADs show different temperature characteristics, and in all cases a lower density than their surroundings. In the limb flare event studied here, continuous heating likely occurs in both loops and SAF during the gradual flare phase and leads to an extended cooling phase.

It is well known that the class of rotation invariant algorithms are suboptimal even for learning sparse linear problems when the number of examples is below the "dimension" of the problem. This class includes any gradient descent trained neural net with a fully-connected input layer (initialized with a rotationally symmetric distribution). The simplest sparse problem is learning a single feature out of $d$ features. In that case the classification error or regression loss grows with $1-k/n$ where $k$ is the number of examples seen. These lower bounds become vacuous when the number of examples $k$ reaches the dimension $d$. We show that when noise is added to this sparse linear problem, rotation invariant algorithms are still suboptimal after seeing $d$ or more examples. We prove this via a lower bound for the Bayes optimal algorithm on a rotationally symmetrized problem. We then prove much lower upper bounds on the same problem for simple non-rotation invariant algorithms. Finally we analyze the gradient flow trajectories of many standard optimization algorithms in some simple cases and show how they veer toward or away from the sparse targets. We believe that our trajectory categorization will be useful in designing algorithms that can exploit sparse targets and our method for proving lower bounds will be crucial for analyzing other families of algorithms that admit different classes of invariances.

The Solar energetic particle analysis platform for the inner heliosphere (SERPENTINE) project presents it's new multi-spacecraft SEP event catalog for events observed in solar cycle 25. Observations from five different viewpoints are utilized, provided by Solar Orbiter, Parker Solar Probe, STEREO A, BepiColombo, and the near-Earth spacecraft Wind and SOHO. The catalog contains key SEP parameters for 25-40 MeV protons, 1 MeV electrons, and 100 keV electrons. Furthermore, basic parameters of the associated flare and type-II radio burst are listed, as well as the coordinates of the observer and solar source locations. SEP onset times are determined using the Poisson-CUSUM method. SEP peak times and intensities refer to the global intensity maximum. If different viewing directions are available, we use the one with the earliest onset for the onset determination and the one with the highest peak intensity for the peak identification. Associated flares are identified using observations from near Earth and Solar Orbiter. Associated type II radio bursts are determined from ground-based observations in the metric frequency range and from spacecraft observations in the decametric range. The current version of the catalog contains 45 multi-spacecraft events observed in the period from Nov 2020 until May 2023, of which 13 were widespread events and four were classified as narrow-spread events. Using X-ray observations by GOES/XRS and Solar Orbiter/STIX, we were able to identify the associated flare in all but four events. Using ground-based and space-borne radio observations, we found an associated type-II radio burst for 40 events. In total, the catalog contains 142 single event observations, of which 20 (45) have been observed at radial distances below 0.6 AU (0.8 AU).

We report on the detection of the gamma-ray emission above 100 MeV from the solar flare of September 29, 2022, by Fermi LAT with simultaneous coverage in HXR by Solar Orbiter STIX. The Solar Orbiter-Earth separation was 178$^{\circ}$ at the time of the flare as seen from Earth, with Solar Orbiter observing the east limb. Based on STIX imaging, the flare was located 16$^{\circ}$ behind the eastern limb as seen from Earth. The STIX and GBM non-thermal emission and the LAT emission above 100 MeV all show similarly shaped time profiles, and the Fermi profiles peaked only 20 seconds after the STIX signal from the main flare site, setting this flare apart from all the other occulted flares observed by Fermi LAT. The radio spectral imaging based on the Nan\c{c}ay Radioheliograph and ORFEES spectrograph reveal geometries consistent with a magnetic structure that connects the parent active region behind the limb to the visible disk. We studied the basic characteristics of the gamma-ray time profile, in particular, the rise and decay times and the time delay between the gamma-ray and HXR peak fluxes. We compared the characteristics of this event with those of four Fermi LAT behind-the-limb flares and with an on-disk event and found that this event is strikingly similar to the impulsive on-disk flare. Based on multiwavelength observations, we find that the gamma-ray emission above 100 MeV originated from ions accelerated in the parent active region behind the limb and was transported to the visible disk via a large magnetic structure connected to the parent active region behind the limb. Our results strongly suggest that the source of the emission above 100 MeV from the September 29, 2022 flare cannot be the CME-driven shock., Comment: 12 pages, 7 figures, accepted for publication in Astronomy and Astrophysics

Most mathematical distortions used in ML are fundamentally integral in nature: $f$-divergences, Bregman divergences, (regularized) optimal transport distances, integral probability metrics, geodesic distances, etc. In this paper, we unveil a grounded theory and tools which can help improve these distortions to better cope with ML requirements. We start with a generalization of Riemann integration that also encapsulates functions that are not strictly additive but are, more generally, $t$-additive, as in nonextensive statistical mechanics. Notably, this recovers Volterra's product integral as a special case. We then generalize the Fundamental Theorem of calculus using an extension of the (Euclidean) derivative. This, along with a series of more specific Theorems, serves as a basis for results showing how one can specifically design, alter, or change fundamental properties of distortion measures in a simple way, with a special emphasis on geometric- and ML-related properties that are the metricity, hyperbolicity, and encoding. We show how to apply it to a problem that has recently gained traction in ML: hyperbolic embeddings with a "cheap" and accurate encoding along the hyperbolic vs Euclidean scale. We unveil a new application for which the Poincar\'e disk model has very appealing features, and our theory comes in handy: \textit{model} embeddings for boosted combinations of decision trees, trained using the log-loss (trees) and logistic loss (combinations)., Comment: Subsumed by paper "Hyperbolic Embeddings of Supervised Models" by Richard Nock, Ehsan Amid, Frank Nielsen, Alexander Soen and Manfred K. Warmuth, appearing at NeurIPS'24

Tempered Exponential Measures (TEMs) are a parametric generalization of the exponential family of distributions maximizing the tempered entropy function among positive measures subject to a probability normalization of their power densities. Calculus on TEMs relies on a deformed algebra of arithmetic operators induced by the deformed logarithms used to define the tempered entropy. In this work, we introduce three different parameterizations of finite discrete TEMs via Legendre functions of the negative tempered entropy function. In particular, we establish an isometry between such parameterizations in terms of a generalization of the Hilbert log cross-ratio simplex distance to a tempered Hilbert co-simplex distance. Similar to the Hilbert geometry, the tempered Hilbert distance is characterized as a $t$-symmetrization of the oriented tempered Funk distance. We motivate our construction by introducing the notion of $t$-lengths of smooth curves in a tautological Finsler manifold. We then demonstrate the properties of our generalized structure in different settings and numerically examine the quality of its differentiable approximations for optimization in machine learning settings.

It is well known among the scientific community that solar flare activity often begins well before the main impulsive energy release. Our aim is to investigate the earliest phase of four distinct flares observed by Solar Orbiter/STIX and determine the relationships of the newly heated plasma to flare structure and dynamics. The analysis focuses on four events that were observed from both Earth and Solar Orbiter, which allows for a comparison of STIX observations with those of GOES/XRS and SDO/AIA. The early phases of the events were studied using STIX and GOES spectroscopic analysis to investigate the evolution of the physical parameters of the plasma, including the isothermal temperature and emission measure. Furthermore, to determine the location of the heated plasma, STIX observations were combined with AIA images. The events with clear emission prior to the impulsive phase show elevated temperatures ($> 10\,\mathrm{MK}$) from the very beginning, which indicates that energy release started before any detection by STIX. Although the temperature shows little variation during the initial phase, the emission measure increases by about two orders of magnitude, implying a series of incrementally greater energy releases. The spectral analysis of STIX and GOES from the very first time bins suggests that the emission has a multi-thermal nature, with a hot component of more than $10\,\mathrm{MK}$. This analysis confirms the existence of "hot onsets," with STIX detecting the hot onset pattern even earlier than GOES. These elevated temperatures imply that energy release actually begins well before any detection by STIX. Therefore, hot onsets may be significant in the initiation, early development, or even prediction of solar flares., Comment: 14 pages, 11 figures

In the field of optimal transport, two prominent subfields face each other: (i) unregularized optimal transport, "\`a-la-Kantorovich", which leads to extremely sparse plans but with algorithms that scale poorly, and (ii) entropic-regularized optimal transport, "\`a-la-Sinkhorn-Cuturi", which gets near-linear approximation algorithms but leads to maximally un-sparse plans. In this paper, we show that an extension of the latter to tempered exponential measures, a generalization of exponential families with indirect measure normalization, gets to a very convenient middle ground, with both very fast approximation algorithms and sparsity, which is under control up to sparsity patterns. In addition, our formulation fits naturally in the unbalanced optimal transport problem setting.

Stellar flares cannot be spatially resolved, which complicates ascertaining the physical processes behind particular spectral signatures. Due to their proximity to Earth, solar flares can serve as a stepping stone for understanding their stellar counterparts, especially when using a Sun-as-a-star instrument and in combination with spatially resolved observations. We aim to understand the disk-integrated spectral behaviors of a confined X2.2 solar flare and its eruptive X9.3 successor as measured by HARPS-N. The behavior of multiple photospheric and chromospheric spectral lines are investigated by means of activity indices and contrast profiles. A number of different photospheric lines were also investigated by means of equivalent widths, and radial velocity measures, which are then related to physical processes directly observed in high-resolution observations made with the Swedish 1-meter Solar Telescope and SDO Our findings suggest a relationship between the evolving shapes of contrast profile time and the flare locations, which assists in constraining flare locations in disk-integrated observations. In addition, an upward bias was found in flare statistics based on activity indices derived from the Ca II H & K lines. In this case, much smaller flares cause a similar increase in the activity index as that produced by larger flares. H$\alpha$-based activity indices do not show this bias and are therefore less susceptible to activity jitter. Sodium line profiles show a strongly asymmetric response during flare activity, which is best captured with a newly defined asymmetrical sodium activity index. A strong flare response was detected in Mn I line profiles, which is unexpected and calls for further exploration. Intensity increases in H$\alpha$, H$\beta$, and certain spectral windows of AIA before the flare onset suggest their potential use as short-term flare predictors., Comment: Published in A&A; Volume 682, A46 (20 pages, 14 figures) doi:10.1051/0004-6361/202347895

The continued success of Large Language Models (LLMs) and other generative artificial intelligence approaches highlights the advantages that large information corpora can have over rigidly defined symbolic models, but also serves as a proof-point of the challenges that purely statistics-based approaches have in terms of safety and trustworthiness. As a framework for contextualizing the potential, as well as the limitations of LLMs and other foundation model-based technologies, we propose the concept of a Large Process Model (LPM) that combines the correlation power of LLMs with the analytical precision and reliability of knowledge-based systems and automated reasoning approaches. LPMs are envisioned to directly utilize the wealth of process management experience that experts have accumulated, as well as process performance data of organizations with diverse characteristics, e.g., regarding size, region, or industry. In this vision, the proposed LPM would allow organizations to receive context-specific (tailored) process and other business models, analytical deep-dives, and improvement recommendations. As such, they would allow to substantially decrease the time and effort required for business transformation, while also allowing for deeper, more impactful, and more actionable insights than previously possible. We argue that implementing an LPM is feasible, but also highlight limitations and research challenges that need to be solved to implement particular aspects of the LPM vision.

The Solar Orbiter mission completed its first remote-sensing observation windows in the spring of 2022. On 2/4/2022, an M-class flare followed by a filament eruption was seen both by the instruments on board the mission and from several observatories in Earth's orbit. The complexity of the observed features is compared with the predictions given by the standard flare model in 3D. We use the observations from a multi-view dataset, which includes EUV imaging to spectroscopy and magnetic field measurements. These data come from IRIS, SDO, Hinode, as well as several instruments on Solar Orbiter. Information given by SDO/HMI and Solar Orbiter PHI/HRT shows that a parasitic polarity emerging underneath the filament is responsible for bringing the flux rope to an unstable state. As the flux rope erupts, Hinode/EIS captures blue-shifted emission in the transition region and coronal lines in the northern leg of the flux rope prior to the flare peak. Solar Orbiter SPICE captures the whole region, complementing the Doppler diagnostics of the filament eruption. Analyses of the formation and evolution of a complex set of flare ribbons and loops show that the parasitic emerging bipole plays an important role in the evolution of the flaring region. While the analysed data are overall consistent with the standard flare model, the present particular magnetic configuration shows that surrounding magnetic activity such as nearby emergence needs to be taken into account to fully understand the processes at work. This filament eruption is the first to be covered from different angles by spectroscopic instruments, and provides an unprecedented diagnostic of the multi-thermal structures present before and during the flare. This dataset of an eruptive event showcases the capabilities of coordinated observations with the Solar Orbiter mission., Comment: Accepted for publication in the Astronomy & Astrophysics special edition "Solar Orbiter First Results (Nominal Mission Phase)" (23/05/2023)

We investigate observational signatures of explosive chromospheric evaporation during a white-light flare (WLF) that occurred on 2022 August 27. Using the moment analysis, bisector techniques, and the Gaussian fitting method, red-shifted velocities of less than 20 km/s are detected in low-temperature spectral lines of Ha, C I and Si IV at the conjugated flare kernels, which could be regarded as downflows caused by chromospheric condensation. Blue-shifted velocities of about 30-40 km/s are found in the high-temperature line of Fe XXI, which can be interpreted as upflows driven by chromospheric evaporation. A nonthermal hard X-ray (HXR) source is co-spatial with one of the flare kernels, and the Doppler velocities are temporally correlated with the HXR fluxes. The nonthermal energy flux is estimated to be at least (1.3+-0.2)*10^10 erg/s/cm^2. The radiation enhancement at Fe I 6569.2 A and 6173 A suggests that the flare is a WLF. Moreover, the while-light emission at Fe I 6569.2 A is temporally and spatially correlated with the blue shift of Fe XXI line, suggesting that both the white-light enhancement and the chromospheric evaporation are triggered and driven by nonthermal electrons. All our observations support the scenario of an electron-driven explosive chromospheric evaporation in the WLF., Comment: The manuscript was accepted for publication in ApJ, and it was a part of the ApJ Focus Issue "Early results from the Chinese Ha Solar Explorer (CHASE)"

FOXSI is a direct-imaging, hard X-ray (HXR) telescope optimized for solar flare observations. It detects hot plasma and energetic electrons in and near energy release sites in the solar corona via bremsstrahlung emission, measuring both spatial structure and particle energy distributions. It provides two orders of magnitude faster imaging spectroscopy than previously available, probing physically relevant timescales (<1s) never before accessible to address fundamental questions of energy release and efficient particle acceleration that have importance far beyond their solar application (e.g., planetary magnetospheres, flaring stars, accretion disks). FOXSI measures not only the bright chromospheric X-ray emission where electrons lose most of their energy, but also simultaneous emission from electrons as they are accelerated in the corona and propagate along magnetic field lines. FOXSI detects emission from high in the tenuous corona, where previous instruments have been blinded by nearby bright features and will fully characterizes the accelerated electrons and hottest plasmas as they evolve in energy, space, and time to solve the mystery of how impulsive energy release leads to solar eruptions, the primary drivers of space weather at Earth, and how those eruptions are energized and evolve., Comment: White paper submitted to the Decadal Survey for Solar and Space Physics (Heliophysics) 2024-2033; 14 pages, 4 figures, 1 table

One of the most popular ML algorithms, AdaBoost, can be derived from the dual of a relative entropy minimization problem subject to the fact that the positive weights on the examples sum to one. Essentially, harder examples receive higher probabilities. We generalize this setup to the recently introduced {\it tempered exponential measure}s (TEMs) where normalization is enforced on a specific power of the measure and not the measure itself. TEMs are indexed by a parameter $t$ and generalize exponential families ($t=1$). Our algorithm, $t$-AdaBoost, recovers AdaBoost~as a special case ($t=1$). We show that $t$-AdaBoost retains AdaBoost's celebrated exponential convergence rate when $t\in [0,1)$ while allowing a slight improvement of the rate's hidden constant compared to $t=1$. $t$-AdaBoost partially computes on a generalization of classical arithmetic over the reals and brings notable properties like guaranteed bounded leveraging coefficients for $t\in [0,1)$. From the loss that $t$-AdaBoost minimizes (a generalization of the exponential loss), we show how to derive a new family of {\it tempered} losses for the induction of domain-partitioning classifiers like decision trees. Crucially, strict properness is ensured for all while their boosting rates span the full known spectrum. Experiments using $t$-AdaBoost+trees display that significant leverage can be achieved by tuning $t$.

We study the phenomenon of \textit{in-context learning} (ICL) exhibited by large language models, where they can adapt to a new learning task, given a handful of labeled examples, without any explicit parameter optimization. Our goal is to explain how a pre-trained transformer model is able to perform ICL under reasonable assumptions on the pre-training process and the downstream tasks. We posit a mechanism whereby a transformer can achieve the following: (a) receive an i.i.d. sequence of examples which have been converted into a prompt using potentially-ambiguous delimiters, (b) correctly segment the prompt into examples and labels, (c) infer from the data a \textit{sparse linear regressor} hypothesis, and finally (d) apply this hypothesis on the given test example and return a predicted label. We establish that this entire procedure is implementable using the transformer mechanism, and we give sample complexity guarantees for this learning framework. Our empirical findings validate the challenge of segmentation, and we show a correspondence between our posited mechanisms and observed attention maps for step (c).

Solar flares and coronal mass ejections (CMEs) are thought to be the most powerful events on the Sun. They can release energy as high as 10^32 erg in tens of minutes,and could produce solar energetic particles (SEPs) in the interplanetary space. We explore global energy budgets of solar major eruptions on 6 September 2017, including the energy partition of a powerful solar flare, the energy budget of the accompanied CME and SEPs. In the wavelength range shortward of 222 nm, a major contribution of the flare radiated energy is in the soft X-ray (SXR) 0.1-7 nm domain. The flare energy radiated at wavelengths of Ly-alpha and middle ultraviolet is larger than that radiated in the extreme ultraviolet wavelength, but it is much less than that radiated in the SXR waveband. The total flare radiated energy could be comparable to the thermal and nonthermal energies. The energies carried by the major flare and its accompanied CME are roughly equal, and they are both powered by the magnetic free energy in the AR NOAA 12673. Moreover, the CME is efficient in accelerating SEPs, and that the prompt component (whether it comes from the solar flare or the CME) contributes only a negligible fraction., Comment: accepted for publication in Research in Astronomy and Astrophysics

Context. A solar eruption on 17 April 2021 produced a widespread Solar Energetic Particle (SEP) event that was observed by five longitudinally well-separated observers in the inner heliosphere at heliocentric distances of 0.42 to 1 au: BepiColombo, Parker Solar Probe, Solar Orbiter, STEREO A, and near-Earth spacecraft. The event produced relativistic electrons and protons. It was associated with a long-lasting solar hard X-ray flare and a medium fast Coronal Mass Ejection (CME) with a speed of 880 km/s driving a shock, an EUV wave as well as long-lasting radio burst activity showing four distinct type III burst. Methods. A multi-spacecraft analysis of remote-sensing and in-situ observations is applied to attribute the SEP observations at the different locations to the various potential source regions at the Sun. An ENLIL simulation is used to characterize the interplanetary state and its role for the energetic particle transport. The magnetic connection between each spacecraft and the Sun is determined. Based on a reconstruction of the coronal shock front we determine the times when the shock establishes magnetic connections with the different observers. Radio observations are used to characterize the directivity of the four main injection episodes, which are then employed in a 2D SEP transport simulation. Results. Timing analysis of the inferred SEP solar injection suggests different source processes being important for the electron and the proton event. Comparison among the characteristics and timing of the potential particle sources, such as the CME-driven shock or the flare, suggests a stronger shock contribution for the proton event and a more likely flare-related source of the electron event. Conclusions. We find that in this event an important ingredient for the wide SEP spread was the wide longitudinal range of about 110 degrees covered by distinct SEP injections.

Coronal and interplanetary shock waves produced by coronal mass ejections (CMEs) are major drivers of space-weather phenomena, inducing major changes in the heliospheric radiation environment and directly perturbing the near-Earth environment, including its magnetosphere. A better understanding of how these shock waves evolve from the corona to the interplanetary medium can therefore contribute to improving nowcasting and forecasting of space weather. Early warnings from these shock waves can come from radio measurements as well as coronagraphic observations that can be exploited to characterise the dynamical evolution of these structures. Our aim is to analyse the geometrical and kinematic properties of 32 CME shock waves derived from multi-point white-light and ultraviolet imagery taken by the Solar Dynamics Observatory (SDO), Solar and Heliospheric Observatory (SoHO), and Solar-Terrestrial Relations Observatory (STEREO) to improve our understanding of how shock waves evolve in 3D during the eruption of a CME. We use our catalogue to search for relations between the shock wave's kinematic properties and the flaring activity associated with the underlying genesis of the CME piston. Past studies have shown that shock waves observed from multiple vantage points can be aptly reproduced geometrically by simple ellipsoids. The catalogue of reconstructed shock waves provides the time-dependent evolution of these ellipsoidal parameters. From these parameters, we deduced the lateral and radial expansion speeds of the shocks evolving over time. We compared these kinematic properties with those obtained from a single viewpoint by SoHO in order to evaluate projection effects. Finally, we examined the relationships between the shock wave and the associated flare when the latter was observed on the disc by considering the measurements of soft and hard X-rays., Comment: 11 pages, 12 figures, accepted for publication in A&A

Aims. To provide a schematic mathematical description of the imaging concept of the Spectrometer/Telescope for Imaging X-rays (STIX) on board Solar Orbiter. The derived model is the fundamental starting point for both the interpretation of STIX data and the description of the data calibration process. Methods. We describe the STIX indirect imaging technique which is based on spatial modulation of the X-ray photon flux by means of tungsten grids. We show that each of 30 STIX imaging sub-collimators measures a complex Fourier component of the flaring X-ray source corresponding to a specific angular frequency. We also provide details about the count distribution model, which describes the relationship between the photon flux and the measured pixel counts. Results. We define the image reconstruction problem for STIX from both visibilities and photon counts. We provide an overview of the algorithms implemented for the solution of the imaging problem, and a comparison of the results obtained with these different methods in the case of the SOL2022-03-31T18 flaring event.

We combine observations from different vantage points to perform a detailed study of a long duration eruptive C7 class flare that occurred on 17 April 2021 and was partially occulted from Earth view. The dynamics and thermal properties of the flare-related plasma flows, the flaring arcade, and the energy releases and particle acceleration are studied together with the kinematic evolution of the associated CME in order to place this long duration event in context of previous eruptive flare studies. The flare showed hard X-ray (HXR) bursts over the duration of an hour in two phases lasting from 16:04 UT to 17:05 UT. During the first phase, a strong increase in emission from hot plasma and impulsive acceleration of the CME was observed. The CME acceleration profile shows a three-part evolution of slow rise, acceleration, and propagation in line with the first STIX HXR burst phase, which is triggered by a rising hot (14 MK) plasmoid. During the CME acceleration phase, we find signatures of ongoing magnetic reconnection behind the erupting structure, in agreement with the standard eruptive flare scenario. The subsequent HXR bursts that occur about 30 minutes after the primary CME acceleration show a spectral hardening (from $\delta \approx $ 7 to $\delta \approx $ 4) but do not correspond to further CME acceleration and chromospheric evaporation. Therefore, the CME-flare feedback relationship may only be of significance within the first 25 minutes of the event under study, as thereafter the flare and the CME eruption evolve independently of each other.

The Spectrometer and Telescope for Imaging X-rays (STIX) on board Solar Orbiter observes solar X-ray emission in the range of 4 – 150 keV and produces spectra and images of solar flares over a wide range of flare magnitudes. During nominal operation, STIX continuously generates data. A constant data flow requires fully automated data-processing pipelines to process and analyze the data, and a data platform to manage, visualize, and distribute the data products to the scientific community. The STIX Data Center has been built to fulfill these needs. In this paper, we outline its main components to help the community better understand the tools and data it provides. The STIX Data Center is operated at the University of Applied Sciences and Arts Northwestern Switzerland (FHNW) and consists of automated processing pipelines and a data platform. The pipelines process STIX telemetry data, perform common analysis tasks, and generate data products at different processing levels. They have been designed to operate fully automatically with minimal human intervention. The data platform provides web-based user interfaces and application programmable interfaces for searching and downloading STIX data products. The STIX Data Center has been operating successfully for more than two years. The platform facilitates instrument operations and provides vital support to STIX data users., Comment: 18 pages, Accepted by A&A, https://doi.org/10.1051/0004-6361/202346031

We study the solar energetic particle (SEP) event observed on 9 October 2021, by multiple spacecraft including Solar Orbiter (SolO). The event was associated with an M1.6 flare, a coronal mass ejection (CME) and a shock wave. During the event, high-energy protons and electrons were recorded by multiple instruments located within a narrow longitudinal cone. An interesting aspect of the event was the multi-stage particle energization during the flare impulsive phase and also what appears to be a separate phase of electron acceleration detected at SolO after the flare maximum. We aim to investigate and identify the multiple sources of energetic electron acceleration. We utilize SEP electron observations from the Energetic Particle Detector (EPD) and hard X-ray (HXR) observations from the Spectrometer/Telescope for Imaging X-rays (STIX) on-board SolO, in combination with radio observations at a broad frequency range. We focus on establishing an association between the energetic electrons and the different HXR and radio emissions associated with the multiple acceleration episodes. We have found that the flare was able to accelerate electrons for at least 20 minutes during the nonthermal phase observed in the form of five discrete HXR pulses. We also show evidence that the shock wave has contributed to the electron acceleration during and after the impulsive flare phase. The detailed analysis of EPD electron data shows that there was a time difference in the release of low- and high-energy electrons, with the high-energy release delayed. Also, the observed electron anisotropy characteristics suggest different connectivity during the two phases of acceleration., Comment: 17 pages, 10 figures (main text). 3 pages, 3 figures (appendix). In review

One of the main science questions of the Solar Orbiter and Parker Solar Probe missions deals with understanding how electrons in the lower solar corona are accelerated and how they subsequently access interplanetary space. We aim to investigate the electron acceleration and energy release sites as well as the manner in which accelerated electrons access the interplanetary space in the case of the SOL2021-02-18T18:05 event, a GOES A8 class microflare associated with a coronal jet. This study takes advantage of three different vantage points, Solar Orbiter, STEREO-A, and Earth, with observations ranging from radio to X-ray. Multi-wavelength timing analysis combined with UV/EUV imagery and X-ray spectroscopy by Solar Orbiter/STIX (Spectrometer/Telescope for Imaging X-rays) is used to investigate the origin of the observed emission during different flare phases. The event under investigation satisfies the classical picture of the onset time of the acceleration of electrons coinciding with the jet and the radio type III bursts. This microflare features prominent hard X-ray nonthermal emission down to at least 10 keV and a spectrum that is much harder than usual for a microflare with a spectral index of 2.9. From Earth's vantage point, the microflare is seen near the limb, revealing the coronal energy release site above the flare loop in EUV, which, from STIX spectroscopic analysis, turns out to be hot (at roughly the same temperature of the flare). Moreover, this region is moving toward higher altitudes over time (about 30 km/s). During the flare, the same region spatially coincides with the origin of the coronal jet. We conclude that the energy release site observed above-the-loop corresponds to the electron acceleration site, corroborating that interchange reconnection is a viable candidate for particle acceleration in the low corona on field lines open to interplanetary space., Comment: 9 pages, 6 figures

The Sun is the source of different types of radio bursts that are associated with solar flares, for example. Among the most frequently observed phenomena are type III solar bursts. Their radio images at low frequencies (below 100 MHz) are relatively poorly studied due to the limitations of legacy radio telescopes. We study the general characteristics of types IIIb and U with stria structure solar radio bursts in the frequency range of 20 - 80 MHz, in particular the source size and evolution in different altitudes, as well as the velocity and energy of electron beams responsible for their generation. In this work types IIIb and U with stria structure radio bursts are analyzed using data from the LOFAR telescope including dynamic spectra and imaging observations, as well as data taken in the X-ray range (GOES and RHESSI satellites) and in the extreme ultraviolet (SDO satellite). In this study we determined the source size limited by the actual shape of the contour at particular frequencies of type IIIb and U solar bursts in a relatively wide frequency band from 20 to 80 MHz. Two of the bursts seem to appear at roughly the same place in the studied active region and their source sizes are similar. It is different in the case of another burst, which seems to be related to another part of the magnetic field structure in this active region. The velocities of the electron beams responsible for the generation of the three bursts studied here were also found to be different., Comment: Astronomy & Astrophysics, in press; 9 pages, 6 figures

The link with exponential families has allowed $k$-means clustering to be generalized to a wide variety of data generating distributions in exponential families and clustering distortions among Bregman divergences. Getting the framework to work above exponential families is important to lift roadblocks like the lack of robustness of some population minimizers carved in their axiomatization. Current generalisations of exponential families like $q$-exponential families or even deformed exponential families fail at achieving the goal. In this paper, we provide a new attempt at getting the complete framework, grounded in a new generalisation of exponential families that we introduce, tempered exponential measures (TEM). TEMs keep the maximum entropy axiomatization framework of $q$-exponential families, but instead of normalizing the measure, normalize a dual called a co-distribution. Numerous interesting properties arise for clustering such as improved and controllable robustness for population minimizers, that keep a simple analytic form.

In this work, we propose a novel approach for layerwise representation learning of a trained neural network. In particular, we form a Bregman divergence based on the layer's transfer function and construct an extension of the original Bregman PCA formulation by incorporating a mean vector and normalizing the principal directions with respect to the geometry of the local convex function around the mean. This generalization allows exporting the learned representation as a fixed layer with a non-linearity. As an application to knowledge distillation, we cast the learning problem for the student network as predicting the compression coefficients of the teacher's representations, which are passed as the input to the imported layer. Our empirical findings indicate that our approach is substantially more effective for transferring information between networks than typical teacher-student training using the teacher's penultimate layer representations and soft labels.

In this paper, we introduce the SAP Signavio Academic Models (SAP-SAM) dataset, a collection of hundreds of thousands of business models, mainly process models in BPMN notation. The model collection is a subset of the models that were created over the course of roughly a decade on academic.signavio.com, a free-of-charge software-as-a-service platform that researchers, teachers, and students can use to create business (process) models. We provide a preliminary analysis of the model collection, as well as recommendations on how to work with it. In addition, we discuss potential use cases and limitations of the model collection from academic and industry perspectives.

Hexagonal Si2Te2 monolayers (ML-Si2Te2) were predicted to show strain-dependent band-crossover between semiconducting and room-temperature quantum spin Hall phases. However, investigations on this artificial two-dimensional (2D) material have mainly been restricted to theoretical calculations because its bulk counterpart does not exist naturally. Here, we report on the successful epitaxial growth of ML-Si2Te2 films on Sb2Te3 thin film substrates. High-quality (1*1) ML-Si2Te2 films with a coverage as high as 95% were obtained as revealed by scanning tunneling microscopy. X-ray photoelectron spectroscopy confirms the absence of intermixing between Si2Te2 and Sb2Te3 at the interface. By combining scanning tunneling spectroscopy with density functional theory calculations, we demonstrate the semiconducting band structure of ML-Si2Te2 on Sb2Te3. Furthermore, it is theoretically predicted that the system can be driven into the nontrivial phase via reducing the strain by 4.4% using strain engineering. Our results pave the way for in-depth investigations on this 2D topological insulator candidate.

We present the Fermi-LAT observations of the behind-the-limb (BTL) flare of July 17, 2021 and the joint detection of this flare by STIX onboard Solar Orbiter. The separation between Earth and the Solar Orbiter was 99.2$^{\circ}$ at 05:00 UT, allowing STIX to have a front view of the flare. The location of the flare was ~S20E140 in Stonyhurst heliographic coordinates making this the most distant behind-the-limb flare ever detected in $>$100 MeV gamma-rays. The LAT detection lasted for $\sim$16 minutes, the peak flux was $ 3.6 \pm 0.8 $ (10$^{-5}$) ph cm$^{-2}$ s$^{-1}$ with a significance $>$15$\sigma$. A coronal wave was observed from both STEREO-A and SDO in extreme ultraviolet (EUV) with an onset on the visible disk in coincidence with the LAT onset. A complex type II radio burst was observed by GLOSS also in coincidence with the onset of the LAT emission indicating the presence of a shock wave. We discuss the relation between the time derivative of the EUV wave intensity profile at 193\angstrom\ as observed by STEREO-A and the LAT flux to show that the appearance of the coronal wave at the visible disk and the acceleration of protons as traced by the observed $>$100 MeV gamma-ray emission are coupled. We also report how this coupling is present in the data from 3 other BTL flares detected by Fermi-LAT suggesting that the protons driving the gamma-ray emission of BTL solar flares and the coronal wave share a common origin., Comment: 10 pages, 9 figures

Context. The Spectrometer/Telescope for Imaging X-rays (STIX) is one of 6 remote sensing instruments on-board Solar Orbiter. It provides hard X-ray imaging spectroscopy of solar flares by sampling the Fourier transform of the incoming flux. Aims. To show that the visibility amplitude and phase calibration of 24 out of 30 STIX sub-collimators is well advanced and that a set of imaging methods is able to provide the first hard X-ray images of the flaring Sun from Solar Orbiter. Methods. We applied four visibility-based image reconstruction methods and a count-based one to calibrated STIX observations. The resulting reconstructions are compared to those provided by an optimization algorithm used for fitting the amplitudes of STIX visibilities. Results. When applied to six flares with GOES class between C4 and M4 which occurred in May 2021, the five imaging methods produce results morphologically consistent with the ones provided by the Atmospheric Imaging Assembly on-board the Solar Dynamic Observatory (SDO/AIA) in UV wavelengths. The $\chi^2$ values and the parameters of the reconstructed sources are comparable between methods, thus confirming their robustness. Conclusions. This paper shows that the current calibration of the main part of STIX sub-collimators has reached a satisfactory level for scientific data exploitation, and that the imaging algorithms already available in the STIX data analysis software provide reliable and robust reconstructions of the morphology of solar flares.

We present the surprising result that randomly initialized neural networks are good feature extractors in expectation. These random features correspond to finite-sample realizations of what we call Neural Network Prior Kernel (NNPK), which is inherently infinite-dimensional. We conduct ablations across multiple architectures of varying sizes as well as initializations and activation functions. Our analysis suggests that certain structures that manifest in a trained model are already present at initialization. Therefore, NNPK may provide further insight into why neural networks are so effective in learning such structures.