List of abstracts
2026

Tidal disruption events (TDEs) occur when stars are torn apart by supermassive black holes, producing distinctive optical transients that are also promising electromagnetic counterparts to high-energy neutrinos and ultra-high-energy cosmic rays. The Vera C. Rubin Observatory's LSST will detect thousands of TDEs, motivating efficient photometric classifiers to distinguish them from supernovae, AGN, and superluminous supernovae. We developed a machine-learning pipeline that fits Gaussian processes (GPs) to multi-band light curves and feeds the resulting color, rise/fade time, and GP length-scale features into an XGBoost classifier optimized for high precision. Trained on the ELAsTiCC2 simulation, our classifier reaches up to 95% precision at ~72% recall, with post-peak colors as the dominant predictive features. We are now generalizing this pipeline to the MALLORN Challenge dataset and adapting the feature set for early-alert classification on truncated light curves to enable timely multi-messenger follow-up.

In this new era of high-energy spectro-polarimetry, our understanding of high-energy sources has the potential to be greatly improved. This is especially the case for Black Hole X-ray Binaries (BHXB), also known as Microquasars, for which many unresolved challenges remain. In particular acceleration mechanisms and how they relate to accretion-ejection processes. Combined with spectroscopy, polarization is a powerful tool to reveal the geometry of the system and the configuration of the magnetic fields. The persistent Microquasar Cygnus X-1 is particularly well suited for our analysis. In soft X-rays, the polarization is low and aligned with the general direction of the jet, while the polarization angle in the soft gamma-rays is found close to 70° away from the jet. We use a novel method to combine high-energy spectra and polarization of this bright source from 1 keV to 2 MeV in a single fit, including observations from IXPE, PoGO+, XL-Calibur, AstroSat, INTEGRAL/SPI and INTEGRAL/IBIS. Our study involves energy-dependent polarization models for both emission processes, in part inspired by similar studies on Blazars. Preliminary results already reveal new constraints about the non-thermal hard-tail origin (above 100 keV), in particular its high-energy spectral cut-off and the degree of order of the magnetic field.

The Andromeda galaxy (M31) is a promising target for the indirect search of dark matter (DM) due to its proximity and expected massive DM halo. It functions as test case for a Milky Way (MW) like galaxy as the isotropic emission from the MW halo itself cannot be detected with a coded mask telescope like INTEGRAL/SPI. MeV data can put strong limits on DM models from the MeV up to the TeV mass range since weakly interacting massive particles also produce a significant flux of secondary MeV photons from inverse Compton scattering and positron annihilation. We analyze the spectrum of M31 obtained with the SPI instrument to constrain DM models. From the 511 keV emission from positron annihilation, we estimate the pair production rate in M31 to derive a lower mass limit on thermal DM. We further constrain the parameter space for DM annihilation and decay into electron–positron pairs, and place limits on primordial black hole DM.

The Cosmic Photon Background (CPB), an isotropic radiation field spanning the full electromagnetic spectrum, provides a potential source of positrons through photon–photon interactions (pair production). To model the CPB, it is decomposed into a set of gray-body components that evolve with redshift according to source-specific luminosity functions. The pair-production rate is obtained by integrating the angle- and energy-dependent cross section over the evolving photon field. The generated electron–positron pairs are then propagated to the present epoch (z = 0), accounting for cosmological expansion and energy losses in the intergalactic medium. The resulting photon emission is computed at each redshift and integrated along the line of sight to determine its contribution to the present-day Cosmic Gamma-Ray Background (CGB). The pair-production emissivity rises steeply from ~2.0 × 10⁻³⁶ s⁻¹ cm⁻³ at z = 0 to a peak of ~1.8 × 10⁻³¹ s⁻¹ cm⁻³ at z ≈ 2.7, followed by a decline at higher redshifts. This corresponds to a total cosmic pair-production rate of order 10⁵⁴ positrons per second up to z = 10. The secondary emission from inverse Compton scattering of these pairs off the CPB yields a significant contribution to the CGB. In particular, in the 1–10 MeV energy range, this component may account for approximately 10–20% of the total observed gamma-ray background. These results indicate that cosmological γ–γ absorption sets a minimum level of secondary emission that must be included in models of the CGB, especially in the MeV regime.

Originally, the IBWS (INTEGRAL/BART) workshops focused on the work of High energy astrophysics group (at that dime dominated by young research fellows and students) in Astronomical Institute of the Academy of Sciences of the Czech Republic and relevant national and international collaborators from the field, with intensive student participation. During the early years, these activities were focussed on the ESA INTEGRAL satellite and on the related ground-based robotic telescopes, e.g. the small robotic BART telescope at the Ondrejov Observatory. Nowadays, the IBWS workshops promote regional collaboration in galactic and extragalactic high-energy astrophysics, both experimental as well as theoretical, with an emphasis on the interface between satellite projects and ground-based experiments (e.g. robotic telescopes). We continue our emphasis on wide participation and presentations of students and young research fellows.

Astrophysics with digitized archival astronomical photographic plates An Upgrade. In this contribution, I will present the recent status of the digitization of archival astronomical plate archives, both direct and spectral. Astronomical photoplates capture the state of the sky over the past 100+ years. Digitization allows modern algorithms to track variable objects, including transients of various types and those related to high-energy astrophysics targets, asteroids, and other unique phenomena that would otherwise remain forgotten in the archives, albeit representing important data sources for recent astrophysics. Despite of recent progress over last years, many plate archives still remain without digitization and hence with difficult access.

Confronting observational measurements with theoretical models of neutron star spacetimes presents several challenges. These are particularly pronounced due to the differences in the definitions of measurable physical quantities and the quantities commonly used to define the spacetime metric, with the dependence of these quantities on the neutron star equation of state playing a major role. Here, we make use of the universal relations between neutron star parameters that were previously found through empirical means. We demonstrate how these relations can be utilised to obtain neutron star parameters from observed kilohertz quasi-periodic oscillations, provided the neutron star spin frequency is known.

Galaxy evolution over time remains unclear, with ongoing debate about how collisions affect star formation and metallicity. The role of the local environment shaped by the large-scale structure of the Universe in galaxy mergers may be significant, yet it has not been thoroughly examined. Using the IllustrisTNG cosmological simulation, we processed the catalog data and merger tree files of the TNG300-1 run. We calculated the average star formation rate (SFR) and stellar mass of galaxies over the redshift range 0 < z < 15 to trace the cosmic star formation history and galaxy growth. We investigated the environments of galaxy mergers with a focus on local density within the cosmic web, and found that interactions with gas-rich dwarf galaxies can trigger a resurgence in gas supply, highlighting the importance of gas dynamics in sustaining star formation. We compared our results with recent JWST observations and identified differences in the star formation rate density (SFRD) history between simulations and observations, providing new insights into early galaxy formation and evolution. See Koncz et. al. Universe 2025, 11(9), 286.

The origin of the Galactic 511 keV positron annihilation line has been a mystery for five decades. One proposed explanation is positron production in stellar flares, motivated by the detection of the 511 keV line in solar flares and by the association of this emission with old stellar populations. In this work, we explore this scenario using two complementary approaches. First, we build a theoretical model to estimate the quasi-persistent 511 keV emission from flaring stars. Starting from solar flare observations, we construct empirical scaling relations between flare energy and 511 keV luminosity and extend them to Galactic stellar populations using flare-frequency-energy distributions for different spectral types. In parallel, we analyze INTEGRAL/SPI data in the 511 keV band using combinations of known point sources and simple spatial templates, such as disk and bulge components modeled as two-dimensional Gaussians. We also test alternative descriptions in which no bulge template is assumed and the emission is instead described by a disk component together with a population of globular clusters, scaled by their masses and distances. This ongoing work aims to assess whether stellar flares can plausibly account for the observed Galactic 511 keV emission.

Interest in strange stars has increased recently with the observations of HESS J1731−347. This source may be interpreted as a strange star with an estimated mass of ~0.77 M☉, and a radius of ~10.4 km. This light object, together with the observations of heavy pulsars with masses of ~2 M☉, challenges the equations of state over a wide parameter space. In this work we will study whether various astronomical observations of neutron stars can also be explained by strange stars. We will focus on the maximum mass currently observed and on tidal deformability, both of which serve as possible constraints on the equations of state of strange stars.

Galactic cosmic rays are widely believed to originate from shock acceleration in supernova remnants; however, observations of ionization rates in the interstellar medium (ISM) in the inner Galaxy suggest the presence of additional, potentially unknown, acceleration sources. Low-energy cosmic rays (LECRs) excite nuclei in the ISM, producing MeV gamma-ray line emission through subsequent de-excitation, and observations of this emission enable robust constraints on the LECR flux and its spatial distribution. We model the MeV gamma-ray spatial distribution using three-dimensional computation based on cosmic-ray spectra predicted by GALPROP. We then evaluate their detectability relative to non-thermal leptonic emission components and identify the locations where it can be detected by current and future MeV gamma-ray missions.

Many stellar high-energy sources are manifestations of one of the stages of binary star evolution. In this talk, I will discuss the astrophysical processes influencing binary star evolution such as mass transfer and the response of mass-losing and mass-accreting stars, formation of circumbinary medium, explosions, and violent events such as common envelope evolution, as well as their relation to end products such as gravitational wave sources. I will cover in greater detail two types of transient events: luminous red novae, which accompany stellar mergers and potentially hold the key to connecting many different evolutionary pathways , and classical novae, where recent GeV Fermi detections as well as ground, space, and interferometric observations suggest a much more complicated picture of mass ejection than previously thought.

One of the major tasks of astrophysics is to understand the emission mechanisms of observed sources and regions in the sky. Only by pinpointing down these mechanisms, it is possible to derive physical parameters and learn about the evolution of astrophysical objects. Alas, many observations of high-energy phenomena are ambiguous, requiring more and orthogonal information. The nature of several sources, among others accreting X-ray binary systems, core-collapse and thermonuclear supernovae, cosmic-rays, stellar flares and potentially dark matter, all show signatures of positron production and annihilation. Utilising this underrated emission mechanism can shed light on unsolved problems in astrophysics and cosmology. In this talk, I will show examples of how we can learn from these gamma-ray signatures already now, and what might be possible in the context of new gamma-ray satellite missions, such as the accepted NASA mission COSI.

This project focuses on expanding the source catalogue derived from observations with the Spectrometer on INTEGRAL (SPI). While SPI has provided valuable insights into the gamma-ray sky, its current source catalogue remains limited by sensitivity constraints, background modeling challenges, and source confusion in crowded regions. The aim of this work is to improve source detection and characterization by applying enhanced data analysis techniques, including refined background estimation and optimized event selection. By systematically reprocessing archival data from the past 22 years of observation and incorporating longer exposure times, the project seeks to identify previously undetected or marginal sources and to improve the significance and localization of known ones. The resulting expanded catalogue will contribute to a more complete understanding of high-energy astrophysical populations and diffuse emission components, thereby maximizing the scientific return of the INTEGRAL mission.

High-frequency quasi-periodic oscillations (QPOs) are observed in the X-ray variability of accreting black holes (BHs) and neutron stars (NSs). These oscillations provide some of the most compelling insights into the physics of accretion flows in strong gravity and the equation of state of dense matter in astrophysics. In this talk, we will discuss the implications of various QPO models for both groups of sources, with a particular focus on comparing two QPO models based on the precession of accreted matter. The first of these is the relativistic precession model, which associates QPO frequencies with the frequencies of general relativistic test particle motion in the vicinity of the innermost stable circular orbit. The second model makes the same QPO frequency assignment but considers fluid precession rather than particle precession. While it clearly matched the NS data better, we found that the fluid flow precession model implied a lower NS mass and was consistent with higher BH spins. Furthermore, by making use of the CompOSE library, we demonstrate that the fluid flow precession model, unlike the former model, is compatible with a wide range of neutron star equations of state. The NS mass implied by the model typically ranges from 1.7 to 2.1 solar masses.

The detection and monitoring of high-energy emission from astrophysical sources provide key insights into energetic processes such as accretion, magnetic activity, and feedback in star-forming environments. Wide-field high-energy survey missions, such as the proposed THESEUS mission, are designed to perform repeated observations of large fractions of the sky, primarily targeting transient phenomena including gamma-ray bursts. As a consequence of their survey strategy, such missions also enable systematic, time-domain coverage in the soft X-ray regime, opening new opportunities for studying Galactic sources, including nearby star-forming regions and young stellar objects. We aim to assess the observability of nearby star-forming regions within the baseline survey strategy of wide-field high-energy missions and to explore how their temporal sampling can be used to characterize variability and feedback processes. In addition, we consider the synergy with ultraviolet facilities, such as ULTRASAT, which provide complementary constraints on stellar radiation fields and winds. By combining high-energy and UV time-domain coverage, it becomes possible to investigate the coupling between energetic processes in young stars and their impact on the surrounding interstellar medium.

The Karl Henize Mt. Wilson Michigan Southern Sky Hα survey comprises 290 photographic plates obtained at Lamont-Hussey Observatory, South Africa (1950–1952), now held at PARI. On such Hα low-dispersion spectral plates, normal stars produce elongated spectral traces, while compact non-dispersed features represent a distinct population hypothesised to include high-redshift galaxies, quasars, and GRB afterglows shifted beyond the 480–680 nm plate bandpass. We present a computational pipeline using OpenCV contour detection and Photutils DAOStarFinder to classify candidates by circularity (C ≥ 0.7) and aspect ratio (AR < 2.0). Applied to four pilot plates, 119 compact candidates (5.1%) were identified from 2,332 detections, with astrometric calibration enabling cross-matching against Gaia DR3 and LAMOST for validation.

The isotopes 26Al and 60Fe show short lifetimes (∼ 106 yr) with respect to the time scales of Galactic evolution and are therefore used as messengers of ongoing nucleosynthesis and star formation in the Milky Way. The production of both isotopes is commonly associated with massive stars. They are thought to be ejected in their supernovae, and in the case of 26Al partially in the wind phase of the star. Because of their similar origin and lifetimes , the ratio 60Fe/26Al of their γ-ray emission would be independent of the true location and distribution of their source, and yields a chance to test the outcome of stellar evolution models. The expected results of 60Fe/26Al for core-collapse supernovae, however, show discrepancies with measurements (Spyrou et al. 2024). This method only works, if 26Al and 60Fe have the same origin throughout the Galaxy. At the same time recent results from 26 Al measurements suggest a star formation rate (SFR) of ≳ 5 M⊙/yr (Siegert et al. 2023), which exceeds other literature values. Classical novae are also believed to produce 26 Al, which, if the contribution is large enough, would decrease the SFR from the correlation of 26 Al with massive stars. We aim to determine the 26 Al mass as a function of Galactocentric radius and compare it with two Galactic chemical evolution (GCE) models (Vasini et al. 2025; Martinet et al. 2022) to estimate the contribution of (very) massive stars and classical novae. We are using 20 yr of INTEGRAL/SPI observations of the decay gamma-ray line at 1. 809 MeV to test different assumptions on the radial structure of 26 Al. We present the total mass, the Galactic distribution of the 26Al, and the relative contribution of 26Al assuming different Galactic 3D models. We find significant contributions from classical novae which exceed prior studies for all tested models. A larger contribution of 26Al from classical novae directly decreases the effective 60Fe/26Al ratio affects the interpretation as a probe of massive-star nucleosynthesis and would lower the SFR from 26Al γ-ray measurements accordingly to 1-2 M⊙/yr

Since the typical lengths of the recurrence times Tc (cycle lengths) of outbursts in soft X-ray transients (SXTs) range from months to decades, X-ray monitors are needed to detect them and investigate their evolution. Evolution of Tc shows a complex curve with episodes of increase and decrease. Significant results can be obtained even when different monitors with various energy bands, available for the individual time segments (years or decades), are used (e.g., ASM/RXTE, MAXI/ISS, BAT/Swift). The character of the O-C curves of SXTs bears a striking similarity to that of dwarf novae observed in the optical band. It means that variations in Tc are large, but generally not chaotic, and long-term trends in the O-C curves can be resolved. Observing the variety of the X-ray evolution of dwarf novae is still limited by the low sensitivity of the current X-ray monitors.

In neutron star low-mass X-ray binaries, the boundary layer beneath the accretion disk is likely a major source of irradiation of the inner disk. Because this illuminating geometry differs from the compact corona usually assumed in relativistic reflection models, the radial profile of disk-incident flux may also depart from standard prescriptions. We investigate the illumination of a thin accretion disk by a latitude-dependent spreading layer on the surface of a weakly magnetized neutron star and find that the resulting radial dependence is well described by a smoothly broken power law, with parameters that depend systematically on the spreading layer geometry.

The internal composition of neutron stars remains one of the most compelling mysteries in modern astrophysics, as the extreme densities found within their cores cannot be replicated in terrestrial laboratories. Understanding the state of matter under such conditions is only possible through the systematic study of macroscopic stellar properties derived from multi-messenger observations. A fundamental question in this field is whether deconfined quark matter exists deep within the high-density stellar interior. This work investigates the complex phase transitions from hadronic nuclear matter to quark matter, specifically analyzing their impact on the stiffness of the equation of state. By examining tidal deformability measurements from the landmark GW170817 gravitational wave event, we aim to determine if these signals provide definitive signatures of quark cores, offering new insights into the fundamental nature of matter and the stability of hybrid stars.

Low-frequency quasi-periodic oscillations (LF QPOs) in neutron star (NS) X-ray binaries are commonly interpreted as manifestations of Lense–Thirring (LT) precession of the inner accretion flow. We test this hypothesis using models of geometrically thick accretion tori in the Hartle–Thorne spacetime, focusing on the source 4U 1608–52, for which both LF QPO frequencies and the NS spin are observationally constrained. By comparing predicted LT precession frequencies with those allowed by realistic NS models, we find that parameter combinations capable of reproducing the observed LF QPOs are not compatible with realistic NS configurations. In particular, such configurations imply stellar radii exceeding the innermost stable circular orbit, thereby limiting the maximum achievable precession frequency. We therefore conclude that a direct identification of LF QPOs with the fundamental LT precession frequency is not viable for this source. Instead, our results suggest that the observed variability may be associated with harmonics of the LT precession frequency, consistent with some earlier suggestions in the literature.

Relativistic accreting stars often exhibit strong quasi-periodic oscillations in their radiation. We examined the modulation of X-ray radiation from the neutron star boundary layer through the obscuration of an oscillating, thick inner accretion flow (Torok et al., ApJ, 2025). To do this, we adopted the full relativistic ray-tracing code LSD of Bakala et al. (A&A, 2015). We found that the investigated mechanism adequately explains the high amplitudes of the observed rapid neutron star X-ray variability. Further, it reproduces the differences observed in this variability between neutron star and black hole cases. Here, we briefly recall these results and present several extensions. These, among others, include synthetic data of Sgr~A$^*$ that could be verified or falsified by the NewAthena X-ray observatory in the future.

Twin-peak quasi-periodic oscillations (QPOs) observed in the X-ray variability of accreting neutron stars (NSs) offer a unique window into the physics of dense matter. We model QPO frequencies in the atoll source 4U 1608–52 using two precession-based models within the Hartle–Thorne spacetime framework, adopting the NS spin frequency of 620 Hz determined from X-ray burst measurements. Using 135 equations of state (EoSs) from the CompOSE library, we systematically confront a wide range of NS models with the observational data. Compared to the relativistic precession model, the fluid precession model clearly provides a better match to the data, while implying a NS mass that is approximately 10% lower. The favoured model constrains the NS mass to M = (1.8–2.0) Msun and the dimensionless spin parameter to j∼0.2–0.4.

Most models used to fit Fe-K lines in neutron star systems assume that such emission lines come from the accretion disk alone and neglect the presence of stellar surface. This simplification omits obscuration effects and possible iron line emission from the surface, where shear-heated material cools and spreads. Here, we present a relativistic model that incorporates both components consistently. Using full relativistic ray tracing code LSD, we compute photon trajectories from the disk and the neutron star surface. We demonstrate that neglecting the presence of the neutron star itself can lead to biased interpretations of relativistic iron emission line profiles. We show that such profiles are degenerated with pure disk emission for different disk parameters.

Using Preconditioned Spherical Generalized Additive Model for redshift completeness sky maps we study the spatial compatibility of Gamma-Ray Burst redshift subpopulations in the Swift GRB catalog. The redshifts of GRBs are measured using either rapid afterglow spectroscopy or delayed host-galaxy observations. We use several spherical spatial statistics, such as Monte Carlo log-likelihood permutation tests, Maximum Mean Discrepancy, the spherical Ripley's K-function, and the two-dimensional Fasano-Franceschini test, to measure the possible spatial differences. Our analysis finds no statistically significant difference between the two subpopulations. This result justifies the application of a combined GRB sky map during the reconstruction of the large-scale cosmological structures.

How efficiently does a relativistic collisionless shock accelerate electrons, and what fraction of the electron energy remains in a low-energy Maxwellian tail rather than being channelled into a non-thermal power law? These questions are central to shock physics in many astrophysical settings: pulsar wind nebulae, AGN jets, and the blast waves of gamma-ray bursts. Particle-in-cell simulations have predicted a partially-thermalized electron component at the forward shock for over fifteen years, but direct observational evidence has been lacking. On 2025 July 2, the Ondřejov D50 robotic telescope began observing GRB 250702F (z = 1.520) just 27.8 seconds after the Fermi/Swift trigger, providing continuous high-cadence optical coverage through the prompt-to-afterglow transition. After a bright flare of clearly internal-shock origin, a second component rises rapidly and turns over smoothly, decaying with a continuously steepening slope over more than 20 minutes (100 to 1400 s). At its steepest the decay is too fast for a forward shock, even though the component otherwise resembles one. Around 1400 s it transitions into a normal power-law afterglow. We interpret the steep decay as the synchrotron peak of the Maxwellian electron population sweeping through the optical band as the shock decelerates, following ν_th ∝ t^{-3/2}. A hybrid Maxwellian-plus-power-law forward-shock model fits the full optical evolution and yields a non-thermal energy fraction δ = 0.84 ± 0.02 and a characteristic Lorentz factor γ_th ≈ 900 for the Maxwellian population. This is well above the bulk Lorentz factor Γ₀ ≃ 160. These observations provide direct optical evidence of a partially-thermalized electron distribution at a relativistic forward shock. GRB 250702F is a clean case: negligible host extinction, dense cadence, and an early start combine to constrain the intrinsic spectral slope directly, removing the degeneracy that complicates interpretation in most afterglows. The resulting constraints on the Maxwellian fraction and the non-thermal acceleration efficiency match what particle-in-cell simulations have been predicting since 2008.

The FRAM robotic telescopes have been in operation for nearly two decades, producing a unique long-term dataset of wide-field sky images. The data includes observations of targeted astronomical sources, such as gamma-ray bursts, variable stars, and comets, as well as survey fields and atmospheric calibration images. The latter, although originally acquired for calibration purposes, also constitute a valuable serendipitous survey of the large regions of the sky. To make this dataset available to the scientific community in accordance with the FAIR principles - to make it Findable, Accessible, Interoperable, and Reusable - we have developed the FRAM Data Archive, a dedicated portal providing open access to more than six million images covering the whole sky. The archive allows users to search and retrieve original images, generate cutouts around specified sky positions, and access automatically derived photometric measurements for all objects detected in the images. We describe the architecture and current capabilities of the archive, including automated data ingestion and maintenance, calibration-frame production, astrometric calibration, photometric reduction, and generation of higher-level data products. We discuss the challenges associated with this heterogeneous long term dataset and outline possible applications in time-domain astronomy, archival searches, education, and public scientific use.

The transition sharpness of the optical forward-shock onset in gamma-ray bursts (GRBs) has not been systematically quantified. I introduce a hyperbola parametrisation for afterglow light curves that measures the transition width logR at the onset peak, offering more stable MCMC convergence than standard smooth broken power laws. The method is applied to eight GRB afterglows: three events observed in detail with the D50, SBT, and FRAM at La Palma; three archival afterglows reprocessed from published data, including observations from BART, D50, and FRAM at Auger; and two further bursts from recent group work. A linear correlation between logR and p, the electron power-law index, is found across seven of these bursts: harder spectra correspond to sharper onset transitions. An errors-in-variables MCMC fit confirms a significant negative slope with small intrinsic scatter. The parametrisation provides a stable and extensible framework for characterising the afterglow onset across large archival samples.

In the era of wide-field surveys such as the Zwicky Transient Facility and the Vera C. Rubin Observatory, the role of small robotic telescopes has shifted from discovery to follow-up observations of newly detected transients. Their scientific use depends on the ability to adapt observing strategies in real time as data are acquired. I present an automated pipeline for transient detection and follow-up optimisation developed for the D50 and SBT telescopes at the Ondřejov Observatory. The system integrates real-time photometric reduction (pyrt), presented at this conference before, with afterglow modelling, transient detection and alerts from other facilities. This approach has enabled early optical monitoring of several gamma-ray bursts and supernovae, including GRB 250702F (also discussed in the talk by Martin Jelínek).

This study presents a systematic analysis of the neutral hydrogen (NH) column density in Gamma-Ray Burst (GRB) environments by combining legacy data with the latest observations. We extend our previous analysis of the 2012–2017 high-resolution VLT/X-shooter spectroscopic data with the newest observations. The primary objective is to investigate the potential discrepancies between X-ray derived (from Swift/XRT) and optical Lyman-alpha column densities across a wide redshift range. By integrating the newest X-shooter measurements, we aim to refine the constraints on the redshift dependence of NH and investigate the role of photo-ionization and circumburst clearing effects. Our results provide critical calibration for future high-energy missions, such as THESEUS, highlighting the necessity of multi-wavelength approaches in understanding the interstellar medium of GRB host galaxies.

The landscape of of high-energy astrophysical transients keeps expanding, owing to a long and growing list of successful space- and earth-based observatories gathering data across the broadband. A common thread connecting various phenomena from gamma-ray bursts and their afterglows to X-ray flashes and optical transients, is the release of fast-moving plasma producing some form of non-thermal emission. In order to understand the physics behind a given event, modelling therefore typically requires a combination of jet dynamics, radiative process modelling and a coherent framework to translate both to an observational prediction. In this presentation, I will describe a few recent transients where I have been involved in modelling (GRB 170817A, GRB221009A, GRB230307A, EP240408A, EP241021A). These will be used to illustrate the various approaches, both numerical and (semi-)analytical, that are taken in practice to interpret extreme transients.

Complete spatial randomness (CSR) describes a point process whereby point events occur within a given study area in a completely random fashion. It is synonymous with a homogeneous spatial Poisson process. Such a process is modeled using only one parameter $nu$ , i.e. the density of points within the defined area. According to observations, the angular distribution of GRBs follows this property within the limits of statistical inference. We studied the question, however, whether this property is also valid for a 3D space distribution consisting GRBs having measured redshifts. Slicing sample into quantiles by redshift we tested their CSR, separately.

The Jochen Greiner’s GRB table represents one of the most comprehensive, although subjective, repository of Gamma-Ray Burst localizations, spanning several decades of observations from BeppoSAX to the latest missions like Swift, Fermi, and the Einstein Probe (EP). Due to the diverse nature of the instruments and localization procedures involved the statistical analysis of this dataset requires careful consideration of selection effects and instrumental biases. In this study, we perform a multi-dimensional statistical analysis of the Greiner table to uncover underlying patterns in the global properties of GRBs. By employing advanced statistical methods, we evaluate how the inclusion of various mission data affects the homogeneity of the catalog. Our results provide insights into the evolutionary effects of the GRB population and offer a critical assessment of the catalog's utility for cosmological studies. This work serves as a foundational statistical characterization of the dataset, supporting the synergy between legacy catalogs and future high-energy surveys.

The FRAM robotic telescopes started as experimental tools for atmospheric monitoring at astroparticle observatories over two decades ago. Over time, both their purpose and methods have greatly changed, sometimes following unexpectedly twisting paths. The original design target of the FRAM at the Pierre Auger Observatory was the measurement of wavelength dependence of aerosol extinction using photoelectric measurements of bright stars. After this has proven prohibitively difficult, the program re-focused on rapid cloud detection along the apparent trajectory of interesting cosmic ray showers - and during these observations, the breakthrough method of aerosol measurement using wide-field photometry was developed, leading to the expansion of the FRAM project to support the upcoming Cherenkov Telescope Array Observatory. Eventually, we went full circle: using the exceptionally stable conditions at Roque de los Muchachos, we can now show that the wide-field method is actually the right way for the wavelength-dependence measurement. As this quantity is directly correlated with the size of the aerosol particles, we are here opening a new window for atmospheric physics at night, with possibly broad application far beyond simple atmospheric monitoring.

Spacemanic has developed and operated a broad portfolio of small satellite missions spanning scientific, educational, and technology-demonstration objectives. This talk will provide an overview of selected missions focused on space weather, amateur radio experimentation, and gamma-ray burst detection, and will outline the common engineering and operational principles behind them. Particular emphasis will be placed on the mission architectures, payload functions, and the constraints imposed by nanosatellite platforms. The presentation will also discuss how these missions contribute to in-orbit validation, scientific data acquisition, and community engagement through compact and cost-efficient satellite systems. Together, they illustrate how small satellites can support a diverse range of objectives while remaining responsive to evolving scientific and technological needs.

The CORVUS IOV mission is a 6U in-orbit validation flight of Spacemanic’s CORVUS platform, developed under the ESA Pioneer Programme. The mission is designed to demonstrate the platform’s modularity, scalability, and suitability for next-generation small satellite applications. It carries two demonstration payloads: a monochromatic camera for Earth observation and space-debris monitoring, and an SDR receiver for radio-environment measurements across multiple frequency bands. The talk will present the mission concept, platform architecture, and key subsystem features, including communications, power, onboard computing, and attitude control. It will also summarize the technical objectives of the IOV mission and discuss how CORVUS provides a flexible foundation for future EO, IoT, and technology-demonstration missions.

We present the Phase B planning of the Transient High-Energy Sky and Early Universe Surveyor (THESEUS) Infrared Telescope’s Calibration Unit, carried out during Phase A, with emphasis on LED-based calibration concepts and optical design considerations. The work outlines a structured testing strategy that combines individual device characterization with batch-level validation to ensure consistent and reliable performance. In parallel, we investigate how different LED configurations and mounting approaches influence the optical behavior of the system. Ray-tracing simulations are used to explore light propagation and assess potential systematic effects impacting calibration accuracy. By integrating planned experimental activities with simulation-driven studies, this work establishes a solid foundation for the development of a robust and well-characterized calibration unit in Phase B.

Stratospheric balloons are frequently utilized experimental platforms within various research domains including meteorology, radiation physics, and astronomy. Thereby, the balloon gondola can be adapted to withstand multiple conditions and environmental requirements, ranging from an elementary construction for school projects up to complex designs for scientific NASA missions. The project ASTRABAX ("Aschaffenburg Stratospheric Balloon Experiment") investigated the radiation exposures of biological cells and of biomaterials in the upper atmosphere during a series of three stratospheric balloon flights. The continuously enhanced multimodal platform allowed to accommodate numerous device configurations within the UV- and temperature-controlled gondola. Here we show insights into the construction process, including achievements, present limitations and potential areas for improvement. Furthermore, the characterization of the stratospheric radiation environment during the multiple flights with different cosmic ray dosimeters is presented. The results of the feasibility flights indicate the broadening potential for educational and research purposes where low costs, reduced work force, and 3D printing technology can enhance the implementation of experiments. An easy assembly facilitates the expeditious involvement of students in the fields of radiation research, astrophysics, modern materials, and astrobiology.

The Compton Spectrometer and Imager (COSI) is scheduled for launch in 2027. It will cover a bandwidth from 0.2 to 5 MeV with its four main goals of shedding light into the origin of galactic positrons , enhancing our knowledge of stellar nucleosynthesis, giving insight into the most extreme environments via polarization and probing the physics of multimessenger events. The baseline data-pipeline relies on optimized conventional methods, with substantial heritage from prior gamma-ray missions. These are amended with machine learning based alternatives developed by multiple groups in the collaboration. The talk gives an overview of the upcoming mission and the activities in Mainz aimed at helping COSI to reach its full potential.

Accelerometers are devices that sense acceleration and transmit it as an electrical signal to measurement units. Recently, accelerometers have made an important contributionto several fields such as aerospace, space, health monitoring, and other industries.This study is dedicated to the mathematical modelling and development of a newly developed three-axis piezoelectric accelerometer for space applications based on LEO CubeSat missions. The modelling, development, and required simulations are carried out using COMSOL Multiphysics software. The main objective of this work is to develop an easily manufacturable, cost-effective, and sensitive three-axis piezoelectric accelerometer for attitude determination systems in low Earth orbit based space applications.

QUVIK is a Czech-led project of a small UV space telescope that has been selected and funded to be launched in year30/31. Masaryk University will be primarily responsible for developing and running a Science Operation Center that has 2 main tasks: defining most scientifically valuable observation plan (taking into account constraints of Low Earth Orbit and the power budget) and creating the most performing data processing pipeline (with a very limited baseline of on-board data handling). Our experience of operations of small telescopes on ground (mostly run in automatized regime) is the principal ingredience in this development. First steps taken in the process closely supervised by ESA support team will be presented.

Clearance measurements of various rooms in nuclear power plants is an important step in the decommissioning of these facilities. Currently, wall sections are divided into 1 m² areas, and their gamma spectra are measured individually to detect possible contaminations with different radioactive isotopes. Of particular interest here are ¹³⁷Cs (662 keV) and ⁶⁰Co (1173 and 1332 keV). To simplify this process by measuring whole rooms at once, we are developing a Compton camera (Radiation Source Locator, RSL7) for these clearance measurements in collaboration with project partners from Hellma Materials, Brenk Systemplanung, the Fraunhofer Institute for Scientific and Technical Trend Analysis (INT), and the Johannes Gutenberg University of Mainz. Based on the COSIpy analysis framework, developed for the NASA-SMEX mission COmpton Spectrometer and Imager (COSI), we are developing the software also for the RSL7. This involves fitting the activity of different radionuclides of previously created 3D models of all components of a room (template fit). In this talk, I will show simulations of the Compton camera, the processing of the response, and the fitting process using simulated datasets.

Results of tests of prototype of Schmidt lobster eye based on new technology are presented. Used new technological concept promises precise assembly of optical mirrors, which is the key aspect to obtain sharp focal image. This concept is subject of patent and utility model. Outer dimensions as well as focal length were chosen to allow this prototype be boarded on nano-satellite of CubeSat type. The prototype is designed for soft X-ray energies around 1keV. The module was tested using visible light, which is possible because glass plates coated with gold are used as mirrors. Focal image is taken by optical camera. FWHM is evaluated. The results show good accordance with simulations. It means that the technological concept offers precise assembly of optical mirrors, which is the key aspect to obtain sharp focal image.

GRBAlpha and GRBBeta are CubeSat missions equipped with gamma-ray detectors based on CsI(Tl) scintillators read out by silicon photomultipliers (SiPMs), primarily designed for the detection of gamma-ray bursts (GRBs). Masaryk University is strongly involved in the development, operation, and science case of these missions. GRBAlpha, a 1U CubeSat, was the first GRB-detecting satellite of this size. It was launched into low Earth orbit (LEO) in 2021 and operated successfully for more than four years, until June 9, 2025. During its mission, it detected more than 200 gamma-ray transients, of which approximately 124 were GRBs, including both long and short events. Among these were the most intense GRBs ever recorded, GRB 221009A, and the second brightest, GRB 230307A. GRBBeta, a 2U CubeSat integrated at Masaryk University and launched in 2024, continues to operate successfully and contribute to GRB detection and analysis. In addition to its gamma-ray detector, GRBBeta is equipped with a UV camera (LUVCam), sensitive in the near-ultraviolet range (240–310 nm), which serves as a technology demonstrator for the performance of the CMOS Gpixel GSENSE4040 imaging sensor in LEO. In this talk, I will present the scientific results obtained with GRBAlpha, as well as the latest results from GRBBeta, including LUVCam observations.

We show that a Soft X-ray Imager (SXI) onboard the ESA-CAS satellite SMILE will be able to observe various types of mass accreting compact objects in the universe. We show that mainly neutron stars accreting matter from their companions in X-ray binaries are promising targets for evaluating the possibilities of SXI in astronomical research. We present the typical features of the long-term activity of various X-ray binaries in the SXI/SMILE field of view. We also show that binaries with steady-state thermonuclear reactions on board accreting white dwarfs in the Magellanic Clouds are promising targets for SXI observations. We discuss how SXI can contribute to this branch.

The increasing use of CubeSat platforms requires reliable and cost-effective radiation monitoring solutions. This work presents the development of a COTS-based ionizing radiation monitoring payload for small satellite applications, targeting continuous in-orbit environmental characterization without the use of expensive rad-hard components. The payload is designed within a modular CubeSat architecture and utilizes volatile and non-volatile memory elements as radiation-sensitive components. The measurement concept is based on correlating radiation flux with error rate variations, enabling indirect dosimetry. The work focuses on the system architecture, payload design, and the definition of calibration and testing procedures. The presentation introduces the developed payload concept and outlines its integration into a CubeSat platform, highlighting key design considerations and challenges relevant to university-led small satellite missions.

This work presents a student-built 1U CubeSat project of the University of Debrecen, Hungary, based on the open-source LibreCube framework. The current phase focuses on the development and laboratory-level demonstration (up to TRL3) of core subsystems, with particular emphasis on the Attitude and Orbit Control System (AOCS). While the overall concept includes multiple student payloads. Due to the high cost of commercial space-grade actuators, the project focuses on the design and in-house development of key AOCS components, specifically magnetorquers and a reaction wheel. The proposed hybrid, low-cost AOCS integrates commercial-grade sensors (including magnetometers, MEMS gyroscopes, and sun sensors) with this custom actuator architecture. Validation is performed through simulation and laboratory measurements, including coil current characterization, magnetic dipole estimation, and sensor–actuator interaction analysis. The results demonstrate that a functional, low-cost AOCS can be realized and verified at laboratory level, providing a scalable foundation for future higher-TRL CubeSat development.

This will be the concluding address for the 20th IBWS conference including selected photographs taken during the event. We remind that the IBWS (INTEGRAL/BART) workshops originally focused on the work of High energy astrophysics group (at that dime dominated by young research fellows and students) in Astronomical Institute of the Academy of Sciences of the Czech Republic and relevant national and international collaborators from the field, with intensive student participation. During the early years, these activities were focussed on the ESA INTEGRAL satellite and on the related ground-based robotic telescopes, e.g. the small robotic BART telescope at the Ondrejov Observatory. Nowadays, the IBWS workshops promote regional collaboration in galactic and extragalactic high-energy astrophysics, both experimental as well as theoretical, with an emphasis on the interface between satellite projects and ground-based experiments (e.g. robotic telescopes). At the conference there continued our emphasis on wide participation and presentations of students and young research fellows.

Design and optimization of components operating in complex environments typically face two main challenges: the high complexity of the problem requires investigation of a large number of design points, while multi-objective criteria complicate the search for optimal solutions. The presented optimization approach combines efficient AI-assisted surrogate modeling and statistical data analysis to accelerate exploration of the design space. By focusing on probabilistic ranges, the method improves robustness, uncertainty awareness, and engineering decision-making under complex multi-objective constraints. The proposed workflow also enables interactive visualization of trade-offs between competing objectives, supporting more informed engineering decisions. The presented case studies demonstrate the potential of this approach to reduce computational costs and overcome limitations of traditional optimization methods.

In the operational environment of space missions, the boundary between peak performance and physiological failure may be defined by the efficiency of a single breath. Despite decades of development, respiratory dead space in oxygen masks such as the MBU-20/P oxygen mask could be a critical bottleneck under specific pressure regimes, leading to CO₂ rebreathing and a reduction in cognitive performance. This study addresses these challenges by reformulating classical fluid dynamics into a probabilistic optimization framework. A transient CFD approach is implemented to capture the full complexity of the breathing cycle. A key pillar of the research is the deployment of surrogate modeling via UptimAI, effectively bridging the gap between high-fidelity simulation and time-efficient design optimization within constrained development timelines. This approach enables high-resolution mapping of non-linear gas mixing phenomena and supports the identification of practically relevant design improvements. Preliminary results indicate specific geometric modifications of the mask that significantly enhance flushing efficiency, thereby ensuring a continuous supply of fresh breathing mixture to the pilot.

Cognitive performance determines not only how astronauts and scientists work, but also how well we think, create, and make decisions. Even highly trained individuals experience measurable fluctuations in cognitive performance over time. This contribution presents the ICARUS ARMOR Next Gen experiment, prepared for a future International Space Station mission of Czech astronaut Aleš Svoboda, which aims to predict short-term cognitive performance based on the temporal sequence of monitored stressors. The measurement concept combines cognitive testing, physiological biosignals, and biochemical stress markers into a unified multimodal dataset. These data are used to develop a personalized AI-based surrogate model, conceived as an individual software tool capable of forecasting cognitive performance over the following hours to days. While the experiment is motivated by the extreme conditions of human spaceflight, the approach is directly relevant to everyday personal and professional life. Participants will gain insight into how their own cognitive performance evolves, what factors shape it, and how it may be predicted. Selected tools used for astronaut cognitive testing will be available for hands-on experience, with some already available or becoming accessible for personal use.

Accurate Point Spread Function characterization is fundamental in image processing. This work presents a pipeline for decomposing the PSF in astronomical images into its constituent components: atmospheric seeing (Moffat function), optical aberrations (Zernike polynomials), and tracking error. As a preprocessing step, we developed a tool to minimize distortions from sensor tilt and field curvature in the camera setup. We then create a physics-based simulator that generates synthetic star fields with known distortion and aberration parameters, providing ground truth training data. Classical parameter estimation validates the forward model, while a machine learning stage is trained on the simulator output to recover distortion coefficients directly from observed stellar profiles, enabling automated PSF characterization for astronomical imaging systems.