Journées de l'AT-PEM 2026 - Défis numériques en astrophysique des hautes énergies
Amphithéâtre Charpak
Sorbonne Université, campus Jussieu
Numerical Challenges in High-Energy Astrophysics
Registration deadline: 25/09/2026
The "Action Thématique Phénomènes Extrêmes et Multi-messagers" (AT-PEM, formerly PNHE) has been an INSU thematic action since 2011 and is also supported by IN2P3, INP, CNES, and CEA/IRFU.
Rationale:
Multi-messenger astrophysics has opened unprecedented windows onto extreme astrophysical phenomena. Processing and interpreting these observations require increasingly sophisticated models capable of bridging vast spatial and temporal scales, coupling complex physical processes, and exploiting modern high-performance computing architectures.
The 2026 edition of the AT-PEM days aims to highlight these challenges by bringing together observers, theorists, and numerical modelers to discuss the tools and approaches used within the High-Energy Astrophysics community in France.
A key objective of the meeting is to foster discussion across communities, identify current limitations of existing approaches, and outline future directions for numerical modeling.
Structure:
The event will be structured around five broadly defined themes, each corresponding to one full session:
- Transport and diffuse backgrounds,
- Relativistic jets,
- Supernovae,
- Accreting black holes,
- Magnétospheres.
Each session will start with a one-hour invited observational review talk, followed by a series of invited, contributed, and flash talks, and a concluding session, all focusing on numerical aspects of the topic.
Practical informations:
- Registration required (free) - Deadline: 25/09/2026
- In person
- Start: 18 November 2026 at 2 PM
- End: 20 November 2026 at 5.30 PM
- Location: Amphithéâtre Charpak, Sorbonne Université, campus Jussieu, Paris
Organizing Committee: Arno Vanthieghem (LUX, arno.vanthieghem@obspm.fr), Benoît Cerutti (IPAG), Astrid Lamberts (OCA), Alexandre Marcowith (LUPM), Pierre-Olivier Petrucci (IPAG), Peggy Varnière (APC), Cristina Volpe (APC)
-
-
Welcome and Opening Remarks
-
Introduction: Transport and diffuse backgroundPrésident de session: Jonathan Biteau ({UNIV PARIS-SACLAY}UMR9012)
-
1
Introduction - Transport and diffuse backgrounds: Probing Extragalactic Electromagnetic Fields with Astroparticles
The extragalactic sky at the highest energies is shaped by the complex interplay between relativistic messengers and the diffuse electromagnetic fields through which they propagate. Grounded in multi-messenger frameworks and empirical models of cosmic-ray transport, this talk provides a comprehensive synthesis of very-high-energy gamma-ray and ultra-high-energy cosmic-ray propagation across cosmological distances, including the production of cosmogenic neutrinos.
Orateur: Jonathan Biteau ({UNIV PARIS-SACLAY}UMR9012)
-
1
-
Invited talks: Transport and diffuse backgrounds
-
2
Diffuse Galactic gamma-ray emission at very high energies, and TeV halos around pulsars
Gamma-ray observations provide important information on cosmic-ray (CR) propagation in our Galaxy.First, we present a new model of CR propagation in the Milky Way, where CRs are injected at discrete transient sources in the disc. We then calculate the corresponding diffuse Galactic gamma-ray emission. We find that the diffuse gamma-ray emission above 100 TeV is very clumpy, and does not correlate with the gas density along the line of sight. It is substantially different from the relatively smoother emission detected by Fermi at GeV energies. We also discuss how many hadronic and leptonic PeVatrons would be detectable in our simulations, and compare with LHAASO data. We show that this allows to place new constraints on PeV CR sources.Second, we discuss TeV gamma-ray halos around old pulsars. We show that current gamma-ray measurements place interesting constraints on the turbulent magnetic fields around these pulsars, and we examine the implications for CR transport. In particular, we study the case of LHAASO's "peanut-shaped" source.
Orateur: Gwenael GIACINTI (APC Paris) -
3
Astrophysical gravitational-wave backgrounds
The detection of gravitational waves from compact-binary coalescences by LIGO–Virgo–KAGRA has opened a new window onto the Universe. As detector sensitivities improve, and with the advent of future observatories such as LISA and the Einstein Telescope, the number and diversity of individually detectable gravitational-wave sources will increase dramatically. Yet many more sources will remain too faint to be detected individually, or will overlap in time and frequency, giving rise to gravitational-wave backgrounds. In this talk, I will discuss various astrophysical gravitational-wave backgrounds and the source populations that generate them, including white-dwarf binaries, binary neutron stars, stellar-mass and supermassive black-hole binaries. I will also highlight some of the challenges involved in detecting and characterizing these backgrounds, which can themselves constitute an additional noise source in searches for cosmological gravitational-wave signals.
Orateur: Irina Dvorkin (Institut d'Astrophysique de Paris) -
4
Field line & particle transport in magnetic turbulence : synthetic vs MHD
Over the past decade, a wealth of new data has ushered in an era of precision cosmic-ray physics. By revealing previously unanticipated features in cosmic-ray spectra, these observations challenge current paradigms of Galactic cosmic-ray acceleration and transport. When it comes to modeling cosmic-ray transport in magnetic turbulence, existing theoretical models still struggle to accurately reproduce simulation results, even in highly simplified turbulent magnetic-field configurations. In this talk, I will present our ongoing efforts to investigate the fundamental properties of magnetic field lines and cosmic-ray transport in synthetic turbulence, and show how these results can help us interpret transport in MHD turbulence. I will highlight remarkable similarities between the two cases and discuss their differences.
Orateur: Yoann GÉNOLINI (LAPTh, Annecy)
-
2
-
15:45
Coffee break
-
Contributed talks: Transport and diffuse backgrounds
-
5
Blowing PeVatrons: Wind Bubbles and Jets from Stellar-Mass Compact Objects
The origin of Galactic PeV cosmic rays remains one of the central questions in high-energy astrophysics. Stellar-mass compact objects, such as ultraluminous X-ray sources and microquasars, launch powerful outflows that inflate wind bubbles or produce collimated jets, naturally giving rise to long-lived shock fronts where diffusive shock acceleration can operate efficiently. In this contribution, I will present models of particle acceleration in these environments and show that protons can be accelerated up to PeV energies, and under favorable conditions even beyond 10 PeV. In the context of recent LHAASO detections and the forthcoming CTA Observatory, these systems emerge as promising candidates for a previously overlooked population of Galactic PeVatrons.
Orateur: Enrico Peretti (APC) -
6
Millisecond pulsars in the Galactic bulge: from simulations to new discoveries
The Galactic bulge is expected to host a large population of old millisecond pulsars (MSPs), which could contribute substantially to unresolved X-ray emission and the elusive Fermi-LAT Galactic-centre gamma-ray excess (GCE). Yet few pulsars have been detected toward the Galactic centre, where particularly severe observational biases make MSPs more difficult to find.
Propulsing is an ANR-funded project that combines radio and gamma-ray searches for MSPs, population inference, and analyses of the GCE to investigate its potential MSP origin. Our population-synthesis framework simulates MSP populations using alternative models for their spatial distribution, intrinsic properties, and signal propagation through the interstellar medium, then applies survey selection effects to predict the detectable MSP sample. Initial simulations were used to inform the design of our deep MeerKAT and Parkes surveys of the Galactic bulge. These surveys have now begun and have already led to new pulsar discoveries.
In this talk, I will present our simulation framework for inferring the bulge MSP population, including its application to these latest survey results. Comparing the predicted detectable population with observations tests whether MSP detections toward the Galactic bulge are compatible with the Galactic disk population alone or whether an additional bulge population is required. Initial applications indicate that current disk models cannot reproduce the detections at the expected rate, supporting an enhanced MSP density toward the Galactic centre. Our framework thus connects the design of deep radio surveys, constraints on the bulge population, and interpretations of the GCE.
Orateur: Emilie Parent (Institut de Planétologie et d'Astrophysique de Grenoble) -
7
Cosmic Ray Transport and Acceleration in Merging Galaxy Clusters
Merging galaxy clusters exhibit strong observational evidence for efficient cosmic ray acceleration, particularly in the form of synchrotron-emitting radio relics and halos. These particles are likely accelerated at merger and accretion shocks. However, in the presence of the large diffusion coefficients one would expect in the relatively unmagnetized intracluster medium (ICM), long acceleration times limit the maximum cosmic ray energy. On the other hand, recent work on cosmic ray transport suggests that the diffusion coefficient can be suppressed in ICM-like environments. In this picture, deviations from local thermodynamic equilibrium can trigger the mirror instability, creating plasma-scale magnetic structures, or "micromirrors," that efficiently scatter particles. I will briefly discuss how these micromirrors enforce a baseline value of the maximum proton energy that does not rely on cosmic ray driven magnetic field amplification. I will also introduce cosmological simulations that self-consistently include cosmic ray acceleration at shocks, with an effective diffusion coefficient set by micromirrors.
Orateur: Rebecca Diesing (University of Chicago)
-
5
-
Flash talks: Transport and diffuse backgrounds
-
8
Probing Diffuse Radio Emission from Pulsars in Terzan 5 with Simulation-Based Inference
Globular clusters (GCs) are known to host an important population of millisecond pulsars (MSPs). Terzan 5, located at 5.5 kpc from Earth, contains 49 known MSPs, with additional sources expected to remain undetected due to dispersion and instrumental sensitivity limits. These unresolved MSPs are expected to contribute to the diffuse radio emission observed towards the cluster. Using Simulation-Based Inference (SBI) to constrain the MSP luminosity function, we find that these unresolved MSPs are unlikely to account for the dominant fraction of the observed diffuse emission.
Orateur: Dr Adrian Faucher (IPAG) -
9
GRAND and HERON projects
I will present the status and perspectives of the two projects for radio detection of UHE cosmic particles in which CNRS is involved.
Orateur: Olivier Martineau (LPNHE)
-
8
-
Conclusion: Transport and diffuse backgrounds
- 10
-
Cocktail - Observatoire de Paris - Salle Cassini
Salle Cassini
77 avenue Denfert-Rochereau, 75014 Paris
-
-
-
Introduction: Relativistic jets
-
11
Introduction: Relativistic jetsOrateurs: Matteo Cerruti (Université de Paris - APC), Susanna Diana VERGANI (CNRS-Observatoire de Paris)
-
11
-
Invited talks: Relativistic jets
-
12
The impact of pair production on Gamma Ray Burst prompt emission spectra
Gamma Ray Burst prompt emission is widely attributed to radiation from accelerated electrons within ultra-relativistic ejecta. The detection of high energy components in several GRBs, along with the theoretical predictions of inverse Compton scattering, suggest the presence of abundant photons above the pair production threshold in the comoving frame. Under such conditions, photon-photon interactions lead to significant production of electron-positron pairs. By modelling numerically the coupled evolution of photons and leptons, we investigate the impact of secondary pairs on the emergent spectrum in the optically thin regime. We find that the emission from secondary pairs can drive substantial deviation from the standard synchrotron prediction in the soft gamma-ray range, including changes of both low and high energy photon indices. In rare cases, the annihilation of these pairs can produce a narrow emission line above 1 MeV, similar to the one reported for the first time in the BOAT GRB221009A. We further explore the physical conditions required for the emergence of this annihilation line and show that it can only be reproduced under strict conditions. The recent launch of SVOM and Einstein Probe opens up new perspectives for the study of prompt low energy emission. In particular, the spectral shape above 4 keV can be precisely measured by the combination of ECLAIRS and GRM onboard SVOM. This could prove crucial in measuring these effects and improving our understanding of the prompt emission mechanisms.
Orateur: Mouad Gnaoui (IAP, Sorbonne Université) -
13
Multi-messenger modeling of binary neutron star mergers
I will discuss the key ingredients required to model the multi-messenger emission from binary neutron star mergers, from gravitational waves to electromagnetic radiation, highlighting the main degeneracies and open questions. I will then move from individual events to populations, describing how synthetic populations can be constructed, from astrophysical assumptions to observational predictions.
Orateur: Sofia Bisero (GEPI, Observatoire de Paris, Université PSL, CNRS) -
14
From numerical simulations to Bayesian inference: surrogate models for time-resolved blazar SEDs
Time-resolved spectral energy distributions (SEDs) provide a powerful probe of the physical evolution of highenergy sources, but extacting this information with detailed numerical models is computationally challenging. This becomes particularly prohibitive when full posterior sampling is required. To address this challenge, we developed a surrogate-modeling framework, based on convolutional neural networks, that emulate the output of state-of-the-art time-dependent numerical simulations with sufficient accurately for statistical inference. Trained and validated against the underlying physical simulations, the surrogate reproduces the predicted spectra while reducing model evaluation costs by orders of magnitude. Applying the framework to OJ 287, PKS 2155-304, and 1ES 1959+650, I will demonstrate how time-resolved posterior inference characterizes the evolution of the jet magnetic fields, particle injection and acceleration, and dynamical parameters across different emission state.
Orateur: Damien BEGUE (APC)
-
12
-
10:45
Coffee
-
Contributed talks: Relativistic jets
-
15
From Photons to Neutrinos: Advancing Multi-Messenger Blazar Modeling with Gammapy
The origin of astrophysical neutrinos and ultra-high-energy cosmic rays remains one of the central questions of the multi-messenger era. While numerous classes of astrophysical sources have been proposed as potential accelerators, the underlying particle acceleration and emission mechanisms remain uncertain. Combining observations across multiple messengers with advanced modeling tools is key to unlocking the secrets of these extreme sources.
In this contribution, we present a novel, user-friendly multi-messenger modeling approach built on the open-source package Gammapy. The framework combines observations from the optical band to very-high-energy gamma rays and performs analyses directly at the event list level. By enabling the direct fitting of theoretical source models within Gammapy, it reduces intermediate data products, minimizes associated assumptions and reduces parameters biases and errors underestimation..
Focusing on blazars as prime multi-messenger candidates, we demonstrate how this approach improves constraints on astrophysical models. We further present ongoing developments extending the framework to include neutrino observations and (lepto-)hadronic emission models. Since hadronic models are computationally expensive and and therefore cannot be easily fit on the fly, we discuss our current solutions for hadronic fits and outline future developments toward fully integrated hadronic multi-messenger analyses.Orateur: Lea Heckmann (APC) -
16
Fast Leptohadronic SED Modeling of Relativistic Jets with Neural Network Emulators
Leptohadronic modeling of jet emission is central to interpreting multimessenger observations, but time-dependent kinetic codes such as AM3 and LeHaMoC are too slow for Bayesian fitting over high-dimensional parameter spaces. We present GRU-based neural network emulators trained on large AM3 and LeHaMoC SED libraries. They reproduce the full broadband spectra with high accuracy at millisecond evaluation times. For blazars, this enables full posterior sampling of leptonic and leptohadronic models against multiwavelength data. We also outline the next step like extending the framework to GRBs.
Orateur: Georgios Vasilopoulos (National and Kapodistrian University of Athens) -
17
RECENT RESULTS ON THE PROMPT GRB EMISSION MODELLING: PROPERTIES OF THE X-RAY EMISSION
The multi-wavelength observations of gamma-ray bursts provide insights into the physical conditions in the relativistic jets responsible for their emission. I will discuss new X-ray observations and present the properties of X-ray emission computed in the standard GRB framework of internal shocks. Numerical simulations of the emitted spectrum in the comoving frame performed for a large parameter space, and the spectra/light curves modelled for several representative cases will be shown. I will discuss how these results may contribute to solve some puzzling problems raised by recent GRB observations in X-ray range.
Orateur: Zeljka Bosnjak (IAP)
-
15
-
Flash talks: Relativistic jets
-
18
Synchrotron emission from cooling internal shocks
Gamma-ray bursts (GRBs) release a star's lifetime energy in seconds through an ultra relativistic jet, but the exact process behind the jet prompt emission remains unknown. In the internal shocks model, variations in the jet velocity causes the material to collide at a large distance from the source. These collisions accelerates electrons which emits this energy as gamma-rays through efficient synchrotron cooling. However, observations have shown more photons at low energies than usual synchrotron models can predict, pushing towards alternate explanations.
In this work, we evolve electron distributions across the entirety of a shell's hydrodynamical evolution after an internal shock. Controling the relative importance of synchrotron and adiabatic cooling through a global cooling parameter, we compute the observed emission across a wide range of cooling regimes. We find that effects from an extended emission region become relevant even in radiatively efficient regimes. Moreover, material that keeps on radiating after shock crossing and subsequent rarefaction wave significantly contribute to the low-energy spectra of internal shocks. Finally, we set the basis for a GRB spectral model controlled by a low number of physical parameters.Orateur: Arthur Charlet -
19
GRB afterglows from radio to very high energies
In this flash talk, I will present key features of gamma-ray burst (GRB) afterglows across the electromagnetic spectrum (X-ray plateaus, optical rebrightenings, radio variability). I will show examples of observations with the SVOM mission and discuss how they help gain physical insights into the progenitor properties (central engine and environment characteristics, particle acceleration and radiative processes).
Orateur: Clément Pellouin (CEA - AIM) -
20
Jets from gamma-ray AGN analyzed via radio VLBI and Gaia optical data
Analysis of a sample of 1233 active galactic nuclei (AGN) common to the ICRF3 (radio VLBI), Gaia (optical), and Fermi/4FGL-DR4 (gamma-ray) catalogs makes it possible to identify the regions of these sources that are brightest in the optical and likely also in the gamma-ray bands. For the majority of them, this active zone is indiscernible from the AGN core (~ 80%), while for a significant number of sources, it corresponds to the base of the jet (~ 8%) or to a knot further along the jet (~12%). Blue FSRQs, red FSRQs and BL Lacs show different behavior, but in all cases the gamma-ray fluxes appear to decrease with the distance along the VLBI jets. The case of a few emblematic blazars will be presented and the consequences that can be drawn from these results will be discussed.
(from Lambert, S. & Sol, H., 2026, preprint)Orateur: HELENE SOL (LUTH)
-
18
-
Conclusion: Relativistic jets
-
21
Conclusion: Relativistic jetsOrateur: Matteo Bugli (Institut d'astrophysique de Paris - CNRS)
-
21
-
13:00
Lunch
-
Introduction: Supernovae
-
22
Introduction - Supernovae: What we can learn from supernovae and supernova remnants observations?
Supernovae and supernova remnants (SNRs) play a fundamental role in galactic ecosystems. They are primary drivers of chemical enrichment, injecting heavy elements synthesized during stellar evolution and explosive nucleosynthesis into the interstellar medium (ISM). Their shocks heat and structure the ISM. They accelerate particles, contribute to turbulence and magnetic field amplification, and are sources of Galactic cosmic rays. In this presentation, I will review recent advances in the observations of supernova and supernova remnants, with a particular focus on the high-energy band, on the new results from XRISM spatially-resolved high-resolution spectroscopy data and IXPE polarization data.
Orateur: Anne Decourchelle (Dap/AIM, CEA Paris-Saclay)
-
22
-
Invited talks: Supernovae
-
23
Neutrino flavor instabilities: from theory to observables
Accurately modeling neutrino flavor oscillations in global simulations of core-collapse supernovae remains a major challenge, albeit a potentially crucial one for making reliable predictions. Indeed, it is now widely recognized that flavor instabilities, in which neutrino distributions computed in classical transport frameworks can be dramatically reshaped by flavor conversion, are ubiquitous in such environments.
In this talk, I will describe recent progress in understanding and modeling these instabilities, reviewing the various approaches developed in the community. From parametric studies to physics-informed subgrid models, we are currently making key steps toward fully incorporating neutrino physics into large-scale simulations and assessing its wide-ranging implications.Orateur: Julien Froustey (IFIC, Universitat de València) -
24
Cosmic-ray transport from large to small scales
I will present current progress on Galactic cosmic-ray transport with a strong focus on the escape of cosmic rays around supernova remnants studied using particle-in-cell simulations. Some results concerning the transport of cosmic rays in their self-generated turbulence will be discussed together with potential observational consequences.
Orateur: Phan Vo Hong Minh (LUX) -
25
Particle acceleration in novae
Classical and recurrent novae are powered by thermonuclear runaways on the surfaces of accreting white dwarfs. Novae have emerged as an important class of astrophysical particle accelerators, as demonstrated in particular by the discovery of GeV gamma-ray emission with Fermi-LAT. In classical novae, particle acceleration is thought to occur primarily at internal shocks within the ejecta, which are often radiative owing to the high densities of the shocked material. In symbiotic recurrent novae such as RS Ophiuchi, particles are most likely accelerated at the blast wave driven by the ejecta into the dense wind of a red giant companion. Recent detections of TeV gamma rays from RS Ophiuchi have provided compelling evidence for proton acceleration to very high energies. I will review these observations, discuss the physical mechanisms of particle acceleration in nova shocks, and highlight the implications for our understanding of cosmic-ray acceleration.
Orateur: Vincent Tatischeff (IJCLab)
-
23
-
16:15
Coffee
-
Contributed talks: Supernovae
-
26
Studying astrophysical double-shock structures with particle-in-cell simulations
Shocks are abundant in astrophysical systems, and are promising candidates for extremely efficient particle acceleration. As shocks can travel in different-density media and have a range of properties, they can give rise to colliding flows. Such may create initially double structures, bound by the forward and reverse propagating shocks. The aim of our study is to investigate the extent of possible energisation, and what is the effect on the particle trajectories, that one may expect in the inter-shock region, while the separation of the two shocks remains not too distant. We use 1D particle-in-cell simulations to trace the particles and characterise the electron energisation when confined in the inter-shock region, while exploring which parameters defining the plasma and shock properties may be more favourable to a more efficient electron pre-acceleration. We discuss our preliminary findings and the imminent future steps of our study.
Orateur: Michelle Tsirou (Institute of Physics of the Czech Academy of Sciences (FZU)) -
27
Cosmic-Ray Electron Flux from Dynamically Evolving Supernova Remnants
Supernova remnants (SNRs) are generally recognized as the primary sources of Galactic cosmic-ray electrons. Models usually approximate particle injection into the interstellar medium as an instantaneous burst with a phenomenological analytical spectrum. In this work, we consider a time-dependent electron injection framework based on the Sedov-Taylor model of SNR evolution. This allows us to directly incorporate the dynamical processes of shock microphysics, where magnetic field evolution and energy-dependent escape govern the timing and spectral shape of released particles. We explicitly include early escape mechanisms as well as the adiabatic expansion and radiative cooling experienced by trapped particles before their time-dependent release. These injection spectra are subsequently propagated through the Galaxy, subject to spatial diffusion and radiative energy losses. We present preliminary results on the expected electron fluxes at Earth from the modelled SNRs.
Orateur: Lorenzo Stefanuto (Università di Torino) -
28
Laboratory investigation of laser-driven stochastic Fermi acceleration
The origin of high-energy cosmic rays remains one of the central open questions in astrophysics. Collisionless shocks are widely recognized as efficient particle accelerators, but the microphysics of early-stage ion energization -- before a fully nonlinear shock structure develops -- is still poorly constrained, particularly in regimes where large-scale interchange-type instabilities such as the magnetic Rayleigh–Taylor instability (MRTI) dominate the turbulent dynamics, as expected at supernova-remnant contact discontinuities. We present a laboratory platform that produces and characterizes the MRTI in an astrophysically relevant regime, by coupling high-power laser-driven plasma flows with strong externally applied magnetic fields. We observe a clear energization of protons originating from the ambient gas; magnetohydrodynamic (MHD) simulations supplemented with test-particle tracking identify stochastic (second-order) Fermi acceleration, driven by the MRTI-generated turbulent fluctuations, as the dominant mechanism. These measurements provide the first direct experimental evidence of early-stage ion energization by the MRTI. Achieving quantitative agreement between the laboratory data and the simulations requires a careful treatment of the extended Ohm's law, demonstrating that non-ideal MHD effects play a leading role in shaping the acceleration efficiency and energy spectrum.
Orateur: Weipeng Yao (LUX)
-
26
-
Flash talks: Supernovae
-
29
Resonance-induced modulations of multimessenger emission in core collapse supernovae
Core-collapse supernovae are among the most powerful explosions in the universe, and they are sources of multimessenger emission, as they emit gravitational waves, neutrinos and electromagnetic radiation. In the present study, we investigate the rotation contribution in a series of state-of-the-art two dimensional core collapse supernova simulations. We identify a range of pre-collapse rotation rates in which a resonance emerges between the proto-neutron star core normal modes and rotationally driven oscillations. This resonance amplifies the gravitational-wave signal and induces a correlated modulation in the neutrino emission. Our analysis reveals that the excited oscillations can be traced back to the fundamental mode within the proto-neutron star. Resonance-driven bursts produce high-amplitude enhancements lasting hundreds of milliseconds at kilohertz frequencies, making them key observables for current and next-generation interferometers.
Orateur: Dr Marco Cusinato (CEA-Saclay) -
30
Electron-Ion Energy Partition in High Mach-Number Perpendicular Shocks
Observations, numerical simulations, and laboratory experiments relevant to supernova remnant shock waves show a significant deviation of the energy partition between electrons and ions from the naive prediction based on adiabatic compression of the electron population at the shock front. High-Mach number perpendicular shock waves are mediated by plasma instabilities driven by a beam of reflected particles at the shock transition, such as the Weibel instability. Modeling of particle heating, injection, and acceleration at the shock transition requires, at least, a proper description of their scattering by a self-generated microturbulence, their coherent gyromotion in the background magnetic field, and the interspecies coupling through the longitudinal electrostatic potential. The latter arises from the difference in inertia between electrons and ions decelerated across the shock transition.
Based on a Monte Carlo-PIC method, we explore particle transport and energization in weakly magnetized perpendicular non-relativistic collisionless shocks. Feedback on particle transport from a quasi-magnetostatic microturbulence is reduced to a Monte carlo process, while the background electromagnetic field is self-consistently solved in a 1D PIC framework. Using this 1D3V method, we investigate electron heating and non-thermal particle acceleration with increasing Alfvenic Mach number ≳ 40. Our results are benchmarked against large-scale fully kinetic 2D3V PIC simulations.
Orateur: Xingwei Gong (Observatoire de Paris) -
31
Impact of the Si/O interface on supernova explosions
Predicting the outcome of core collapse in massive stars (whether the star explodes as a supernova or collapses into a black hole) remains an open problem. The complex physics involved and the uncertainties in progenitor structure make it difficult to identify which stars are more likely to explode. Recent studies suggest that the density gradient at the interface between the silicon and oxygen shells may play an important role in determining the explosion outcome. To investigate this effect, we perform simulations of synthetic progenitors in which the density contrast at this interface is systematically varied while all other parameters are kept fixed. The simulations are carried out in two dimensions with the code CoCoNuT, including general relativistic gravity and M1 neutrino transport.
Orateur: Nicolas Houry
-
29
-
Conclusion: Supernovae
-
32
Conclusion: SupernovaeOrateur: Raphaël Raynaud (CEA Saclay)
-
32
-
-
-
Introduction: Accreting black holes
-
33
Introduction - Accreting black holes
Accreting black holes are powerful astrophysical sources whose emission provides a unique probe of the extreme environments surrounding compact objects. Their broadband radiation, from radio to X-rays and gamma rays, reveals the complex interplay between accretion disks, relativistic jets, and outflows. I will show how we can investigate the structure and dynamics of the inner accretion flows through spectral and polarisation properties, temporal variability, and multi-wavelength observations.
Orateur: pierre-olivier petrucci (IPAG)
-
33
-
Invited talks: Accreting black holes
-
34
The Growth of Supermassive Black Holes with Cosmological Hydrodynamical Simulations
Cosmological hydrodynamical simulations are a powerful tool for modelling the complexity of galaxy formation, along with the supermassive black holes (SMBHs) at their centres. I will begin by reviewing how modern simulations model SMBH growth, including gas accretion, active galactic nucleus feedback, and black hole mergers. I will then present the resulting predicted SMBH populations and show how they translate into gravitational-wave signals from SMBH coalescence, together with their possible electromagnetic counterparts. The talk will focus on two recent challenges in building a robust model of BH evolution: how to account for SMBH mergers that occur below the resolution scale of the simulation, and whether SMBHs in dwarf galaxies should be rapid accretors.
Orateur: Dr Yohan Dubois (Institut d'Astrophysique de Paris) -
35
Black Hole Accretion: Simulating Electromagnetic Observables
Three computational steps are required to construct synthetic electromagnetic observations of black hole accretion flows, specifically source modelling, general relativistic ray tracing, and general relativistic radiative transfer (GRRT). In this talk, we review various existing approaches, highlighting their respective advantages and limitations. For source modelling, we discuss semi-analytical treatments, alongside self-consistent plasma evolution through general relativistic magnetohydrodynamic (GRMHD) simulations, which describe the macroscopic fluid dynamics and magnetic-field topology of the accretion flow, and first-principles kinetic modelling through general relativistic particle-in-cell (GRPIC) methods, which capture kinetic plasma processes and non-thermal particle acceleration. We then present the GYOTO code, which combines these ingredients to produce strong-gravity observables, including those in polarised light, for arbitrary spacetimes and emitting structures.
Orateur: Irene Urso (LIRA, Paris Observatory) -
36
Turbulence in extreme environments
This talk will review some recent progress in numerical modelling of turbulence in extreme environments, its relevance for energy dissipation, particle tranport, and acceleration, and its connections to high-energy multimessenger astrophysics.
Orateur: Martin Lemoine (APC (Astroparticule & Cosmologie, CNRS - Université de Paris))
-
34
-
10:45
Coffee
-
Contributed talks: Accreting black holes
-
37
The dialectic of numerical simulations and accretion-ejection theory
It is now well established that whenever a large scale magnetic field is threading an accretion disk, super fast-magnetosonic outflows are launched transporting mass, angular momentum and energy from the disk, and allowing thereby accretion. This interdependent accretion-ejection process is nowadays theoretically understood, at least in a qualitative way.
There are however several aspects that hinder a more quantitative understanding. Global 3D numerical experiments are highly valuable, as they allow to investigate complex time-dependent situations, including the coupling with the central object. However, they are biased by their initial (and sometimes boundary) conditions and their limited duration forbids to see them as representing astrophysical objects.
These caveats should encourage us to confront numerical experiments with accretion-ejection theory. Several examples of such fruitful comparisons, on disks as well as jets, will be presented. We therefore advocate a more informed use of 3D global MHD simulations, geared more towards the study of specific processes. MHD turbulence in accreting systems, whose effects extend far beyond the scope of a simple viscosity law, is one such example.
Orateur: Jonathan Ferreira (IPAG) -
38
DYPLO–MONK: Coupling Global Radiative Transfer with Accretion–Ejection Physics
Accretion flows around black holes result from a complex combination between radiative cooling, advection and ejection. The Jet Emitting Disk–Standard Accretion Disk (JED–SAD) model provides a self-consistent description of these processes. Its thermal structure is computed with DYPLO, a radial solver that determines the local thermal equilibrium of the accretion flow, including quantities such as the electron and ion temperatures, density and disk scale height. However, the radiative cooling in DYPLO is computed locally, treating each radial ring independently and therefore neglecting photons produced in other regions of the accretion flow. To overcome this limitation, we couple DYPLO with MONK, a general-relativistic Monte-Carlo radiative-transfer code. By following photons throughout the accretion flow, MONK naturally treats both local and non-local radiative interactions, while accounting for the geometry of the JED–SAD structure. The cooling rates computed by MONK are iteratively fed back into DYPLO until a self-consistent thermal and radiative solution is reached.
We present this new DYPLO–MONK coupling and investigate how global radiative cooling modifies the accretion flow, from its thermal structure and physical parameters to the resulting X-ray spectra. The coupling provides a self-consistent treatment of global radiative transfer, accretion (including advection), and ejection, linking the physical structure of the flow directly to its emerging spectrum.Orateur: Rafan Muhammad -
39
The Fig Code: A GPU Radiative Transfer Code for GRPIC Simulations and More
The geometry of black hole coronae, which is responsible for the hard X-ray emission in X-ray binaries, is still poorly constrained. Open-source radiative transfer codes are key to bridge the gap between observed spectra and semi-analytical models or first-principles plasma simulations.
In this talk, I present Fig: a new ray-tracing code usable as a stand-alone and for post-processing simulations such as GRPIC simulations. Fig handles the Inverse Compton phenomena for polarized photons with Monte Carlo methods.
I present the benchmarks and study cases of the code, such as a polarized image of a Kerr black hole with a Novikov-Thorne disk. Finally, I propose a few cases of applications for Fig.Orateur: M. Élie Sévègnes (Institut de Recherche en Astrophysique et Planétologie)
-
37
-
Flash talks: Accreting black holes
-
40
How do the first Massive Black Holes Grow?
JWST observations up to z ≈ 11 have uncovered unexpected populations of high-redshift AGN, suggesting a much higher abundance of massive black holes (MBH) in the early Universe than predicted by prior extrapolations of observational datasets. However, state-of-the-art simulations of MBH accretion yield results inconsistent with these findings: they suggest that massive black holes grow too inefficiently in low-mass galaxies—systems similar in mass to the hosts of high-redshift AGN detected by JWST—to explain such observed abundance at these early times. Indeed, star formation and supernova feedback cause gas to become very turbulent, and it is continuously stirred and ejected by stellar explosions, thus suppressing MBH growth.
Hence, the goal of my PhD thesis is to assess how stellar feedback influence MBH growth. This type of feedback is not only SNe : Cosmic Rays, pre-heating by OB stars and modeling of star formation itself are important. To try to comprehensively study these phenomena, I run zoom-in cosmological simulations focusing on individual galaxies at z ≥ 4 in their environment, with increasing physical complexity, which will lead to “full-physics” models with radiative and magneto-hydrodynamical simulations. I realize “agnostic” simulations, i.e. I do not artificially enhance MBH growth to make it consistent with JWST observations, but instead try to independently understand the concerned parameter space. Depending on the conclusions, it could indicate that perhaps some observational results on high-z black hole masses and the number of AGN have to be taken with a grain of salt.Orateur: Emile Dosso (Institut d'Astrophysique de Paris) -
41
Detailed Modeling of Stochastic Particle Acceleration and Multi-Messenger Emission in AGN Coronae
Stochastic particle acceleration in magnetized turbulent plasmas, such as those found in the coronae of Active Galactic Nuclei, has emerged as a key mechanism for explaining their multi-messenger emission. However, a major challenge lies in bridging the gap between the microphysical processes that govern particle acceleration and the macrophysical scales of these compact environments.
In this talk, we present a hybrid numerical framework that self-consistently models non-linear kinetic processes, including turbulence-driven particle acceleration and its feedback on the turbulent cascade, as well as hadronic and radiative losses. The code solves transport equations that incorporate recent theoretical developments in stochastic acceleration.
This approach enables robust multi-messenger predictions. By comparing the resulting emission in the coronal context with neutrino observations of Seyfert galaxies from the IceCube neutrino detector, we can gain new insights into the physical conditions in AGN coronae and place constraints on both particle acceleration mechanisms and the underlying plasma environment.Orateur: Sébastien Le Bihan (APC - CNRS,UPC) -
42
Hunting for Steady, Luminous, Thermal Black Hole Disks in angle-dependent RGRMHD simulations
Classical thin-disk theory explains the spectra observed in Black hole accretion disks but predicts thermal instability in the luminous regime, in sharp contrast with the long-lived soft states of X-ray binaries and the sustained luminous activity of bright quasars. Recently, the addition of full angle-dependent radiation transport in general relativistic radiation-magnetohydrodynamic simulations has revealed two stable thermal disk solutions with distinct magnetic structures and radiation-production channels, namely a magnetic puffy disk and a newly identified current-sheet thin disk.
Here, we show that the current-sheet disk is the only solution consistent with the observational properties of soft-state X-ray binaries. Most notably, it reproduces the observed low variability, with variability power below 10^-5 and a very weak dependence on mean flux, as most of the radiation is produced near a thin current-sheet layer at the midplane. By contrast, the alternative puffy disk solution produces much stronger variability than observed. These results suggest that the new current-sheet disk solution may represent a common structure of luminous black hole accretion disks, and establish timing variability as a probe of the otherwise inaccessible magnetic structure of accretion flows.
Orateur: Maxime Parra (Ehime University Japon) -
43
General Relativistic Simulation of Quasi-Spherical Accretion Flows in Dark Matter Halo
We have investigated the influence of dark matter halos on quasi-spherical accretion onto supermassive black holes (SMBHs) using general relativistic hydrodynamic (GRHD) simulations. By incorporating Hernquist and NFW halo profiles into the BHAC code, we model accretion flows with zero and low angular momentum across a range of halo-to-BH mass ratios up to 10. Our results show that the presence of a DM halo can enhance the mass accretion rate and significantly alter the flow morphology, including density, velocity, and temperature structure. We also explore how rotation interacts with DM gravity to regulate inflow and outflow dynamics. These findings provide insights into SMBH feeding in DM-rich environments, especially relevant for high-redshift quasars, and emphasize the need to account for halo gravity in modeling black hole growth.
Orateur: Razieh Ranjbar (IPM - Institute for Research in Fundamental Sciences)
-
40
-
Conclusion: Accreting black holes
- 44
-
13:00
Lunch
-
14:00
Coffee
-
Introduction: Magnetospheres
-
45
Introduction: MagnetospheresOrateur: Lucas Guillemot (LPC2E/CNRS and Université d'Orléans)
-
45
-
Invited talks: Magnetospheres
-
46
Population synthesis of pulsars and magnetars
Population synthesis is a powerful numerical tool that allows one to infer the parameters characterizing a given population.
Within this framework, pulsars and magnetars are generated at birth and their evolution is followed up to the present epoch.
Once this evolution is modeled, detection criteria are applied in order to determine whether each object in the simulated population would be observable with current or future instruments. In this presentation I will especially show results of population synthesis of recycled pulsars and a few results about magnetars.Orateur: Mattéo Sautron (ObAS) -
47
A kinetic plasma perspective on black hole magnetospheres
Recent high-angular-resolution observations of black holes have revealed magnetospheric structures in their immediate vicinity, which may drive active galactic nuclei jets and particle acceleration. Fundamental questions persist regarding jet launching and the variability of electromagnetic emission. In this talk, I present how general relativistic kinetic plasma simulations have transformed our understanding of black hole magnetosphere dynamics. I will discuss the robustness of their predictions and the current computational challenges.
Orateur: Enzo Figueiredo (IPAG, CNRS, Université Grenoble Alpes) -
48
Fast radio bursts as precursor radio emission from monster shocks
Fast radio bursts (FRBs) are among the most luminous coherent transients in the Universe, yet the physical mechanisms responsible for their emission remain poorly understood. In this talk, I will explore the kinetic physics underlying a theoretical scenario in which low-frequency fast magnetosonic waves, launched by magnetar activity, steepen as they propagate through the declining background magnetic field, eventually forming relativistic “monster shocks.”
Using first-principles particle-in-cell simulations, we show that these shocks efficiently dissipate the driving-wave energy while generating coherent electromagnetic precursors, with frequencies reaching the GHz band under conditions expected in magnetar magnetospheres. To understand the underlying dissipation and emission mechanisms, we developed a stationary multifluid model that describes the kinetic-scale shock transition as a sequence of nonlinear solitonic structures. We show that shock dissipation and thermalization arise primarily from chaotic particle dynamics and phase mixing.
Building on this framework, we investigate the role of radiation reaction in the extreme cooling regime relevant to magnetars, where particles can lose their energy within a single gyration period. Combining an extension of the multifluid model with kinetic simulations, we identify the conditions for coherent emission and derive the scaling of its characteristic frequency with radiation reaction. Remarkably, we find that monster shocks remain efficient sources of coherent radiation even under strong cooling.
These results establish a physical connection between large-scale magnetospheric disturbances, nonlinear collisionless shock dynamics in extreme regimes, and fast radio burst generation.
Orateur: Arno Vanthieghem (Observatoire de Paris)
-
46
-
Contributed talks: Magnetospheres
-
49
Fast radio bursts from the non-linear evolution of low frequency plasma waves in the magnetosphere of magnetars.
Magnetars are highly magnetized neutron stars characterized by their frequent, energetic X-ray bursts. Dynamic motion of magnetar crust has been theorized to fill the magnetosphere with low frequency plasma waves that eventually power the transients we observe. It was recently realized that fast magnetosonic (fast) waves, one of the two low frequency wave modes in the highly magnetized limit, can become nonlinear within the magnetosphere and steepen into some of the strongest shocks in the universe. In certain parameter regimes, these shocks can simultaneously produce hard X-rays and a GHZ precursor wave suggestive of fast radio bursts. We present the first global Particle-in-Cell simulations of the non-linear steepening of fast magnetosonic waves in a dipolar magnetosphere, and quantitatively demonstrate the strong plasma acceleration in the upstream of these shocks. We comment on both the high energy emission and precursor emission that will be produced from such extreme shocks. We predict the expected frequency, power, angular extent, and duration of the precursor emission and describe the properties of FRB observations that can be reproduced by these precursor waves.
Orateur: Dominic Bernardi (IPAG) -
50
Modeling radiation-rich magnetar magnetospheres using Particle-in-Cell simulations
Magnetars are highly magnetized neutron stars whose persistent hard X-ray emission is thought to arise from resonant inverse Compton scattering of photons by relativistic particles in their magnetospheres. Unlike pulsar magnetospheres, radiation-rich magnetar magnetospheres remain largely unexplored with self-consistent kinetic simulations. Global Particle-in-Cell (PIC) simulations are an ideal tool for such environments; however, they require very high resolutions and consequently significant computing power. These constraints result from the extreme surface magnetic field leading to large spatial and temporal scale separation (kinetic scales vs. system size and radiative cooling vs. dynamical timescales). To make such simulations numerically tractable, the common practice is to employ rescaled parameters to artificially reduce the scale separation. This procedure can be delicate as the radiative treatment is also affected by the rescaling and could lead to numerically driven radiation instead of the physically driven one.
In this talk, I will describe how we implement resonant scattering in our PIC simulation. I will show results from 1D simulations where we model the plasma along different twisted field lines, and discuss the ongoing development of global 2D simulations and their numerical challenges.
Orateur: Adrien Soudais (Dartmouth College) -
51
Influence of black hole spin and statistical properties of magnetospheric eruptions in GRPIC simulations
Recent observations of Sagittarius A* have revealed the presence of a hotspot orbiting close to the black hole during near-infrared and X-ray flares.
Magnetospheric eruptions have recently been proposed as a possible explanation for these flares. Using axisymmetric General Relativistic Particle-In-Cell (GRPIC) simulations, Figueiredo & al., 2026, identified quasi-periodic cycles associated with successive phases of plasma accretion and ejection. In this work, we investigated the influence of black hole spin on these cycles and we examined the statistical properties of the eruptions.Our simulations showed no clear dependence of the main characteristics of the cycles on black hole spin. The statistical analysis revealed a correlation between flare amplitude and the maximum energy of accelerated particles. These results suggest that the eruptive behaviour is primarily driven by the accumulation and release of magnetic flux, rather than directly by black hole spin.
This work contributes to our understanding of the physical mechanisms driving magnetospheric eruptions and provide a basis for future three-dimensional and radiative studies of eruptions around black holes.Orateur: Maïa Villemin (IPAG)
-
49
-
Flash talks: Magnetospheres
-
52
Fast Radio Burst Generation by Relativistic Magnetized Shocks in Magnetar Magnetospheres
Fast Radio Bursts (FRBs) are extremely bright millisecond-duration radio transients whose physical origin remains poorly understood. Among the leading progenitor candidates are magnetars: highly magnetized neutron stars capable of powering large-amplitude disturbances in their magnetospheres. In particular, starquakes can excite fast magnetosonic waves propagating across the background magnetic field. As these waves travel outward in a declining magnetic field, nonlinear effects can lead to progressive steepening and eventually to the formation of relativistic magnetized shocks. Such shocks are promising sites for the generation of coherent electromagnetic emission in the GHz band, potentially associated with FRBs. Motivated by these considerations, the nonlinear evolution and dissipation of strongly magnetized waves are investigated through the development of a fully kinetic theoretical framework supported by large-scale particle-in-cell simulations encompassing both kinetic and MHD scales.
Orateur: Gaia Bilosi (LUX) -
53
Pulsar ARoMa: a code for global simulations of neutron star magnetospheres within realistic parameters
Neutron star magnetospheres are the environments of magnetised and rapidly rotating neutron stars which often manisfest through the phenomenon of pulsar. From the electron Larmor radius to the radius of the neutron star, the tremendous scale separation has made simulations of such environments extremely challenging. While resistive force-free codes provide a good description of the magnetosphere plasma but cannot afford radiative processes, Particle-In-Cell (PIC) codes have to truncate physical quantities by many orders of magnitudes to resolve the plasma microphysics and radiative processes. Recently, Mottez (2024) presented simulations of electrospheres, i.e. the environments of weakly energetic magnetospheres (not pulsars), with a new simulation code: Pulsar ARoMa. Thanks to the combination of methods from PIC and Vlasov codes, with a spectral solver for Maxwell equations, it is able to use realistic parameters and scale separations in global simulations of neutron star magnetospheres, at a reduce computational cost.
Two years after this article, Pulsar ARoMa has undergone updates and faced new challenges. We present the current state of the particle and electromagnetic field solver. Radiative processes related to the physics in electrospheres have been successfully implemented, using similar methods as for particles, and we were able to produce the first gamma-ray spectra of electrospheres. To extend the code to the simulations of active pulsar magnetospheres, numerical methods for electron-positron pair creations have also been designed.Orateur: Théo Francez (LUX - Observatoire de Paris - PSL) -
54
Unravelling the emission of old pulsars at the highest energies with Fermi-LAT
Pulsars are highly magnetized, fast-spinning neutron stars, among which millisecond pulsars (MSPs) form the oldest population. Pulsars are powerful particle accelerators, and thus efficient gamma-ray emitters, with some expected to reach the very-high energy (VHE, >100 GeV) scale. Nonetheless, VHE detections have, to date, been restricted to “normal”, young pulsars.
Unraveling the VHE emission from MSPs remains an important challenge in high-energy astrophysics. In this presentation, I will present an updated analysis of 15 years of gamma-ray data from the Fermi Large Area Telescope (LAT) towards known MSPs, along with a stacking analysis of the MSP sample to derive the combined source-class spectral energy distribution. I will also assess the reliability of the Fermi-LAT pipeline in reconstructing possible high-energy tails in MSPs spectra through dedicated simulations.Orateur: Léo Lam Kon Seng (LAPTh) -
55
Spin-orbit alignment hypothesis in millisecond pulsars
Millisecond pulsars are spun up during their accretion phase in a binary system. The exchange of angular momentum between the accretion disk and the star tends to align the spin and orbital angular momenta on a very short time scale compared to the accretion stage. This alignment can also be seen as the equality between two angles : the orbital inclination angle i, corresponding to the angle between the line-of-sight of an observer and the orbital angular momentum ; and ζ the viewing angle, corresponding to the angle between the line-of-sight and the rotation axis. In this work, we studied a subset of gamma-ray millisecond pulsars in binaries for which the orbital inclination angle i has been accurately constrained by observations. Our goal is to constrain the observer viewing angle ζ and to check whether it agrees with the orbital inclination angle i. We use a Bayesian inference technique to fit the millisecond pulsars gamma-ray light curves based on the third gamma-ray pulsar catalogue (3PC). The emission model relies on the striped wind model deduced from force-free neutron star magnetosphere simulations. We found good agreement between the two angles i and ζ for a significant fraction of our sample, about four fifth, confirming the spin-orbit alignment scenario during the accretion stage. However about one fifth of our sample deviates significantly from this alignment. The reasons are manifold: either the gamma-ray fit is not reliable or some precession and external torque avoid an almost perfect alignment.
Orateur: Alexandra Lorange
-
52
-
Conclusion: Magnetospheres
-
56
Conclusion: MagnetospheresOrateur: Guillaume Voisin (LUTH, Observatoire de Paris, CNRS)
-
56
-