E5: Numerical Observatory of Violent Accretion systems NOVAs strong gravity and beyond

> Read the articles connected to the project.



    first GR-HD simulation of a thin disk equilibrium GYOTO view of a GR accretion disk

    Collaboration between AIM, APC and LUTH


    The advent of high energy observation facilities in the last decades has proven the existence of powerful mechanisms emitting photons up to gamma-rays. It is now commonly admitted that the most energetic events are associated with compact objects believed to be relics of massive stars. These objects are prone to the most extreme gravity fields and are likely efficient attractors of the plasma present in their vicinity. The motion of plasma in the close neighborhood of compact objects is only properly described in the framework of general relativistic magnetohydrodynamic (GR- MHD). The equations governing GR-MHD are so complex that the only way to solve them is trough large-scale numerical simulations.

    The topic of the project is to sustain a computational effort dedicated to GRMHD simulations of accretion flows near compact objects and to link them to synthetic observations of the associated violent events, one of the major themes within LabeX UnivEarths.


    General Objective of the Project


    The quest to unveil the nature of the compact objects detected in the center of radio-loud galaxies, in X-ray binaries and also at the center of the Milky Way, is at the crossroads between high resolution, time-dependent, multi-wavelength observations and the coming of age of numerical GR-MHD codes. The project plans to exploit this opportunity by setting a long- standing collaboration between members of three teams involved in the LabeX UnivEarths which are already tackling, individually, the problem from the different angles of analytical and numerical studies, observations, and high-performance code development.


    While the french community is, at the moment, lacking a numerical code able to describe the dynamics of a magnetized plasma in a full GR framework, the LabeX UnivEarths benefits from a unique conjunction where plasma numericists, high-energy observers and GR numericists are present in the same area and have had a good track-record on common past projects.


    By joining forces, our aim is to sustain the development of such a GR-MHD code in order to perform large-scale simulations of plasma accretion onto compact objects and subsequently create an ”observation” from it. Another innovative feature of our code is stem from its ability to handle any kind of GR metrics, not only conventional metric such as Schwarzschild or Kerr, but also metrics of alternative, non-GR, gravitation theories which generally do not have any analytical expression.


  • The three teams and interaction

    Our team, while approaching a new subject linking their respective research, is composed of people used to working together on a variety of topics such as GRBs and planet formation. For this project we plan on using our distinct approaches to tackle the creation the first full GR- MHD code coupled to a ray-tracing code able to provide synthetic observations of the environment of compact objects.

    WP leader CASSE Fabien APC MCF (Univ. Paris Diderot)
    WP co-leader RODRIGUEZ Jérôme AIM Ing. Chercheur, CEA
    WP co-leader MELIANI Zakaria LUTh AA, CNAP
    WP member VINCENT Frédéric LUTh CR, CNRS
    WP member DEMIDEM Camilia APC PhD student (started 2016/10)
    WP member CANGEMI Floriane AIM PhD student (student 2017/10)


    AIM – high-energy observers axis led by J.Rodriguez. They are expert of multi-wavelengths observations of compact objects and have access (through dedicated and collaborative programs) to a wide range of observations. This broad range of data have allowed them to tackle the problems of accretion-ejection connections and mechanisms prevailing in those systems, including also the study of rapid X-ray variability.


    APC – ADAMIS simulation axis led by F. Casse and P. Varniere. Their expertise ranges from analytical to MHD numerical studies of the accretion-ejection system. They also do the extra step of linking their results with observations. They are developers in the mpi-amrvac project.

    LUTH – Relativistic fluid simulations axis, led by Z.Meliani and E.Gourgoulhon. ZM is long term developer on the mpi-amrvac project is now fully concentrating on implementing and using GR-MHD to study compact objects. EG is a leader in strong gravity calculation.



    Over the last four years, we have managed to develop a new general relativistic (GR) fluid code aiming at studying the behavior of plasmas prone to extreme gravitational fields, namely in the vicinity of any kind of compact objects, and fully coupled it with ray-tracing to get spectral and timing synthetic observations. The numerical progress we made during the first years have opened the door to new astrophysical fluid studies while we carry on efforts in data processing in order to access the physical conditions prevailing in accretion flows orbiting around compact objects. It is noteworthy that we are now harvesting scientific results as the number of refereed papers stemming from WP NOVAs has reached 27 over the last 5 years.

    This WP focus on all aspects of compact objects’ fast variability, from observations to numerical simulations of our models. Indeed, we lead on two models that cater to the low and high frequency Quasi-Periodic Oscillations observed in the Power Density Spectrum of those sources. Using numerical observation of our theoretical models we are using those QPOs to infer what is happening in the source.

    Exploring alternate compact object models

    While microquasars are thought to be black-hole binaries, the exact nature of the compact object remain open. The detections of gravitational waves whose signal is consistent with coalescent binary black holes strongly supports such assumption but does not yet entirely rule out other possibilities such as the boson star model. We have been running various type of simulations of fluid dynamics orbiting around such objects and showed they share similar features with black holes. However, we have also pointed out distinctive accretion/ejection features. Meliani et al. (CQG 2016a,2016b,2017).

    Instability in accretion flows around black holes

    In the context of accretion disks orbiting around black holes we focused on the physics of a fluid instability which may explain observational features occurring in the vicinity of compact object systems. Indeed, the Rossby wave instability (RWI) has been proposed to explain radiative emission variability in microquasars as well as supermassive black hole such as SgrA*. Using our newly developed code we performed the first in-depth study of this instability in a general relativistic framework by first following the instability from a Newtonian distance and getting closer and closer to the last stable orbit of the disk in the case of a null spin and then continued on for higher spin. Such study allows us to investigate the minute changes in the behaviour of the instability and see if it could lead to observational differences. We demonstrated that high-spin black holes do influence the development of the instability within the disk. This influence is twofold as it first induces a higher saturation level of the instability (enhancing then the detected rms), and also impact the perception of time, hence, in order to “observe” our simulation, we need to go through a full GR ray-tracing of the emission. As a result, we now have fully coupled our GR code with the GR raytracing Gyoto and are able to produce synthetic observations of our simulation. (Casse et al MNRAS 2017, Casse & Varniere MNRAS 2018).

    Quasi-Periodic Oscillations (QPOs) in X-ray binary systems

    Shading a new light on known behavior: One of the strength of NOVAs is its ability to perform test, for example, can we detect QPO in the energy spectrum?  Up to now the QPO are solely a timing-feature, i.e. absent from the spectral analysis, even while they can represent up to 40% of the flux. Using NOVAs we were able to compute the simulated energy spectra of the {full system + QPO} and then use the data reduction software XSPEC to study the impact of this QPO on the energy spectrum and determine when, if ever, it is acceptable to neglect them in the spectral fit. The simulated data we obtained is coherent with the observed, and up to now unexplained, departure from correlation seen in the Tin-rin diagram of XTE J1550-564 in presence of HFQPO and LFQPOs B or A. Thanks to our simulations we were able to explain this departure.  Indeed, in the case of a QPO with an amplitude as small as 5% rms we cannot neglect the presence of the hot structure as it leads to systematic errors in the fits. Therefore there is a need to improve the disk fitting by taking into account the structures at the origin of the QPOs if we want to constraint the disk parameters in their presence (Varniere et al., A&A 2016). Following this publication, there has been a demand to make an extended version of our XSPEC model public, allowing it to be used to fit observations. This was developed and delivered to members of the NICER team. This will not be a distinct publication but the model is given on demand until the website is ready.

    . Another aspect of NOVAs is its ability to search for explanation, such as finding the origin of the surprising change of correlation, exhibited by all sources, between the frequency of the QPO and its rms strength. During one outburst the source start on the left (low frequency) with the rms amplitude increasing as the frequency increases, as one would expect from a growing instability. Around a few Hz, the frequency continue to increases but the rms now decrease, well pass its start rms. Using simulation of our QPO model developed within the NOVAs framework we were able to reproduce this behavior and link it to the simultaneous spectral behavior of the source, thus being the first QPO model able to explain such change in the correlation (Varniere & Vincent, A&A 2016, Varniere & Vincent, ApJ 2017).

    Exploring new behavior: While the Kerr version of our GR code was being tested we worked on exploring in more details the observations of the elusive HF-QPO. We were in particular interested in the 3:2 ratio that has been said to be a preferred ratio for the HFQPOs in several sources but not always detected because of the detection limit. In order to lower the detection limit we created a new method to add-up observation having similar timing and energy spectrum. This lead us to five groups out of which we detected two pairs of HFQPOs, and lowered by a factor of four the limit on the presence of an undetected QPO on the three remaining groups. Interestingly none of the pair we detected where in the  3:2, which lead us to a modified list of requirement for a HFQPO models. (Varniere & Rodriguez, ApJ 2018).

    In the spirit of the NOVAs project we then turn to our newly obtained GR-HD simulations of the RWI which gave us simulated lightcurve through the coupling with Gyoto. Using those we created the synthetic PDS associated with different parameters in the disk. Here the aim is dual, first we demonstrate the ability of the RWI to produce timing features similar to the observed one, and secondly we are exploring in which cases we obtain the different ratio observed between the QPO peaks. This two paper together translate the observed constraints into the physical state of the system. (Varniere et al., subm. to ApJ).

    As a side project we encountered some direct spin effects that could be at the origin of some systematic fitting difficulties that exist for all high-spin systems. We are now working on observational checks.

    Non-thermal particle acceleration in relativistic plasmas

    In the last two years, a particle-in-cell (PIC) treatment of supra-thermal particles have been incorporated in NOVAs. Coupling PIC and MHD methods is an innovative way to depict the mutual interaction between these particles and the background relativistic fluid hence allowing us to compute non-thermal radiative emission. In early 2018 we have published the first-ever fully consistent PIC/MHD simulations describing the acceleration of cosmic-ray altogether with magnetic field amplification occurring near planar astrophysical shocks. This study has brought unprecedented results regarding the impact of the obliquity of the magnetic field with respect to the shock (van Marle et al. MNRAS 2018). Using the increase in numerical performances of the PIC/MHD framework, we performed the first full 3D generalization of such process (van Marle et al., subm. to MNRAS). For the first time in the literature, we have a complete picture of the turbulence process as well as a clear understanding of the particle acceleration mechanism near planar front shocks. In Demidem et al. (MNRAS 2018) we took one step beyond the previous planar shock assumption by studying the dynamical response of a relativistic shock fronts to incoming perturbations. We have demonstrated that relativistic shocks, as the ones occurring near compact objects, are prone to corrugation when turbulent perturbations reaches the shock front. Such shock oscillations are of particular interest as they open the door to study of particle acceleration near non-planar shocks, namely in more realistic physical conditions.





    Relativistic magnetohydrodynamical simulations of the resonant corrugation of a fast shock front
    Demidem, M. Lemoine & F. Casse, 2018, MNRAS Vol. 475, 2713

    Impact of the gravity of a Schwarzschild black hole upon the Rossby wave instability
    Casse, P.Varniere & Z. Meliani, 2017, MNRAS Vol. 464, 3704

    Looking for the Elusive 3:2 Ratio of High-frequency Quasi-periodic Oscillations in the Microquasar XTE J1550564
    Varniere & J. Rodriguez, 2018, ApJ Vol. 865, 113

    On the Rossby Wave Instability in accretion discs surrounding spinning black holes Casse & P. Varniere, 2018, MNRAS Vol. 481, 2736


    Shocks in relativistic transverse stratified jets, a new paradigm for radio-loud AGN.
    Hervet, Z. Meliani et al., 2017, A&A (in press)


    On tidal disruption of clouds and disk formation near boson stars

    Meliani, F. Casse, P. Grandclement, E. Gourgoulhon, 2017, Class. & Quant. Gravity (in press)


    On magnetic field amplification and particle acceleration near non-relativistic astrophysical shocks:  Particles in MHD Cells simulations

    A.J. van Marle, F.Casse & A. Marcowith, 2017, MNRAS (in press)


    Single-dish and VLBI observations of Cygnus X-3 during the 2016 giant flare episode

    Egron, E. et al, 2017, MNRAS Vol. 471, 2703


    Reproducing the Correlations of Type C Low-frequency Quasi-periodic Oscillation Parameters in

    XTE J1550-564 with a Spiral Structure

    Varniere & F. Vincent, 2017, ApJ Vol. 834, 188


    Refereed proceedings :


    On magnetic field amplification and particle acceleration near non-relativistic collisionless shocks:

    Particles in MHD Cells simulations

    Casse, A.J. van Marle, A. Marcowith, 2017, Plasma Physics & Controlled Fusion (in press), Invited talk at the 44th euopean conference on Plasma Physics (Belfast, UK), June 2017.


    Using a combined PIC-MHD code to simulate particle acceleration in astrophysical shocks

    A.J. van Marle, F. Casse & A. Marcowith, 2017, Proceedings of Science (in press), Talk at the 35th International Cosmic Ray Conference (ICRC) , Busan, South Korea, July 2017