SHERPAS Team

The High Energy Group

of LAOG

In the eigthies, the study of non-thermal phenomena in AGNs was the main research activity of Guy Pelletier, who came from the plasma physics community. This was essentially the only theoretical work in the laboratory (Dir. Alain Omont) which, at that time, was originally oriented towards Radio-Astronomy with a focus on the interstellar medium. This AGN activity then gave birth to a group in the early nineties when Gilles Henri, coming from the University of Paris XI got a permanent position at the University of Grenoble and when two brilliant students, Francoise Rosso and Jonathan Ferreira, were hired for a thesis program on Astrophysical Magnetohydrodynamics (MHD) with G. Pelletier. We thus started a larger activity devoted to accretion-ejection for both AGNs and Young Stellar Objects (hereafter YSOs) with the students and a new activity devoted to the High Energy aspect of AGNs with Gilles Henri (the first gamma spectra of some AGNs were about to be produced by EGRET).

The laboratory was about to increase dramatically with the opening of Infra Red activity and the coming of a group involved on Young Stellar Objects was under discussion. We were particularly in favor of welcoming this new component because this was about to open a new field of collaboration in the laboratory and we warmly supported the arrival of Claude Bertout and his team. Then the laboratory was metamorphosed and our group, which hired new permanent researchers from Toulouse (Pierre-Yves Longaretti and Didier Fraix-Burnet), took the name SHERPAS (Sources of High Energy Relativistic Plasmas and Accretion-ejection Structures). It emphasizes both the type of objects we deal with, namely sources of high energy powered by accretion-ejection, and the type of physics we develop, namely, MHD, relativistic plasma physics and transfer of high energy radiation.

Table 1 lists the current permanent staff with their main activity. In 2005, the group hosts Yael Fuchs with an ATER position (one year) and three graduate students: Clement Cabanac (started in 2003), Geoffroy Lesur (started in 2004) and Nicolas Bessolaz (started in 2004, in co-direction with the FOST team). In Fall 2005 a post-doc, Claudio Zanni, has joined the group for 2 or 3 years, as well as another PhD student Timothe Boutelier.

Table 1: SHERPAS group members in 2005.

Among the teams working on MHD astrophysics, high energy cosmic phenomena and the environment of compact objects, we adopted the following attitude. Instead of running after scoops, we firmly decided to develop ground researches about the physical issues and the interrelations between the gross phenomena governed by gravitation and MHD, the kinetic level of relativistic plasmas, including the physics of particle acceleration, and the physics of the high energy photon emission. Therefore, we are mostly involved in theoretical developments together with numerical simulations. However, we also provide our expertise by participating in large collaborations organized for the exploitation of facilities in the whole energy range from millimeter to TeV gamma range.

Accretion-ejection phenomena are common-place in astrophysics. They are present in the cores of active galaxies (AGNs) and quasars but also around compact objects such as X-ray binaries or even some cataclysmic variables within our galaxy. It is the major physical process governing the growth rate of young stellar objects (YSOs). Accretion-ejection is therefore a process of major importance. It has long been recognized that the high degree of collimation exhibited by jets from AGNs or YSOs requires a self-confinement process which can only be provided by large scale magnetic fields carried in along the jet. This gave rise to MHD models of jet formation and collimation. On the other hand, all these objects display a correlation between accretion and ejection observational signatures (Hartigan, Edwards & Ghandour 1995, ApJ, 452, 736). These correlations gave birth to the idea that accretion and ejection were actually interdependent processes.

Our team was the first to identify the concept of a Magnetized Accretion-Ejection Structure (MAES, Ferreira & Pelletier 1993, A&A, 276, 625). In such a structure a large scale magnetic field of bipolar topology is threading the disk. The field exerts a torque which takes away the disk angular momentum thereby allowing it to accrete towards the central object. A turbulence is needed in order to allow mass to steadily diffuse through the magnetic field. This angular momentum and energy is then transferred back to a small fraction of the disk material which gives rise to a self-confined MHD jet. In contrast with other teams, the MAES has been computed by solving the exact equations of both the disk and the jets: usually, the disk is either ignored (taken as a mere boundary condition, e.g. Blandford & Payne 1982, MNRAS, 199, 883, Shu et al. 1994, ApJ, 429, 781) or its vertical structure is crudely approximated (e.g. Wardle & Konigl 1993, 410, 218, Li 1995, ApJ, 444, 848). By taking into account all terms (which was possible thanks to a self-similar ansatz), we were able to provide the full parameter space of MAES with strong consequences on the level of the required MHD turbulence (Casse & Ferreira 2000a, A&A, 353, 1115, Casse & Ferreira 2000b, A&A, 361, 1178). This work has been recently extended by producing the only solutions of jets from accretion disks that cross the three MHD critical points (Ferreira & Casse 2004).

We are now endowed with the only available model in the literature of disk-driven jets which provides all fields (density, velocity and magnetic fields) as functions of the disk physical conditions in a consistent way.

Since jets from YSOs are cooling by optically thin emission lines it is possible to derive strong constraints on their dynamics and discriminate models. Thus, most past work has been devoted to YSOs. The application of the MAES model to compact objects is only beginning, with a focus on microquasars .

Using the self-similar MAES model, we were able to compute synthetic observations and compare them to real observations. A previous work done in collaboration with observers (Garcia et al. 2001a, A&A, 377, 589, Garcia et al. 2001b, 377, 609) showed that most of the T-Tauri optical jet properties (line profiles, flux, jet velocities, evolution of the jet diameter along the distance) could be easily explained by disk-driven jets. However, observations require a rather large ejection to accretion ratio which can only be attained when there is some heat deposition at the disk surface layers. This was again confirmed by near-IR modeling of the jet emission (Pesenti et al. 2003). Finally, when taking into account observational biases in the detection of jet rotation, disk-driven models with heat deposition are the best candidates (Pesenti et al. 2004, Ferreira et al. 2005, submitted). This work has been done in collaboration with members of the FOST team. Such heat deposition has to come from local dissipation of the accretion energy, even in the presence of illumination by stellar UV and X-ray fluxes (Garcia et al. 2005, submitted).

Microquasars are X-ray binaries where the primary is a black hole, displaying jets with an intriguing time variability. Indeed, they change their spectral state from a High/Soft (dominated by a soft disk component) to a Low/Hard (dominated by a hard power-law emission) state on time scales of hours. Jets are only seen during the Low/Hard state. Moreover, it has been recently shown that this evolution is following an hysteresis cycle. Therefore, these objects provide valuable clues on the secular evolution of the accretion-ejection process (their dynamical time scale is the millisecond). Such transitions between spectral states would be unobservable for AGNs.

Within our framework, a MAES is settled in the innermost disk regions surrounded by a standard accretion disk as illustrated in Fig. 1 (Ferreira et al. 2005). This picture allows to explain several aspects of the microquasar phenomenology: jet production in the Low/Hard state, jet quenching in the High/Soft state, superluminal flares (pair plasma) under certain circumstances. Although each spectral state can be explained by varying the relative importance of each component, several crucial questions remain to be investigated. This requires a coupling between MHD and high energy physics and is therefore a central theme of our group.

Figure 1: A standard accretion disk (SAD) fed with an accretion rate of 0.01L_Edd/c² is established down to a radius r_J which marks the transition towards a jet emitting disk (JED), settled down to the last stable orbit. The JED is driving a mildly relativistic self-collimated electron-proton jet (MAES) which, when suitable conditions are met, is confining and inner ultra-relativistic electron-positron beam (the so-called two-flow model). Field lines are drawn in red solid lines and the number density is shown in greyscale (log₁₀ n/m³).

The exceptional propagation length of jets, as compared to their radii, raises the question of the physical mechanisms responsible for their stability. This question has at least two different aspects:

(1) Jet global stability properties. Purely hydrodynamical jets are quickly destroyed due to the development of the Kelvin- Helmholtz instability. MHD jets seem to be more stable with respect to Kelvin-Helmholtz modes. However, such MHD jet are prone to be unstable with respect to purely MHD (current- and pressure-driven) instabilities, the outcome of which is still unknown on theoretical grounds. These instabilities are well-known to be quite disruptive in the fusion context, so that the very small level of research activity in the astrophysics community on these issues is all the more surprising.

(2) Particle acceleration. However, one certainly does not want to quench every possible mode of instability in MHD jets as some turbulence is required to accelerate the non-thermal particle populations which are responsible for the high energy emission of these objects. In particular, in the framework of the two-flow model developed by the SHERPAS team, it is essential to understand and characterize the processes responsible for the turbulent stirring of the pair plasma, which tap the energy reservoir of the large scale MHD jet.

In order to make progress on these issues, the effect of pressure- and current-driven instabilities in jets are examined ab initio. Pressure-driven instabilities are expected to be most disruptive in jets confined by the hoop-stress, and a linear analysis of the problem has been undertaken (Kersale, Longaretti & Pelletier, 2000, A&A, 363,1166; Longaretti 2003; Longaretti & Baty in preparation). A complete review of the question of jet structure and stability, based on both the astrophysical and fusion literature, has also been written (Longaretti 2005).

4.2 Transport in accretion disks

The question of mass and angular momentum transport in accretion disks is one of the oldest issues in this branch of astrophysics, and has not yet been resolved in a satisfying way. In spite of the remarkable progress undergone in the last fifteen years, with the (re)discovery of the Magneto-Rotational Instability (MRI, Balbus & Hawley 1991, ApJ, 376, 214), and the (essentially numerical) characterization of the induced turbulent transport in the nonlinear regime (e.g. Stone et al. 1996, ApJ, 463, 656), many issues are still acutely open:

(1) Not all disks, or disk regions, are ionized enough to sustain MHD activity (e.g. Gammie 1996, ApJ, 457, 355, Matsumura & Pudritz 2003, ApJ, 598, 645). The transport in these regions must therefore be sustained through non-MHD mechanisms.

(2) The accretion-ejection structures most actively studied in our group do require a fairly high level of turbulent resistivity to be self-consistently maintained (Ferreira & Pelletier 1995, A&A, 295, 807). It is unclear whether the MRI can provide it.

To settle these questions, we have first addressed and completely reinvestigated the old issue of the existence of subcritical turbulence in keplerian flows. The underlying idea is that all linearly stable flows accessible to laboratory experiments are observed to undergo a transition to turbulence (called subcritical). The initial proposal by Shakura & Sunyaev (1973, A&A, 24, 337) was that such a mechanism was at work in keplerian disks (which are hydrodynamically stable). This picture has given rise to controversial points of view in the astrophysics literature (Balbus, Hawley & Stone 1996, ApJ, 467, 76; Richard & Zahn 1999, A&A, 347, 734). Due to the formidable complexity of the problem, little progress had been accomplished on this issue at a fundamental level, until the last decade, where an interesting breakthrough has been operated in the fluid mechanics community, for non-rotating shear flows. Based on this new understanding, the question has been reinvestigated in rotating flows (such as the keplerian flows), through i/ a phenomenological analysis (Longaretti 2002); ii/ a reinvestigation of all the relevant laboratory experimental data (Longaretti and Dauchot 2005; Longaretti and Dauchot, in preparation; Dubrulle et al. 2005); and iii/ numerical simulations (Lesur and Longaretti 2005). This work has lead to the conclusion that a stabilizing rotation does not quench the transition to turbulence, but nevertheless considerably reduces the efficiency of the subcritical turbulent transport, thereby settling this old dispute.

Geoffroy Lesur has started a PhD in September 2004, under the supervision of Pierre-Yves Longaretti, with the aim to investigate in a more systematic way the question of turbulent transport in disks and jets, both from analytic and numerical points of view.

Seyfert galaxies are strong X-ray emitters with a characteristic X-ray spectra. It is roughly power-law like above 2 keV, with the presence of a high energy cut-off near 100 keV (e.g. Petrucci et al. 2001, ApJ, 556, 716; Zdziarski et al. 2000, ApJ, 542, 703). Secondary components, like a fluorescent iron line near 6.4 keV and a bump in the 10-50 keV range are also generally present. The X-ray emission is generally supposed to be produced by a hot and optically thin thermal plasma (a corona) localized above the accretion disk and comptonizing the disk photons. The secondary components are thought to result from the Compton reflection of the X-rays on the disk surface. Due to the lack of sensitivity in the hard X-ray/soft gamma ray energy range (0.1-1 MeV) of the detectors, the real nature of the high energy continuum of Seyfert galaxies remains unclear.

We have recently tested two different geometries of the disk-corona configuration using the most well adapted data for the study of the Seyfert high energy continuum i.e. the BeppoSAX brightest Seyfert sample (Petrucci & Dadina 2005 submitted). This subsample contains 28 objects. It appears that both geometries agree with the data but more importantly, we show that the behaviors of the physical parameters of the models (mainly the temperature and optical depth of the X-ray corona) are strongly geometry dependent, resulting in completely different physical interpretations. This work is an extension of previous works done on a smaller sample (Petrucci et al, 2000, ApJ, 540, 131; Petrucci et al. 2001, ApJ, 556, 716) and underlines the weakness of the actual constraints.

Stronger and less ambiguous constraints on the nature of the coronal plasma can be obtained from variability studies. For example, thermal models, where hot and cold phases are in radiative equilibrium, predict that the X-ray spectrum of the sources should harden when the corona temperature increases (Haardt et al. 1997, ApJ, 476, 620) while non-thermal models predict the reverse (Petrucci et al. 2001, A&A, 374, 719). The analysis of the 1 month simultaneous IUE/RXTE monitoring campaign on NGC 7469, performed in 1996, appears to be in agreement with thermal comptonization emission (Nandra et al. 2000, ApJ, 544, 734). We have performed a more recent reanalysis of these data with realistic comptonization codes that gives another support to this interpretation (Petrucci et al. 2004). Missions with X-ray/gamma-ray broad band capabilities and higher sensitivity, like the ASTROE-2 mission, are required to progress in this field.

On the other hand, the highly sensitive instruments of the XMM-Newton satellite enable very new and exciting results concerning the fluorescent iron line. Broad iron lines are clearly present in some Seyferts and, in objects like MCG-6-30-15, their broad profiles suggest the presence of spinning black hole, close to the extreme Kerr value (Fabian et al., 2000, PASP, 112, 1145) and require very steep disk emissivity laws (Wilms et al. 2001, MNRAS, 328, L27; Fabian et al. 2002, MNRAS, 335, L1), steeper than standard accretion disk one. This is in agreement with the model we proposed some years ago where the irradiating X-ray source was concentrated on the disk axis, above the black hole (Henri & Petrucci 1997, A&A, 326, 87; Petrucci & Henri 1997, 326, 99 but see also Martocchia & Matt 1996, A&A, 282, L53; Martocchia et al. 2002, A&A, 383, L23). More recently, redshifted narrow iron lines have been also observed in an increase number of objects (e.g. Yaqoob et al. 2003, ApJ, 596, 85; Longinotti et al. 2004; Porquet et al. 2004, A&A, 427, 101; Turner et al. 2004, ApJ, 603, 62; Iwasawa et al. 2004, MNRAS, 335, 1073). These lines can be interpreted as signatures of small magnetic flares illuminating small part of the accretion disk. We are investigating the presence of such lines in the Seyfert 1 galaxy Mkn 841. In a first XMM observation, we observed a rapidly variable narrow iron line in this source (Petrucci et al. 2002). Due to the short exposure (2×15 ks), the statistic was not high enough to contrain the origin of the variability. A new XMM observation, longer (45 ks+25 ks) than the previous one, has been performed recently (January/July 2005). Work is in progress to carefully analyse these data and to better determine the origin of the line variability.

A subclass of AGNs is formed by the radio-louds objects, characterized by an intense radio luminosity : they comprise radiogalaxies and quasars. When imaged through interferometric techniques, the radio emission is found to arise from powerful jets emitted from the central engine in the galaxy core. The radio emission is highly polarized and rapidly variable, and is thought to be the result of incoherent synchrotron emission, produced by non thermal, highly relativistic particles in the presence of magnetic field, emitted in relativistic jets with high Lorentz factors. The most powerful and variable objects are called blazars. This property is also responsible for the appearance of superluminal motions at parcec scale, i.e. apparent velocity larger than the velocity of light, which imply a lower limit on the value of the Lorentz factor, which is usually about 10.

Figure 2: Spectral energy distribution of Markarian 501 during the flaring states of April 7 and 16, 1997, fitted by a simple time averaged model (Saugé & Henri 2004). Dashed line : simulated spectrum before intrinsic absorption ; solid line : spectrum after correcting form intrinsic absorption. The grey points are the raw data from CAT observations, and open squares are the data corrected from extragalactic absorption.

In the early 90s, the development of space gamma-ray astronomy (particularly the CGRO observatory) and ground based Atmospheric Cerenkov Telescopes (ACT), has revealed that many blazars are also strong gamma-ray emitters, the highest energy photons being detected above the TeV energy through ACT like Whipple. A new generation of ACT is currently growing, the most remarkable and probably best instrument being the HESS telescopes in Namibia (see international collaborations). Gamma-ray photons, like radio emission, are thought to be produced in the relativistic jets, explaining the high luminosity without gamma-gamma absorption.

There are three processes, still discussed, to explain the generation of extragalactic jets : i) the jet launching by a magnetized accretion disk, ii) the Poynting flux generation by a magnetized spinning Black Hole, iii) the Compton rocket propulsion in the anisotropic radiation field of an accretion disk (purely hydrodynamical processes have been discarded). All the three scenarii have problems to account for observations. The first one turns out to be the most powerful, the most collimating, and able to explain the power input in the FR2 hot spots, which is a sizable fraction of the accretion power. To be powerful, it requires that the jet extracts most of the angular momentum of the disk, which, in turn, deeply modifies the accretion regime (u_accr ~ h/r V_kep instead of h²/r² V_kep like in standard accretion disk) and it can be shown that the accretion flow is neither a SAD nor an ADAF; we call it "JED" for Jet Emitting Disk. But it turns out that this process cannot generate highly relativistic flow because of its baryon load. The second one is less powerful, has a much less collimating property and particularly suffers of the efficient Compton drag exerted by the accretion disk radiation field. The third one leads to a rapid decay when the electron-positron plasma cools by Compton radiation.

A priori there is no reason to assess that only one of these processes is at work. How to avoid the two others? The general solution is a combination of the three. In order to explain radio observations, Guy Pelletier (1985, Congrés de la Société Française de Physique) proposed that the jets could harbour a double structure: a strong, only mildly relativistic jet (v ~ 0.5c) emitted by the accretion disk, through the MHD mechanism described in Sect. 3, and a highly relativistic beam of electronpositron pairs, propagating inside this jet whatever its production process (Pelletier et al. 1988, Phys Rev A, 38, 2552, Sol et al. 1989, A & A, Pelletier & Roland 1989, A&A, 224, 24, etc.). In a second stage, the model has been more elaborated in order to explain the discovery of the gamma emission of Blazars. The pair plasma of the relativistic beam is not only channeled by the MHD jet, but also heated by the MHD turbulence (Henri & Pelletier 1991, ApJ, 383, L7), which allows the Compton Rocket to work. Only this beam would be responsible for all relativistic phenomena, superluminal motions and non thermal radio-to-gamma-rays emission (e.g. Marcowith, Pelletier & Henri 1997, A&A, 323, 271). This two-flow structure solves most theoretical problems associated to the generation of the jet, explaining for exemple the relativistic motion by an anisotropic Inverse Compton effect (Renaud & Henri 1998, MNRAS, 300, 1047). We have developed detailed radiative calculations, showing that the non thermal emission could be very well explained by the model. In the frame of Ludovic Saugé's thesis, defended in 2004, we have investigated the influence of the particle energy distribution function (EDF), exploring the consequences of using a quasi-monoenergetic EDF (more exactly a relativistic quasi-maxwellian, Saugé & Henri 2004) instead of a multi power-law used by most authors. We have shown that such a distribution applies very well to TeV blazars (Fig. 2). We have also developed a full time-dependent model describing the evolution of the pair plasma coupled with the turbulence. This model is undergoing improvements in order to allow detailed comparisons with observations (particularly the recent multiwavelength campaigns led by the HESS collaboration).

This double ejection model has been mostly developed in the environment of Schwarzschild Black Holes. In the case of a Kerr Black Hole, we expect that the Poynting flux generated pair beam will decay under usual conditions in AGNs and that only the Compton Rocket would work. However some situations where the opacity to pair creation would be lower would give more chance to the Blandford-Znajek to contribute. A three process ejection could even be at work when the opacity condition varies in the nucleus, with Blandford-Znajek process alternately at work when the opacity is low, the Compton rocket at work when the opacity is high (see explanations in Black Hole induced ejections Pelletier 2005).

5.3 Microquasars

Another class of compact objects is represented by the so-called X-ray binaries, in which a neutron star or a stellar mass black hole is associated with a normal star in close orbit, accreting matter from its companion. Some objects have shown phenomena very similar to quasars, exhibiting strong variable radio emission, superluminal motion, and possibly gamma-ray emission. As they can be considered as reduced version of quasars (the central object being 10⁸ times lighter than supermassive AGNs black holes), they are often called microquasars . We have undertaken the application of our model of relativistic jets to the ejection observed in microquasars, in connection with a more general scheme (Ferreira et al. 2005). We explain the various states of these objects by a transition between a standard accretion disk (SAD) and a jet emitting disk (JED). In this frame, the relativistic ejections could be associated with the explosive formation of dense pair plasma during so-called intermediate states, combining a still powerful jet and a luminous accretion disk. In these states, disk photons can be scattered up to very high energy by non thermal particles from the jet, which would produce enough gamma-ray photons to trigger pair production. Detailed quantitative calculations are under work.

In another work, with the PhD student Clément Cabanac, we are studying the time behavior of X-ray binaries observed with the hard X and gamma-rays satellite INTEGRAL. We are currently developing numerical tools to extract the time information from the raw data, especially the presence of quasi-periodic oscillations (QPO) that have been commonly detected by the RXTE satellite. The origin of these QPO is still disputed and an important question is their energy spectrum. We are developing in parallel a theoretical model for these QPO based on an instability at the transition region between the SAD and the JED.

6 Relativistic Plasmas and Cosmic Rays

The non-thermal radiation from Black Hole environments is emitted by a relativistic plasma generated in the jets. Therefore high energy radiation (including TeV gamma rays, cosmic rays and possible neutrinos) is an important consequence of accretion-ejection flows around Black Holes. These relativistic plasmas are generated by specific dissipation mechanisms occurring at relativistic shocks and magnetic reconnections that require some theoretical developments. Moreover the dynamics of these accretion-ejection flows cannot be fully mastered without a significant understanding of these microphysics processes. Some progress have been recently done in our team on these topics which are at the forefront of the physics of the so-called Astroparticle Science .

6.1 Relativistic plasmas

Relativistic aspect of the accretion-ejection phenomenon.

The issue of the formation of relativistic jets involves three important intricate problems, namely, the crossing of critical surfaces, collimation and conversion of the Poynting flux into matter-energy flux. These problems have been addressed in a synthesis that will appear in a collective book organized by R. Blandford and M. Lyutikov (Pelletier 2005).

A major problem of the physics of accretion disks is the excitation of a turbulence state able to efficiently transfer the angular momentum of matter in order that it progressively falls on the central Black Hole. In the nineties, a successful solution has been proposed with the so-called Magneto-Rotational Instability . We have extended the analysis of the instability to the case of a Kerr Black Hole and found an expected intensification and a wider window of the instability, that makes it compatible with the intense magnetic field required for launching jets, despite the stabilizing tendency of its tension effect. The analysis has been presented in the same chapter as before and will be developed in a forthcoming paper.

Magnetic reconnection in relativistic regime.

Nowadays the process of magnetic reconnection is invoked in order to convert Poynting flux into matter energy flux and radiation in ultra-relativistic flows. Only very fast reconnection processes can be relevant in this context. Now standard reconnection processes based on the generalization of the Sweet-Parker scheme are too slow. New investigations in Tokamaks and space plasma physics revealed a new mechanism that turns out to be independent on dissipation and fast. We are extending this approach to relativistic plasmas. A fair description of the phenomenon can be obtained in the frame of an appropriate modification of relativistic MHD for both baryon loaded plasmas and pair-plasmas. This has been presented in a first publication (Pelletier 2005) and a paper on relativistic reconnection is to be submitted (Pelletier & Longaretti 2005).

6.2 Cosmic Rays

Transport of Cosmic Rays in chaotic magnetic fields.

The properties of the transport of Cosmic Rays in weak turbulence theory are known since the seventies. Because we crucially need to know them in the strong turbulence regime, especially in relation with the new investigations concerning the Ultra-High-Energy Cosmic Rays, we have undergone a systematical study of the transport, by combining theoretical and semi-analytical approaches with Monte-Carlo numerical simulations (Casse, Lemoine & Pelletier 2002). We have generalized the result of weak turbulence for the pitch angle diffusion and for the spatial diffusion along the magnetic field, as a function of the characteristics of the turbulence spectrum and the particle rigidity. But, as for the transverse diffusion with respect to the mean field, the behavior is deeply different and controlled by the chaotic aspect of the field lines. Furthermore, the diffusion at Larmor radii larger than the coherence length of the magnetic field does not stop suddenly, the diffusion coefficient increases proportionally to the square of the particle energy.

Fermi acceleration in relativistic regime with magnetic fronts and shocks.

The enigma of the existence of Ultra High Energy Cosmic rays can be solved along two different but not exclusive ways : either by a bottom up scenario where the UHECRs are generated by an acceleration process in the high energy astrophysical sources, or by a top down scenario where some quantum objects, proposed by a new physics beyond the standard model of particle physics, decay and produce particles in this energy range. The Pierre Auger Observatory will soon provide a crucial answer to this question. Indeed we will soon know whether there exists a particle spectrum beyond the GZK limit (which is the energy threshold beyond which protons loose energy by producing mesons through collisions with the CMB-photons), revealing a generation of Cosmic Rays that would not come from extragalactic sources. We have contributed to the bottom up scenario by studying the relativistic regime of Fermi acceleration, necessary for the Cosmic Rays to reach energies of order 10²⁰ eV in the considered sources (AGNs and Gamma-Ray Bursts or GRBs). We have developed the Fermi acceleration in relativistic regime along two different ways : with relativistic fronts and with relativistic shocks.

Figure 3: The spectrum of cosmic rays at a relativistic shock. This spectrum appears as the envelope of Fermi cycles (i.e. cosmic rays of large mean free path cross the shock front back and forth several times and gain energy at each crossing cycle downstream-upstream-downstream), the first corresponding to an amplification of the cosmic ray energy from its initial value by a factor Gamma², and the subsequent ones to an amplification by a factor 2.

When an ensemble of magnetic relativistic fronts propagating with relative speeds that are mildly relativistic, like in GRBs, the elastic scattering of magnetic fronts generates an efficient Fermi acceleration that we applied to GRBs (see next paragraph). As for the case of a relativistic shock, Martin Lemoine and G.P. have developed a combined theoretical and numerical study. The formation of the energy spectrum has been analyzed confirming a recent analysis proposed by Achterberg and Gallant (1999, MNRAS, 305, L6) and the time scales of the process have been determined. Reliable laws for the generation of Cosmic Rays in relativistic flows have therefore been provided (Fig. 3).

Excitation of turbulence by Cosmic Rays upstream a shock and consequences for the transport and the Fermi acceleration.

The results we obtained on the transport of Cosmic Rays make its dependence on the MHD turbulence spectrum precise. There exist semi-phenomenological theories of inertial turbulence spectra for common situations where the turbulence is excited at large scales and then cascades over several decades towards smaller scales where the dissipation takes place. Moreover the Reynolds numbers in astrophysical media are very large. However, the turbulence excited upstream astrophysical shocks is of a particular nature. Indeed it is excited by an instability resulting from the acceleration of Cosmic Rays at the shock and the instability amplifies the MHD modes over all scales through a resonnant regime (larger scales) and a non-resonnant regime (shorter scales). We have thus undergone the calculation of the turbulence spectra and the transport coefficients. Then we examined the consequences of the excitation of MHD turbulence on the energetic balance and the final spectrum of the Cosmic Rays that is steepened. This theoretical effort was made necessary by the indications delivered by the new observations of supernovae remnants by Chandra and XMM. Our theory incorporates not only the newest developments on anisotropic MHD turbulence but also an important effect: the backscattering of progressive Alfv´en modes off sound waves (more precisely the slow magneto-sonic waves). These studies have been realized with Martin Lemoine, Alexandre Marcowith (CESR) and G.P. (in course of publication).

High energy emission and Cosmic Rays from Gamma-Ray Bursts

Gamma-Ray Bursts (GRBs) are phenomena radiating as much energy as supernovae but in a very short time, a few seconds or even less. The collimation of the ultra relativistic flows involved in GRBs strongly amplifies the energy flux. Under the less favorable hypothesis about the magnetic field and its irregularities, it has been shown that Cosmic Rays undergoing scattering off relativistic magnetized fronts (revealed by the light curves) could indeed reach the UHE range (Gialis & Pelletier 2004). The performances of GRBs as sources of high energy radiation in the form of photons and neutrinos have been estimated (Gialis & Pelletier 2005). A gamma diagnosis of UHECR generation is proposed; the diagnosis in the range of a few tens of GeV appears to be non-ambiguous, because this signal of hadronic origin should not be contaminated by inverse Compton emission by electrons which is deeply in the Klein-Nishina regime. We are eagerly waiting for the observations of GRBs by GLAST that could make this diagnosis.

7 Heavy numerical simulations

The field of theoretical astrophysical dynamics has undergone two major evolutions in the last two decades or so. First, the idea that magnetic fields play a key role has progressively imposed itself to a community in which MHD phenomena had long been regarded as exotic. Secondly, the progressive rise of powerful numerical tools (both software and hardware) has transformed numerical simulations from a gadget to a must in the exploration and understanding of the nonlinear outcome of the physical processes considered as important in our trade. Consequently, MHD has become the central area of study of accretion-ejection related phenomena, and the modern dynamicist working in this field must not only master analytical techniques, but numerical ones as well. To a lesser but growing extent, a similar evolution can be witnessed in the field of high energy astrophysical processes, which is also of direct interest for the team activity.

The French community is still largely lagging behind the international leadership on the numerical front, by lack of both human and material means (despite the information made by the ASSNA). The SHERPAS team has become more and more aware of this deficiency for its own activity, and has therefore put a particular emphasis on numerics by progressively building up an internal expertise, and by defining as a goal to address an evergrowing fraction of problems through (M)HD simulations (in support of more theoretical analyzes). In practice, this numerical activity is only now reaching a production stage, and has been undertaken in several directions:

(M)HD stability and turbulent transport in disks and jets with Geoffroy Lesur (PhD started september 2004) and P.-Y. Longaretti.
2D magnetospheric star/disk interaction with Nicolas Bessolaz (PhD started november 2004) and J. Ferreira, in collaboration with Rony Keppens (Netherlands) and J. Bouvier (FOST). The goals are to study first, the conditions leading to steady accretion funnels and second, the possibility to launch Reconnection X-winds as envisioned by Ferreira, Pelletier & Appl (2000, MNRAS, 312, 387).
3D magnetospheric star/disk interaction with Claudio Zanni (post-doc starting Fall 2005) and J. Ferreira, in collaboration with Christian Fendt (Germany). C. Zanni will study the oblique rotator and try to relate the dynamical situation computed with current observational works carried out by the FOST team (C. Dougados, J. Bouvier and F. Ménard).

8 Astrocladistics

Imagined by Didier Fraix-Burnet in 2001, astrocladistics is a brand new approach toward establishing an evolutionary history of galaxies, from which a new classification might be devised. This methodology, borrowed from phylogenetic systematics, a branch of evolutionary biology, is built on the original idea that galaxy diversity, generated by their evolution, in particular through interactions, could be organized in a hierarchical manner. In a similar way as for living organisms, Didier Fraix-Burnet and two biologists (Philippe Choler and Emmanuel Douzery) have proposed to depict parenthood between galaxy classes on hierarchical trees called cladograms. The underlying notion of complexity for galaxies is thus introduced in extragalactic astrophysics. Up to now, astrocladistics has been successfully used for the first time on a sample of Dwarf Galaxies from the Local Group, then on two samples of simulated galaxies (GALICS database of Bruno Guiderdoni et al., stage de DEA of Anne Verhamme), and currently on samples of galaxies from the Virgo cluster in collaboration with Emmanuel Davoust from the Laboratoire d'Astrophysique de Toulouse. We now understand how the few hundreds of galaxies in our samples could have evolved and how much they resemble each other and why. All this work has been orally presented at two international conferences (Fraix-Burnet et al 2003, Fraix-Burnet 2004), three papers have been submitted (two of which are now accepted) and two others are being written. Cladistics is probably unusual and may be even somewhat revolutionary for astrophysicists, but it is clear today that it brings several concepts and a formalism of major interest for the contemporaneous extragalactic astrophysics.

9 Participation in large collaborations

The SHERPAS team has strong involvements in several big international projects: the JETSET european network, several instrumental collaborations (AMBER, VITRUV, SIMBOL-X, GLAST, ECLAIRS) and the large HESS international collaboration. Moreover, it has strong links with the GdR PCHE and the national programs PNPS and PNG.

Instruments in which LAOG is involved :

AMBER: Didier Fraix-Burnet is a member of the Science Group and is PI or coI of several Guaranteed Time proposals on AGNs.
VITRUV: P.-O. Petrucci and G. Henri are members of the Science Group of this second generation instrument for the VLTI.

Other instruments and projects :

JETSET is an european Marie Curie Research and Training Network gathering 11 european laboratories which has officially began the 1st of february 2005 and will run four years (http://www.jetsets.org). J. Ferreira is sharing with S. Massaglia (Italy) the management of the Work Program MHD models of Jets and Outflows .
SIMBOL-X: P-.O. Petrucci is member of the Science Group of this hard X-ray (0.5-70 keV) instrument that should be launched around 2010.
GLAST: the whole team is involved in the scientific support. With an overlap with the HESS window, this mission should provide invaluable clues on compact objects and GRBs.
ECLAIRS: G. Henri and G. Pelletier are involved in the scientific support of this instrument with the goal of studying the prompt emission of GRBs.
HESS: G. Henri, G. Pelletier, P.-O. Petrucci are deeply involved in the HESS collaboration where they provide a scientific support, especially for the interpretation of Blazar observations and multi-wavelength approaches. This Astroparticle experiment is very successful (cf. papers in Nature, Science...) and is opening a new window of astrophysics.

10 Scientific Highlights (2002-2005)

Transport of cosmic rays in chaotic magnetic fields , Casse Fabien, Lemoine Martin, Pelletier Guy, 2002, Phys. Rev. D. The knowledge of the diffusion coefficients of cosmic rays is crucial for astrophysical applications and especially for astroparticle physics, both for mastering the cosmic ray propagation and the efficiency of Fermi acceleration. They have been computed with a Monte Carlo simulation as a function of the cosmic ray rigidity and the spectrum of the magnetic field (characterized by its index and its degree of irregularities). The Bohm scaling, often assumed in astroparticle physics in the strong turbulence regime, has been ruled out. The scalings derived for the pitch angle diffusion and for the parallel diffusion with the quasi-linear theory are extended in an appropriate way in the strong turbulence regime. However the transverse diffusion coefficient follows a law which is nor Bohm nor quasi-linear, but a law that stems from the analysis of the spatial chaos of field lines. This result is important to estimate the confinement time of cosmic-rays in galaxies and in extragalactic jets.
Particle Transport in Tangled Magnetic Fields and Fermi Acceleration at Relativistic Shocks, Lemoine Martin, Pelletier Guy, 2003, ApJ, 589L. In Monte Carlo simulations, test particle trajectories are followed in a magnetic scattering medium in which a plane is moving at subrelativistic or relativistic speed. The conditional probabilities for a particle to come back to the plane after crossing as a function of the pitch angles have been computed. Then the amplification of the particle energy at each Fermi cycle (downstream-upstream-downstream) across a shock is calculated for arbitrary value of the shock speed. Thus the formation of the cosmic ray spectrum at a relativistic shock is derived and the result displays that the first Fermi cycle produces a strong energy amplification by a factor on the order of the shock Lorentz factor to the square, whereas the other cycles amplify by a factor 2, as predicted by Achterberg and Gallant. The index is computed as a function of the shock Lorentz factor and the results are in excellent agreement with an analytical formula. A precise law for the acceleration time have also been obtained.
Stationary Accretion Disks Launching Super-fast-magnetosonic Magnetohydrodynamic Jets , Ferreira J., Casse F. 2004, ApJ, 601, L139 This is the last paper in the series on Magnetized Accretion-Ejection Structures (MAES), which represents the only available self-consistent model of a self-confined jet driven by an accretion disk (with the disk fully resolved), able to cross all three MHD critical points. Biases induced by the self-similarity contraints are discussed.
Physical interpretation of the NGC 7469 UV/X-ray variability , Petrucci, P. O., Maraschi, L., Haardt, F., & Nandra, K. 2004, A&A, 413, 477 Data fits, with a detailed model of comptonized spectra obtained from simultaneous UV and X observations of a Seyfert galaxy. These fits not only reproduce the spectral shape, but the source variability as well. They allow us to show that the data are consistent with a model where all the accretion power is liberated in the corona (producing X-rays), while the UV emission comes from the reprocessing of the X-ray flow rather than from the energy dissipation in the disk. The observed variability is produced by a geometric modification of the corona.
Astrocladistics: a phylogenetic analysis of galaxy evolution. I. Character evolutions and galaxy histories , D. Fraix- Burnet, P. Choler, E. Douzery, A. Verhamme. 2005, accepted in Journal of Classification. This is a first paper in series; the method of astrocladistics is described. It should allow us to establish a galaxy classificatoin based on physical criteria besides the galaxy morphology.
On the Relevance of Subcritical Turbulence to Accretion Disk Transport , Lesur, G., and Longaretti, P.-Y. 2005, accepted in A&A This paper solves a riddle dating back to the seventies, and which has given rise to an important controversy in the last ten years. Namely, it is shown that a linearly stable, keplerian hydrodynamic flow is indeed turbulent through non linear mechanisms, but that this turbulence has a low efficiency in terms of angular momentum transport, with a Shakura- Sunyaev parameter < 10^-5. The discrepancy between numerical simulations and laboratory experiments is explained away.
The bulk Lorentz factor crisis of TeV Blazars: evidence for an inhomogeneous pile-up energy distribution , Gilles Henri & Ludovic Saug´e, 2005, accepted in ApJ It is shown that the Lorentz factors of Blazar jets, deduced from homogeneous SSC models, are contradicted by the source statisticsn which implies Lorentz factors of order 3, whereas homogeneous models require 50 or more. The only way out of this conundrum is to take the jet stratification into account, so that high energy photons are not produced in the same region as low energy ones. Observations then require mono-energetic distribution functions whereas most models make use of power-law distributions. In contradistinction with the dominant view, this work shows that (1) jets do not need to be highly relativistic, and (2) turbulence rather than shocks is at the origin of the production of high energy particles.

11 References

Clic here to have access to all the references of the group

Name	Grade	Speciality
Jonathan FERREIRA	Assistant Prof.	Accretion-Ejection, MHD
Didier FRAIX-BURNET	Researcher	Astrocladistics
Gilles HENRI	Prof.	High energy phenomena
Pierre-Yves LONGARETTI	Researcher	Instabilities and Transport
Guy PELLETIER	Prof (Group Leader)	Relativistic plasma physics
Pierre-Olivier PETRUCCI	Researcher	High energy phenomena
Peggy Varnière	Researcher	MHD simulations
Nicolas BESSOLAZ	PhD	MHD simulations
Timothe BOUTELIER	PhD	High energy phenomena
Clement CABANNAC	PhD	High energy phenomena
Goeffroy LESUR	PhD	Instabilities and Transport
Claudio ZANNI	Postdoc	MHD simulations