Document AERES
Report 2007 - Mid 2012 / Project 2014 - 2018

Service d’Astrophysique – Laboratoire AIM (SAp) 

Profile of the Unit


Staff members of the Astrophysics division of Irfu (SAp) [1] are part of the “Astrophysics, Instrumentation- Modeling“ (AIM) joint research unit [2]. On the 30th of June 2013, AIM consisted of 189 employees; 105 have a permanent position: 52 researchers (38 from CEA-Irfu, 9 from the Paris-Diderot University, and 5 from CNRS-INSU), 12 instrumentalists (11 CEA-Irfu, 1 CNRS-INSU), 38 engineers and technicians (CEA-Irfu), 5 support staff (4 CEA-Irfu, 1 CNRS-INSU), and 84 have a temporary contract: 28 PhD students, 52 post-docs, 3 engineers, 1 apprentice. The average number of employees for each category of staff over the 2008-mid 2013 period is indicated on the diagram    

[1] Except four physicists of SAp who are part of the APC joint research unit and one physicist who is part of the CELIA joint research unit.
[2] AIM has been very positively evaluated by AERES last year (report available at the AERES web site).

Missions and Strategy


AIM conducts programs concerning two of the four science themes of Irfu: energy content of the Universe and structuration of the Universe.

The missions of the unit are twofold:
• first, to acquire new knowledge about the
Universe and its constituents by :
o developing research programs in Astrophysics at the best international
o developing state of the art instruments required for this research, mainly for space missions, but also for large ground-based telescopes;
• second, to disseminate the new knowledge by training PhD students, lecturing, and making the results accessible to the general public.

The global strategy of AIM is to be along the whole chain of new knowledge acquisition, from instrumental Research & Development to astrophysical discoveries, in a coordinated way such that the various activities reinforce each other. Indeed, having high level scientists interested in projects is a way to orient and to participate in new projects; R&D allows us to be well positioned technically in projects (especially at the heart of an instrument: the sensor). Participating in the realization of an instrument gives us an expert knowledge of it, which allows us to participate in interesting part of the ground-segment and then to be in the best position to exploit it; most often, our participation in the instrumentation brings guaranteed observing time, which benefits the scientists of the laboratory, closing the loop.

The scientific strategic followed by AIM is a multi-wavelengths and multi-scales study of the Universe. Via observations of large samples of galaxies (gas and dust content, morphology, etc.) and galaxy clusters (temperature, mass, etc.), we can build catalogues of objects and look for underlying laws describing the observed properties (self-similarities of galaxy clusters; “main infrared sequence” for galaxies). In parallel, we conduct large numerical simulations to isolate the main physical mechanisms at work. We moreover use observations of galaxy clusters and galaxies (weak lensing) as cosmology tools to constraint the “dark Universe”; structuration of the Universe and observational cosmology are strongly interleaved. Via observations of specific galactic constituents (the interstellar medium, supernova remnants, microquasars) and large numerical simulations, we study key physical processes responsible for the evolution of galaxies, such as star formation, accretion-ejection of matter around black holes, acceleration/propagation of cosmic rays. Via observations and modelling of the Sun and the Saturn system (planet, rings, and satellites), as well as other stars, we make detailed studies of physical processes at work in stellar evolution and dynamics, planet formation, and star-planet interactions.
While the activity of the unit is largely devoted to the study of the structuration of the Universe (80%), we have started to increase our participation in the study of the energy content of the Universe. We have decided to do so by proposing to CNES and then to ESA a space mission devoted to the study of the dark Universe, the DUNE mission, which has evolved into the Euclid mission, the M2 ESA Cosmic Vision mission. We occupy key positions for this mission at the international consortium level (Coordination and Management support, Mission System lead, Survey Implementation lead, Ground Segment scientist) together with key hardware responsibilities (focal plane of the VIS instrument...), as well as in the Science Ground Segment (responsible of high level data analysis); (see highlight US5: ”Adoption of the EUCLID mission by ESA”). The rationale behind our positioning was to be in the best position to constrain dark matter with the weak lensing probe. Note that Euclid will also be an excellent mission to study the structuration of the Universe (legacy aspect).

Organisation and skills


At the head of the unit are a director (P.-O. Lagage until the 31th of December 2013 and then A. Decourchelle), 2 deputy directors (I. Grenier, professor at Paris-Diderot University, and P.-A. Duc, “directeur de recherche” at CNRS), and a technical director (M. Talvard from CEA). M. Talvard is also the security officer, assisted by a security engineer (P. Marlet). Three secretaries are in charge of administrative matters (temporary contracts, missions, orders, etc...); additional support (budget, Human Resources, ...) is provided at the level of CEA-Irfu. IT support is provided by two technicians, at the unit level, and by the general support of Irfu.
AIM is organized into teams (see the organization chart); each team has a leader, who is in charge of the scientific or technical animation and of human management aspects (for example annual performance evaluation). There are regular meetings (about every other month) between the direction and the team leaders. Note that having both engineers/technicians and astrophysicists in the same Unit leads to a strong link between astrophysicists and engineers, which is key at several stages of a project.
Five teams have been organized according to astrophysical themes: Cosmology and Galaxy Evolution, Star Formation and Interstellar Medium, Star Dynamics and Environment, Disks, Rings and Planets, Cosmic Phenomena at High Energy. One team is devoted to modelling. Although there is modeling activity within the thematic teams, there is also a need for modelling activity to take place independently to ensure the development of transverse tools which can be of interest to several teams. A pluri-disciplinary team in Computational Cosmology, named ComoStat and common to AIM and SEDI, has been created in 2010 to acknowledge the need to develop new methods to analyze astronomical data, especially in cosmology which requires powerful statistical methods (PLANCK, EUCLID, etc.), and taking the opportunity of the advanced European Research Council Grant obtained by Jean-Luc Starck.The numerical simulations users/developers from various teams meet regularly in the framework of the COAST program (COmputationnal ASTrophysics) (see, which is transverse to AIM and SEDI.
The technical teams are organized according to expertise: space project management, system, architecture (LSAS); space product assurance, integration (LQIS); space spectro-imager (LSIS); space electronics (LEDES); interface science-instrumentation (LISIS). AIM is one of the 5 French laboratories to have the capacity to manage the development of an instrument for a space mission for Astrophysics. The LSAS and LQSI teams ensure our ability to conduct the realization of a space instrument; we are the only French space astrophysics unit which has been able to keep four permanent staff working in space product assurance. In terms of system engineering, we have supported ESA in the Euclid Mission Definition task by introducing systematic methodology for requirements engineering. This is crucial for a mission as complex as Euclid where ultimate performance is only relevant when taking together space segment, instruments, ground segment, calibration, data processing, and survey operation. The LSIS and LEDES teams are at the heart of our R&T effort, namely, by developing novel focal planes with front-end electronics. The space constraints are taken at the early stages of our R&T developments, accelerating the space readiness. The LISIS has developed a rare skill in radiation effects on detectors and is co-responsible for the radiation facility at Irfu premises on the main Saclay CEA center; it is also leading regular meetings of about 25 radiation experts from several divisions inside Irfu, in strong collaboration with experts of space agencies (CNES, ESA) or other research institutes (ONERA).
The technical staff is allocated to instrumental projects following a classical matrix organization. Engineers and technicians most often work on more than one project, which is one of the arguments in favor of a vertical organization in terms of skills. Regular meetings between the technical team leaders, the technical director and the projects manager have been established to monitor the progress of each project and to compare the resource allocation to the needs. Note that almost all of the projects are made in common with the Irfu SIS or SEDI technical department.
The organization of the projects is the classical organization of space projects: a project scientist (PI, Co-PI or Co-I at the level of the international project), a project manager, an instrument scientist, a system engineer leading a system group with various architects (mechanics, optics, electronics, etc...).


Document AERES

JWST/MIRIM test bench developed by SAp-­‐AIM, SIS, SEDI and SACM in the framework of the MIRI instrument for the JWST, the successor of HST. It features a 4K cryostat, a telescope simulator, control, command…The test bench was used to test the MIRI imager, we are responsible of. The bench is planned to be used to test the detectors for the E-ELT METIS instrument.

For the integration of space instruments, AIM has a clean room (150 m2 class 100 000, 100 m2 class 10 000 with spots at a class 100), whose maintenance is at the charge of the LQIS team. The LEDES team has maintained competences in electromagnetic compatibility and is responsible for the maintenance of the AIM anechoic chamber, which allows easy compatibility tests. Different test benches to characterize detectors (from visible to hard X-ray) have been built over time, so that we have and cover a large range of temperature: from ambient temperature to 50 mK.


We have also built a new conference room, where it is possible to visualize the results of numerical simulations in 3D.



The activities and staff are balanced between detector Research & Technology (R&T), space instrumentation, signal processing, multi-wavelengths observations/interpretation, and multi-scales modeling, especially with numerical simulations.
AIM is one of the very few Astrophysics units in Europe simultaneously conducting large observing programs at the best international level, developing innovative data processing methods, state of the art space and ground-based instrumentation, front line numerical simulations on massively parallel computers. This strategy relies upon a strong interaction with the technical divisions of Irfu (SIS, SEDI and to a lesser extent SACM), which participate in the R&D, instrumental development, signal processing and bring complementary expertise.
The typical time scale of a space mission is of the order of 20 years, so that we are working in parallel on projects at different phases. We have been at the starting point of projects by participating in the prospective exercises proposing scientific themes; for example, back in 2004, we led the team of scientists proposing to ESA to dedicate a space mission to the study of dark energy; we have recently (May 2013) been very active in answering to the ESA call about white papers for the next L2 and L3 missions, co-leading a white paper about high angular resolution in the far-infrared, and participating in the scientific interest of the X-ray mission ATHENA+. We led the study of mission concepts, such as the DUNE concept for dark Universe, which has evolved into Euclid. We are then leading or participating in competitive preliminary studies of future space missions, following with the realization of the instrumentation of accepted space missions and of the associated ground segment, and at the end in the scientific exploitation of the launched missions; in addition we have to conduct R&D program to be ready for future calls for missions.
The activity in numerical simulations consists in:
• developing original numerical codes for massively parallel computers, such as the RAMSES adaptive mesh code, and making them available to the community. Note that improving codes contributes as much to the global performance increase as the hardware improvement;
• participating in Grand challenges during the commissioning of new machines, (Mare Nostrum, CCRT,...);
• using the codes in relation to observations, either to interpret or predict the result of observations; • using them standalone to discover the main mechanisms at work in the structuration of the Universe (e.g. cold flows, disk instabilities, disc/planet interactions, planetesimal accretion,...);
• using them to predict or optimize the performances of future space missions, such as Euclid.
The way to obtain computing time on a super computer is now similar to the way to obtain time on a telescope: proposals evaluated by a peer review committee. In 2011, we have obtained, as Principal Investigator (PI), 25 million hours of computing time (galaxy evolution, star formation, magnetism of the stars, turbulence in protoplanetary disks). This is 55% of the total computing time distributed for Astrophysics by the French Agency for computing (GENCI). In addition, we have obtained 9 million hours of time in the Framework of the European Prace Agency.


Main results

Scientific Exploitation

In the 2008 – mid 2013 period, our first priority has been the scientific exploitation of the ESA Herschel mission launched in 2009; two science themes have been concerned: galaxy evolution (star formation rate history) and star formation (pre-stellar phase). We have benefited from the launch of three other observing facilities: Planck, Fermi, Kepler, as well as of continuation of the operation of CASSINI, XMM, SOHO, COROT in space, VLT, IRAM, CFHT, Hess on ground. We have also benefited from the availability of up-to-date massively parallel computers through GENCI in France and Prace at the European level.
We have been at the origin of several changes of paradigm: in the domain of galaxy evolution (galaxy mergers are no longer the dominant source of triggering star formation; see highlight SU2: “New views on galaxy evolution“), of star formation (stars form in filaments; see highlight SU1: “Star formation as revealed by the ESA Herschel mission”), of planet satellites (see highlight N°SU5: “The Saturn satellites made from the Saturn disk”),.... The interior of stars has been revealed thanks to asteroseismology, which has triggered the need for more detailed modeling, taking into account dynamical effects (rotation, magnetic field) (see highlight SU6: “Probing the interior of stars”). A new important component of the interstellar medium (dark gas) has been discovered interpreting the diffuse gamma-ray emission generated by cosmic-rays interacting with the insterstellar medium (see highlight SU3: “Ten times more γ sources with the Fermi Telescope”). We have elucidated the physical mechanism responsible for asymmetric supernovae explosions and the natal kicks of pulsars (see highlight SU9: “The SASI mechanism in core-collapse supernovae)... More details can be found in sections 7.2 (energetic content of the Universe) and 7.3 (structuration of the Universe) of the document.



Instrumentation (with SEDI, SIS, SACM)

Space Instrumentation

Our space projects are developed in partnership with CNES according to the agreement signed between CNES and CEA (sharing the full cost on a fifty/fifty basis). During the 2008 - mid 2013 period, we have achieved the construction of a major instrument: the JWST-MIRI imager (French Co-Principal Investigator level (Co-PI)); the JWST launch is now scheduled in 2018 (see highlight SU4: ”Delivery of the MIRI instrument for the NASA JWST mission”).

Document AERES

A lot of documents. set of documents provided to ESA in the framework of the definition phase of the Euclid mission by the Euclid consortium; we have been leading the study in terms of science, technique and management.

In parallel to the construction of the JWST-MIRI camera, we had to study projects that will enter the science exploitation phase in the 2019-2024 timeframe or even beyond. Furthermore, given the ESA policy to have competing studies for several missions in the definition and assessment phases, and given our broad technical and scientific coverage, we participated in numerous instrumental studies, heavily loading the technical staff. Our participation is always twofold: science definition and instrumental definition. The level of participation varies according to the project, ranging from leading role to an expertise role. We have played a leading role in two space missions: the ESA DUNE-EUCLID mission and the French- Chinese SVOM mission, which has been in a “frozen” state since early 2012, waiting for the Chinese decision about the platform; the decision has now been taken and the project could start again beginning of 2014. We have had a key role in the study of the STIX instrument of the ESA Solar Orbiter M1-mission by studying the detection unit; it consists in 32 detection modules derived from the MACSI concept that we have developed in our R&T programs.

We contributed to the definition phase of SPICA, Plato, IXO, ATHENA, ECHO, LOFT. We have also participated to several proposals for future missions (ESA M3: NEAT, SPICES, COSPIX; ESA Small missions: PLAVI, microNeat, A-Star; ESA L1 : instrument ECHOES for JUICE with NASA-JPL), which have not been selected.

Ground-based Instrumentation

Regarding ground-based projects, the main achievement has been the building of the APEX-ArTeMiS camera (PI), which will enter in full scientific exploitation in 2014 on the APEX telescope in Chile (see highlight SU8: ”ArTeMiS camera on the sub-millimetric APEX telescope in Chile”).
We have prepared the mid-term future by participating in the phase A study of the METIS instrument for the European Extremely Large Telescope (E-ELT) planned to be installed in 2025. The time scale for developing ground-based instrumentation on large telescopes can now be similar to that for space projects. We are also participating in the preparatory phase of the CTA project (site selection, ground-segment).
Research & Technology
At the level of R&T and demonstrators, we have concentrated our activity on focal plane arrays from hard X-rays to mm wavelengths.
•    We have developed with Irfu-SEDI (ASICS skills) and with the know-how of the 3D+ company, a novel 2048-pixel fine-pitch camera for hard X rays (MACSI modules; (see highlight SU7: ”Novel hard X-ray spectro-imager for Astrophysics”)).
•    We have passed crucial steps in the development of novel ultra-high resolution X-ray imaging spectrometer realized at the INAC Grenoble technological Platform; (see highlight SU10: ”Novel micro-calorimeters for X-ray Astronomy”)
•    We have collaborated with the Le?ti at Grenoble to define the future generation of sub-mm bolometer arrays, capitalizing on the very successful development of the novel PACS bolometer arrays (in the framework of the FOCUS LabEx, “Laboratoire d’Excellence”, FOCal plane arrays for Universe Sensing, we are co-coordinating with IPAG laboratory in Grenoble.
•    We have developed cryogenic (50 mK) electronics for these arrays under an FP7 contract we are leading.
•    We are characterizing IR detectors made at LETI, under ESA contracts following calls for tenders.


Document AERES 2012

See also the review for the AERES (Agence d’évaluation de la recherche et de l’enseignement supérieur) evaluation, describing the activities and projets

Clefs CEA
I.    Report 2007 - Mid 2012
II.   Report by Teams
III. Project 2014 - 2018
IV . Project by Teams
V.   Appendix
V.   Full document (24 Mo)

Maj : 13/01/2013 (3416)


Retour en haut