1

Timetable

09:15 Welcoming09:30 Laura Bernard09:50 Clotilde Laigle10:10 Federico Mogavero10:25 Erwan Allys10:40 Thomas Lacroix11:00 Coffee11:20 Oscar Ramos11:35 Caterina Umilta11:50 Alba Vidal Garcia12:10 Rebekka Bieri12:30 Sebastien Carassou12:45 Lunch14:45 Tanguy Marchand15:00 Jean-Baptiste Fouvry15:20 Jesse Palmerio15:35 Luciana Ziino15:50 Coffee16:15 Florian Sarron16:30 Celine Gouin16:45 Nicolas Cornuault17:00 Melanie Habouzit17:20 Siwei Zou17:35 Tilman Hartwig18:00 Cocktail

2

09:30 – 09:50 Laura Bernard (3rd yr.)

Dynamics of non-spinning compact binaries atthe fourth post-Newtonian order

The coalescence of compact binary systems are among the mostpromising sources of gravitational waves for the current and nextgenerations of interferometers. Due to the faintness of the sig-nal, both the detection and precise determination of the physicalparameters require highly accurate template banks of theoreticalwaveforms, which are then filtered against the detector data.

During the inspiral phase of the coalescence, when the two ob-jects are widely separated, the perturbative post-Newtonian ap-proach allows one to describe the dynamics of the binary systemand to compute the radiation energy flux, from which the orbitalphase evolution can be derived.

In this talk, I will address the question of the equations of mo-tion at the fourth post-Newtonian order in harmonic coordinates.I will present a method based on a Fokker action, adapted to thespecificities of the post-Newtonian formalism in harmonic coordi-nates. In particular, I will focus on the treatment of the so-calledtail effects which appear for the first time in the dynamics at 4PN.

Supervisor: Luc Blanchet

3

Clotilde Laigle (3rd yr.) 09:50 – 10:10

Understanding galaxy populations throughoutCosmic Time: Joint analysis of observed and

simulated lightconesThe study of the co-evolution of the cosmic web and galaxies

embedded within it, from the epoch of cosmic dawn (formation ofthe first galaxies) to today relies heavily on comparisons of realobservations on large field with virtual galaxy surveys produced bystate-of-the-art cosmological hydrodynamics simulations. They al-low to carry out meaningful statistical measurements of the differ-ences between models and observations. In other words, analysingimages of galaxies, whether captured by a telescope or generatedwith a supercomputer tuned to have the same limitations as thistelescope, ensures that we are comparing (millions of) apples toapples. I will present results of such comparative analyses basedon COSMOS field and Horizon-AGN, which provide invaluabledetailed insight on the evolution of galaxy stellar masses and mor-phological transformations as a function of environment and cos-mic time.

Supervisors: Henry McCracken & Christophe Pichon

4

10:10 – 10:25 Federico Mogavero (2nd yr.)

Microlensing planet detection viageosynchronous and low Earth orbit satellites

The gravitational lensing of light has lead to the discovery ofdozens of extrasolar planetary systems during the last decade.Moreover, it provides one of the very few methods to measuremass and abundance of free-floating planets and isolated browndwarfs in the Galaxy. However, planet detection through mi-crolensing is usually limited by a well-known degeneracy in theEinstein timescale tE, which prevents mass and distance of thelens to be univocally determined. It has been shown that a satel-lite in geosynchronous orbit could provide masses and distancesfor most standard planetary events (tE ≈ 20 days) via a microlensparallax measurement. We extend the analysis to shorter Einsteintimescales, tE ≈ 1 day, when dealing with the case of Jupiter-masslenses. We then study the capabilities of a low Earth orbit satel-lite on even shorter timescales, tE ≈0.1 days. We show that ageosynchronous satellite could detect parallaxes for Jupiter-massfree floaters and discover planetary systems around very low-massbrown dwarfs. Moreover, a low Earth orbit satellite could lead tothe discovery of Earth-mass free-floating planets.

Supervisor: Jean-Philippe Beaulieu

5

Erwan Allys (2nd yr.) 10:25 – 10:40

Bosonic structure of supersymmetric cosmicstrings in a realistic GUT context

The most advanced observations in cosmology can probe the earlyuniverse up to energies around 1015 GeV, where we expect a grandunification theory to describe particle physics. At this scale, thedata are compatible with a phase of inflation, which would createthe primordial fluctuations that seed the large scale structures. Ina realistic particle physics model, the production of topologicaldefects like cosmic strings is unavoidable, and could be observedin principle. However, their microscopic structure is rarely consid-ered, and usually only toy-models are used. We will describe howthe complete microscopic structure of such objects can be studiedin this context, using the most recent cosmological and high en-ergy physics theories, such as hybrid inflation and super-symmetricgrand unified theory. We will then discuss the properties of thoserealistic strings, and the cosmological consequences they imply.

Supervisor: Peter Patrick

6

10:40 – 11:00 Thomas Lacroix (4th yr.)

Interpretations of new gamma-ray emissionsfrom the center of the Milky Way and

Centaurus AGamma rays are an extremely valuable source of information onthe high-energy processes taking place at the center of galaxies.In this presentation I will describe two very intriguing features re-cently reported in the gamma-ray spectra of the central region ofthe Milky Way and the Centaurus A radio galaxy. I will discussthe interpretation of these features in terms of the high-energyemission of electrons and positrons produced either by the annihi-lation of dark matter particles distributed in a very steep densityspike in the central region of galactic halos, or by a population ofmillisecond pulsars.

Supervisors: Joseph Silk (IAP) & Celine Boehm (Institute forParticle physics Phenomenology, Durham, UK)

7

Oscar Ramos (1st yr.) 11:20 – 11:35

Testing Lorentz Invariance by binary black holesThe recent detection of gravitational waves has opened a newavenue towards the highly-relativistic and dynamical strong-fieldregime of the Einstein equations. It’s therefore possible to testGR in a new arena characterized by velocities comparable to thespeed of light and by large, dynamical curvatures. After describ-ing how gravitational waves and binary black holes can be used totest GR, we will focus upon one particular alternative: Lorentz-violating gravity. Although Lorentz-invariance is at the heart oftheoretical physics, it’s violation in the gravitational sector maylead to a quantum theory of gravitation, for instance. Thus, wewill show the first steps towards the ambitious project of usingGW signals to test Lorentz-violating theories.

Supervisors: Luc Blanchet & Enrico Barausse

8

11:35 – 11:50 Caterina Umilta (2nd yr.)

Improving Planck cosmological results withcomponent separation

The Planck mission provided whole sky maps in the microwaverange, which have opened a window on precision cosmic microwavebackground (CMB) science in cosmology. For this purpose, accu-racy in the separation technique of foreground astrophysical emis-sions becomes crucial. A lot of effort has been done within thePlanck Collaboration, however, the CMB maps they delivered stillhave a residual foreground contamination too high for cosmologicalexploitation. Among these residuals, the unresolved point sourcesbackground is one of the most perturbing, since this is a not wellknown emission ad since it is degenerate with the small scales ofthe CMB angular power spectrum. This degeneracy can be re-moved either by assuming a model for the CMB, as is the case forthe Planck 2015 likelihood, or by introducing some physical knowl-edge about the point sources background. I will present our workto tackle this problem, by modeling the unresolved point sourcesas two independent families, a radio and an infrared one. Thisconstraint breaks the degeneracy and allows, through the blindcomponent separation algorithm SMICA, to better characterizethe foreground residual, which is a crucial step towards estimat-ing cosmological parameters.

Supervisors: Karim Benabed & Jean-Francois Cardoso

9

Alba Vidal Garcia (3rd yr.) 11:50 – 12:10

Modeling ultraviolet-line diagnostics of stars,the ionized and the neutral interstellar medium

in star-forming galaxiesI will present a combination of state-of-the-art models for the pro-duction of radiation by young stellar populations and its transferthrough the ionized and the neutral ISM to investigate the ultravi-olet spectral diagnostics of these three components in star-forminggalaxies. We investigate the ultraviolet spectral features individ-ually most sensitive to the properties of stars, the ionized andthe neutral gas. Beyond an a posteriori justification of the mainspectral diagnostics useful to characterize stellar populations, theionized and the neutral ISM, our models provide a means of sim-ulating and interpreting in a versatile and physically consistentway the contributions by these three components to the ultravio-let emission from star-forming galaxies.

Supervisors: Stephane Charlot

10

12:10 – 12:30 Rebekka Bieri (3rd yr.)

Feedback from Radiatively-driven AGN WindsThe observed co-evolution between the properties of galaxies andthe mass of supermassive black holes (SMBHs) in their centeris commonly explained via a self- regulating mechanism involv-ing feedback from Active Galactic Nuclei (AGN), whereby phasesof rapid gas accretion onto black holes can lead to a release oflarge amounts of energy. Radiation driven winds are dominantin the high-redshift Universe when galaxies are gas-rich and feedthe central SMBH engine at close to the Eddington limit. Usu-ally, sub-grid models for AGN-feedback in simulations rely ona number of assumptions. I will talk about a more ad-hoc ap-proach where we self-consistently modelled the radiation from aSMBH via radiation-hydrodynamics (RHD). I will talk about oursimulations that allow us to examine the detailed physics of thequasar feedback by observing the interaction of the photons with atwo-phase interstellar medium. The simulations allow us a betterquantification of the efficiency of the matter-radiation coupling aswell as the dominant effect responsible in driving an outflow.

Supervisors: Joseph Silk, Gary A. Mamon & Yohan Dubois

11

Sebastien Carassou (2nd yr.) 12:30 – 12:45

Inferring the spectro-morphometric evolution ofgalaxies from image simulations

After 20 years of large and small scale galaxy surveys (SDSS,CFHTLS, UDF etc.), the field of galaxy evolution is now wellestablished in the era of Big Data. We can extract the spectropho-tometric and morphometric properties of millions of galaxies, overa period that covers more than 9 billion years of cosmic history.But current constrains on models of galaxy evolution suffer fromselection biases that, if not taken into account carefully, can leadto contradictory predictions (e.g. the size evolution of bulges fromz ∼ 2).

To adress this issue, we are developing a new approach combin-ing sampling techniques (in a Bayesian framework) and empiricalmodeling with realistic image simulations that reproduce a largefraction of these selection effects. This will allow us to performa direct comparison between observed and simulated images andtherefore infer robust constraints on model parameters predictingthe evolution of bulges and disks from z ∼ 2 to z ∼ 0.

Supervisors: Valerie de Lapparent, Emmanuel Bertin & Damienle Borgne

12

14:45 – 15:00 Tanguy Marchand (1st yr.)

4.5 Post-Newtonian order project:tails-of-tails-of-tails terms

Post-Newtonian theory enables us to predict the waveform of thegravitational waves emitted in the inspiral phase by a system oftwo compact objects coalescing. State-of-the-art computationsprovide the waveform up to 3.5PN (i.e. a precision of 1/c7). Inthe current project of 4.5PN (i.e. obtaining formulae up to 1/c9),the so-called tails terms, which are non-linear interactions betweenthe mass and source multipole of a coalescing system and its massneeds to be computed to high order. After having briefly presentedthe role of Post-Newtonian works in the landscape of gravitationalwave physics today, we’ll present the 4.5PN project focusing specif-ically on the tails-of-tails-of-tails terms.

Supervisor: Luc Blanchet

13

Jean-Baptiste Fouvry (3rd yr.) 15:00 – 15:20

Secular diffusion in discrete self-gravitatingtepid discs

Fluctuations in a stellar system’s gravitational field cause the or-bits of stars to evolve. Fluctuations can either originate fromexternal perturbers (collisionless framework) or from the intrin-sic discreteness of the system (collisional framework). When ac-counting for finite-N effects, one may rely on the inhomogeneousBalescu-Lenard equation, to capture the induced secular diffusion.I will present this formalism and show how one can account for thesystem’s susceptibility via the matrix method. When applied toa tepid stellar disc, it predicts the formation of narrow ridge-likestructures in action-space, in agreement with numerical simula-tions. In astrophysics, the inhomogeneous Balescu-Lenard equa-tion is a rich framework, which may describe the secular diffusionof giant molecular clouds in galactic discs, the secular migrationand segregation of planetesimals in proto-planetary discs, or eventhe long-term evolution of population of stars within the Galacticcentre. It could also be used as a valuable check of the accuracyof N-body integrators on secular timescales.

Supervisors: Christophe Pichon (IAP) & James Binney (Oxford).

14

15:20 – 15:35 Jesse Palmerio (1st yr.)

Gamma ray bursts as a tool for cosmology :population models and host galaxies

Gamma Ray Bursts (GRBs) are brief flashes of gamma rays thatthe are associated with relativistic jets produced by stellar-massblack holes formed after the collapse of massive star or the mergerof two neutron stars/black holes. These events are among the mostluminous in our universe and can be detected up to very high red-shift. Therefore gamma-ray bursts are not only studied for theirinteresting physics but can also provide very useful tools to ex-plore the distant Universe. The Sino-French mission SVOM, to belaunched in 2021, is devoted to these two aspects. In this context,my PhD project is focused on constraining GRB progenitors usingtwo complementary approaches : (1) observational studies of GRBhost galaxies ; (2) GRB population models.

Supervisors: Frederic Daigne & Susanna Vergani (GEPI/IAP)

15

Luciana Ziino (1st yr.) 15:35 – 15:50

Study of exoplanet atmospheres through highresolution spectroscopy

During a transit, when the light from a star is partially blockedby a planet, we can perform a differential observation that willmeasure a transmission spectrum of the thin atmospheric ring ofthe planet at its terminator. Such measurements are being suc-cessfully made today from the ground using for example CRIRESon the VLT. Snellen et al. (2010) studied the high resolutionCRIRES spectra of the star HD209458, revealing absorption linesfrom carbon monoxide produced in the planet atmosphere. Theobserved spectra were completely dominated by numerous telluricabsortion lines, caused mainly by methane and water vapour in theEarth’s atmosphere. To model the Earth’s atmosphere I will usea different approach, first working on the same data set analysedby Snellen. The new method will make use of Molecfit, a softwaretool (created by ESO) that corrects astronomical observations foratmospheric absorption features, based on fitting synthetic trans-mission spectra to the astronomical data.

Supervisor: Jean-Philippe Beaulieu

16

16:15 – 16:30 Florian Sarron (1st yr.)

Clusters of galaxies in the cosmic webNumerical simulations of the formation of large scale structuresin the universe paint the picture of galaxy clusters being locatedat the intersection of cosmic filaments and accreting galaxies andgroups of galaxies along these preferred directions throughout cos-mic time.

These filaments have proven to be hard to detect due to theirintrinsic low density contrast with respect to the field. Detec-tion of such structure has been limited to relatively low redshifts(z . 0.3) and usually through the use of spectroscopic redshifts.Since clusters of galaxies are thought to form around redhifts ofz ∼ 1, being able to study their environment up to such a rangeof redshifts is of critical importance to put constraints on theirformation and evolution scenarios.

We are searching for such filaments around the thousands ofclusters detected in the Canada France Hawaii Telescope LegacySurvey (up to redshift z = 1.5 in the Deep fields, z = 1.2 in theWide fields) through the use of photometric redshifts. For this,we draw 2D galaxy density maps in redshift slices, applying anadaptive kernel technique. This allows us to detect the structuresin the physical environment of each cluster. We are currently de-veloping tools to study the characteristics of such structures, andquantify a possible evolution of these with cluster redshift andmass/richness, and will present preliminary results.

Supervisors: Florence Durret (IAP) & Christophe Adami (LAM)

17

Celine Gouin (1st yr.) 16:30 – 16:45

Gravitational lensing through numericalsimulations: signature of the cosmic web

through the azimuthal shape of galaxy clustersWith the advent of large volume hydrodynamical cosmologicalsimulation and in the context of near future precision lensing ex-periments like the Euclid mission, we are in a good position formaking accurate predictions, identifying new systematics and ex-ploring new probes of the mass content in the Universe. This is themain goal of my thesis work. As a preambule, I used the PLUSN-body simulation to measure the shear signal around massivegalaxy clusters and focused on multipolar moments of this signalas a probe of the connectivity of clusters since they make up thenodes of the cosmic web. We make predictions on the signal am-plitude as a function of cluster-centric distance, mass and redshiftand present the prospects for measuring it in current and futurelensing data sets.

Supervisors: Christophe Pichon & Raphael Gavazzi

18

16:45 – 17:00 Nicolas Cornuault (2nd yr.)

How do cosmological filaments feed galaxies?As matter gravitationally condenses at galactic scales, its maincomponent, namely dark matter (DM), settles in the form of halosand its baryonic component may then decouple from DM becauseelectromagnetics provide it with specific thermodynamic proper-ties. Following this decoupling, the halo gas is able to condense togive birth to a galaxy at the center of its host halo. The furthergrowth of the galaxy results from this ability to maintain efficientcondensation of the gas. Interactions between the galactic out-flows and radiations and the accreting gas regulate this efficiencythrough processes gathered under the name of ’feedback’. How-ever, Those interactions are strongly depend on the nature of theaccreting flows.Apart from the specific case of halo mergers, at haloscale, we can divide those flows into two main categories. The firstone, that has been extensively studied, is that of spherically in-falling gas, idealized as isotropic and referred to as ’smooth’ accre-tion. The second one has been given interest recently mainly dueto numerical simulations and concerns relatively confined streamsof gas that is denser than its surroundings, in the form of filamen-tary structures. The fact that a strong shock wave may propagatealong a filament raised the question of its state when it reachesthe galaxy – is it able to radiate away enough of its internal heatto efficiently accrete on the galaxy? On the other hand, recentobservations tend to show that halo gases may be multiphase andtheoretical studies now account for this. Moreover, other observa-tions and simple physical considerations tend to show that almostall cosmological gas flows are highly turbulent. We argue herethat taking into account both thermal fragmentation and turbu-lent motions in the possibly shocked streams may strongly affectthe nature of the accreting filaments as well as their interactionswith galactic feedback.

Supervisors: Matt Lehnert (IAP), Francois Boulanger (IAS) &Pierre Guillard (IAP)

19

Melanie Habouzit (3rd yr.) 17:00 – 17:20

Formation of supermassive black holesMassive black holes (MBHs) of millions of solar masses and abovereside in the center of most local galaxies, including our own MilkyWay. They are known to power active galactic nuclei and quasars,and are able to acquire 109 solar masses within a span of lessthan 1 Gyr. However, as they formed in the very early universe,large uncertainties on their formation still remain. Direct collapseof pristine gas in primordial halos is one of the most promisingscenarios addressing MBH formation as it predicts the formationof 104−5 solar mass seeds. However, some physical prescriptionsare required for the pristine gas to collapse instead of fragmentingand forming stars, namely the absence of metals and moleculardissociating radiation coming from nearby star-forming galaxies.Cosmological hydrodynamical simulations provide us with a verygood tool to track metal enrichment as well as direct collapse can-didates environment. Using several sets of simulations (differentresolution, supernova feedback recipes), we have investigated thenumber density of potential direct collapse candidates through cos-mic time.

Supervisor: Marta Volonteri

20

17:20 – 17:35 Siwei Zou (1st yr.)

The gas in and around high redshift galaxiesWe study a complete sample of neutral atomic carbon quasar ab-sorption systems, observed with the Very Large Telescope. Suchabsorbers probe highly shielded gas in which molecules are ex-pected to form. We will get information on the depletion of thedust, metallicity, molecular content and physical state of the gasin the ISM of high redshift galaxies.

Supervisors: Patrick Petitjean & Pasquier Noterdaeme

21

Tilman Hartwig (2nd yr.) 17:35 – 17:50

Gravitational Waves from the Remnants of theFirst Stars

Gravitational waves (GWs) provide a revolutionary tool to in-vestigate yet unobserved astrophysical objects. Especially the firststars, which are believed to be more massive than present-daystars, might be indirectly observable via the merger of their com-pact remnants. I have developed a self-consistent, cosmologicallyrepresentative, semi-analytical model to simulate the formation ofthe first stars to track the binary stellar evolution of the individualsystems until the coalescence of the compact remnants. I presentestimates of the contribution of primordial stars to the intrinsicmerger rate density and to the detection rate of the AdvancedLaser Interferometer Gravitational-Wave Observatory (aLIGO).Owing to their higher masses, the remnants of primordial starsproduce strong GW signals, even if their contribution in numberis relatively small. I find a probability of ∼ 1% that the currentdetection GW150914 is of primordial origin and we can expectthat aLIGO will detect roughly 1 primordial BH-BH merger peryear for the final design sensitivity, although this rate dependssensitively on the primordial initial mass function. Turning thisaround, the detection of black hole mergers with a total binarymass of ∼ 300M� would enable us to constrain the primordial ini-tial mass function.

Supervisor: Marta Volonteri

22