Dealing with large amount of data is a new problematic task in astrophysics. One may distinguish the management of these data (astroinformatics) and their scientific use (astrostatistics) even if the border is rather fuzzy. Dimensionality reduction in both the number of observations and the number of variables (observables) is necessary for an easier physical understanding. This is the purpose of classification which has been traditionally eye-based and essentially still is but this becomes not possible anymore. In this talk, I present a general overview of machine learning approaches for unsupervised classification, with applications to stars (chemical abundances) and galaxies (spectra).
During the 90s, the keystones of the celestial reference system took distance since the community left a stellar realisation for an extragalactic realisation. Very long baseline interferometry [VLBI] is used in this purpose because it determines the absolute astrometric positions of thousands of active galactic nuclei with an accuracy of tens of microsecond of arc. The realisation of the extragalactic celestial reference frame by a well-chosen set of sources is at the basis of modern geodesy for wich scientific and societal challenges are regularly highlighted.
The VLBI astrometric accuracy stayed unrivaled for the 40 last years. Only the Gaia space mission competes VLBI nowadays. By skiping technical and technological challenges that allowed this feat from the ground, I will explain that this precision makes us sensitive to perturbations linked to the complex and animated physical structure of the active galactic nuclei. Until now, the adopted strategy for the realisation of a hyper-stable celestial reference frame is to put aside the seemingly most turbulent sources. I will give some elements that let us think this strategy will not be good enough at medium term. The challenges of the future for ever more accurate celestial frame will require the study of those sources and their regular monitoring in collaboration with the astrophysical community in order to understand (i) on which sources can we rely on to realize our celestial reference frame and (ii) given a source, can it be sufficiently stable on a finite time to be useful for the realisation of a celestial reference frame.
Better understanding Solar System Giant Planet formation and evolution requires in situ measurements, remote sensing observations either with telescopes or planetary missions, and modeling. While more and more exoplanets are discovered every day and while we will better characterize them with new observatories like JWST, the planets of the Solar System remain our local laboratory for studying formation and evolution of such bodies. The (sub)millimeter domain, owing to the very high spectral resolution of the heterodyne technique and to the ever increasing spatial resolution and sensitivity of new observatories like ALMA, is suitable for determining planetary atmospheric composition and dynamics when coupled with appropriate radiative transfer, photochemical or thermochemical modeling.
In this seminar, I will present observations and modeling of the Solar System Giant Planets I have led or been involved in.
I will show how Herschel and ALMA observations, and time-dependent 1D or 2D photochemical modeling have enabled us to improve our understanding of how the composition and chemistry in the stratospheres of the Giant Planets are altered by seasons and external sources. I will also introduce the Submillimetre Wave Instrument of the Jupiter Icy Moons Explorer (JUICE) mission, which will allow us, in about a decade from now, to monitor Jupiter’s atmosphere, both in terms of chemistry and dynamics, and with spectral and spatial resolutions and temporal coverage never achieved before.
I will finally show that thermochemical modeling of the deep tropospheres of the Giant Planets can help us constrain their deep composition and thus their formation processes. The next step is the participation in an atmospheric probe proposal for the Ice Giants, and the development of its mass spectrometer, in preparation for a NASA-ESA joint mission to these distant worlds.
The recent detection of Gravitational Waves (GWs) by LIGO and VIRGO opened a new observation window on the Universe and started the era of Gravitational Astronomy. Atom interferometry has been proposed to extend the detection bandwidth of GW detectors in the infrasound band (10 mHz – 10 Hz) [1], where actual ground based detectors are limited by low frequency gravity noise. Adopting as probes arrays of atomic ensembles in free fall, and tracking their motion on geodesics with atom interferometry allows the suppression of Newtonian Noise [2], enables low frequency sensitivity, and opens the way toward the realization of low frequency GW detectors on Earth. I will report on the « Matter wave – laser based Interferometer Gravitation Antenna » (MIGA) project [3], whose target is to build an atom interferometry based demonstrator for GW detection in the underground environment of LSBB (Rustrel, France).
[1] S Dimopoulos et al, Phys Lett B 678, 37 (2009)
[2] W Chaibi et al, Phys Rev D 93 (2), 021101 (2009)
[3] B Canuel, A. Bertoldi et al, Sci. Rep. 8, 14064 (2018)
The ESA Gaia mission is revolutionising our understanding of the Milky Way by providing precise proper motions and parallaxes for over a billion stars, as well as excellent photometry. Among the many aspects of Galactic astronomy that Gaia can tackle, the unprecedented size, depth and quality of this dataset allows us to better characterise stellar clusters and to discover new objects. In this talk I will review some of the major results in Galactic cluster science recently obtained from Gaia data and discuss the ongoing and future work that Gaia makes possible.
The ellipsoid of random motions is a fundamental ingredient of galaxy dynamics. In particular, it can be used to constrain the shape of orbits of stars or gas in galactic disks. First, I will present results from a recent study that constrained the structure of stellar orbits in nearby galactic disks. A new correlation between the stellar mass and the shape of the orbits has been evidenced: stellar orbits in more massive galaxies are radially biased, while less massive disks have more tangential to isotropic orbits. A discrepancy with expectations from the epicycle theory of orbits has also been evidenced. Then, results from ongoing studies of the velocity ellipsoid of HI and CO gas in other samples of galaxies will be shown. Contrary to the common assumption that random motions of gas are isotropic, it is shown a large diversity of orbits in nearby spiral galaxies. Disagreement with the epicyclic approximation are again observed in the atomic gas component.
The new generation of interferometers provide unprecedented constraints on the protostellar disk formation process. Observations indicate that most disks have a small extent at the Class 0 stage and that disks grow in size at latter stages. I will present the results of 3D protostellar collapse calculations that cover a wide range of initial mass (from 0.5 to 100 solar mass), as well as different initial rotation and/or turbulence support. The calculations are performed using the RAMSES code, including the effect of non-ideal MHD with the ambipolar diffusion and radiative transfer. I will show how ambipolar diffusion is regulating the disk and outflow formation at the early stages of the class 0 phase. I will discuss the disk properties: magnetisation level, magnetic field lines topology, stability. In a second part, I will present recent work done in the context of the protostar formation (second collapse) where the effects of non-ideal MHD (ambipolar and Ohmic diffusion) are taken into account. I will highlight the differences with previous results obtained with ideal MHD and show to what extent these kind of models can provide constraints on the protostellar evolution (disk, protostar). I will finally present preliminary results of protostellar collapse models which include coupled dust and gas dynamics.
Eratosthenes of Cyrene (276-194 BCE) was one of the great scholars of the Hellenistic period. Responsible for the Library of Alexandria,