This project is enclosed in the theoretical astrophysics framework. We propose to confront theoretical results to observations in order to improve our current understanding of star formation. The theoretical results will be obtained using state-of-the-art international numerical codes that integrate dynamical, chemical and radiative processes. We will consider the collapse and the evolution of large turbulent and magnetized molecular clouds (up to a few thousand of solar masses over scales of a few parsec) which give birth to star clusters. Using the radiation-magneto-hydrodynamics code RAMSES, which is to date unique in the community, we will study the interaction between star formation and the interstellar medium (ISM) and account for the dynamical and thermal feedbacks of the forming protostars on the chemical evolution of the molecular clouds. We will have a particular focus on the regions where matter is being ejected from the protostars in the form of outflows and jets. The simulated 3D physical cloud structures will be used to model chemical evolution using the robust chemical code ALCHEMIC and its state-of-the-art chemical network that includes hundreds of molecules and ices. The chemical and physical structures of the molecular cloud will then be post-processed using the 3D radiative transfer code RADMC-3D, to produce synthetic molecular line emission and dust emission maps. These maps will form the basis of a comprehensive physical understanding for observations, which are expected to make considerable progress within the next few years with the release of HERSCHEL data and the forthcoming Atacama Large Millimetre Array (ALMA) radio-interferometer. This project will make close connection between theory and observations and will give rise to a better understanding of the star-forming ISM physics.

Galaxies grow in mass by accreting matter flowing along streams and filaments of the cosmic web. But they would become too massive, if there was not a super-massive black hole in their nucleus to inject energy and matter back into the intergalactic medium. The details of such a feedback mechanism are however still mysterious. Giant galaxies in cluster centers (BCGs) are a unique example of Black Hole/Intergalactic gas interaction. The goal of LYRICS is to understand the life cycle of gas in the presence of AGN-feedback by studying the large filamentary network surrounding those galaxies. LYRICS builds up on i) large datasets and archival observations (MUSE, ALMA, NOEMA) ii) new implementation of physical models of gas excitation and iii) mature hydrodynamical simulations including AGN-feedback. Organised around a very complementary team, LYRICS will provide the very first comprehensive picture of the origin and state of the gas circulating around BCGs.

Star formation in galaxies is the engine that drives cosmic evolution. Decades of observations have yielded diverse empirical rules about how it operates, but we still have no comprehensive, quantitative theory that encompasses the full range of physics and spatial scales involved. The 4-year project DAOISM (Deep Analysis Of the InterStellar Medium) is a collaborative project between observational astronomers, theoretical astrophysicists and data scientists that is developing advanced statistical methods and data analysis techniques for deployment on observations of the cold star-forming interstellar medium (ISM) in galaxies. The project combines state-of-the-art modelling, data from world-class facilities (ALMA, MUSE, IRAM), and original data science methods in order to answer key outstanding questions in star formation: How best to estimate local physical conditions? Which spectral line diagnostics best capture the star-forming capability of molecular gas in clouds and galaxies? How does galactic environment affect their physics and chemistry? The essence of DAOISM is to build a detailed comprehensive understanding of the local Orion B cloud and use it to improve the interpretation of poorly resolved extragalactic observations that map how star formation proceeds across the diversity of galactic environments in the local Universe. Our approach combines statistical analysis, machine learning and Bayesian inversion of data based on state-of-the-art models.These new strategies are urgently needed since the latest astronomical data and physico-chemical models of the ISM now regularly attain a volume and complexity that are intractable to the simple correlation analysis methods employed in the field to date. DAOISM leverages our team’s complementary state-of-the-art observational programs (co-PIs of ORION-B and PHANGS), ISM models (Meudon PDR code, time-dependent astrochemistry code) and data science expertise. The ORION-B project is an IRAM-30m Large Program imaging the Orion-B giant molecular cloud (GMC). The resulting unique wide-field hyper-spectral data cube (>30 molecular lines detected, ~820,000 pixels, ~240,000 spectral channels per pixel) enables an unprecedented characterization of the physical structure, chemistry and dynamics of a GMC, and their link to its star formation activity. PHANGS combines high resolution ALMA observations of molecular emission lines, VLT/MUSE observations of ionised gas lines, and Hubble Space Telescope imaging to deliver the first statistically robust description of the evolution of ~100,000 star-forming regions throughout the nearby galaxy population. A dedicated joint analysis of the PHANGS and ORION-B surveys is an exceptional opportunity to bridge Galactic and extragalactic star formation studies. The DAOISM project is structured along three scientific work packages, progressing from a detailed description of the ISM physical and chemical properties that locally control star formation, to the development of advanced methods for revealing the best diagnostics for these parameters and processes in local as well as galaxy-scale measurements. By the end of the ANR-funding period, DAOISM will deliver: i) a comprehensive characterization of the physical, kinematic, chemical and magnetic structure of a typical GMC and of their relation to star formation and feedback processes, as well as new statistical methods adapted to the description and analysis of the magneto-turbulent cold ISM; ii) a new observing mode for the NOEMA interferometer allowing efficient large wide-field high resolution observations, iii) new tools for disentangling the emission contributed by physically distinct regions within a single resolution element, and for comparing resolved Galactic observations to spatially unresolved extragalactic data, a major roadblock for current extragalactic star formation studies. All tools and codes developed by DAOISM will be made public.

We are entering a golden age for astrochemistry, with the simultaneous increase in detector sensitivity, band pass and telescope collecting area. The analysis of astronomical spectra, and the detection of new interstellar species is nevertheless hampered by the lack of knowledge molecular spectroscopy, both for stable species and for ions and radicals. Despite the hostile environment conditions, the interstellar and circumstellar chemistry is remarkably rich, with many species, their isomers and isotopologues (notably those where a D atom is substituted for an H). Another issue is the lack of understanding of the chemical formation pathways to interstellar molecules, leading to a low efficiency in the search for new species. This interdisciplinary project aims at improving both points by developing jointly : i) new techniques for more sensitive molecular spectroscopy based on instrumental developments in the THz domain, ii) new production techniques for possible interstellar molecules, iii) new observation and modeling programs. The team gathers well known experts in astrophysics, microwave instrumentation, molecular spectroscopy and organic synthesis who will share their knowledge.

Magnetism strongly impacts the formation of stars and planets. Thanks to state-of-the-art numerical simulations of molecular cloud collapses, developped by us, we can now predict the structure and magnetisation of the baby protostar at the end of the collapse phase. During the latest stages of star formation, once the protostar has grown and got rid of its envelope, magnetism and its impact on star and disk evolution, and planet formation, have now well been investigated. The intermediate embedded protostellar phase has however been poorly studied, and models of magnetic embedded protostars are remarkably absent. We have therefore no idea of how a newly formed magnetic protostar evolves for about 1 Myr before revealing itself out of her dusty cocoon, and how its magnetic field is involved in the long-standing problem of accretion/ejection in protostars. To fill this important gap, we will build a theoretically and observationally consistent MHD model of a young magnetic protostar.