High-redshift Dusty Star-Forming Galaxies: a panchromatic approach to constrain massive galaxy evolution

The goal of this thesis is to investigate the early stages of massive galaxy evolution by defining an overall view of their physical properties combining the information extracted by all the details of their spectral behaviour. To this aim, I focused on the population of Dusty Star-Forming Galaxies at the Cosmic Noon ($zsim 2$). In my thesis project, I first deal with the modelling of the spatially-averaged time evolution of galaxy baryonic components, namely gas, stars, metals and dust, on the basis of a simple but effective approach that allows to solve analytically the equations that describe their evolution. Contrariwise to most of the analytic models on the market, the one developed during this PhD thesis self-consistently compute the metal and dust enrichment histories of the cold gas and stellar mass using as input the solutions for the evolution of the mass components. The solutions are coupled to specific prescriptions for parameter setting (inspired by extit{in-situ} galaxy-black hole co-evolution) and merger rates (based on numerical simulations) and, as such, reproduce the main statistical relationships followed by high-z massive star-forming galaxies and local ellipticals, that are thought to be their quiescent counterparts at $zsim0$. The analytic solutions are then exploited to interpret the spatially-averaged astrophysical properties of a pilot sample of (sub-)millimeter selected Dusty Star-Forming Galaxies in the multi-wavelength GOODS-S field and spectroscopically confirmed to be at the peak of Cosmic Star Formation History. Ultimately, they are used to disentangle the main physical processes regulating the evolution of these galaxies. The study highlights the importance of multi-wavelength broad-band and spectroscopic data to constrain dusty galaxy evolution at high-z and their role in the formation of spheroids, along with the need of a complete theoretical scenario that allows to self-consistently interpret the outcomes obtained from observational analyses. One possible framework is the one provided by the extit{in-situ} scenario for galaxy-black hole co-evolution, that has been used in this work to interpret the reconstructed panchromatic view combining spatially integrated (i.e. galaxy age, Star Formation Rate, stellar mass, dust mass, dust attenuation), spatially resolved (multi-wavelength sizes) and spectral (i.e. molecular gas content, kinematics and AGN/stellar driven outflows) properties of the aforementioned pilot sample of DSFG. The analysis is performed under specific requirements (e.g. spectroscopic measurement of galaxy redshift, complete sampling of galaxy multi-band emission) in order to unbiasedly constrain galaxy integral properties by performing an energy-balanced fit of the SED from the UV/optical to the radio band, including also galaxy X-ray emission, with the Code Investigating GALaxy Emission. Galaxy optical, far-infrared and radio sizes are measured from continuum maps at the highest spatial resolution currently available ($Delta hetalesssim1$ arcsec). CO spectral emission lines are extracted from publicly available data cubes in the Atacama Large Millimeter/sub-millimeter Array Archive and allow to measure the molecular gas content and to disentangle between a disk dominated configuration of the gaseous component and molecular outflows possibly driven by the central active nucleus. The multiple pieces of information coming from such a panchromatic study offer a clear description of the properties of individual galaxies and, once each of them is inscribed in the evolutionary context, offer a general view of the evolutionary mechanisms.