Ever since we first laid eyes on the twinkling lights in the night sky, our species began its age-old quest to understand how we came into existence as a planet and what the future holds for it. Most of the traditional formation theories of planets were anchored on the examples drawn from our own solar system. With the surprising and emerging trends among the yet incomplete exoplanet demographics, we are at the wake of a rigorous revision of our theoretical understanding of how planets form and evolve. To form accurate theories however, it is necessary to base them on a planet population that spans the complete range of parameter space not only in terms of its physical properties like mass and orbital separation, but also with respect to the type of stars that host these planets and their age. In this regard, direct detection, whereby you measure photons coming from the planet, helps one get closer to the whole picture since the ideal target population for this technique are young, giant planets in wide orbits that are generally difficult to observe with other detection techniques. Over the last few years, the sensitivity reached by direct imaging observations has seen tremendous improvement owing to the use of high-contrast tools like coronagraphy and adaptive optics. The development of high-resolution spectrographs together with advanced post-processing techniques have recently, for the first time, enabled witnessing planets while in the process of being born, helping us understand how they grow by devouring material from the planetary nursery — a mechanism known as accretion. This is an exciting era for planetary science, with many ongoing as well as planned future surveys with both ground and space-based telescopes dedicated to unravelling the mysteries surrounding the origin of planets. 

In this thesis, I provide an overview of direct detection as a tool to study sub-stellar objects – a categorisation that includes both planets and brown dwarfs, and whose blurred lines of distinction is a point of contention in astronomy today. I concentrate my discussion on two techniques, high-contrast imaging and high-resolution spectroscopy, both of which have proven significant in the race for planet detection and characterisation. Three scientific research works are carried out as a part of this thesis, using which I highlight the benefits of these techniques in constraining the physical properties of planets and brown dwarfs, as well as obtaining clues to their formation mechanism. In Paper I, I search for a Jupiter-like planet around a nearby Sun-like star that has long eluded imaging surveys, revealing its presence only via its influence on the parent star. I show how the brightness constraints at various separations and multiple wavelengths from the parent star help set a lower limit on the vaguely defined age of the system, in the absence of detection of the planet in our observations. In Paper II, I report the discovery of two low-mass companions to a massive, bright, young star, infer their orbital dynamics from multi-epoch imaging data, and constrain their physical properties using simultaneous low-resolution spectroscopy. In Paper III, I use a high-resolution spectrograph to observe for the first time, resolved Hydrogen and Helium emission lines from a young, isolated planetary-mass object in the midst of formation. Based on analysis of these line profiles, I obtain clues to the possible accretion mechanism at play in this nebulous cosmic phenomenon.

The thesis explores the interactions between cosmological gravitational waves (GWs) and large-scale magnetic fields. GWs are radiation produced by spacetime variations of the stress-energy tensor. Due to the weak coupling between gravity and the matter sector, GWs are a unique messenger from the early Universe before the Cosmic Microwave Background (CMB). Magnetic fields are observed across the Universe from the scales of planets and stars to galaxies and clusters, as well as the voids beyond the clusters. The present-day large-scale magnetic fields are believed to have evolved from primordial seed fields via magnetogenesis mechanisms active during the cosmic inflation and reheating epochs or the cosmological phase transitions occurring at the electroweak (EW) or quantum chromodynamic (QCD) scales in the early radiation-dominated (RD) era. The production of stochastic GW backgrounds (SGWBs) can be expected from the primordial electromagnetic (EM) fields or magnetohydrodynamic (MHD) turbulence around the time of EW and QCD phase transitions. The SGWBs then propagate through the pre-CMB Universe until the present day, carrying with them essential imprints of the corresponding sources as well as the underlying gravity theory at the early times. Since MHD turbulence is ubiquitously expected in the RD era, their induced SGWB spectrum is extensively studied.In one aspect of the thesis, we demonstrate that the MHD-GW system exhibits features of modified gravity (MG) in terms of the spectral slopes and amplitudes of the relic SGWB. We compute the spectra of GWs produced by MHD turbulence at the EW and QCD phase transitions, assuming massive gravity and scalar-tensor theories as two MG examples. Then we comparatively analyze these modified GW spectra with their counterparts in general relativity, and determine their qualitative and quantitative differences due to three effective MG parameters – graviton mass, GW friction, and GW speed. These spectral features are compared against the existing pulsar timing array (PTA) measurement in the nHz band as well as the expected GW detection sensitivities of upcoming instruments such as the Laser Interferometer Space Antenna (LISA) in the mHz band and Square Kilometer Array (SKA) as a PTA. The framework is general and can be applied to non-MHD sourced GW spectra. However, fixing MHD sources yields concrete constraints on the effective MG parameters.The other aspect of the thesis concerns the interaction between SGWBs from the early Universe and large-scale magnetic fields in the post-CMB Universe, which would convert a fraction of the gravitons in the SGWBs into photons of the same frequency via the inverse Gertsenshtein effect. The graviton-induced photons could then be a source of spectral distortions of the blackbody CMB.This, in turn, would allow us to constrain the GW amplitudes in the MHz-THz frequency regime. The high-frequency GWs (HFGWs) correspond to new physics in the early Universe such as beyond-Standard-Model mechanisms or sub-stellar mass primordial black holes. We place constraints on the HFGW energy densities by exploiting the existing direct observations of the radio sky, measurements of the 21-cm signal upper limits, the kinematic Sunyaev-Zeldovich observations, and assuming that graviton-induced photons saturate all of the reported radio excess over the CMB. We also forecast the potential of SKA and proposed future CMB surveys as novel HFGW detectors, and show that they will significantly tighten the current constraints and bring us a step closer to detecting HFGWs.

The first billion years of the Universe is a unique era, marked by the formation of the first stars, galaxies, and accreting black holes, which release ionising radiation into the intergalactic medium (IGM). As a result, these luminous sources initiate a period during which the cold and dense IGM, primarily consisting of neutral hydrogen (H_I), is heated and ionised. We refer to this era as the Epoch of Reionization (EoR). The EoR is a global phase transition that is not trivial to observe or model computationally. It is a multi-scale event that evolves with time and depends on the nature of the astrophysical processes that govern the formation of stars and galaxies, as well as the fundamental cosmology that defines the properties of the large-scale IGM. While various measurements of cosmic reionization exist, presently they are too few to constrain the entirety of the process. However, observations from the James Webb Space Telescope and the Square Kilometre Array (SKA), among others, will provide new insight into the process. Particularly, the SKA will observe the power spectrum (PS) of the 21 cm signal from the EoR, which originates from the hyperfine transition of neutral hydrogen atoms H_I in the IGM that can emit 21 cm photons. 

In Paper I, we investigate the evolution of the 21 cm PS across the EoR by perturbing the signal and studying its composing terms. We highlight the importance higher-order terms play in shaping the PS on large scales and quantify its evolution. Crucially, we find a characteristic length scale within the 21 cm PS, determined by the mean free path ionising photons travel in the IGM (MFP). Hence, the 21 cm PS has two regimes. We show that the large-scale signal is a biased version of the cosmological density field, and the small-scale PS is determined by the astrophysics of reionization. In Paper II, we use the decomposition of the 21 cm PS and relate it to the PS of the free electron density field. Thus, we analytically connect the 21 cm observable to a probe of the free electron density field. Such a probe is the patchy kinetic Sunyaev-Zel'dovich effect (pkSZ), observed as a foreground to the primary cosmic microwave background temperature anisotropies on small scales. The pkSZ is an integrated probe sensitive to the duration of the EoR and the characteristic size of ionised bubbles. We construct a forecast study of both probes. We show that inferences from 21 cm PS from the SKA can be verified when combined with the pkSZ observation, as each data set is influenced by different systematics. In Paper III, we focus on the modelling of the MFP within large-scale simulations, focusing on the end of reionization (EndEoR). The MFP of ionising photons is inferred from quasar data and depends on several factors. In the post-EoR era, it depends on the distribution and evolution of Lyman Limit systems (LLS), small-scale absorbers that are typically not resolved in large-scale simulations. We investigate the assumptions needed to accurately model the LLS in simulations, and we study their impact on the observables at the EndEoR. We find that LLS modelling has a profound impact on the duration of the final stages of the EoR, the shape of the 21 cm PS as well as other observables of the ionised IGM inferred from quasar spectra, such as the Ultraviolet Background of ionising photons, the effective optical depth of Lyman alpha photons, and the MFP of ionising photons.

Stellar spectroscopy is a powerful tool for analysing everything from the physics of stars to the evolution of galaxies. Dark absorption lines in the light received here on Earth offer direct insight into physical processes in the stars themselves, as they uniquely reflect the atomic structure of the elements present. These lines can be used to yield useful information about the stellar atmosphere, such as its temperature, density, and chemical composition, representing the abundance of the elements in the star.

Furthermore, we can use abundance information in stars to trace the formation of elements and differentiate between separate nucleosynthesis paths. A good understanding of the formation process within stars during their lives and explosive deaths can, in turn, help us better understand the creation of galactic structures.

The inference of stellar abundances requires a physical model of the stellar atmosphere, and only by accurately modelling the interactions between photons and matter can we measure the abundance of the element that causes the spectral line. However, we are often held back by simplistic assumptions to reduce the complexity of the model which, in the past, has been a necessity. However, with the advancement in computational power, these assumptions can be removed. First, local thermodynamic equilibrium (LTE) is a common simplification that is only valid in deeper layers of stars at higher densities. Second, it is often assumed that the atmospheres of solar-like stars are hydrostatic and can be represented by one-dimensional (1D) models, instead of running expensive three-dimensional (3D) hydrodynamic simulations. Both simplifications cause incorrect predictions of observed strengths and shapes of spectral lines. I illustrate the impact of moving from LTE to non-LTE by replacing Saha-Boltzmann equations with statistical equilibrium calculations, producing a significant impact on titanium, a key galactic evolution tracer due to both its number of spectral lines and unique link between the alpha and iron-peak elements. Previous studies of titanium have revealed a substantial ionisation imbalance between titanium abundance estimates based on Ti I and Ti II spectral lines when using classic LTE models. The LTE assumption leads to insufficient ionisation via ultraviolet radiation, which has a pronounced impact on the Ti I population due to its relatively low abundance. However, the lack of accurate atomic data for, e.g., inelastic collisional cross-sections, has made non-LTE studies challenging and subject to astrophysical calibration. In this work, I present a new model atom based on the latest atomic data and apply it to the largest data set yet examined for 1D non-LTE titanium abundances of 70,000 stars. A significant non-LTE effect is found for metal-poor stars, in particular. Metal-poor dwarfs are not brought to balance by 1D non-LTE calculations, but giants have their imbalance greatly reduced, bringing the galactic chemical evolution of titanium using neutral and ionised lines into agreement. Finally, we follow this with the first 3D non-LTE simulation of titanium in any star, which in our case is the Sun. We find that 3D modelling is necessary to reproduce spatially resolved observations of sunlight and that non-LTE effects of neutral titanium are boosted in 3D, raising the measurement of the solar Ti abundance.

Giant planets have had a long history of radically overturning our expectations of how they form and where they are likely to be found around other stars. In 1995, the first exoplanet detected around a Sun-like star was not found further out from the star, as expected from the locations of Jupiter and Saturn and then current formation theories, but rather on a 4-day orbit with a surface temperature just above the melting point of silver and a radius nearly twice that of Jupiter. Since then we have detected thousands of exoplanets, which have shown remarkable diversity, and imaged the discs around young stars where baby planets are being born. Although there are many common characteristics of these exoplanets and discs, some stand out as outliers. There are systems that are thought ‘too old’ to form planets, or planetary-mass companions that are ‘too big’ in relation to their host stars or should not have had the time to grow that massive to begin with. These are some of the (many) outstanding questions on the frontier of research into planet formation, and in just the past few years we have finally been able to directly observe a few planets that are in the process of forming. In an almost parallel development to the rapid expansion of research into exoplanets, we have also come to realise that brown dwarfs can be excellent analogues to giant planets and contribute significantly to our understanding of both the atmospheres and the formation process of giant planets.

This thesis explores several aspects of the dynamics of substellar atmospheres and the accretion process at planetary masses. It discusses the observing methods, which provide the foundations of the photometric and spectroscopic observations that produced the data for the included papers. This is followed by a chapter on star and planet formation and one discussing the variability of substellar atmospheres. The final chapter delves more directly into the observational features of accretion and the tracers and diagnostics which enable us to start qualitatively characterise the accretion process at planetary masses.

The first paper presents a NOT/NOTCam photometric survey of ten brown dwarfs, where the goal was to identify new high-amplitude variables that could be suitable for deeper studies. A large fraction was found to be variable, significantly adding to the number of known variable brown dwarfs.

In the second paper, integral field spectroscopy obtained with VLT/MUSE of the planetary-mass companion Delorme 1 (AB)b and its host binary star is presented. Very strong hydrogen line emission was detected from the companion, indicative of active accretion in this 40-myr-old system. 

In the third paper, Delorme 1 (AB)b was further studied by VLT/UVES and R = 50000 spectroscopy. As a result, near-UV hydrogen emission lines were resolved in a planetary-mass companion for the first time. The analysis of these lines strengthened the case for active accretion in the companion.

Neutral hydrogen (HI) is the most abundant baryonic element and a crucial component of galaxies. In the early universe, the interaction between neutral gas and the light produced by galaxies is thought to have given rise to cosmological reionization. During this last major phase transition of the Universe, the bulk of HI within the intergalactic medium (IGM) was ionized. However, observational constraints limit our understanding of the interplay between the radiation produced by galaxies and their neutral gas, and that of the physical processes that caused this important cosmological period. Another poorly understood mechanism is the one that drives the escape of Lyman-α (Lyα) emission from star-forming galaxies and shapes the line profile of this well-used tracer of galaxies at high redshift. To fully answer these questions, direct observations of the neutral gas content and distribution of Lyα emitters are needed.

This thesis presents 21cm observations of the neutral gas reservoirs of rare local galaxies that are analogous to objects in the early universe. We have observed the neutral gas of Haro 11, the closest known ionizing radiation (Lyman Continuum, LyC) leaking galaxy. This 21cm observation was the first to successfully map the neutral gas distribution of a confirmed LyC emitter. We observed a strongly asymmetric neutral gas distribution resulting from a merger event, with the bulk of the HI mass offset from the regions producing LyC radiation in the galaxy. By decreasing the column density of gas on large scales, this HI distribution has facilitated the escape of ionizing radiation from the center of the galaxy to the IGM.

We have also observed the neutral gas content and distribution of local Lyα-emitting galaxies on a variety of angular scales. We mapped the neutral gas at scales that can be compared to those characterizing Lyα emission in two galaxies, and supplemented the data with integral field spectroscopic observations tracing ionized gas and dust. Additionally, we reduced low angular resolution 21cm observations of 37 galaxies in the Lyman Alpha Reference Samples. We found clear evidence of interaction in most of the galaxies in the samples (~60%), indicating that mergers play an important role in the Lyα escape from galaxies. We found that global HI properties did not correlate well with any metric quantifying Lyα emission, although different classes of Lyα-emitters were distributed differently around HI scaling relations. This indicates that star-forming galaxies with low neutral gas content have a higher chance of emitting strongly in Lyα, but ultimately, neutral gas regulates the escape of Lyα emission on small scales.

21cm studies of both Lyα and LyC emitters conducted in this thesis have demonstrated that galaxy interactions play an important role in the escape of ultraviolet (UV) radiation from local galaxies. By perturbing the neutral gas content of galaxies and enhancing their star formation rate, these interactions are interesting mechanisms that facilitate the escape of UV radiation. According to cosmological models, galaxy mergers occurred more frequently in the early universe. Assessing the contribution of environment and galaxy interactions at high redshift will be essential to fully understand the first billion years of our universe.

The astrophysical origin of rapid neutron capture (r-process) elements has long remained a puzzle and been the object of scientific debate. Neutron star (NS) mergers have historically been suggested as an ideal site for the creation of these elements, and were propelled into focus following the detection of the first binary neutron star (BNS) merger in 2017. The gravitational wave (GW) signal GW170817 was accompanied by a short gamma-ray burst (sGRB) GRB170817A, and a radioactively powered electromagnetic (EM) transient AT2017gfo, known as a kilonova (KN). Since this detection, the study of NS mergers has greatly expanded across the diverse fields that model the various stages of the merger, from GW signal modelling, to radiative transfer studies predicting the emergent KN lightcurves (LCs) and spectra.

One main goal of studying NS mergers and the associated KNe is to establish the importance of compact object mergers as key sites of r-process nucleosynthesis in the Universe. As such, identification of elements and their abundances within the merger ejecta represents a critical objective. LC and spectral analyses of KNe provide promising channels to do so, and require detailed models in order to interpret observational data. With complete GW and multi-band EM data only available for a single object thus far, the importance of detailed models regarding every aspect of KN physics is paramount. KN simulations typically make use of radiative transfer (RT) codes that propagate photons through the expanding ejecta, in order to provide LC and spectral outputs. These often model the early, photospheric times of the KN, when the ejecta are still dense enough such that the gas state is well described by Local Thermodynamic Equilibrium (LTE) conditions, which requires thermal collisional processes to dominate within the ejecta.

Since the ejecta are expanding rapidly however, these conditions cease to apply after several days, and the KN transitions to the Non-Local Thermodynamic Equilibrium (NLTE) regime, where thermal collisional processes are no longer dominant in establishing the gas state of the ejecta. This now requires the detailed modelling of various NLTE processes which increases the complexity, yet modelling of this regime can also provide great rewards. Notably, as times goes on and the ejecta continue to expand, they will eventually become optically thin to most wavelengths and enter the nebular phase. There, the spectra are expected to be emission line dominated, providing an excellent opportunity for element identification by spectral analysis.

This doctoral thesis conducts RT modelling in order to explore the NLTE regime of the KN in a systematic, physically accurate way. To this end, the spectral synthesis code SUMO (SUpernova MOnte Carlo Code) was adapted to model KNe, and used to investigate the spectral emission in the NLTE regime. The work in this doctoral thesis provides a first step into fully consistent modelling and analysis of KNe at later times, and a solid foundation from which to move forwards.

Feedback in star forming galaxies is the key process that regulates how many stars form given the available gas reservoir. The radiation, energy and momentum released by the stellar activity and the active galactic nuclei change the physical properties of the gas in the galaxy such as the ionisation state, the density and the kinematics. In the most extreme scenarios, feedback is able to launch powerful outflows that can bring significant portions of gas out into the intergalactic medium, directly suppressing the star formation in the galaxies and therefore influencing their evolution. For non-active galaxies the dominant source of feedback is represented by the ionising radiation, stellar winds and supernova explosions of massive stars. Young star clusters (YSCs) are the natural habitat of massive stars and inject large amounts of radiation, energy and momentum into the surrounding interstellar medium (ISM). The fractal geometry of stellar clustering amplifies the impact of the stellar feedback on the surrounding gas in the ISM. In fact, the total outward momentum acting on the gas around a star cluster is given by the sum of the contributions from the single stars, which being at the centre of the physical system never cancel each other outward momenta. 

The main goal of this thesis is to investigate how stellar feedback originates at small scales, as well as to quantify its impact on the ISM gas surrounding young stellar populations. We use far-ultraviolet (FUV) spectroscopy of YSCs to study both the physical properties of the stellar population and the gas kinematics in the ISM. The first study presented in the thesis focuses on the starburst galaxy Haro 11, characterized by three knots of star formation, each populated by young star clusters of slightly different ages. We measure the stellar feedback in terms of photo-ionisation rate, mechanical energy and mechanical luminosity and we study its relation to the outflows of ionised gas traced by optical emission lines. The second and third works are both part of a survey with a sample of 20 YSCs named CLUES (CLusters in the Uv as EngineS). We consistently derive with multiple methods the ages, metallicites, masses, attenuation for all clusters. We model the gas kinematics tracing both the neutral and ionised phase, and we detect an outflow in most targets. This survey reveals for the first time the properties of outflows driven by YSCs at scales between tens and a few hundreds pc. We find that the relation found in previous works between galactic outflows and the star-formation properties of host galaxies extends to smaller scales. Both the work on Haro 11 and the CLUES survey serve as benchmark studies for future investigations of high-redshift galaxies, based on observations of upcoming facilities, that will open a window on the star formation at different epochs in the history of the Universe.

The chromospheric and coronal heating problems remain one of the foremost open questions in solar and stellar physics: how are the outer layers of the Sun heated from a few thousand kelvin in the photosphere to the million-degree corona? Radiation alone is not capable of transporting the required energy to explain observations from the inner layers. Phenomena associated with the presence of magnetic fields could provide the missing energy. However, the reconstruction of the magnetic field vector from observations is complex because radiation must be modelled under non-local thermodynamical equilibrium conditions. Additionally, it is hard to achieve high spatio-temporal cadence and a sufficiently high signal-to-noise ratio simultaneously. Therefore, observational datasets are greatly affected by noise. 

The aim of this thesis was to improve the fidelity of an efficient technique to derive the magnetic field vector (a spatially-constrained weak-field approximation in Paper I), to derive the full stratification of the magnetic field in plage regions, and to study the chromospheric radiative losses and their relation to the magnetic field stratification. By studying the spatio-temporal distribution of the radiative losses in Paper II, we could discern the constribution from some heating mechanisms in the chromosphere from an observational perspective. In Paper III, we study the magnetic field strength of low-lying chromospheric loops in order to set constraints on the heating mechanisms that could be at work in these structures.

In August 2017 a gravitational wave signal from a binary neutron star merger was detected for the first time, and it was followed by several electromagnetic detections that spanned the whole electromagnetic spectrum. Those observations proved the ejection of two dynamical components: mildly-relativistic neutron rich ejecta and an ultra-relativistic jet. The first one produced a rapidly evolving thermal transient powered by the radioactive decay of the freshly synthesized r-process nuclei and subsequent recombination, named "macronova" or "kilonova".  The second component powered instead a gamma-ray burst consisting of an early (and anomalous) gamma-ray signal ≈1.7 second after the merger followed by a long lasting multi-wavelength afterglow. These two dynamical components are typically observed and modelled independently, but when a jet is produced and propagates through the surrounding environment the properties of both may change significantly because of the interaction. However the observational consequences of such interaction are still unclear.

In this thesis I present a comprehensive introduction to the properties of jets and post-merger ejecta, with the aim to show how 3D hydrodynamic simulations form a key instrument in investigating their interaction. I present the results of some of the first simulations of jets propagating within a realistic post-merger wind, and how the properties of the macronova are significantly affected. As an example, this work shows how a jet can "punch-away" a fraction of high-opacity material at early times, causing the macronova becoming brighter and bluer for on-axis observers in the first few days following the merger. Moreover such interaction represents the main contribution to the final shape of the jet, that has a crucial impact in the late time emission for off-axis observers. While looking forward for future detections, this thesis aims to show how crucial is a detailed understanding of the environment surrounding the merger and how, in case of a joint observation of a macronova and a jet, should be taken into account how the interaction might have an impact on the inferred properties of the system.

Core-Collapse Supernovae (CCSNe) are important phenomena in the scope of nucleosynthesis and, as the final stage of massive stars' life, they are key processes in the understanding of stellar evolution. They also are the birthplace of neutron stars and black holes, therefore they play a major role in the modelling and understanding of compact object mergers. While CCSNe have been observed for a long time, it is mainly through electromagnetic radiation. This channel gives us precious information about the explosion energy and nucleosynthesis, but fails to inform us about the collapse and initial explosion mechanism. While other observational channels are becoming available, through neutrino detection and gravitational waves, we are still waiting for a galactic CCSN to get an appropriate signal giving us insight on the explosion mechanism. We, therefore, have to rely on simulations for now. CCSN simulations have been performed for 60 years, improving decade after decade, and are now able to produce systematic self-consistent explosions. Several parameters impact the final outcome of our simulations, originating from different physics treatments, such as the gravity, neutrino transport and interactions, micro-physics through the equation of state, or magnetic fields. To understand the explosion mechanism behind a CCSN, we need to study the impact of each of these uncertain pieces of physics. In this thesis, I focused on the impact of the emission of heavy-lepton neutrinos and axions on the explosion, concentrated on the early proto-neutron star cooling. I explain details of the CCSN process, as well as some of the particle physics I focused on. I show how a change in heavy-lepton neutrino and axion emissions can accelerate the early proto-neutron star cooling and subsequently help the explosion.

The chromosphere is a complex and dynamic layer of the solar atmosphere, largely dominated by the local magnetic field configuration. It acts as an important interface between the photosphere below it and the hot corona above. However, studying this layer is not straightforward, as it is largely transparent in optical wavelengths. On top of that most of its observable radiation is formed in conditions far from thermodynamic equilibrium, and thus only partially sensitive to local plasma conditions. Observations of the active features found in the chromosphere such as plage, fibrils, and jets, are therefore more difficult to interpret than emission from active features in the photosphere.

This thesis focuses on plage and peacock-jets, two types of chromospheric features. Additionally, I study the quiet solar atmosphere for calibration purposes. In all three cases, I utilize high-resolution spectral and spectro-polarimetric data from the Swedish 1-m Solar Telescope (SST) in order to constrain the physical parameters of these regions and to create high-resolution reference profiles of the quiet regions.

In the first paper, the magnetic field vector of a plage region is inferred using STiC, a spectro-polarimetric inversion code, which is achieved after applying several methods to improve the signal-to-noise ratio.

In the second paper, a peacock jet near an X9.3-class flare is studied. The expanding flare ribbon moves under the jet and inhibits new material from being accelerated upwards. This coupled with back-lighting from the heavily broadened line profile of the flare ribbon that can be approximated as quasi-continuum, allowed us to estimate its density and mass by using a cloud model.  

The third paper is an observational study of the center-to-limb variations of ten spectral lines commonly used for solar diagnostics.

The Lyman alpha (Lyα) emission line from the recombination of hydrogen atoms is predicted to be the strongest spectral feature of galaxies. This emission line is regularly used to detect and characterise high redshift galaxies. However, studying this spectral line is challenging. Lyα is a resonant line, meaning it interacts strongly with neutral hydrogen. Consequently, the interpretation of Lyα observations of galaxies is very complicated. Nonetheless, this complexity provides a wealth of information. One way to learn how to extract this information is to focus on low-z star-forming galaxies. The characteristics of these galaxies, such as their gas kinematics and dust, affect Lyα escape from galaxies. This is the primary motivation for the projects included in this thesis, where we focus on Lyα observations in the low-z Universe.

We studied the morphology of Lyα emission in galaxies from the largest galaxy sample with available Lyα imaging, the Lyman Alpha Reference Sample (LARS). We investigated the Lyα light distribution and how it affects the Lyα global observables, such as Lyα luminosity, Lyα escape fraction and Lyα equivalent width. We found that the Lyα luminosity anti-correlate with the Lyα halo luminosity fraction. In other words, in galaxies that are faint in Lyα, most of the Lyα luminosity originates from their Lyα halos. Because of surface brightness dimming, this could introduce a detection bias against low luminosity Lyα emitters at high redshift, which explains why observing faint Lyα emitters at high redshift is hard. Hence, the conclusions drawn from studying data that only sample bright LAEs may have been biased.

We also investigated the origin of the extended Lyα halo emission using line-of-sight kinematic information in the LARS sample. We found that the gas kinematics in the centres of galaxies plays a vital role. We discovered that Lyα flux in the central regions of galaxies varies with the intensity of the outflow of gas in the centre. If the outflow is weak, galaxies show Lyα absorption or lower Lyα flux than expected (based on the observed Hα flux), which ultimately affects the measured Lyα halo luminosity fraction.

Finally, we studied a sample of six intensely star-forming galaxies in the low-z Universe. Due to the strong emission of Hβ + [O III] nebular emission lines and the compact nature of these galaxies, they are called Green Pea galaxies. We looked for any relationship between the Lyα properties and well-established diagnostics for determining whether a galaxy is a Lyman continuum emitter. We found that one of the primary Lyman continuum emitter diagnostics -- Lyα spectral peak separation -- correlates with the Lyα halo luminosity fraction. This result suggests that the Lyα halo luminosity fraction could be used as a diagnostic of Lyman continuum escape.

The field of extragalactic astronomy is progressing rapidly but there is still much to understand and many questions that remain open. During recent years the Lyα emission line has come to the fore as a potentially very powerful astrophysical tool and is now routinely used to find galaxies at the very highest redshifts. However, using Lyα is complicated by the fact that it is a resonant line, which means that it undergoes radiative transfer as it travels through neutral hydrogen when escaping from galaxies. This makes Lyα observations very difficult to interpret, but it also means that Lyα can provide information about the neutral hydrogen in the universe, giving it the potential to, for instance, map the progression of the Epoch of Reionization—the large scale phase transition during which the universe went from being completely neutral to being dominated by ionized gas. 

In order to make the most of Lyα observations and extend its usefulness even further we need to understand exactly what kind of galaxies emit Lyα, and, by extension, which physical processes control its escape.  Even though Lyα is relatively easy to detect at high redshift we cannot study the details of the escape process there, due to the lack of additional information about the emitting galaxies.  In order to understand Lyα in detail then, we need to observe galaxies at much lower redshift where we can get more information. This is the main driver of the projects included in this thesis which focuses on furthering our understanding of Lyα using local universe observations. 

We find that Lyα escape is a strongly multivariate issue and that using simple machine learning techniques can both help us predict Lyα and determine what the main drivers of Lyα emission are. We present two studies focusing on multivariate prediction of both imaging and spectral observations of Lyα showing that it is, in general, possible to predict the total Lyα luminosity very well but that the equivalent width and the escape fraction still remain somewhat elusive. We show that the primary variables controlling of Lyα seem to be the production rate of Lyα photons and the ionization state of the surrounding gas. 

We also study the spatial distribution of extended Lyα halo emission in an effort to determine how important spatial scattering of Lyα is and whether the properties of halos change between low and high redshift. We find that halos at low redshift are remarkably consistent with high redshift results but that the extended emission most likely is not solely due to scattering of Lyα photons produced in the central galaxy but is also produced by faint stellar components at large radii, something not demonstrated before.

Lastly, we present a database of Lyα spectral profiles and use that dataset to examine the evolution of Lyα spectra as a function of redshift. The profiles show clear trends, with blueshifted emission becoming markedly less prominent at high redshifts. Using a prescription for the expected attenuation of the intergalactic medium, we show that the evolution in the profiles is consistent with the intrinsic line shape not evolving between redshift 0 and redshift 6. We also present work analyzing correlations of Lyα with both stellar population  and nebular gas properties in extensive detail shedding further light on what properties makes a galaxy a Lyα emitter.

In cosmic history, some of the major changes such as reionization were driven by baryons (i.e. the stars and gas in galaxies), despite the fact that they contribute only few percent to the total mass-energy budget in the Universe. This thesis is about the interplay between gas and stars in highly star-forming galaxies and aims to explore the physics that dictates transformation processes that took place at various stages in cosmic history.

Using panchromatic observations ranging from the 21cm H I line in the radio regime to the extreme ultraviolet (UV), we studied ionizing radiation from massive stars (direct and through hydrogen recombination lines) as well as the atomic and molecular gas content in 15 highly star-forming local galaxies. The results are brought into cosmological context, taking a step forward towards finding answers to the following open questions in galaxy evolution: Which physical conditions enable galaxies to leak ionizing radiation (and power reionization)? What drives the high Lyman-alpha escape fractions observed in the early Universe? How did the massive stellar clumps found in high redshift galaxies have possibly formed?

One of the galaxies we studied is Tololo 1247-232. Our results show that ionizing photons (i.e. Lyman continuum) escape from the region around two central massive stellar clusters. From UV absorption lines we further conclude that bulk of the gas in the galaxy must be ionized and clumpy. Moreover, the 21cm H I data reveal a low upper limit neutral gas fraction. We thus argue that the Lyman continuum escape in Tololo 1247-232 is facilitated by the large amount of ionizing radiation that is produced in the central region and then escapes from clumpy, density bounded regions. This scenario may also explain how early galaxies at z>6 have powered cosmic reionization.

Additionally, we performed infrared and molecular gas (traced by CO) observations of galaxies drawn from the "Lyman Alpha Reference Sample'' (LARS). The galaxies were selected as analogues of high-redshift galaxies. Our main discovery is a roughly linear trend between the Lyman-alpha escape fraction and the total gas depletion time. This finding is counter-intuitive, because given the resonant scattering nature of Lyman-alpha photons, an increase in atomic gas should result in longer path lengths out of the galaxy, making photons more prone to absorption. Some other process seems to facilitate Lyman-alpha escape. We speculate that gas accretion enhances the turbulence of the cold gas and shifts the Lyman-alpha photons out of resonance. This scenario would naturally explain elevated Lyman-alpha escape fractions during the phases in cosmic history when galaxies were still accretion-dominated (at high-z) rather than defined by gas depletion.

Finally, we present high-resolution interferometric observations of a single galaxy, LARS 8. The galaxy is a proto-typical analogue of normal star-forming galaxies at z~1-2, i.e. it is massive, has a large gas fraction, is rotationally supported and its morphology is dominated by massive clumps. We show that these clumps are the result of an extremely gravitationally unstable gas disc. Large scale instabilities are found across the whole extent of the rotating disc, with only the innermost 500pc being stabilized by its bulgelike structure. Our findings prove that gravitational instabilities may play a significant role in galaxy evolution, in particular at z≃1-3, when galaxies are characterized by massive clumps.

Feedback from young, massive stars plays an essential role in the self-regulation of star formation in galaxies, and in shaping the galaxies' global properties. This phenomenon originates at small scales, surrounding the stars, but has been observed to be effective up to galactic-wide scales. The exact mechanism which allows the ionising radiation to escape the star-forming regions (HII regions), initially still embedded in their natal molecular hydrogen gas, is still unknown. Constraining the escape of ionising photons from HII regions is also relevant in order to explain the origin of the diffuse ionised gas (DIG) that is observed to contribute up to 50% to the Ha luminosity of nearby galaxies. 

I present the results of the study of stellar feedback in two nearby galaxies (NGC 7793 and M83), at spatial scales that critically connect the sources of ionisation with their immediate surroundings. We determine the fraction of DIG and study its properties and origin. We find that in NGC 7793 ionising sources located in the DIG are producing a sufficient amount of hydrogen-ionising (LyC) photons to explain the diffuse gas emission. In M83, on the other hand, the DIG is ionised by a mixed contribution of photoionisation and shocks. We investigate the link between LyC leakage from HII regions and their stellar and gas properties. We find that the age spread of the stellar population in the region does not seem to imply a higher leakage. Also the ionisation structure of the regions (e.g. the presence of "channels" that are transparent to the LyC photons) appears to be uncorrelated with escape in our sample. In M83, we also study the relative importance of different types of stellar feedback. We find that the pressure exerted by the ionised gas is always dominant over the direct radiation pressure. When the total HII region pressure is compared to the environmental pressure, we observe that regions near the galactic centre are in equilibrium with the surroundings, whereas regions in the disk are overpressured and are therefore expanding. We also find that changes in the local environmental conditions are the dominant factor in setting the ionised gas pressure, and that the pressure terms are linked to the physical properties (age and mass) of the young star clusters powering the regions. In the near future, observations from the James Webb Space Telescope will allow us to study the most embedded star-forming regions with a resolution comparable to the present one.

The detection of a blue macronova following the event GW170817 has emphasized the role that neutrinos play in merging neutron stars. In particular, neutrinos are able to drive mass ejection, the so-called neutrino-driven winds, and change the neutron richness of the matter by absorption. Since the amount of neutrons in the ejecta sets the r-process nucleosynthesis and the matter opacity, the macronova signal arising from the decay of unstable r-process nuclei in the wind carries the signature of weak interactions in mergers as it shines in the optical wavelength band. However, other mass ejection channels have been shown to potentially contribute to this optical counterpart of the macronova. Looking forward to future, new macronovae detections, it is therefore important to systematically explore the impact of neutrino-driven winds in shaping macronovae light curves. For this purpose, in this thesis we introduce a computationally efficient neutrino scheme, called Advanced Spectral Leakage (ASL), that, together with hydrodynamic simulations of binary neutron star mergers, will allow to characterize macronovae and link the physics of binary neutron star mergers with observations. 

The chromosphere is an intermediate layer of the Sun's atmosphere where radiative equilibrium breaks down. The standard chromospheric diagnostics such as the Mg II h and k and Ca II H and K spectral lines are formed under nonlocal thermodynamic equilibrium (NLTE) and they are only partially sensitive to the local conditions. Consequently, the interpretation of their profiles is not straightforward. In contrast, millimeter (mm) continuum radiation is produced by thermal free-free collisional interactions in the chromosphere under most solar conditions, and the observed brightness temperatures are better proxies for plasma temperatures. Observations at these long wavelengths have been recently enabled thanks to the Atacama Large Millimeter/submillimeter Array (ALMA), but the Sun remains largely unexplored in this spectral range.

In this thesis I explore the diagnostic potential of the mm continuum to study the solar chromosphere using inversions and radiation-magnetohydrodynamics (r-MHD) simulations. In particular, this work takes an unprecedented look at solar active-regions in the mm using some of the first solar ALMA observations.

In Paper I, we investigated whether the mm continuum helps to constrain temperatures in NLTE inversions of the MgII and CaII resonance lines using synthetic data from a 3D r-MHD simulation. In Paper II, we applied the same inversion technique to observational data in order to constrain temperature and microturbulence in plage, and we detected signatures of wave heating in coordinated observations with the IRIS satellite. In Paper III, we reported the first results of a comprehensive effort to characterize the visibility of small-scale heating events in an active-region using multiwavelength observations from the mm to the extreme-ultraviolet. We detected multiple, dynamic, transient brightenings -- we called them "millimeter bursts", and we investigated magnetic reconnection using a simulation.

This thesis shows that ALMA offers a complementary spectral diagnostic to the existing ones at visible and ultraviolet wavelengths and it underscores the importance of mm continuum observations for constraining models of the solar atmosphere.

Stars are involved in most research fields of astronomy, ranging from studies of faraway galaxies, exploding supernovae, to more nearby exoplanets and even our own Sun. As such, it is paramount that our physical interpretation of stars is accurate. By observing stars at different epochs, we can fashion evolutionary models to predict important events that occur at different phases during their life-cycle. Thus, exemplary stars where properties including mass, age and luminosity can be observed become increasingly valuable as benchmarks for calibrating said models with. Sometimes, all of these essential properties can be measured for a single system. For instance, for a binary star which circles a common centre of mass we can from its orbital motion calculate the dynamical mass of the system. If the stellar system also has a well-determined age we may use it as a benchmark for our models, and hence refer to it as an emblematic binary system.

In this thesis we are searching for exactly these emblematic binaries, both among lowmass stars and substellar brown dwarfs. We also show how to measure the different characteristics that make the systems into exemplary touchstones. We provide an overview over the different types of stellar binaries, how mass and age estimates are performed, as well as discuss the implications multiplicity has for the formation and evolution of stars and brown dwarfs. In Paper I we present the results from an orbital fit we constrained for a low-mass binary with a known age, making into a valuable and relatively rare benchmark. We also show in Paper II how long baseline astrometry can be exploited in order to place better constraints for orbital fits and dynamical masses for low-mass companions to stars by measuring the perturbation in proper motion over time. The dynamical masses are sequentially tested against evolutionary models, which at these low masses display several discrepancies compared to the observables, and are thus questioned. We explore more uncharted mass-regimes in Paper III, where we employ laser guide star assisted adaptive optics to search for multiplicity among faint substellar objects in young moving groups, detecting 3 new young brown dwarf binary systems. These new binaries will prove to be highly valuable systems for future research of brown dwarfs, and will be able to be studied further with for instance the Extremely Large Telescope or James Webb Space Telescope, which also makes them into prominent benchmarks for substellar evolutionary models. Furthermore, age estimation typically dominates the error budget for low-mass stars and brown dwarfs, requiring several different approaches for a robust assessment. In Paper IV we test and compare different techniques for age determination of 7 low-mass binary stars. These binaries have had their orbital motion monitored for a longer time, and will soon be constrained well enough that dynamical masses may be procured. As such, these low-mass binaries will extend the so far scarce number of exemplary systems where both mass, luminosity and age can be determined, to later be used to calibrate theoretical evolutionary models.

Blue compact galaxies (BCGs) are compact, metal-poor, starbursting galaxies with characteristics similar to what is expected for the young high-redshifted galaxies. BCGs are among the most active in producing a large number of massive star clusters, each containing thousands of massive stars. During their short life, massive stars are continuously injecting energy, heat and momentum into the ISM via their intense radiation, stellar winds, and later on supernova explosions. These feedback mechanisms impact directly the star's surroundings, but when this feedback originates from a concentration of massive star clusters, it can strongly affect the condition of the gas of the entire galaxy.

This thesis presents a detailed analysis of the ionized gas condition and the effect of strong feedback in Haro 11, an extreme starbursting BCG and the closest Lyman continuum (LyC) leaking galaxy. We exploit the spectro-photometric capabilities of the MUSE instrument, by slicing the galaxy spectra in a sequence of maps in velocity bins, in order to obtain a 3D information of the galaxy. Haro 11 has a rich population of massive and predominantly young star clusters, concentrated in three compact knots within its 4 x 4 kpc$^2$ centre. We find that the localised stellar feedback is strongly impacting the global kinematics and the condition of the gas up to further distances in the halo. Many kpc-scale structures such as filaments, shells and bubbles were traced in our data. Moreover, the strong feedback seems to have developed kpc-scale bubbles, outflows and galactic ionized cones with drastic consequences for the likely escape of Ly$\alpha$ and LyC photons, gas and metals out of the galaxy. The extended halo around Haro 11 is governed by photoionization processes and/or shocks from recurrent supernovae originated in the central starburst region. Due to the galaxy's extreme ISM condition, commonly used emission lines diagnostics produce, in part, large discrepancies in the ionized gas properties.

The results presented in this work highlight: a) the strong impact of stellar feedback affecting the ISM at all scales in starburst systems; b) the fact that traditional relations drawn up from averaged measurements of emission lines or from simplified models, fail in probing the condition of the gas in extreme environments. This is an appeal to revisit the standard relations by including more realistic models where several physical processes are simultaneously at work; c) the method applied here can be used to explore in detail the high sensitive, high spatial-resolution data from future facilities such as JWST/ELT.

Solitary stars wandering too close to the supermassive black hole at the centre of their galaxy may become tidally disrupted, if the tidal forces due to the black hole overcome the self-gravity holding the star together. Depending on the strength of the encounter, the star may be partially disrupted, resulting in a surviving stellar core and two tidal arms, or may be completely disrupted, resulting in a long and thin tidal stream expected to fall back and circularize into an accretion disc (the two cases are illustrated on the cover of this thesis).

While some aspects of a tidal disruption can be described analytically with reasonable accuracy, such an event is the highly non-linear outcome of the interplay between the stellar hydrodynamics and self-gravity, tidal accelerations from the black hole, radiation, potentially magnetic fields and, in extreme cases, nuclear reactions. In the vicinity of the black hole, general relativistic effects become important in determining both the fate of the star and the subsequent evolution of the debris stream.

In this thesis we present a new approach for studying the relativistic regime of tidal disruptions. It combines an exact relativistic description of the hydrodynamical evolution of a test fluid in a fixed curved spacetime with a Newtonian treatment of the fluid's self-gravity. The method, though trivial to incorporate into existing Newtonian codes, yields very accurate results at minimal additional computational expense.

Equipped with this new tool, we set out to systematically explore the parameter space of tidal disruptions, focusing on the effects of the impact parameter (describing the strength of the disruption) and of the black hole spin on the morphology and energetics of the resulting debris stream. We also study the effects of general relativity on partial disruptions, in order to determine the range of impact parameters at which partial disruptions occur for various black hole masses, and the effects of general relativity on the velocity kick imparted to the surviving core. Finally, we simulate the first part of a tidal disruption with our code and then use the resulting debris distribution as input for a grid-based, general relativistic magnetohydrodynamics code, with which we follow the formation and evolution of the resulting accretion disc.

The search for exoplanets, i.e., planets orbiting other stars than the Sun, is a relatively new research field, but has already established itself as one of the most prolific and intriguing areas of astronomy. By now we are in a situation where the focus is not only on finding companions to stars, but also on characterising their atmospheres and physical properties, which overall allows us to put our Solar System into context. In the near future, these efforts could potentially lead to the first confirmation of a life-bearing planet besides the Earth. 

The great majority of these exoplanet studies have been carried out indirectly, where the presence and characterisation of the companions are inferred solely from the observation of the host star. In the last decade, however, high-contrast direct imaging has been continuously developed to get rid of the starlight and reveal the existence of low-mass companions. Although this technique is currently limited to giant planets orbiting at large separations, it is able to directly detect the light emitted or scattered off the planet’s atmosphere at high signal to noise, which makes it the most promising planet-hunting method to characterise new worlds. Moreover, its capability to image faint objects close to the parent star allows for not only the detection of planetary-mass companions, but also low-mass stars, brown dwarfs, and circumstellar disks where planet formation takes place. This opens up a broad range of science cases where direct observations can be used to understand planet formation, atmospheric physics and stellar evolution.      

In this PhD thesis I provide an up-to-date introduction to the basis of the direct imaging technique, and explain the star and planet formation mechanisms. Three publications are attached to this introduction, each of them dealing with distinct science cases that can be  assessed with high-contrast observations. In Paper I we resolve and model the aftermath of star formation, the so-called debris disk phase analogue to the asteroid and Kuiper belts in our Solar System, around the HD 32297 star with Subaru/HiCIAO.  We reveal an edge-on disk and find the first indications of a double-ring scenario. We also present the first polarimetric study of this system, constraining the properties of the dust around the star.  In Paper II we focus on the planetary-mass regime, and conduct the first direct imaging survey searching for circumbinary planets orbiting tight binary systems (SPOTS: Search for Planets Orbiting Two Stars). We present the results of the observations of 62 targets with VLT/NaCo and VLT/SPHERE, and perform a statistical analysis on the findings, placing constraints on the population of giant planets and brown dwarfs on wide orbits. Finally, in Paper III we resolve a triple stellar system with the newly-commissioned SCExAO/CHARIS integral field spectrograph. Taking advantage of the coeval nature of the system and the different range of masses involved, we use the data to reaffirm a previously suggested isochronal age discrepancy between the low- and the intermediate-mass population of stars.

This thesis presents the modeling of radio and X-ray emissions from supernova (SN) shock fronts and hydrodynamical simulations of SN-circumstellar medium (CSM) interaction. The interaction of SN ejecta with the CSM drives a strong shock wave into the CSM. These shocks are ideal places where effective particle acceleration and magnetic field amplification can take place. The accelerated relativistic particles, in the presence of magnetic field, could emit part of their energy via synchrotron radiation in radio wavelengths. The flux of this radiation, when compared with observations, gives an estimate of the CSM density. This could either be the particle density (n_ISM) in case of the SN exploding in a constant density medium, characteristic of interstellar medium, or pre-SN mass loss rate (dM/dt) of the progenitor system for a wind medium. In Paper I we have modeled the synchrotron emission and compared that with the radio upper limits measured for the Type Ia SNe 2011fe and 2014J. Assuming equipartition of energy between electric and magnetic fields, with 10% of the thermal shock energy in each field, we obtain a very low density medium, having n_ISM <~ 0.35 cm^-3, around both the SNe. In terms of dM/dt this implies an upper limit of 10^-9 M_sun yr^-1 for a wind velocity, v_w, of 100 km s^-1. This study suggests that in SN shocks it is more likely that the amplification efficiency of magnetic fields is less than that for the electric fields. In Paper II, we carry out the hydrodynamical simulations of the interaction between SN ejecta and CSM for SN 1993J and SN 2011dh. Subsequently, the radio and X-ray emission have been calculated from the shocked gas encapsulated between the forward and reverse shocks. Considering the ejecta profile of these SNe from multi-group radiation hydrodynamics simulation (STELLA), it is found from our investigation that for a wind velocity of 10 km/s around 6500 years prior to the explosion of SN 1993J a change in mass loss rate occurred in the system. For a binary system this may imply that the change in dM/dt could be due to a change in the mass accretion efficiency of the companion star. In case of SN 2011dh the late time emission is turned up to be consistent with a wind medium with (dM/dt)/v_w = 4 × 10^-6 M_sun yr^-1/10 km s^-1. Paper III focuses on the radio emission from four young SNe Type Ia, SN 2013dy, SN 2016coj, SN 2018pv and SN 2018gv. Using the same model for radio emission as in Paper I, the upper limits on dM/dt and n_ISM are estimated. We found tenuous media around these SNe, which put tight constrain on their progenitor systems.