The secondary stars in cataclysmic variables (CVs) and close binaries are crucial to our understanding of the origin, evolution and behaviour of these systems. Magnetic activity of the secondaries has been invoked to explain many of their observed characteristics, and is thought to influence their evolution to shorter orbital periods through magnetic braking. Studies of such binaries provide unique tests of stellar dynamo theories, allowing the impact of tidal forces on the dynamo to be assessed. In this thesis, the indirect imaging technique of Roche tomography is used to map surface features and to determine binary parameters. Differential rotation (DR) of the convective envelopes of low-mass stars is a key ingredient in stellar dynamo theory. However, observational measurements of DR rates are still greatly lacking. In this thesis, the DR rate of the secondary in the CV, AE Aqr, has been measured for the first time in a mass-transferring binary. The surface was found to be not tidally locked, overturning the assumptions held by both theoreticians and observers for decades - that tidal distortion by the primary star would suppress any DR. Comparisons of the long-term magnetic activity of different stellar types is crucial to understanding the nature of the stellar dynamo, how it evolves, and what system parameters are most important in its operation. In this thesis, the long-term magnetic activity of the secondary in AE Aqr is explored by imaging starspots on seven surface maps, spanning eight years. Variations in starspot distributions due to DR and the underlying stellar dynamo mechanism are discussed; two distinct spot bands may form part of a magnetic activity cycle. Lastly, the magnetic activity of the near fully-convective close-binary, QS Vir is explored. Long- lived starspots and large prominences are found, remaining stable on timescales of over a year. Furthermore, the secondary star is found to under-fill its Roche lobe, and the system appears to be a pre-CV, yet to have initiated mass transfer via the inner Lagrangian point.

In this thesis, I present new ultraviolet/optical/infrared photometric and spectroscopic observations of pre-main sequence stars (PMS) that have formed either in metal-poor conditions, or in the vicinity of strong ionising radiation. This includes observations of 235 Classical T Tauri stars in the Lagoon Nebula; 63 Classical T Tauril Herbig Ae stars in the Carina Nebula open cluster Trumpler 14; 24 intermediate mass T Tauri stars in the low-Z Sh 2-284 star-forming region; and one Herbig 8[e] PMS candidate in the metal-poor 30 Doradus region. I measure the accretion rates of these PMS stars using the intensities of the U/Halpha band excess measured through either optical spectra of imaging. Where possible, I use archive infrared photometry in the 1.2-8 micron wavelength range to estimate the PMS disc evolutionary stage. The influence of the surrounding environment on the accretion rate evolution of pre-main sequence stars in these regions is explored using the spatial, and temporal distributions of accretion rate, mass, age and disc stage of PMS stars. In the wide-field photometric data of the Lagoon Nebula, I find that the spatial distributions of PMS stars is a continuum, ranging from dense clustering to relative isolation. Strongly accretion PMS stars are generally clumped together, in close proximity to their natal molecular cloud, whereas weaker, older accretors are relatively space apart. Ionising radiation from early-type stars appears to positively affect accretion rates on scales of 2-3 pc, but no evidence for triggered star formation is found. In addition, the accretion rates measured from Halpha imaging correlate well to those estimated from U-band photometry. In wide-field photometric data of Trumpler 14, I discover a population of PMS candidates nearly 25 Myrs old. I argue that these PMS candidates are a foreground population, approximately 5 Myr old that belong to the Carina Nebula cluster Trumpler 16. Using Halpha spectra of 24 intermediate mass T Tauri stars in Sh 2-284 (Z-0.004), I demonstrate that there is little evidence for a systematic change in accretion rates with metallicity, contrary to previous literature results at Z-0.006-0.002 in the Magellanic Clouds. I suggest that previous studies are likely affected by detection limits and biases. I also present ultraviolet/optical spectra of the Herbig 8[e] PMS candidate VFTS 822 located in the 30 Doradus region of the Large Magellanic Cloud. I discuss the impact of the discovery of VFTS 822 for star formation studies in the Magellanic Clouds, external Galaxies .

No description available.

No description available.

The properties which define a molecular cloud core — the evolutionary phase bridging between a molecular cloud and a protostar — are extensive. These properties include the initial density profile; velocity field; and magnetic field strength and geometry (and the alignment of this with other fields). These properties have a major effect on the nature of the protostar or protostars ultimately produced when the core collapses. We present a series of calculations using smoothed particle radiation magnetohydrodynamics of the collapse of a molecular cloud core to the first hydrostatic core phase. Before this, we describe and analyse our numerical method, including exploring historical difficulties and the limits of the stability. We explore the role of the geometry of the magnetic field, and showing that the nature of any outflows produced from a first hydrostatic core is closely related to the inclination angle of the field. We continue this analysis into the role of the field strength and geometry. We find that highly misaligned fields do not form bipolar outflows and discuss the cause of this, and additionally find that the angular momentum transport in weak field calculations is insufficient to prevent fragmentation and the formation of binary systems. When an outflow is formed, even in the most idealised initial conditions, the velocity is never |v_z| > 10 km · s^{−1}. We consider next the role of turbulent and rotational kinetic energy, and find that transonic turbulence can prevent the formation of an outflow unless a critical ratio of rotation to turbulent energy is exceeded. Even so, we observe that outflows produced in non–laminar calculations are slower (|v_z| ∼ 1 km · s^{−1}) than those from laminar configurations. We then show that a Bonnor–Ebert density profile can produce a stable binary system with a helical outflow without the prolific fragmentation seen in fast rotating uniform density distributions.

This thesis presents a study of the physical and chemical properties of star forming regions, both in the Milky Way and in the distant Universe, building on the existing astrochem- ical models developed by the group at UCL. Observations of the nearby star-forming region, L134A, which were carried out with the James Clark Maxwell Telescope (JCMT) in Hawai'i are compared to the predictions of a model of star formation from gas rich in atomic (rather than molecular) hydrogen. A similar model is used to investigate the effect of non-equilibrium chemistry on the derivation of the cosmic-ray ionization rate, an important parameter in controlling both the chemistry and the physics of star forming clumps. A collapse faster than free-fall is proposed as an explanation for differences be tween the distribution of CS and N2H+ in such regions. Moving beyond the Milky Way, JCMT observations of sulphur-bearing species in the nearby starburst galaxy, M82, are presented and compared with existing molecular observations of similar systems. M82 is a local anlogue for star forming systems in the early Universe, many of which have star formation rates several thousand times that of the Milky Way. A model which treats the molecular gas in such systems as an assembly of 'hot cores' (protostellar cores which have a distinctive chemical signature) has been developed, and is used to predict the abundance of many species. An application of this model is used to explain the observed deviation in the early Universe from the otherwise tight relation between infrared and HCN luminosity via relatively recent star formation from near-primordial gas. Many of the stars formed in the early Universe must now be in massive elliptical systems, and work on the structure of these systems is presented. Data from the Sloan Digital Sky Survey is analysed to show that such galaxies have cores dominated by baryons rather than dark matter, and the dark matter profile is constrained by adiabatic contraction.

This thesis is concerned with obtaining black brane solutions to Fayet-Iliopoulos gauged N=2 supergravity that obey the strong, Planckian version of the third law of black hole mechanics. We first construct a new two-parameter family of black brane solutions to gauged N=2 supergravity in four dimensions using time-like dimensional reduction as a solution generating technique. The solutions we obtain have zero entropy density in the zero temperature limit and hence satisfy the strong, Planckian version of the third law of black hole mechanics. Therefore, these "Nernst branes" could be holographically dual to (2+1)-dimensional systems in condensed matter physics where such behaviour is considered generic. Whilst the spacetime interpolates between different hyperscaling-violating Lifshitz geometries and thus correctly captures the scaling behaviour of such condensed matter systems, we observe singular behaviour in both the near horizon and asymptotic regimes. For the "very special" class of four-dimensional models under consideration, it is natural to try to resolve such behaviour by lifting the solution to five dimensions. Doing so, we find a family of boosted AdS-Schwarzschild black branes that continue to satisfy the third law. With AdS asymptotics comes access to techniques that allow for a more complete thermodynamic analysis. At the same time, this geometry fits naturally into gauge-gravity duality and resolves all asymptotic singular behaviour, suggesting the four-dimensional solution was unable to access the full degrees of freedom of the system. Interestingly however, the near horizon singularity persists which may suggest that a unique ground state is always accompanied by singular behaviour of the horizon.

This thesis studies the propagation of fundamental fields on black hole and black hole analogue spacetimes. We consider the scalar, electromagnetic, gravitational and Dirac fields, and their governing equations, in various scenarios. We initially consider an analogue gravity model, the draining bathtub vortex, that shares features with the Kerr black hole, such as a horizon and an ergoregion. We solve the wave equation approximately, via the eikonal approximation, and numerically, using the method of lines, and show that a point-like disturbance maps out the lightcone of the effective spacetime. The Schwarzschild and Kerr black hole spacetimes are then introduced and we discuss their key features. We solve the scalar wave equation for the black hole spacetimes and compare with the analogue spacetime. We then introduce the self force, the back reaction of a body's own field on its motion. The scalar self force on Kerr spacetime is calculated using the worldline integration method. This involves solving the scalar wave equation to find the Green function via the Kirchhoff representation and integrating over the entire past history of the worldline. The electromagnetic (EM) self force is calculated via the mode sum method. We use both analytical and numerical techniques to calculate EM self force for a particle held static outside of a Schwarzschild black hole. The gauge freedom of the gravitational self force is also discussed. We construct for eccentric orbits on Schwarzschild the spin precession invariant, a gauge invariant quantity. We compare the spin precession invariant calculated using numerical self force data with a post-Newtonian calculation. Finally we investigate the Dirac (fermionic) field in searching for the existence of bound states. We find that the solutions which satisfy the boundary conditions, obey a three-term recurrence-relation. Using continued-fraction methods we find a spectrum of quasi-bound states of the Dirac field exists.

I present a magnitude and proper motion limited catalogue of ~ 10,000 white dwarf candidates, obtained from the SuperCOSMOS Sky Survey by means of reduced proper motion selection. This catalogue extends to R ~ 19.75 and μ ~ 0".05 yr⁻¹, and covers more than three quarters of the sky. Spectroscopic follow-up observations have been obtained for a subsample of objects, and are used to assess the reliability of the selection procedure and contamination of the final catalogue. Photometric parallaxes provide distance estimates accurate to ~ 50%. This catalogue is used to measure the luminosity function (LF) for white dwarfs in the Solar neighbourhood over the range 4 ≤ Mbol ≤ 18. A new technique is devised to separate the LFs for disk and spheroid stars, which allows all stars to contribute to the LF even at tangential velocities where the populations overlap. The disk LF shows a sharp decline at Mbol = 15.75, in agreement with other studies, and extends over a magnitude fainter than previously determined. The spheroid LF has a sharp peak at Mbol = 15.75, then levels out and shows no sign of a drop off at faint magnitudes, as expected for a considerably older population. By simulating white dwarf LFs for stellar populations with a range of star formation histories, I measure the age of the local disk and spheroid. The disk age of 9.13⁺⁰ ⁶⁶₋₀.₆₇ Gyr agrees with that measured in similar studies, and with the basic picture of galaxy formation in a cosmological context. No reliable solution is found for the spheroid age. By integrating the LFs, I investigate the local mass density of spheroid white dwarfs, with particular reference to their contribution to the baryonic dark matter content of the Galaxy. I also place an upper limit on their contribution to the possible MACHO population and the microlensing events observed towards the LMC.

We report our investigation into the stellar populations of early-type galaxies (ETGs) in the local universe. Massive ETGs are thought to form in a manner very different to our own Milky Way, with their cores formed via a swiftly-quenched rapid starburst >10 Gyr ago and their outskirts assembled later through cumulative mergers with predominantly less-massive systems. This formation history is encoded in the properties of their stellar populations. Various evidence indicates that in the cores of massive ETGs, the stellar initial mass function differs from that of the Milky Way, with an increased fraction of dwarf stars formed. This effect, along with signatures in the chemical abundance properties of these stars, has been linked to ETG assembly histories. We use a combination of infrared and optical spatially-resolved spectroscopy of a sample of eight nearby massive ETGs (σ_average ~ 230 km/s) to measure empirical gradients in the strengths of spectroscopic absorption features, some of which are not hitherto well-explored. Using state-of-the-art stellar population models and robust methods for statistical inference, we link these to the underlying properties of the stars in these galaxies. We measure strong gradients in spectroscopic features linked to Na I (at 0.82μm, 1.14μm, and 2.21μm) as well as the Ca II 0.86μm triplet. We measure weak/no gradient in several features that trace the IMF, such as the FeH 0.99μm Wing-Ford band. We use models to interpret these measurements, inferring e.g. an [Fe/H] gradient of -0.16±0.05 per dex in fractional radius and an average [Na/Fe] gradient of -0.35±0.09. We find a large but radially-constant enhancement to [Mg/Fe] of ~0.4 and a much lower [Ca/Fe] enhancement of ~0.1. Finally, we find no significant IMF gradient either on average or even in individual galaxies where such gradients have been previously reported.