This thesis is concerned with a number of problems relating to mass outflow in active galactic nucleii. A number of authors have discussed radiation pressure driven winds, and some have discussed the spectral evolution in such a wind, but under restricted conditions, e.g. the radiation spectrum is thermal at high frequencies. A number of authors have discussed the Comptonization of an arbitrary input spectrum in a stationary medium. In Chapter 2, I consider the evolution of an arbitrary input spectrum by a supercritical outflow. I consider outflows that contain a number of e pairs, these flows are supercritical for modest mass outflows. I find that the input spectrum is not significantly distorted, but that the high frequency cut off moves to a lower frequency. There is a major flaw in the model: most of the radiation energy is carried by photons of energy larger than 0.5 MeV; however, the Kompaneets equation, which is central to the model, is not valid for photons of this energy. There is also an indication that pair production is an important process in the inner regions of the system, a process which has been ignored. The second problem is concerned with the broad line region. The 'standard model' requires a two-phase equilibrium between a hot intercloud medium and cool clouds. This is incompatible with the radiative heating implied by observed spectra. In Chapter 3, I introduce generallised two body heating into the intercloud wind. I find that both the dynamics and thermal equilibria of the system are compatible with the observed velocities of broad line clouds confined by an outflowing medium at 109 K. Many BLR theories require the presence of a wind, probably created in the nuclear region. Variability observations suggest that the X-ray spectrum is also created in this region, this spectrum is often described as a universal power law. It is unreasonable to assume that this ubiquitous slope can be produced by arbitrary tuning of the input parameters. In Chapter 4, I describe a model where the wind dynamics and spectral slope are related: the disk atmosphere is heated by UV radiation and by injected e pairs, causing it to form a wind. The mass loss limits the optical depth of the atmosphere and hence the evolution of the X-ray spectrum, so that the system is tuned by the dynamics to produce the canonical power law. Unfortunately, some of the approximations used in the model produce a set of equations that severely limit the range of physical parameter space that may be sensibly investigated, one or more of these approximations must be relaxed for the model to be of greater use.

The purpose of this thesis is to investigate the properties of semiconductor quantum dots by applying a quantum Monte Carlo technique. In the quantum mechanical regime such systems exhibit a range of fascinating and potentially useful phenomena. It is found that path integral Monte Carlo is generally a powerful technique for evaluating finite temperature properties of quantum many-body systems. The method is outlined for the treatment of a single-particle system. The generalisation to N particles is explained while the appropriate symmetry of the wave functions is incorporated for identical particles. Inherent numerical problems which arise for fermions are considered. An efficient new method is introduced which significantly reduces the statistical errors for large numbers of fermions. The Monte Carlo procedure is adapted to allow the calculation of magnetic field dependent quantities. The question of the existence of a molecular analogue to a Wigner crystal in a quantum dot is investigated. The phase diagram for the six-electron system studied exhibits an ordered phase in the regime of weak electrostatic confinement and low temperature. The melting temperature of this phase is found to be enhanced by the presence of a perpendicularly applied magnetic field. The dimensionality of quantum dots is considered. The two-electron ground state undergoes a transition as the crossover between three and two dimensions is effected.

The purpose of this work is to investigate the electronic structure of two-dimensional structurally disordered solids and quasicrystals by means of the Beeby and Hayes method which is based on multiple scattering theory. Using this method, the formal expressions for a two-dimensional system, such as the dynamical matrix and the spectral density, are determined in terms of a self-consistent function. The expressions are then applied to two-dimensional models to evaluate the densities of electronic states. Two principal issues are investigated in the calculations of the density of states. One is to illustrate the convergence of the method through the results of calculations. The other issue is to observe the variation of densities of electronic states with- respect to varying the disorder in the structure. For the first case, different sizes of the dynamical matrix, which correspond to the atomic structure of the system, are solved, and for each matrix the density of states is calculated. It is found that the densities of states for a disordered system do not change after the matrix size (NxN) exceeds 101x101. In the second case, some particular models are chosen, such as 3-fold, 4-fold, 5-fold and 6-fold coordinated sites, and the densities of states are calculated for each model. It is shown that the densities of states vary significantly with increasing disorder in the structure. The electronic structure of a two-dimensional quasicrystal, i.e. a Penrose lattice, has also been investigated in the tight binding limit using this approach. We have in particular studied the vertex model of the lattice and calculated the density of electronic states and integrated density of electronic states. The method used is effectively for an infinite structure and produces no distortions from edge effects or periodic continuation. The vertex model, which is based on fat and thin rhombi, has three nearest interatomic distances, the short diagonal of a thin rhombus, the edge of a rhombus, and the short diagonal of a fat rhombus. We have found the effect of interactions involving these various distances in the DOS. For example, there exists a very high central peak at zero energy if only the shortest is taken, the peak disappears if the first and second shortest distances are taken, and when all three distances are included there are no gaps and no central peak. The density of electronic states is asymmetric in all three cases.

Amorphous materials offer not only novel technological applications, but a valuable insight into the condensed state. Interest in this relatively new field has created a demand for new theories to describe the subtle electronic, optical and physical properties associated with disorder necessary if new devices are to be exploited. Metal-metalloid alloys in particular have found use in the electronics industry because of their wide-ranging electronic properties, from metallic to semiconducting, via the so-called metal-insulator transition. The underlying mechanism of this process is still widely disputed and a comprehensive theory remains to be found. Most researchers agree that to fully understand these materials a knowledge of the atomic structure must be obtained. Furthermore, recent work has shown that this information should be considered alongside a picture of the general homogeneity of any material being investigated. We present in this work the results of a structural study of four metal-metalloid systems; a-Ge1-xTix, a-Si1-xTix, a-Si1-xNix and a-Ge1-xNix, prepared in thin-film form across a wide composition range by RF Sputtering, encompassing the metal-insulator transition. These samples have been subjected to an optical study to determine the extent of the band gap and hence the composition of the MIT. An EXAFS study has been performed to determine the atomic structure, such as interatomic distances and the number and type of near-neighbours. A relatively new atomic simulation code, RMC, has been applied to the analysis of EXAFS data with the hope that information such as partial radial distribution functions may be directly obtained. SAXS measurements have been made to probe any medium-range structure and to assess the homogeneity of the samples. Optical results show that the MIT occurs at compositions in the 0-20 at.% metal range. We find no obvious structural changes to accompany this event, but observe the high coodination of metal atoms at all compositions, in three of the four systems (a-Si1-xTix, a-Si1-xNix, a-Ge1-xNix). This is characteristic of a close-packed, conducting structure. Metalloid atoms, meanwhile, appear to exist within a tetrahedral random network at low metal content, rising to close-packed at metal rich compositions. SAXS results clarify these findings by revealing the presence of phase separation, suggesting that the homogeneity of samples should not automatically be assumed. We conclude that our samples contain regions of a conducting phase embedded in a semiconducting host network and suggest that the MIT proceeds not through the traditional Anderson mechanism but by the percolation of these regions at some threshold composition.

Geomagnetic storms are well observed phenomena that enhance the plasma of the inner magnetosphere to high energies. They are defined by the characteristic trace in indices that measure the variation of the north-south component of the Earth’s magnetic field, such as the Dst or SYM-H. These indices are not purely measures of the symmetric ring current but include components of other current systems within the magnetosphere, primarily the tail and magnetopause currents. Using the methodology of Asikainen et al. [2010] the SMR index has been deconstructed to observe the evolution of the aforementioned current systems over the storm durations. Reeves et al. [2003] showed that only half of all storms caused an increase in the relativistic electron flux at geosynchronous orbit. For the remaining half the electron flux either does not change or decreases. It has been shown that the ring current decays faster for flux decrease storms than flux increase storms. Using a superposed epoch analysis, of geomagnetic indices and solar wind parameters, it has also been shown that although flux increase storms tend to have faster, less dense solar wind in the recovery phase of storms, it appears that it is the orientation of the IMF, which remains more southward in the recovery phase, that is the key parameter. This allows for the continued injection of plasma sheet particles into the inner magnetosphere. Further evidence to support this has been shown with the hydrogen and helium fluxes mirroring that of the electron flux. Finally, potential wave modes were evaluated over storm durations and potential acceleration mechanisms were noted as being more intense during flux increase storms than flux decrease storms; this is most likely due to the increase in the seed particles necessary for their generation.

In this thesis I have used infrared observations of Jupiter to investigate the flows of ions in the ionosphere and how they are coupled to the ionospheric heating in the auroral regions, determining the drivers of the heating and how they are related to the thermosphere and the magnetosphere. I investigated the H3+ line-of-sight velocity in the mid-to-low latitude region, derived from the Doppler shift of the Q(1,0-) emission line taken by IRTF-CSHELL. No evidence of flows in the region of the H Ly-α bulge predicted by a global circulation model were measured, and the H3+ ions in the mid-to-low latitude region were found to be corotating. Using observations taken by VLT-CRIRES, polar projections of the intensity and line-of-sight velocity of the H3+ ions in Jupiter’s northern auroral region were created. This revealed the ionospheric flows and how they relate to different morphological regions of the northern aurora. These flows vary from extremely sub-rotational to super-rotational, and the drivers of the flows range from the solar wind and magnetospheric interaction to a potential thermospheric driver. The same set of VLT-CRIRES observations are then used to derive the rotational temperature, column density, and total emission of the H3+ ions in the northern auroral regions. These properties were mapped onto polar projections, which revealed changes in temperature during the observations (over a short period of ~80 minutes). The changes in temperature could be caused by local time changes in particle precipitation energy, or they could be caused by the thermospheric response to a transient enhancement of solar wind dynamic pressure, as predicted by models. By comparing all of the H3+ properties, the complex interplay between heating by impact from particle precipitation and Joule heating, as well as cooling by the H3+ thermostat effect was revealed.

Over the past two decades, the potential of Raman spectroscopy as a tool for planetary exploration has been explored in detail and greatly advocated. It is ideally suited for in situ measurement as it provides rapid, non-destructive, unambiguous molecular identification, without any need for mechanical or chemical sample preparation. Developments in the miniaturisation of lasers, charge-coupled device detectors and other instrument components has, for the first time, enabled the development of Raman instruments for space missions. The first to be deployed on another planet will be the Raman Laser Spectrometer instrument onboard the ExoMars rover, a joint mission between the European Space Agency and the Roscosmos State Corporation for Space Activities, which will be launched in 2020. Two further Raman instruments, SuperCam and SHERLOC, will be included in the payload of NASA’s Mars 2020 rover. Prior to the deployment of these instruments, it is necessary to conduct analogue studies using flight-representative hardware in order to optimise instrument configuration, mode of operation, data extraction and analysis protocols. The programme of research presented in this thesis constitutes a series of such studies. The capabilities of two flight-representative, portable Raman spectrometers, one using 532 nm excitation and the other 785 nm, have been evaluated through a series of Mars analogue studies. Spectra have been acquired from a range of relevant target materials, including silica, haematite and calcium sulphate of varying levels of hydration. Caution is urged in the interpretation of spectra from portable Raman systems, since limitations introduced by their miniaturisation make band misassignment possible. As a result of this research, it is recommended that instruments are designed with a minimum spectral range from 100 to 4000 cm-1 and a spectral resolution of at least 3 cm-1, in order to avoid the misinterpretation of spectra. Several sets of analogue samples that are rich in reduced carbon have also been studied. It has been demonstrated that reduced carbon can not only be detected in concentrations as low as 0.08%, but distinct carbon populations can be differentiated by the measurement of certain spectral parameters. Furthermore, this analysis enables the qualitative comparison of the thermal maturity of different samples containing reduced carbon. These analytical techniques will be highly valuable when analysing spectra returned by planetary instruments.

Finding Neptune and Super-Earth sized planets from the ground poses a serious technical challenge. The Next Generation Transit Survey (NGTS) is a groundbased transit survey designed to produce precision photometry down to 13th magnitude to enable these kinds of discoveries. The production of high signal to noise calibration frames is discussed as part of the NGTS reduction pipeline. Next, the red noise properties of the data are looked at in detail in order to understand if any of the pipeline components are contributing correlated noise to the lightcurves. The NGTS data is search for single transit events using change point analysis, producing one strong single transit candidate. Finally, photometric follow-up observations from the South Africa Astronomical Observatory are undertaken with the aim of characterising and vetting NGTS candidates.

The magnetopause is the boundary that separates the Earth's magnetic field from the interplanetary magnetic field (IMF) and largely prevents solar wind plasma from entering the magnetosphere. It shields the Earth from space weather and understanding what affects its location is vital as we become more dependent on ground-, air- and space-based technologies. To study influences on the location of the magnetopause, an automated magnetopause crossing detection routine is developed which can determine the location of the magnetopause using a combination of plasma and magnetic field data. The technique is applied to almost two solar cycles of data (1996 - 2015) from the Geotail spacecraft, producing a database of over 8000 magnetopause crossings. The crossings are normalised for solar wind dynamic pressure and the magnetopause is modelled with the functional form of the Shue et al. [1997, 1998] empirical model. Solar cycle effects on the shape and location of the magnetopause are investigated and the model is compared to models defined by previous authors. Magnetopause location varies significantly throughout the solar cycle. We find that the model developed in this thesis characterises magnetopause location most accurately during solar minima but is less accurate during the increased solar activity observed in the declining phase of solar cycle 23. Finally, we compare the model magnetopause predictions with observations for a variety of solar wind and magnetospheric conditions. We find that the direction of the BZ component of the IMF has a stronger influence on the dayside magnetopause when the solar wind dynamic pressure is weaker. The quantity of open magnetic flux in the magnetosphere orders the dayside magnetopause location. We also examine the effect of the ring current on magnetopause location and results indicate that the dayside magnetopause is eroded and magnetotail is more inflated when the ring current is stronger.

Gamma-ray bursts (GRBs) are the most luminous transient events in the Universe. The population of observed GRBs is organised into two categories: long and short, separated by a two second divide in gamma-ray emission duration. The short type (lasting less than two seconds) have been shown to originate from the merger of two neutron stars, whereas as long bursts (lasting longer than two seconds) originate from the collapse of massive stars. There are subtypes within both classes that challenge the standard model for GRBs. For shorts, some bursts exhibit a re-brightening in their high-energy emission becoming dominant shortly after the initial emission spike known as extended emission bursts. For long bursts, some exhibit flares in their X-ray afterglows that contain a comparable amount of energy to the prompt emission. These are so-called giant X-ray flares. This thesis examines the central engine that drives these extreme types of bursts since they have the potential to discern between various proposed GRB models. A potential explanation for these events may be a highly magnetised, rapidly rotating neutron star (magnetar) fed by fallback accretion. The motivation for using this model is the late-time plateaux seen in some short GRBs that can be interpreted as a long-lived magnetar losing angular momentum along magnetic field lines. The fallback accretion component extends the global energy budget of the system and allows the rotational energy reservoir of the magnetar to be refreshed.