The next generation of balloon-borne and ground-based mm/sub-mm astronomy experiments will require operating temperatures near or below 0.1 K. When these experiments are operated remotely on platforms or at sites with limited infrastructure and maintenance sup port, a compact and reliable dilution refrigerator becomes essential. We have investigated two different dilution refrigerators in order to evaluate which system is most suitable for these applications. We have carried out a feasibility study of the simplest of the two technologies, a single-shot dilution refrigerator. A thermal model for predicting its performance has been developed, and a first prototype which achieved temperatures of about 70 mK was built. We discuss advantages and disadvantages of a single-shot system and show how minor changes to the current design can make it useful for many astronomy applications. The second dilution refrigerator is based on the principle of condensation pumping. We have built and integrated such a refrigerator with a pulse-tube cooler in order to create a completely cryogen-free system. Temperatures below 50 mK have been achieved, and temperatures below 100 mK have been maintained for more than 10 hours with several micro-Watt of cooling power. Using two 3 He sorption coolers and gas-gap heat switches we have also demonstrated how this cooler can be operated in a continuous mode. The entire system is fully automatic in operation and can be controlled and monitored remotely through a standard http protocol. We show how existing thermal models can be used to predict the cooling power and lowest achievable temperatures of the refrigerator. Experimental results are analysed and used to estimate the condensation efficiency, the performance of the heat exchangers and the 3 He circulation rate.

The results of the deep HI survey AGES are presented for two regions in the Virgo Cluster. Covering a total of 15 square degrees to a sensitivity of 0.6 mJy/beam, 73 objects are detected within the cluster, with 109 detections in the background. A new automatic extractor is described which gives greater completeness and reliability than previous methods. The majority of the cluster detections are associated with galaxies previously identified in the optical Virgo Cluster Catalogue, but nearly 30% of the cluster detections are new objects below the VCC completeness limit. Their optical and HI characteristics are described. No definite optically dark galaxies are identified, however 4 intriguing candidates are reported. All have possible optical counterparts but these are extremely faint, and their HI velocity widths appear inconsistent for such objects when compared to the more certain associations. The likelyhood that these are really dark galaxies is discussed, but further observations are required for a more definitive analysis. Cluster galaxies are found to be significantly HI deficient but it is not clear where their missing gas has gone. An automated algorithm is described to try to recover faint extended HI features, but no detections are made. The HI may be ionised by the intracluster medium or the sensitivity of the survey insufficient - the relative merits of these views are assessed. A small fraction (10%) of the early-type galaxies identified in the VCC are detected in HI. Evidence that some of these are morphologically evolving via gas loss, while others are recent additions to the cluster as yet unaffected by their new environment, is discussed. I also compare and contrast the two separate areas studied, describe the cluster's HI mass-function from AGES and other surveys, and discuss the overall influence of the cluster environment on galaxy properties.

In this work we analyze two possible observational manifestations of gravitational waves. We consider the effects of gravitational waves on ground based laser interferometric detectors, and the imprints of relic gravitational waves on the Cosmic Microwave Background (CMB) radiation. In order to study the effect of a gravitational wave on a laser interferometer it is crucial to understand the movement of free test particles. The detailed knowledge of this motion is important conceptually and practically, because the mirrors of laser interferometric detectors of gravitational waves are essentially free test masses. A gravitational wave bring about the relative motion of free test masses. In particular, analogous to movement of free charges in a field of an electromagnetic wave, a gravitational wave drives the masses in the plane of the wave-front and also, to a smaller extent, back and forth in the direction of the wave's propagation. To describe this motion, we introduce the notion of "electric" and "magnetic" components of the gravitational force. Using different methods, we demonstrate the presence and importance of the "magnetic" component of motion of free masses. We then explicitly derive the full response function of a 2-arm laser interferometer to a gravitational wave of arbitrary polarization. We give a convenient description of the response function in terms of the spin-weighted spherical harmonics. We show that the previously ignored "magnetic" component may provide a correction of up to 10%, or so, to the usual "electric" component of the response function. Another promising venue for detecting gravitational waves are the anisotropics in temperature and polarization of the CMB radiation. A strong variable gravitational field of the very early Universe inevitably generates relic gravitational waves by amplifying their zero-point quantum oscillations. These relic gravitational waves leave their imprint on the anisotropics of the CMB. We explain and summarize the properties of relic gravitational waves that are needed to derive their effects on CMB temperature and polarization anisotropics. Analyzing the radiative transfer equations, we reduce them to a single integral equation of Voltairre type and solve it analytically as well as numerically. We formulate the possible correlation functions Cfx> and derive their amplitudes, shapes and oscillatory features. We show that the TE correlation at lower ts must be negative, if it is caused by gravitational waves, and positive if it is caused by density perturbations. This difference in TE correlation may be a signature more valuable observationally than the lack or presence of the BB correlation, since the TE signal is about 100 times stronger than the expected BB signal. We discuss the detection by WMAP of the TE anti-correlation at t 30 and show that such an anti- correlation is possible only in the presence of a significant amount of relic gravitational waves (within the framework of all other common assumptions). We propose models containing considerable amounts of relic gravitational waves that are consistent with the measured TT, TE and EE correlations.

Radiatively driven winds are a key component in a number of astrophysical settings, most notably the wind of hot massive stars. In this thesis we present the mechanism by which line radiation can drive hot star winds. In order to become familiar with the dynamics of numerical wind solutions in one and two dimensions, we also make calculations for winds driven by perfect gas pressure and continuum radiation pressure. The main emphasis of this work is to approximate the form of the line-driving as given by Owocki, Castor & Rybicki (1986), with an efficient algorithm, in a pure absorption (no scattering) case. We construct an efficient model for the solution to the line-driving problem and use this to investigate the dynamics of perturbations in the wind. Once we have shown that our method can reproduce the results of previous authors we add more physics to the model. We firstly change the shape of the line-profile function and observe that although this has a stabilizing effect on the flow the response of the wind to perturbations is largely unchanged. As separate refinement we move away from the pure absorption model and include the line drag phenomenon in the radiative driving calculation. This phenomenon also stabilizes the flow but its effect is decreases further from the central star and the downstream flow feels little of its influence. We use the results from all of our models of hot star winds to create synthetic spectral diagrams. The spectra created from the perturbed models resemble structures which are observed in the spectra of hot star winds called Discrete Absorption Components, the origin of which is unclear. We show that our models can be used to explain astrophysically observed structure and better understand the nature of hot star winds.

The Spectral and Photometric Imaging Receiver (SPIRE) is one of the three instruments on-board the European Space Agency's Herschel Space Observatory, due for launch in 2008. SPIRE is a dual instrument comprising a photometer, and imaging Fourier transform spectrometer. This thesis deals with the design and operation of a software simulator for the SPIRE photometer. The simulator architecture and modelling methods are described, and the fidelity of its output verified. This simulation software is then used to optimise and characterise data from the SPIRE photometer. The optimum observing parameters are derived, in order to maximise observing efficiency, and data quality. The impact of uncorrelated 1/f noise on the extraction of sources of arbitrary scale is assessed, and quantified. This work is also extended to include the impact of uncorrected 1/f noise on observations of sources in a confused environment. These results provide important information regarding the quality of SPIRE photometer data for the planning of large survey observations. The simulator is also an active tool within the SPIRE Instrument Control Centre team, and its use in the selection of the SPIRE map making algorithm is described. This thesis also contains an analysis of observations of the Cassiopeia A supernova remnant made with the Balloon-borne Large Aperture submillimetre Telescope (BLAST), an instrument based on the SPIRE photometer design. This analysis assesses the hypothesis that supernovae might be a significant dust formation mechanism in the universe, as proposed in recent literature. Results from this study suggest that this hypothesis may be correct, but that evidence from previous observations might in fact be upper limits to the total mass of dust, rather than an absolute measurement.

Studying the Universe in a time-resolved manner, including both high-speed variability, and transient events, allows us to explore the most extreme astrophysical environments. The study of transient astronomy and high-speed variability requires specialist instrumentation to explore the fastest timescales, and follow-up rapidly fading events. For this reason I have helped to develop two facilities for time-resolved astronomy, and used them to conduct two research programmes. In this thesis I present the lab-based and on-sky commissioning of the high-speed imager ultraspec. I include a theoretical analysis of the instrument throughput, and compare this with the observed throughput, finding that ultraspec is performing to within ∼ 20% of expectations in all filters. I also present a method for flux calibrating the non-standard KG5 filter, using bolometric corrections and model spectra of stellar atmospheres. I describe the design and operation of the 0.5m robotic telescope pt5m. I have developed a collection of software programmes which enable pt5m to function as an automated transient follow-up facility, which listens for new events and acts on them immediately, collecting and reducing the data. I present the results of performance tests, and scientific results from follow-up observations of transients such as GRBs, supernovae, and an outbursting X-ray binary. I then used pt5m to conduct a search for eclipsing cataclysmic variables, and further investigated the structure of their eclipses with ultraspec. I found or confirmed 13 new eclipsing systems, of which two are definitely suitable and six are possibly suitable for detailed modelling of their eclipse structure. I also used ultraspec, and later pt5m, to search for possible optical counterparts to recently discovered fast radio bursts (FRBs). FRBs are bright, short pulses of radio emission from unknown sources. I followed-up two FRB detections in detail, finding no evidence for counterparts in one field, but 5 variable sources in the other, all of which require further study.

In this thesis I have studied hydrodynamical models of cometary HII regions and distributions of UCHII regions in simulated surveys, for comparison with CORNISH. I present the numerical method used to model the evolution of cometary HII regions produced by ZAMS stars of O and B spectral types, which are driving strong winds and are born off-centre from spherically symmetric cores with power-law (alpha = 2) density slopes. A model parameter grid was produced that spans stellar mass, age and core density. Exploring this parameter space I investigated limb-brightening, a feature commonly seen in cometary HII regions. It was found that all of the models produce this feature. The models have a cavity, bounded by a contact discontinuity separating hot shocked wind and ionised ambient gas, that is similar in size to the surrounding HII region. Due to early pressure confinement, shocks outside of the contact discontinuity were not seen, but the cavities were found to continue to grow. The cavity size in each model plateaus as the expansion of the HII region stagnates, which could be due to the Kelvin-Helmholtz instabilities at the interface mixing in cooler gas. SEDs of the models are similar to those from identical stars evolving in uniform density fields. The turn-over frequency is lower in the power-law models due to a higher proportion of low density gas covered by the HII regions. Following from this I have simulated CORNISH surveys for stars, varying the local density at the location of stellar birth. I have shown that the models used can reproduce the observed size and flux distributions in the CORNISH survey. Higher density environments generally lead to better fits to the observed size and flux distributions. A good match between the overall number of UCHII regions in the simulated surveys for a SFR = 1.5 solar masses per year can be achieved if it is considered that stars are born in a distribution of local densities rather than a single density.

No description available.

Numerical relativity has proven to be a successful and robust tool for non-perturbative studies of gravitational phenomena in the highly dynamical and/or non-linear regime. Perhaps the most prominent achievement in the field is the breakthrough success in simulating the merger of binary black hole systems. Gravitational waveforms resulting from these simulations serve as precise theoretical predictions of general relativity, which can be tested against observational data, such as those recently made by the LIGO experiment. This dissertation explores applications of numerical relativity which lie beyond the realm of astrophysics. One motivation for this comes from the AdS/CFT correspondence, which allows us to study strongly coupled quantum field theories by considering classical gravity with a negative cosmological constant. More concretely, we construct stationary asymptotically anti-de Sitter spacetimes by numerically solving the Einstein equations in a strongly elliptic form, subject to various boundary conditions corresponding to the physical setting of interest. Three applications of this technique are presented here. 1) A toroidal “black ring” in global AdS5, which provides a more complete phase diagram for AdS5 black holes. 2) A black hole on an AdS soliton background, which is dual to a localised ball of deconfined plasma surrounded by confined matter. 3) A rotating horizon extending to the AdS boundary, which allows us to the study the behaviour of the CFT in the presence of a rotating black hole. Outside of AdS/CFT, time-dependent numerical relativity in higher dimensions can also inform inquiries into the mathematical properties of general relativity as a theory of gravity. In particular, long, thin black hole horizons are known to be subject to the Gregory–Laflamme instability, and this is expected to result in an eventual violation of the weak cosmic censorship conjecture. A landmark simulation of the black string confirmed this in the Kaluza–Klein setting, however the generalisation of this setup to asymptotically flat black rings poses new challenges for numerical relativity. Even after a successful simulation, the resulting apparent horizons possess nontrivial geometries which are problematic for existing horizon finding methods. This dissertation also covers aspects of technical development in the GRChombo adaptive mesh refinement code which were necessary for the successful evolution and analysis of a black ring instability.

Modern society relies substantially on satellite technology as it is involved in vital services like telecommunication services, Earth observation, navigation, and many more. There are more than 1000 operational satellites in Earth orbit and most of these spend at least some of their time in the harsh environment of the Van Allen radiation belts. The radiation belts are usually split into two regions, the inner and the outer radiation belt. While the inner belt is considered stable, the flux of electrons in the outer belt can vary over several orders of magnitude, reaching levels that may disrupt satellite operations. It is therefore important to understand the variability of the outer belt and ultimately to predict its behaviour. In this thesis, the radiation belts are described by the BAS Radiation Belt Model (BAS-RBM) which solves a 3D diffusion equation. The BAS-RBM requires accurate diffusion coefficients that describe the interaction between electrons and plasma waves. The most important plasma waves are chorus, plasmaspheric hiss, and EMIC waves. Here, new statistical models of the diffusion coefficients for these waves are presented, which considerably improve existing models. Among others, they benefit from better global wave models due to improved satellite coverage, and revised wave normal angle and plasma density models. The results show that chorus waves are an important acceleration and loss mechanism at energies up to about 1MeV and for all pitch-angles, while plasmaspheric hiss is found to be an essential loss process in the same energy and pitch-angle range. In contrast, EMIC waves proved to be a relevant loss process for ultrarelativistic electrons, but only at lower pitch-angles. The work presented here has led to a better understanding of the variability of the outer radiation belt and has considerably improved the accuracy and reliability of the modelling and forecasting capabilities.