Particle dynamics in the solar corona are of interest since the behaviour of the coronal plasma is important for the understanding of how the solar corona is heated to such high temperatures compared to the photosphere (≈ 1 million Kelvin, compared to a photospheric temperature of ≈ 6 thousand Kelvin ). This thesis deals with particle behaviour in various forms of magnetic and electric fields. The method via which particles are accelerated at reconnection regions is of particular interest as particle acceleration at a magnetic reconnection region is the basis for many solar flare models. Solar flares are releases of energy in the solar corona. The amounts of energy released range from the very small amounts released by nanoflares, that cannot be observed individually, to large events such as X-class flares and coronal mass ejections. Chapter one provides background information about the structure of the Sun and about various forms of solar activity, including solar flares, sunspots, and the generation of the solar magnetic field. Chapter 2 explores various theories of magnetic reconnection. Magnetic reconnection re- gions are usually characterised as containing a central ’null’, a region where the magnetic field is zero, and particles can be freely accelerated in the presence of an electric field, as they decouple from the magnetic field and move non-adiabatically. Chapter 2 gives examples of how such reconnection regions could be formed. Chapter 3 deals with the construction of a ’noisy’ reconnection region. For the purposes of this work, ’noisy’ fields were created by perturbing the magnetic and electric fields with a superposition of eigenmode oscillations. The method for the calculation of such eigenmodes, and the creation of the electric and magnetic fields is detailed here. Chapter 4 details the consequences for particle behaviour in a noisy reconnection region. The behaviour of electrons and protons in such fields was studied. It was found that adding perturbations to the magnetic field caused many smaller nulls to form, which increased the size of the non-adiabatic region. This increased non-adiabatic region led to greater energisa- tion of particles. The X-ray spectra that could be produced by the accelerated electrons were 4 5 also calculated. In this chapter I also investigate the consequences of altering the distribution of the spectrum of modes, and altering the value of the inertial resistivity. In chapter 5, the effects of collisional scattering on particles was also investigated. Colli- sional scattering was introduced by integrating particle trajectories using a stochastic Runge- Kutta method (which is a form of numerical integration). It was found that adding collisional scattering at a reconnection region causes a significant change in particle dynamics in suffi- ciently small electric fields. Particles which undergo collisional scattering in the presence of a small electric field gain more energy than those which do not undergo collisional scatter- ing. This effect decreases as the size of the electric field is increased. The correct relativistic expressions for particle collisions were derived. It was found that collisions have a negligible effect on relativistic particles. Collisional scattering was also used to simulate the drift of particles across magnetic fields. It was found that adding more scattering caused the trajectories of the particles to change from normal gyromotion around the magnetic field, and that particles instead travelled across the magnetic field. I also developed a diffusion coefficient to allow the calculation of a particle’s drift across a magnetic field using only 1D equations. Chapter 6 discusses the findings made in this thesis, and explores how these findings could be built upon in the near future.

Combined X-ray and radio observations of the Sun provide powerful diagnostics of particle acceleration and transport effects during solar flares. In this thesis we present observations of two solar flares. In the first event we report what we believe to be the first observation of hard X-ray emission formed in a coronal, flare-related jet. Occurring on the 22nd of August 2002, the event was observed by the Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) and the Nobeyama Radioheliograph (NoRH) and Polarimeters (NoRP). During the impulsive phase RHESSI observed significant hard X-ray emission to energies as high as 30-50 keV in the jet. RHESSI spectroscopy shows a powerlaw spectrum with a spectral index of ~4 and NoRH images reveal radio emission at 17 GHz and 34 GHz co-spatial with the hard X-ray emission, thus supporting the evidence for nonthermal emission in the jet. The second event occurred on the 24th of August 2002 and was also observed by RHESSI and NoRH. The size and orientation of the flare, which occurred on the west limb of the Sun, make it particularly interesting to study. At both NoRH frequencies emission is observed at all points along a flare loop such that the looptop and footpoint emission are clearly separated. We present observations of the flare decay phase to investigate the long term evolution of the event. In particular we follow the evolution of relevant plasma parameters which are used as an input to a 3D gyrosynchrotron model in an attempt to reproduce the observed emission at radio wavelengths.

Interaction of magnetohydrodynamic (MHD) waves with various structures in a magnetised plasma was considered theoretically in the context of the interpretation of recently observed phenomena in the corona of the Sun. The main emphasis was put on the development of analytical models, utilising various asymptotic techniques based upon the presence of a small parameter. In the consideration of waves guided by field-aligned plasma non-uniformity, such as coronal jets and plumes, the small parameter was the ratio of the diameter of the guiding non-uniformity to the wavelength, the approach known as the \thin ux-tube approximation". In the consideration of nonlinear effects, the wave amplitude was taken to be finite, but small, and hence could be treated as a small parameter too. In the thesis, we addressed several specific timely problems of modern solar physics: the interpretation of recently discovered transverse waves on soft X-ray coronal jets in terms of a kink fast magnetoacoustic wave; modelling of enigmatic torsional waves (also known as twisting waves or waves of the electric current) guided by cylindrical coronal structures, such as loops, plumes, filaments and jets, accounting for the effects of the magnetic twisting and rotation of the equilibrium plasma configuration; weakly nonlinear effects appearing during the propagation of the torsional waves along coronal magnetic waveguides, concentrating on the nonlinearly induced compressible perturbations; and nonlinear steepening of fast magnetoacoustic waves in the vicinity of a magnetic null-point, in the context of the possible triggering of magnetic reconnection by the deposition of current-driven anomalous resistivity. In the first Chapter, we give an overview of the solar atmosphere and dynamical processes observed there such as MHD waves and plasma ows. Also, the set of MHD equations is introduced, and the main modes of a basic coronal plasma structure, a magnetic cylinder, are considered by the method of dispersion relation. In Chapter 2, we considered the long-wavelength limit in the magnetic cylinder dispersion relations, and derived explicit expressions, which link the phase and group speeds for linear kink magnetoacoustic waves guided by hot plasma jets surrounded by a static plasma. With the use of the derived expressions, we showed that transverse waves recently discovered by Hinode/XRT on coronal jets are the kink waves. In the observationally determined range of parameters, the waves are not found to be subject to either the Kelvin-Helmholtz instability or negative energy wave instabilities, and hence they are likely to be excited at the source of the jet. We also carried out forward modelling of the observables, and demonstrated its consistency with XRT observations. In Chapter 3 we considered long wave-length axisymmetric magnetohydrodynamic waves, and derived asymptotic dispersion relations linking phase speeds with the plasma parameters using the second order thin ux tube approximation. We showed that when uniform twist and rotation are both present, the phase speed of torsional waves depends upon the direction of the wave propagation. In addition, the twist and rotation causes compressibility of the torsional waves. The phase relations show that in a torsional wave the density and azimuthal magnetic field perturbations are in phase with the axial magnetic field perturbations and anti-phase with tube crosssection perturbations. In a zero-β non-rotating plasma cylinder confined by the equilibrium twist, the density perturbation is found to be about 66 percent of the amplitude of the twist perturbation in torsional waves. In Chapter 4, we considered the nonlinear phenomena accompanying long-wavelength torsional waves in an untwisted and non-rotating magnetic ux tube. We showed that propagating torsional waves induce compressible perturbations by nonlinear forces, these compressible perturbations oscillate with double the frequency of the torsional waves. In contrast with plane shear Alfvén waves, the amplitude of compressible perturbations is independent of the plasma-β. But, as in the shear Alfvén wave, the amplitude of compressible perturbations are proportional to the torsional wave amplitude square. It was also shown that standing torsional waves induce compressible perturbations of two kinds, those which grow with the characteristic time inversely proportional to the sound speed, and those which oscillate at double the frequency of the inducing torsional wave. The growing density perturbation saturates at the level, inversely proportional to the sound speed. In Chapter 5, we studied the generation of fast magnetoacoustic shocks in the vicinity of a magnetic null-point. In the weakly nonlinear limit, we derived a simple wave evolutionary equation, which provided us with the qualitative information about the nonlinear evolution of the fast wave-pulse: formation of the shock and deformation of the initial shape of the perturbation depending upon the polar angle. We compared our analytical solutions with numerical solutions and found that the speed of the fast magnetoacoustic pulse depends on the initial amplitude of the pulse. In our parametric studies we showed that although the initial amplitude of the magnetoacoustic pulse is responsible for the time the pulse overturns, the initial width of the pulse should not be ignored. We showed that narrower and higher amplitude pulses overturn at larger distance from the null-point. In the context of the sympathetic flaring a stronger initial pulse does not guarantee a stronger effect.

Close binaries containing a compact object make up a wide variety of objects. The evolution of all close binaries depends crucially on the rate at which angular momentum is extracted from the binary orbit. The two most important sources of angular momentum loss are the common envelope phase and magnetic braking. Both processes have been known for long but are still poorly understood, and significant progress will only be achieved if they can be calibrated using innovative observational input. Post-common-envelope binaries are probably among the best-suited class of objects to improve our understanding of close binary evolution, because (1) they are both numerous and well-understood in terms of their stellar components, and (2) they are not contaminated by the presence of an accretion disc. The Sloan Digital Sky Survey provides the possibility of dramatically improving the observational size of known post-common-envelope binaries, with already more than 1500 white dwarf-main sequence binaries having been identified. The major task is now to identify those systems that have undergone a common envelope and to measure their binary parameters. This new, large sample of well-studied post-common-envelope binaries will then provide the much-needed constraints for further development of binary evolution theory. Through my PhD I dedicated all my efforts towards identifying post-common-envelope binaries, obtaining orbital periods of these new systems, and determining their stellar parameters. For this purpose, I adopted the following strategies: (1) About 10% of the white dwarf-main sequence binaries in the Sloan Digital Sky Survey have more than one survey spectrum available. By measuring radial velocities from the Na I ll 8183.27,8194.81 absorption doublet and/or the Ha emission line in the different spectra from each object, I was able to identify radial velocity variable stars, which are prime candidates for being post-common-envelope binaries. This method resulted in the identification of 18 new post-common-envelope binaries among 130 white dwarf-main sequence binaries with multiple Sloan spectra. In addition, using a spectral decomposition/ model atmosphere analysis I determined the stellar parameters such as mass, radius, and temperature for the white dwarfs, and spectral types of the main sequence stars in these 130 white dwarf-main sequence binaries, along with the distances to the systems. I discussed also an apparent systematic issue with the spectral type-radius relation of the companion stars in those white dwarf-main sequence binaries. (2) Follow-up observations by our team have lead to the identification of 89 postcommon-envelope binaries from Sloan, which triples the number previously known. Intense radial velocity studies have lead to the determination of orbital periods for 42 of these systems, seven of them discussed in detail in this thesis. (3) I have developed a procedure based on c2 template fitting and signal-to-noise ratio constraints to identify white dwarf-main sequence binary candidates in the Sloan Digital Sky Survey Data Release 6 spectroscopic data base. This catalogue contains 1591 white dwarf-main sequence binaries identified in this way. Using a spectral decomposition/model atmosphere analysis, I have derived white dwarf temperatures, masses, companion star spectral types, and distances, and discussed the distributions of these parameters. In addition, I have analysed the selection effects of white dwarf-main sequence binaries in Sloan. This sample is an excellent data base for future follow-up observational studies of white dwarf-main sequence binaries.

Long γ-ray bursts (GRBs) are uniquely powerful explosions at cosmological distances. As they mark the deaths of massive stars, they act as beacons of star formation and point out faint galaxies in the distant universe. Thus, they allow us to probe the conditions and the evolution of galaxy formation and metal enrichment throughout the universe. However promising as these prospects are, they need rely on a firm foundation based on the understanding of how the formation of gamma-ray bursts depend on the galactic environments. That is, do GRBs trace all star formation, or are they biased to metal poor and low mass hosts? Here I will explore the host galaxies of these events in order to understand how they relate to the properties of their galaxy populations. Like gamma ray bursts, core-collapse supernovae (CCSN) are the “grand-finale” of the life of massive stars. Providing a census of all massive star formation, they are an ideal control group to compare GRB hosts with at low redshifts. I employ this method to compare restframe properties of the host populations, concluding that GRB hosts are in comparison to CCSN hosts drawn from a compact, low mass and irregular galaxy population. This suggests an inherent bias amongst GRB progenitors, and that they prefer low metallicity environments. Furthermore, the GRB locations on their hosts have higher surface luminosities than for CCSNe, suggesting that GRB progenitors are more massive and short lived than those of CCSNe. Although the low redshift sample only appear to trace star formation in sub-luminous irregular galaxies, I will also show that this need not be strictly true everywhere: I will study the luminosity-metallicity and mass-metallicity relations of GRB hosts up to z ~ 6, and show that at high redshift where the universal metallicity is lower than in the present day universe, GRB hosts appear to follow the metallicity relations of that era. While GRBs might be biased tracers of star formation in the local universe, this suggests that above z ~> 3, the universal metal enrichment is low enough that GRBs trace all star formation. Even at intermediate redshift, I will show that not all GRB hosts are blue and sub-luminous. The host of the dark burst GRB080207 is extremely red, massive and with high inferred dust and gas content. I will discuss how the difficulties of obtaining accurate positions for highly extinguished bursts may have adversely affected host samples and follow-up strategies, and show that the increasing number of well studied dark burst suggest that many of them are massive and dust rich. This implies that, even at lower redshifts, a complete census of all GRBs may trace a higher fraction of star formation then inferred by only optically bright bursts.

Recent years have seen an explosion in the number of eclipsing binaries containing white dwarfs. In the last few years the number of systems has increased from 7 to over 40, thanks mainly to large surveys such as the Sloan Digital Sky Survey and the Catalina Sky Survey. Many of these systems are survivors of the common envelope phase during which the two stars orbit within a single envelope which is rapidly thrown off through loss of energy and angular momentum. Detailed analysis of these systems can yield extremely precise physical parameters for both the white dwarf primary and its companion star. Stellar masses and radii are some of the most fundamental parameters in astronomy and can be used to test models of stellar structure and evolution. They can also be used to constrain the evolutionary history of the binary system offering us the chance to better understand the common envelope phase itself. In this thesis I present high-precision studies of several eclipsing post common envelope binaries. I use a combination of high-speed photometry and high-resolution spectroscopy to measure the masses and radii of both stars in each system. I compare these results to evolutionary models and theoretical mass-radius relations and find that, on the whole, the measured masses and radii agree well with models. However, the main-sequence companion stars are generally oversized compared to evolutionary models, although this deviation is much less severe at very low masses (< ∼ 0.1M⊙). I also find that the measured masses and radii of carbon-oxygen core white dwarfs are in excellent agreement with theoretical models. Conversely, the first ever precision mass-radius measurement of a low-mass helium core white dwarf appears undersized compared to models. Large scale surveys have also begun to identify double white dwarf eclipsing binaries. In this thesis I present a study of one of these systems and show the potential, as a double-lined spectroscopic binary, of measuring precise parameters for both stars in the future. Finally, I show that the mid-eclipse times of eclipsing binaries containing white dwarfs can be measured to a high enough precision that we can monitor them for evidence of period changes. I find that many systems show complex variations in their eclipse times and in many cases the only mechanism able to produce these changes is one or more sub-stellar objects in orbit around the binary. However, I show that care must be taken when attempting to detect planets in binary systems using eclipse timings.

Over the years, substantial effort has gone into calculating models of the intrinsic population of galactic Cataclysmic Variables (CVs). However, comparison with the observed properties of the known CV sample has consistently failed to match the predictions put forward by the ‘disrupted magnetic braking’ model of CV evolution. Some of the discrepancies have been blamed on the heterogeneous set of known CVs, which are not well-suited for a quantitative test of the population models. The Sloan Digital Sky Survey (SDSS) has dramatically improved the observational side of CV population studies. Sampling a large volume in ugriz colour space and extending deeper than any previous large-scale survey, SDSS provides the most homogeneous and complete sample of CVs to date. At the time of writing, the sample of SDSS CVs contains 213 systems, of which 177 are new discoveries. Establishing the detailed properties of these CV is a major task. The work presented in this thesis is part of a larger effort to derive the orbital period distribution of SDSS CV sample. In this thesis, I present my contribution and results of time-series photometric and spectroscopic observations of a total of 29 SDSS CVs. I provide a discussion of their properties and determine orbital periods for 20 systems. I also provide a quantitative analysis of the new sample of CVs from SDSS and compare their intrinsic qualities with the previously known population of CVs. The results show that the period distribution of the SDSS CVs differs from that of the previously known CVs at a 3-σ level. A substantially larger fraction of below-the-gap to above the-gap systems is observed, accompanied by a prominent accumulation of CVs at the orbital period minimum. This result is important, as the accumulation of systems at short orbital periods has been long theorised but never proved observationally. More specifically, the origin of the 80–86min period spike is entirely due to the new CVs identified in SDSS. The systems in the period spike also differ in spectral morphology and accretion activity from the longer-period CVs. Indeed ∼ 20% of the CVs identified by SDSS have white-dwarf dominated spectra indicating low mass transfer rates and late spectral type donors, the majority of which have been found close to the period minimum. A homogeneously selected sample of CVs, with well defined parameters has been long overdue and the results from SDSS are very promising. With this improved observational evidence, theoretical models can be modified to re-address the discrepancies in the current theories that have long plagued our understanding of CV evolution.

This thesis describes a search for transient phenomena using data from the Wide Angle Search for Planets (WASP) observatories. With images taken of the same region of sky every 15 minutes and a field-of-view of 974 degrees squared, the WASP survey is ideal for identifying both short and long duration transients to a limiting magnitude of 16 in V. A high cadence wide-field survey of this type has not been carried out before so provides a different method of sampling variable populations and has the potential to identify previously unknown phenomena. An algorithm to carry out a full untargeted search of the database was developed, tested and run on data from 2004, 2006, 2007 and 2008. False positives were rejected based on their proximity to similar magnitude stars, the tightness of the clustering of detections in sky coordinates, the focus of the images and the number of corroborating detections. Overall, 44 cataclysmic variable candidates, 144 extreme flare candidates, 63 variable star candidates and 57 transients of unknown type were identified. The cataclysmic variable candidates showed similarities to the currently known population although some selection effects were uncovered. Of particular note were two new WZ Sge systems, an eclipsing system with a period of 540 minutes and an object with an M-star spectrum and broad Balmer lines but no visible white dwarf. The flares are some of the largest ever observed on M-type stars. The largest of these released ~1035erg in the V-band alone. A candidate flare was also identified on a G-type star releasing 1.6×1037erg in the V-band. Alternatively, it may be unrelated to the G-star and could be a gamma ray burst or similar transient without a persistent counterpart. A known supernova and two known novae were also identified.

Sunspots are the most conspicuous feature seen on the solar photosphere and are manifestations of the solar magnetic field. Their study, then, may provide us with a greater understanding of the dynamo mechanism thought to be responsible for the generation of this field. In this thesis, the oscillations of sunspots are studied by making use of observational data from two instruments: the Nobeyama Radioheliograph (NoRH) and the Solar Optical Telescope (SOT) on board the Hinode spacecraft. First, a study of long period oscillations was undertaken in which two long period peaks (P > 10 min) were identified in the power spectra of time series generated from sets of images of 3 sunspots observed with NoRH. In addition, by using the techniques of period, power, correlation and lag mapping, it was found that the power in each of these peaks was concentrated over the umbral regions and that there were two regions of approximately equal size oscillating in anti-phase with each other. It was suggested that these properties could be signatures of a "shallow" sunspot. A follow-up study was then performed, in which the lifetimes of the long period oscillations were investigated over a period of 9 days. These oscillations were seen to dominate the spectra during this interval and the periods and amplitudes were stable during that time. A simple model of a damped, driven simple harmonic oscillator (in which the driving term was nonlinear) was proposed to explain the generation and support of the oscillations. Finally, a study of the spatial properties of the 3 minute oscillations was performed by applying the mapping techniques mentioned above to Hinode/SOT data. The distributions of power and lag of maximum correlation coefficient were found to be non-uniform over the sunspots under study and this may be indicative of inhomogeneities of the physical quantities in the structures.

Cataclysmic variables are close binary systems where mass is accreted onto a white dwarf through an accretion disc. Approximately half the gravitational energy in the disc is released in a boundary layer as X-rays. X-rays originate from matter under the extreme and energetic conditions and provide a unique view of high energy processes. X-ray luminosities are sensitive to the accretion rate through the disc and to the conditions in the inner accretion zone. Accretion discs are wide spread throughout astronomy. The extraction of potential energy from accreted material is known to be the principal source of power in several types of system; quasars, galactic nuclei, binary X-ray sources, cataclysmic variables (CVs) and proto-planetary discs. CVs provide probably the best opportunity to study the accretion process in isolation. Previous X-ray analysis of dwarf novae relied upon relatively short snap shot observations, which are unable to provide a full picture of the outburst cycle evolution. Multiple outbursts with far greater temporal accuracy and coverage than has ever been observed before are presented in this thesis. Pointed observations using the proportional counter array on the Rossi X-ray Timing Explorer of SS Cygni, U Gem and SU UMa are analysed. The behaviour in the optical band is similar for the three systems in this thesis, however, a large distinction is seen in the X-ray band. The hard X-ray outburst flux in SS Cygni and SU UMa are quenched below the quiescent flux, while in U Gem it is unusually faint in quiescence brightening in outburst. The hard quiescent X-ray spectrum is replaced by an intense soft X-ray component in outburst for all dwarf novae. Analysis of U Gem suggests that X-rays originate from the inner accretion disc with a scale height not much greater than the disc thickness. The start of the hard X-ray outburst is delayed behind the optical rise, this delay is roughly consistent for the three systems presented. This indicates that the origin of the heating wave in the accretion disc and the time it takes to propagate to the boundary layer are similar for these systems. The hard X-ray recovery also has a range of times, with the peak occurring as the optical flux reaches quiescence suggesting the cooling front reaches the boundary layer at the same time in relation to the end of the optical outburst. The spectra for all three systems presented in this thesis are well described by a thermal plasma model with sub-solar abundances and are consistent with higher reflection during the hard X-ray suppression. This indicates that the disc is likely to be truncated in quiescence. SS Cygni has a wide range of quiescent accretion rates. However, the X-ray flux in SS Cygni and U Gem always increases when the boundary layer transitions from both optically thick to thin, into outburst, and optically thin to thick, out of outburst. This is surprising, the flux is expected to decrease when the critical accretion rate is reached suggesting that the critical accretion rate when the boundary layer transitions is not fixed. The quiescent X-ray flux in SU UMa decreases and, with SS Cygni, is between 2 − 3 orders of magnitude higher than predictions by the disc instability model.