Magnetic flux emergence is the subject of how magnetic fields from the solar interior can rise and expand into the atmosphere to produce active regions. It is the link that joins dynamics in the convection zone with dynamics in the atmosphere. In this thesis, we study many aspects of magnetic flux emergence through mathematical modelling and computer simulations. Our primary aim is to understand the key physical processes that lie behind emergence. The first chapter introduces flux emergence and the theoretical framework, magnetohydrodynamics (MHD), that describes it. In the second chapter, we discuss the numerical techniques used to solve the highly non-linear problems that arise from flux emergence. The third chapter summarizes the current literature. In the fourth chapter, we consider how changing the geometry and parameter values of the initial magnetic field can affect the dynamic evolution of the emerging magnetic field. For an initial toroidal magnetic field, it is found that its axis can emerge to the corona if the tube’s initial field strength is large enough. The fifth chapter describes how flux emergence models can produce large-scale solar eruptions. A 2.5D model of the breakout model, using only dynamic flux emergence, fails to produce any large scale eruptions. A 3D model of toroidal emergence with an overlying magnetic field does, however, produce multiple large-scale eruptions and the form of these is related to the breakout model. The sixth chapter is concerned with signatures of flux emergence and how to identify emerging twisted magnetic structures correctly. Here, a flux emergence model produces signatures found in observations. The signatures from the model, however, have different underlying physical mechanisms to the original interpretations of the observations. The thesis concludes with some final thoughts on current trends in theoretical magnetic flux emergence and possible future directions.

The structure of the magnetic field is often an important factor in many energetic processes in the solar corona. To determine the topology of the magnetic field features such as null points, separatrix surfaces, and separators must be found. It has been found that these features may be preferred sites for the formation of current sheets associated with the accumulation of free magnetic energy. Over the last decade, it also became clear that the geometrical analogs of the separatrices, the so-called quasi separatrix layers, have similar properties. This thesis has the aim of investigating these properties and to find correlations between these quantities. Our goal is to determine the relation between the geometrical features associated with the QSLs and with current structures, sites of reconnection and topological features. With these aims we conduct three different studies. First, we investigate a non linear force free magnetic field extrapolation from observed magnetogram data taken during a solar flare eruption concentrating our attention on two snapshots, one before the event and one after. We determine the QSLs and related structures and by considering carefully how these change between the two snapshots we are able to propose a possible scenario for how the flare occurred. In our second project we consider potential source distributions. We take different potential point source models: two four sources models already presented in the literature and a random distribution of fifteen sources. From these potential models we conduct a detailed analysis of the relationship between topological features and QSLs. It is found that the maxima of the Q-factor in the photosphere are located near and above the position of the subphotospheric null points (extending part way along their spines) and that their narrow QSLs are associated with the curves defined by the photospheric endpoints of all fan field lines that start from subphotospheric sources. Our last study investigates two different flux rope emergence simulations. In particular, we take one case with and one without an overlying magnetic field. Here, we can identify the QSLs, current, and sites of reconnection and determine the relation between them. From this work we found that not all high-Q regions are associated with current and/or reconnection and vice-versa. We also investigated the geometry of the field lines associated with high-Q regions to determine which geometrical behaviour of the magnetic field they are associated with. Those that are associated with reconnection also coincide with topological features such as separators.

Magnetic fields play a key role in a wide variety of phenomena found on the Sun. One such phenomena is the Coronal Mass Ejection (CME) where a large amount of material is ejected from the Sun. CME’s may directly affect the earth, therefore understanding their origin is of key importance for space weather and the near-Earth environment. In this thesis, the nature and evolution of solar magnetic fields is considered through a combination of Magnetic Flux Transport Simulations and Potential Field Source Surface Models. The Magnetic Flux Transport Simulations produce a realistic description of the evolution and distribution of the radial magnetic field at the level of the solar photosphere. This is then applied as a lower boundary condition for the Potential Field Source Surface Models which prescribe a coronal magnetic field. Using these two techniques, the location and variation of coronal null points, a key element in the Magnetic Breakout Model of CMEs, are determined. Results show that the number of coronal null points follow a cyclic variation in phase with the solar cycle. In addition, they preferentially form at lower latitudes as a result of the complex active latitude field. Although a significant number of coronal nulls may exist at any one time (≈ 17), it is shown that only half may satisfy the necessary condition for breakout. From this it is concluded that while the Magnetic Breakout Model of CMEs is an important model in understanding the origin of the CMEs, other processes must occur in order to explain the observed number of CMEs. Finally, the Magnetic Flux Transport Simulations are applied to stellar magnetic fields and in particular to the fast rotating star HD171488. From this speculative study it is shown that the Magnetic Flux Transport Simulations constructed for the Sun may be applied in very different stellar circumstances and that for HD171488 a significantly higher rate of meridional flow (1200-1400 ms⁻¹) is required to match observed magnetic field distributions.

This thesis considers the properties of MHS equilibria formed through non-resistive MHD relaxation of analytical non-potential magnetic field models, which contain two null points connected by a generic separator. Four types of analytical magnetic fields are formulated, with different forms of current. The magnetic field model which has a uniform current directed along the separator, is used through the rest of this thesis to form MHS equilibria and to study reconnection. This magnetic field, which is not force-free, embedded in a high-beta plasma, relaxes non-resistively using a 3D MHD code. The relaxation causes the field about the separator to collapse leading to a twisted current layer forming along the separator. The MHS equilibrium current layer slowly becomes stronger, longer, wider and thinner with time. Its properties, and the properties of the plasma, are found to depend on the initial parameters of the magnetic field, which control the geometry of the magnetic configuration. Such a MHS equilibria is used in a high plasma-beta reconnection experiment. An anomalous resistivity ensures that only the central strong current in the separator current layer is dissipated. The reconnection occurs in two phases characterised by fast and slow reconnection, respectively. Waves, launched from the diffusion site, communicate the loss of force balance at the current layer and set up flows in the system. The energy transport in this system is dominated by Ohmic dissipation. Several methods are presented which allow a low plasma-beta value to be approached in the single-separator model. One method is chosen and this model is relaxed non-resistively to form a MHS equilibrium. A twisted current layer grows along the separator, containing stronger current than in the high plasma-beta experiments, and has a local enhancement in pressure inside it. The growth rate of this current layer is similar to that found in the high plasma-beta experiments, however, the current layer becomes thinner and narrower over time.

Observations of the surface of the Sun reveal multi-scaled, mixed magnetic features that carpet the entire solar surface. Not surprisingly, the global magnetic fields extrapolated from these observations are highly complex. This thesis explores the topology of the Sun's global coronal magnetic fields. The magnetic skeleton of a magnetic field provides us with a way of examining the magnetic field and quantifying its complexity. Using specialised codes to find the magnetic skeletons which were written during the course of this work, we first examine potential field extrapolations of the global solar coronal magnetic field determined from observed synoptic magnetograms from the Heliospheric Magnetic Imager on the Solar Dynamics Observatory. The resolution of the PFSS models is found to be very important for discovering the true nature of the global magnetic skeleton. By increasing the maximum number of harmonics used in the potential field extrapolations and, therefore, the grid resolution, 60 times more null points may be found in the coronal magnetic field. These high resolution fields also have a large global separator network which connects the coronal magnetic field over large distances and involves between 40 % and 60 % of all the null points in the solar atmosphere. This global separator network exists at both solar minimum and solar maximum and has separators that reach high into the solar atmosphere (> 1R☉) even though they connect null points close to the solar surface. These potential field extrapolations are then compared with magnetohydrostatic (MHS) extrapolations of the coronal magnetic field which also provide us with information about the plasma in the corona. With a small component of electric current density in the direction perpendicular to the radial direction, these MHS fields are found to have a plasma beta and pressure typical of the corona. As this small component of electric current density grows, the heliospheric current sheet is warped significantly and the magnetic field, plasma beta and pressure become unphysical. Torsional spine reconnection is also studied local to a single null point. First using a dynamical relaxation of a spiral null point under non-resistive magnetohydrodynamics (MHD) to a MHS equilibrium is form in which a current layer has built up around the spine lines. Then the reconnection under resistive MHD in this current sheet is studied. The current about the spine lines is dissipated and the magnetic energy is mainly converted into heat directly as the field lines untwist about the spine line.

The dynamics of the solar corona are dominated by the magnetic field which creates its structure. The magnetic field in most of the corona is ‘frozen’ to the plasma very effectively. The exception is in small localised regions of intense current concentrations where the magnetic field can slip through the plasma and a restructuring of the magnetic field can occur. This process is known as magnetic reconnection and is believed to be responsible for a wide variety of phenomena in the corona, from the rapid energy release of solar flares to the heating of the high-temperature corona. The coronal field itself is three-dimensional (3D), but much of our understanding of reconnection has been developed through two-dimensional (2D) models. This thesis describes several models for fully 3D reconnection, with both kinematic and fully dynamic models presented. The reconnective behaviour is shown to be fundamentally different in many respects from the 2D case. In addition a numerical experiment is described which examines the reconnection process in coronal magnetic flux tubes whose photospheric footpoints are spun, one type of motion observed to occur on the Sun. The large-scale coronal field itself is thought to be generated by a magnetohydrodynamic dynamo operating in the solar interior. Although the dynamo effect itself is not usually associated with reconnection, since the essential element of the problem is to account for the presence of large-scale fields, reconnection is essential for the restructuring of the amplified small-scale flux. Here we examine some simple models of the solar-dynamo process, taking advantage of their simplicity to make a full exploration of their behaviour in a variety of parameter regimes. A wide variety of dynamic behaviour is found in each of the models, including aperiodic modulation of cyclic solutions and intermittency that strongly resembles the historic record of solar magnetic activity.

No / A crucial challenge to successful flare prediction is forecasting periods that transition between "flare-quiet" and "flare-active." Building on earlier studies in this series in which we describe the methodology, details, and results of flare forecasting comparison efforts, we focus here on patterns of forecast outcomes (success and failure) over multiday periods. A novel analysis is developed to evaluate forecasting success in the context of catching the first event of flare-active periods and, conversely, correctly predicting declining flare activity. We demonstrate these evaluation methods graphically and quantitatively as they provide both quick comparative evaluations and options for detailed analysis. For the testing interval 2016-2017, we determine the relative frequency distribution of two-day dichotomous forecast outcomes for three different event histories (i.e., event/event, no-event/event, and event/no-event) and use it to highlight performance differences between forecasting methods. A trend is identified across all forecasting methods that a high/low forecast probability on day 1 remains high/low on day 2, even though flaring activity is transitioning. For M-class and larger flares, we find that explicitly including persistence or prior flare history in computing forecasts helps to improve overall forecast performance. It is also found that using magnetic/modern data leads to improvement in catching the first-event/first-no-event transitions. Finally, 15% of major (i.e., M-class or above) flare days over the testing interval were effectively missed due to a lack of observations from instruments away from the Earth-Sun line.