The work conducted in this thesis represents a contribution towards understanding the important processes involved in the formation of filaments and eruptive structures in the solar atmosphere. In particular, the role that ongoing photospheric flux cancellation and associated magnetic reconnection plays in the origin of filament plasma. Also, to study how the magnetic field configuration evolves as flux cancellation proceeds to find indications of the onset of filament eruptions as coronal mass ejections (CMEs). This process is capable of re-configuring the supporting magnetic field from a sheared coronal arcade to a flux rope configuration, building up non-potential field in the atmosphere and storing free magnetic energy along the polarity inversion line. These helical configurations consist of concave-up sections or magnetic dips that provide locations for filament material to be supported against gravity. Flux cancellation observations, whereby small-scale opposite polarity features converge, collide and subsequently dis- appear in line-of-sight magnetic field can therefore, provide a way to investigate how much magnetic flux has been built into the magnetic field configuration that contains the filament before eruption. This process is crucial to the understanding of the onset of CMEs. In this thesis, a small sub-set of active regions have been studied. Observations of the photospheric field have been used to study the field evolution of active regions and its relationship to eruptions. To study the magnetic evolution of these regions an automated algorithm has been developed which tracks magnetic features in line-of-sight magnetic field data to calculate the total quantity and rate of flux cancellation. Chromospheric and coronal plasma observations have been used to study the presence of filament material and plasma flows, in relation to flux cancellation sites. Finally, these observations have been compared to the field configuration and locations of magnetic dips present in non-linear force-free field models.

The auroral substorm is the manifestation of explosive energy release from the rapid and global reconfiguration of the magnetotail. The auroral substorm is marked by a sudden brightening and poleward expansion of the most equatorward auroral arc in the midnight sector of the ionosphere. The temporal sequence of magnetospheric processes which lead to the dynamic auroral substorm display remain disputed to this day. This thesis contains original research on the development and exploitation of novel data analysis techniques in order to analyse ground-based all sky imager data of the aurora, enabling the study of substorm processes in remarkable detail. Fourier analysis techniques are used to find the spatial scales of wave-like signatures (otherwise known as auroral beads/rays), which form along substorm onset arcs. Growth rates of ∼0.05 s−1 are found from the exponential growth of the power spectral density of individual spatial scales. By analysing the dataset in this way, comparisons are made between observations and theoretical predictions of plasma instabilities at the near-Earth edge of the plasma-sheet which have been proposed to play a critical part in the substorm onset process. Auroral arc tracking techniques are developed to automate and increase the size of the database of events analysed. The vast majority of independently identified substorm onsets are preceded by azimuthal structuring along the onset arc with median wavelengths of ∼80 km. These beads grow and develop into a magnetospheric instability around 2 minutes prior to auroral substorm onset. Showing that beads are a common feature along the substorm onset arc provides unprecedented quantitative evidence that a near-Earth instability is a fundamental component of the substorm onset process. Finally, analysis techniques are extended to state-of-the-art high resolution multi-spectral auroral data to investigate the processes driving auroral beads. Beads can be resolved in the green-, blue- and red-line aurora with spatial scales as small as 30 km, which later develop into larger structures of ∼80 km. These observations are consistent with Alfvén wave accelerated auroral particle precipitation and therefore imply that the substorm onset arc and auroral beads are driven unstable by waves.

We study magnetic reconnection events in the solar wind, using observations by the 4 Cluster spacecraft during periods where they are located in this region. Use of data from the 4 Cluster spacecraft allows us to look at the three dimensional structure of events at high time resolution, up to 0.04s for magnetic field data. We present a case study of a magnetic reconnection event that was observed within the solar wind flow on the 2nd March 2006. This event enables us to test the consistency of the temporal and spatial structure of magnetic reconnection from large- to small-scales. We use data from the four Cluster spacecraft to study the three-dimensional structure of the event at sub-second resolution. This showed significant differences in the magnetic field data between spacecraft, despite the fact that they are separated by distances of only order 10^4 km. Thus the structure appears to have variations over relatively small scales. We conclude that magnetic reconnection in the solar wind is not necessarily large-scale but, contrary to previous results, may be somewhat patchy in nature. We also present a sketch for magnetic reconnection which has a more complicated structure than previous sketches of magnetic reconnection events in the solar wind. In addition we have introduced the observed magnetic field and ion velocities over the event depending on the inflow parameters and the strength of the diamagnetic effect of the outflow ions and electrons. We also compare a similarly structured mathematical model (Owen and Cowley 1987b) with 3 case studies in the solar wind and compare the outflows predicted by the model with the observed outflows. Finally we present the workings of an algorithm to find magnetic reconnection events in the solar wind and the subsequent statistical analysis of the events found. The statistical analysis revealed that magnetic reconnection events occur for a wide range of solar wind conditions. We found that the exhaust speed was between 0.1 - 2 x Alfvén speed leading to the conclusion that, for ratios of less than 1, not all the magnetic field energy is converted to kinetic energy and, for those ratios greater than 1, that the reconnection site had a higher local Alfvén speed than where the exhaust was observed by the spacecraft. We also compared the data between Cluster spacecraft and determined that the spacecraft usually witness magnetic field structures that are strongly correlated with each other.

External space is coded in the brain by a network of spatially modulated neurons (including place, grid, border and head direction cells), known as the ‘cognitive map’.This internal map allows flexible and efficient navigation through the external world. These neurons use both self-motion and visual landmark information to update their spatial activity and form an accurate representation of space. Sometimes, the spatial meaning of a landmark can be ambiguous, e.g. when it can be approached from different directions (for example, a tree on the border between two fields). In such cases context information, such as odour, colour or texture, can provide clues to separate one environment from another. Recent work has shown that head direction (HD) cells in the retrosplenial cortex can use these non-metric cues to resolve visually symmetrical spaces with directional landmark ambiguity. In this study, we asked whether animals can also use these nonmetric cues to guide their behaviour, in this case in order to solve spatial tasks across multi-compartment space. Here we show that, behaviourally, rodents can correctly encode relative object positions in visually ambiguous space, and can resolve the directional ambiguity of two visually symmetrical spaces based only on odour information. Electrophysiological recordings of hippocampal place and anterior thalamus HD cells confirmed that both cell types can use odour-context information to discriminate these spaces; therefore, we tested potential involvement of the HD system by temporarily inactivating the anterior thalamus with an awake muscimol infusion. In the behavioural task, HD-disrupted animals show impairment in task performance compared to sham but retain response to novelty. Overall, these results indicate that rodents can use odour-context information to resolve directional ambiguity in otherwise identical multi-compartment environments, and suggest an involvement of the HD system in this process.

In this thesis, a combination of data analysis and test particle simulations is used in order to study several aspects of the complex interaction of Titan with the Saturnian magnetosphere. First, the energetic charged particles environment at the orbital distance of Titan is studied using data from the MIMI/LEMMS instrument. Average fluxes and spectral slopes for energetic ions and electrons are analysed. A large variability is found, and it is interpreted as originated from the high mobility of the energetic ions and electrons, making a simple classification of this environment practically impossible, with only a weak correlation between the ion average fluxes with the plasma environment detectable and an asymmetry between the noon-to-dusk and midnight-to-dawn sectors. Second, the effect of local electromagnetic field disturbances in the access of energetic H+ and O+ ions is studied. By studying the trajectories of individual particles to predict where they will deposit their energy, differences in ionisation rates at different locations around the moon of almost 80% are found for H+ ions and of more than 15% for the case of O+ ions. Finally, the contribution of freshly produced pickup ions to the overall mass loss of the atmosphere is investigated by looking at particular signatures left by these ions in the thermal plasma data from the CAPS/IMS instrument. A statistical survey of all the flybys with available data leads to a constraint of the region around the moon where these ions are detected. Mass losses on the anti-Saturn side of the moon of between 570 kg/day and 1 tonne/day are derived depending on the species, accounting for a small fraction of the total losses estimated from distant tail observations.

Suprathermal electrons are an important kinetic feature of the solar wind; however, their origin and evolution is yet to be fully understood. Suprathermal electrons are divided into two component populations: the field-aligned, beam-like 'strahl' and the approximately isotropic `halo'. Observations have revealed that the strahl component cannot be subject solely to expansion effects as it travels away from the Sun along the interplanetary magnetic field (IMF). Thus, strahl scattering processes must occur within the solar wind and may be at least partly responsible for the existence of the isotopic halo. In this thesis, strahl evolution is examined, in order to constrain the possible strahl scattering mechanisms and investigate strahl utility as an IMF probe. Novel analysis methods are applied to strahl observations to obtain the largest radial range of strahl width observations to date; examine the effect of distance travelled along the IMF on strahl width; and provide estimates of IMF orientation. It is found that strahl scattering likely occurs throughout the solar wind, to at least to 5.5 AU, and strahl electrons are most likely scattered to form part of the halo population by 9 AU. It is also concluded that strahl scattering occurs quasi-continuously and results indicate that scattering may be more intense closer to the Sun, particularly for higher energy electrons. The increase in strahl beam-width with distance is also greater for higher energy strahl, suggesting that there is a dominant scattering mechanism with an inherent energy relation, most likely explained by resonant wave-particle interactions. Finally, observing relatively narrow strahl beams at 1 AU using high resolution Cluster data allowed estimations of strahl width that generally lie within 20° of the magnetometer observations.

During development of the vertebrate hindbrain, the neuroepithelium becomes subdivided into seven morphological units, known as rhombomeres. It is necessary that rhombomeres have sharp, well-defined boundaries, which are established from initially rough gene expression domains during early hindbrain segmentation. The mechanisms involved in early hindbrain segmentation that create sharp segment borders are not well understood. There is evidence to suggest that both regulation of cell identity and Eph/ephrin-mediated cell sorting are required for establishing sharp interfaces between rhombomeres. This thesis investigates the extent to which identity regulation contributes to hindbrain border sharpening in zebrafish. I created a new zebrafish reporter line by CRISPR/Cas9-mediated reporter integration at the egr2b locus, which enables cell identity and cell intermingling to be visualised in live embryos during border sharpening. This new reporter line indicates a contribution of cell identity regulation to border sharpening. I also demonstrate that the contribution of cell identity switching to border refinement is greater in cases where cell intermingling is increased by perturbed Eph/ephrin signalling. To help study the role of Eph/ephrin signalling in border sharpening, I have also created a novel EphrinB3b mutant. The thesis also investigates the mechanisms of identity regulation by community effects and discusses their contribution to border refinement by identity respecification; community effects are suspected to help overcome noise in early gene induction through spatial averaging and thus help establish regions of homogeneous gene expression. The ability of candidate genes to non cell-autonomously regulate the identity of neighbouring cells in the hindbrain is investigated. Of particular focus is the potential involvement of retinoic acid (a morphogen involved in specification of anteroposterior identity) and segmentally-expressed Cyp26 enzymes involved in its metabolism. Analysis of mosaic embryos is used to compare the ability of isolated cells and clustered groups of cells to maintain a different identity to their surroundings. Results presented here are consistent with segmental regulation of retinoic acid signalling contributing to border sharpening.

This thesis is a collection of three separate studies -- but split into four chapters -- which address the underlying issues in the nature and dynamics of markets. The studies investigate price-formation in the presence of noisy asymmetric information flow to a synthetic market, the statistical behaviour of in-play predictive markets and a reformulation of the Markowitz portfolio optimisation for financial market securities into the time-domain. The first study looks to examine modern in-play gambling or predictive markets, in particular, horse racing markets. Since the advent of online sports gambling approximately 15 years ago large amounts of data have been collected for many different sporting events such as football, greyhound racing and cricket. In this study, the focus is on in-play horse racing markets where stylised statistical facts are presented and discussed. Price efficiency is analysed, and statistical arbitrage trading algorithms are developed to evaluate such efficiencies/inefficiencies. We develop a new model for testing the efficiencies of the initial implied odds quoted on the market. Exploring the efficiencies/inefficiencies found in the in-play markets we develop a martingale toy model and a statistical arbitrage trading model. In the second study, we explore price-formation and the pioneering approach to financial asset pricing known as the Brody-Hughston-Macrina framework. The Brody-Hughston-Macrina information-based asset pricing framework is investigated in two parts; the first a development of a trading model and the other a generalisation of the information process that does not assume a linear rate-of-information flow. The trading model developed is a computational agent-based model that allows different configurations of agents to trade and hence create a synthetic market. The different configurations are explored by tracking the market price and times between adjacent trades with respect to changing certain model parameters, such as spread. The generalisation of the rate-of-information does not assume a linear function, as in the original Brody-Hughston-Macrina framework, but instead one that is non-linear in time. We estimate such a function from gambling market data and find it not to be a linear function. The non-linear Brody-Hughston-Macrina framework is fitted to winning horse odds signals. The final study is motivated by recent advances in the spectral theory of auto-covariance matrices, and we are led to revisit a reformulation of Markowitz' mean-variance portfolio optimisation approach in the time domain. In its simplest incarnation, it applies to a single traded asset and allows to find an optimal trading strategy which, for a given return, is minimally exposed to market price fluctuations. The model is initially investigated for a range of synthetic price processes, taken to be either second order stationary, or to exhibit second order stationary increments. Attention is paid to consequences of estimating auto-covariance matrices from small finite samples, and auto-covariance matrix cleaning strategies to mitigate against these are investigated. Finally, we apply our framework to real world data.

With more than 400 million cases a year, Campylobacter jejuni is a leading cause of gastroenteritis worldwide. Farm animals, such as cattle poultry and pigs, act as an environmental reservoir for the bacterium. Although many studies have been performed in order to gain a better understanding of the disease and of the environmental lifestyle of the bacterium, much remains to be uncovered. This work aim is to understand the role of the plasmid pTet in the bacterium lifestyle: firstly, analysing which genes are most represented in the plasmids pTet-like and if a difference in genetic set-up between pTet-like plasmids exists and what is its significance. Secondly, assessing whether the plasmids pTet-like are associated with bacterial characteristics, such as virulence, survival or adaptation to different ecological niches. Several plasmids have previously been described in C. jejuni: pTet is the most widely distributed plasmid amongst the known C. jejuni strains. This plasmid is about 45 kilobase pairs (kbp) long and 52 open reading frames (ORFs) can be identified in its sequence. Before this work, it was known to be found in about 20 % of C. jejuni strains, to encode for tetracycline resistance and a conjugative type four secretion system and to possess several genes involved in horizontal gene transfer. However, the function of this plasmid and the reason why it is so broadly distributed remain largely unknown. In the first part of this work I screened 4005 genome sequenced strains of C. jejuni for the presence of the plasmid. I investigated the genetic make-up of the pTet plasmids, sequencing 19 new plasmids and described a core and accessory sets of genes in this plasmid family using an innovative approach. I found that the core set of genes is predicted to be primarily involved with the maintenance and transfer of the plasmid pTet, while the accessory set of genes are predicted to be associated with metabolism, resistance and regulative functions. In the second part of this work I produced 11 C. jejuni deletion mutants. I tested these mutants for phenotypes associated with environmental survival and host invasion such as biofilm production, motility and infection of the model organism Galleria mellonella larvae. Two of the mutant strains showed an impaired phenotype, when compared to the wild type strain, in motility, production of biofilm and growth in MH broth. These findings indicate that the pTet plasmids have functions other than tetracycline resistance and horizontal transfer, and may help to explain the broad distribution of the plasmids in C. jejuni.

The Twin Arginine Translocase (Tat) is one of two protein translocation mechanisms in E. coli to move proteins across the inner bacterial membrane, from the cytosol to the periplasm. A unique feature of the Tat pathway is its ability to translocate fully folded proteins, indeed, in E. coli the Tat pathway preferentially transports correctly folded proteins. This 'proofreading' mechanism, as it has been dubbed, is of interest to the biopharmaceutical industry, however little is known of the mechanism by which Tat proofreads a substrates conformational state. Initial studies (chapter 3) addressed if the Tat proofreading mechanism sensed the surface charge or hydrophobicity of a substrate. To this end, surface residues of an scFv were mutated to create areas of charge and hydrophobicity without altering tertiary structure. Expression of these variants in E. coli revealed that Tat proofreading is tolerant of surface charge and hydrophobicity, but dependent on conformational flexibility. Further studies utilising a maquette in various folding states, confirmed Tat proofreading is sensitive to the structural rigidity of substrates (chapter 4). Investigations then went on to assess the quality of protein entering the periplasm via the Tat pathway by comparing it to the same protein transported by the General Secretory (Sec) pathway (chapter 5). This revealed, at least for a relatively simple biotheraputic, Tat-translocated protein is of the same quality to Sec-translocated protein. Finally, the question of what is responsible for the proofreading ability of Tat began to be addressed through C-terminal truncation studies of the Tat components that attempted to restore export of export-incompatible substrates (chapter 6).