Bakrania, Mayur Rajesh; (2022) Analysing and Modelling Particle Distributions in Near-Earth Space: Machine Learning. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text PhD_Thesis_Mayur_Bakrania_14006915.pdf - Other Download (6MB) | Preview

Abstract

This thesis contains the analysis of 10 years of ESA Cluster observations using machine learning techniques. In the first study, we investigate solar wind electron populations at 1 au. In the second study, we apply a novel machine learning technique to magnetotail data in order to better characterise particle distribution function. In the third study, we make the first in-situ observations of the tearing instability leading to magnetic reconnection in the magnetotail. Solar wind electron velocity distributions at 1 au consist of three main populations: the thermal `core' population and two suprathermal populations called halo and strahl. We apply unsupervised algorithms to phase space density distributions, to perform a statistical study of how the core/halo and core/strahl breakpoint energies vary. The results of our statistical study show a significant decrease in both breakpoint energies against solar wind speed. By fitting Maxwellians to the core, based on our study, we can discuss the relative importance of the core temperature on halo and strahl electrons. Collisionless space plasma environments are characterised by distinct particle populations that typically do not mix. Although moments of their velocity distributions help in distinguishing different plasma regimes, the distribution functions themselves provide more comprehensive information about the plasma state. By applying dimensionality reduction and clustering methods to electron distributions in pitch angle and energy space, we distinguish between the different plasma regions. We identify several new distinct groups of distributions, that are dependent upon significantly more complex plasma and field dynamics. Magnetic reconnection is a fundamental mechanism responsible for explosive energy release in space and laboratory plasmas. The onset of reconnection is via the tearing instability. Due to its elusive nature, there is an absence of in-situ observations of the tearing instability. We present the first direct observations of the tearing instability and the subsequent evolution of plasma electrons and reconnection, using neural network outlier detection methods. Our analysis of the tearing instability and subsequent reconnection provides new insights into the fundamental understanding of the mechanism responsible for reconnection.

Download activity - last month

Export as XMLJSONCSV

Download activity - last 12 months

Export as CSVJSONXML

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item