Ïã¸ÛÁùºÏ²Ê

XClose

Ïã¸ÛÁùºÏ²Ê's Centre for Data Intensive Science

Home
Menu

I Cheng (Matthew)

There are valuable opportunities throughout the CDT to apply machine learning techniques to science and industry projects.

I Cheng

1 January 2019

Project title: AI-Assisted Detection and Plasma Processes of Saturn’s Magnetospheric BoundariesÌýÌý

Research Group: Astrophysics

Supervisor(s): Prof Nick Achilleos

±õ²Ô³Ù°ù´Ç»å³Ü³¦³Ù¾±´Ç²Ô:Ìý

Before joining Ïã¸ÛÁùºÏ²Ê, I had completed an integrated masters in Physics at Imperial College London. My masters' thesis focused on studying magnetic reconnection exhausts in the solar wind and Magnetosheath with Magnetospheric Multiscale spacecraft. The aim was to understand more precisely the structure and energy distribution in reconnection exhausts. The project involved creating a pipeline from data retrieval and pre-processing to performing statistical data analysis to determine the properties of these exhausts in order to predict how the properties would change under different ambient space plasma conditions. I decided to join Ïã¸ÛÁùºÏ²Ê CDT in Data Intensive Science because of the data-focused training courses that are particularly applicable to both academia and industry. There are valuable opportunities throughout the CDT to apply machine learning techniques to science and industry projects.Ìý


Project description:ÌýÌý

A very challenging aspect of magnetospheric boundary statistical studies and this work was the manual identification of hundreds or even thousands of magnetopause crossings by the Cassini spacecraft, particularly the precise transition across the magnetopause current layer. In order to facilitate further studies of this nature, this PhD project will develop a more automated algorithm for detecting the magnetic and plasma signatures of magnetopause and bow shock crossings of Saturn’s magnetosphere. An efficient implementation of a magnetospheric boundary recognition tool would be a paradigm-changing development for modelling magnetopause and bow shock boundaries, as it would facilitate larger datasets to be processed on much shorter time scales as compared to manually identifying crossing signatures. Planetary boundary studies of this nature are particularly timely in light of the recent end-of-mission of Cassini around Saturn, NASA's Juno mission currently in orbit around Jupiter, and ESA's JUICE mission planned to launch in 2022 to explore Jupiter and three of its largest moons, Ganymede, Callisto and Europa.Ìý

First year group project: ONS (Twitter)

Placement: Wynn MacauÌý


Ìý

Publications:

RPS Widget Placeholder

Ìý
Ìý

Ìý