Comparing ground and space-based datasets with data science and unsupervised AI/machine learning techniques

Space weather is, broadly speaking, the way in which Earth is impacted by coupling with the Sun and its magnetic field. Many key impacts of space weather, such as communications blackouts or power outages, occur due to induced currents in ground-based infrastructure. The focus of this project is not to characterise those induced currents directly but instead to build a better picture of the currents which drive them.

The Sun's magnetic field interacts with and connects to Earth's magnetic field in a series of highly dynamic, time-varying processes. This leads to two distinct but interrelated electric current systems. One set of currents are driven down Earth's magnetic field lines (known as Birkeland currents), which in turn drive currents in the atmosphere (known as Hall and Pedersen currents).

Although we understand the theoretical link between ground-based and space-based measurements, there has been little work exploring the correspondence between the two.

This project will enhance our understanding of how these processes impact the Earth's surface by comparing space-based measurements of the Birkeland currents with ground-based magnetometer measurements. You will use the AMPERE dataset, a space-based dataset with dual-hemisphere coverage providing approximately 13-years of data, in conjunction with ground-based datasets such as SuperMAG, a global collaboration of magnetometer networks.

You will exploit a variety of methods in this project, from data science techniques to unsupervised machine learning approaches such as k-means clustering and self-organising maps, in order to explore the similarity between ground magnetometer data and space-based Birkeland current measurements.