Predicting the next global geomagnetic reversal using machine learning
Dr Phil Livermore (SEE), Chris Davies (SEE); William Brown (BGS); Ciaran Beggan (BGS); Chris Finlay (DTU)Project partner(s): BGS (Edinburgh); DTU (Danish Technical University, Denmark)Contact email: firstname.lastname@example.org
The Earth’s magnetic field, generated by turbulent convection in the liquid outer core, has reversed many times over its 3.5 billion-year history, at a present rate of about 2-3 times per million years. The last global reversal took place 780,000 years ago, leading to speculation that we are “overdue”. This fact, coupled with the observations that the field is weakening in the south Atlantic (the so-called south Atlantic anomaly) and the dipole is presently decaying at a rate of 5% per century, suggests that the magnetic field may be headed for a reversal. However, predicting future magnetic field variations is challenging, in part because we don’t yet have a complete physical description of the geodynamo within the core.
Despite these challenges, at our disposal is a large set of observations of the Earth’s magnetic field, describing its polarity state over millions of years, and more recently, several decades of very high quality satellite data now show the evolution of the geomagnetic field in unprecedented detail. To date, studies using data to constrain the geodynamo process within the core have been reliant on a mix of human subjectivity and physics-based models.
The novel aspect of this project is to apply recent advances in machine learning to the prediction of Earth’s magnetic field. Machine learning is a technique in which computers ‘learn’ to interpret data via an explicit training process, using neural networks for example. Such algorithms have been used with great success in, for example, spotting patterns in consumer spending, speech recognition and in recommending movies within Netflix. In this project, we will train neural networks to learn how the magnetic field has changed, and to assess its predictability. Ultimately, the goal is to assess evidence for whether the geomagnetic field is likely to reverse.
The objectives of the PhD project are as follows:
1. Assess predictability for the million-year evolution of the geomagnetic dipole.
2. Investigate predictability of the global magnetic field using a 400-yr observation-derived model, observatory data sets and the latest satellite data.
3. Assessment of predictability of numerical simulations of Earth’s magnetic field.
The student will learn techniques of machine learning using both Matlab and Python, and will have the opportunity to take relevant specific undergraduate or masters level courses. The student will also have access to a broad spectrum of training workshops at Leeds that include techniques in numerical modelling, through to managing your degree and preparing for your viva. The student will be a part of the deep Earth research group, a vibrant part of the Institute of Geophysics and Tectonics, comprising staff members, postdocs and PhD students. The deep Earth group has a strong portfolio of international collaborators which the student will benefit from.
Although the project will be based at Leeds, there are project partners in both Edinburgh and Copenhagen who the student will visit. There will also be opportunities to attend international conferences (UK, Europe, US and elsewhere), and other possible collaborative visits within Europe.
We seek a highly motivated candidate with a strong background in mathematics, physics, computation, geophysics or another highly numerate discipline. Knowledge of geomagnetism is not required, and training will be given in all aspects of the PhD.
Related undergraduate subjects:
- Applied mathematics
- Computer science
- Earth science
- Geophysical science
- Natural sciences
- Physical science