Alyssa Heggison

In 2023 I completed a mathematics degree in Edinburgh, with a final group project in which we developed algorithms for division of unevenly iced cakes! During this time I completed an internship with the Open Educational Resource Service, and organised educational events as a Maths Outreach Team Lead. I'm interested in the application of maths and computer science in the real world, and joined the volcanology group in Sheffield in 2024 after seeing this project advertised. My work involves implementing computer vision and machine learning techniques, working towards automation of volcanic sulphur dioxide monitoring, and I'm enjoying the opportunity to learn about volcanology and physics in the process. 

Automated Analysis of Volcano Imagery with Machine Learning Techniques

Reliable measurements of the rate at which gas is released by a volcano are key for understanding a system, and for hazard-monitoring. Over the past decade the volcanology group at Sheffield have developed low-energy-use, low-cost UV Cameras for measurement of sulphur dioxide, which have already been installed at multiple locations and are returning large quantities of image data. The aim of this project is to automate the processing of this data, to obtain emission rates from the recorded images, making real-time monitoring using cameras more practical. I’ll also focus on quantifying the uncertainty in estimated gas emission rates, providing more information about the reliability of calculated data. So far, I have focussed on classification of data using transfer learning, for the removal of samples in which weather conditions obscure the view. I hope for this to contribute to the wider use of similar camera and processing systems, increasing the availability of monitoring data for research.

Supervisors: Dr Thomas Wilkes, Dr Tom Pering and Dr Jefersson dos Santos