Mechanical engineers are using maths to fight mice on Marion Island

Posted on June 04, 2021

In the sub-Antarctic Indian Ocean, 2160 km south of Cape Town, lies a pair of islands called the Prince Edward Islands. The larger of the two islands, Marion Island, is ~290 km2 and houses an impressive research base and meteorological station, where world class research is conducted by several local institutions. This South African territory, managed by the SANAP and DFEE, is a haven for thousands of pairs of breeding seabirds. House mice, introduced to Marion Island over 200 years ago by sealers and whalers, have become a serious threat to endangered albatrosses and petrels by predating on defenceless chicks. The daunting project of eradicating these invasive species, the Mouse Free Marion (MFM) project, is currently being planned for 2023. If successful, Marion Island will be the largest island on which rodents have been eradicated.

The Wandering Albatross nests on Marion Island, with mice attacking their defenceless chicks.

A vital part of a successful eradication is sufficient weather data to plan flight operations and aerial baiting. Given the size of Marion Island and the fact that the island can experience “all four seasons in one day”, the single weather station on the leeward side of the island (i.e. sheltered from the dominant wind direction) will not suffice.

This is where Mechanical Engineering came in handy. Under the supervision of Prof Ken Craig and Ms Janine Schoombie, and in collaboration with Prof Peter le Roux (Department of Plant and Soil Sciences), M.Eng student Kyle Goddard joined a group of scientists on Marion Island to install 32 wind sensors across Marion Island, which would aid a numerical campaign to simulate fine-scale wind on Marion Island.

A high-resolution digital elevation model of Marion Island, complete with lava flow and scoria cones (small volcanic hills), was created with a mesh of 140 million cells. Data from the wind sensors on the island provided input values to simulate 16 wind directions using computational fluid dynamics (CFD). The simulations showed good correlation to measured data, with an overall prediction error below 10%. In addition to the 16 wind direction simulations, effective post-processing allowed for the generation of composite wind maps covering the entire surface of the island. Together, these maps are believed to be helpful to many future biological studies and any possible forays into wind energy generation on Marion Island.

Very few CFD projects exist where such a large island is simulated and this approach did have its challenges. Environmental concerns and a harsh climate on the island prevented installing the “perfect” wind measurement campaign, making validation efforts all the more difficult. Besides building and troubleshooting the very dense mesh required for these large-scale simulations, the project also called for some intensive data manipulation techniques when comparing the simulations to reality. These tasks simply could not be handled without leveraging the computational power of the Centre for High-Performance Computing (CHPC), Cape Town. 

CFD analysis allows the creation of graphics like these, which show the wind speeds various parts of the island will experience.

The simulations from this project, while valuable in themselves, also form a baseline from which to further improve our CFD approach and by extension, our understanding of the fine-scale wind patterns on Marion Island. There are currently several under- and post-graduate students working to investigate the fluid dynamics of wind on Marion Island, as well as post-graduates in the Natural and Agricultural Sciences faculty who are using these insights to predict the impacts of changing wind conditions on the region’s fauna and flora.

As both biotic and abiotic systems on Marion Island are shaped by the strong and consistent westerly winds of the sub-Antarctic region, this body of work forms part of a larger multi-disciplinary, multi-institutional study that is investigating wind as an underexplored driver of ecosystem patterns.

Talks will soon commence between the MFM project coordinators and the scientists/engineers on this team to discuss how the mouse eradication project can benefit from the wind investigation.

- Author Janine Schoombie
Published by Bradley Bock

Copyright © University of Pretoria 2021. All rights reserved.

FAQ's Email Us Virtual Campus Share Cookie Preferences