About the Project
Supervisors
Dr Jake Bicknell (University of Kent, Durrell Institute of Conservation and Ecology, School of Anthropology and Conservation)
Matthew Struebig (University of Kent, Durrell Institute of Conservation and Ecology, School of Anthropology and Conservation)
Eleni Matechou (University of Kent, Durrell Institute of Conservation and Ecology, School of Anthropology and Conservation, School of Mathematics, Statistics and Actuarial Science)
Project background
Artisanal gold mining is a major driver of tropical deforestation, accelerating rapidly in the remote tropical forests of the Amazon and West Africa. This leaves behind extensive forest degradation, with associated knock-on effects for biodiversity, climate change, and regional weather patterns. The recovery of abandoned mines is correspondingly critical, but even detecting mines in remote forests represents a challenge. Therefore, this project will first work on a tool to rapidly detect gold mines. Secondly, there is evidence that some abandoned gold mines regenerate naturally, while others do not. The project will thus assess the conditions that may or may not facilitate forest recovery in these gold mines. The tool will enable conservation and government agencies to prioritise actions to restore and recover abandoned gold mines in tropical forests.
Research methodology
Using remote sensing datasets (Sentinel and Landsat) the research will work on a machine learning framework to automate the detection of artisanal gold mines. The initial geographic focus will be Guyana and Suriname, eventually expanding to West Africa and beyond. This will involve software such as R, GIS, Python, and Google Earth Engine. The student will work closely with our partners in Germany and Guyana to develop the framework.
The researcher will then use complimentary remotely sensed datasets to assess the conditions under which mines recover naturally, and when do they do not. This will involve statistical models to investigate relationships between vegetation recovery and the bio-physical attributes of mines. This will be used to predict thresholds at which artisanal gold mines will recover on their own, or may need active restoration, and will form the basis of a regeneration analysis tool.
Last, the project will develop a web-based interface for the regeneration analysis tool. The aim is that this will be used by conservation and mining agencies to inform the prioritisation of forest restoration following mining.
Training
The selected student will have access to the University of Kent’s skills training, including R, Python, GIS, deep learning, and machine learning.
Person specification
Applicants to a PhD programme should hold a good Honours degree (First or 2:1) and a Master’s Degree (at Merit or Distinction) in a relevant discipline, or the equivalent from an internationally recognised institution.
The University of Kent requires all non-native speakers of English to reach a minimum standard of proficiency in written and spoken English before beginning a postgraduate degree. For more information on English language requirements, please visit this page.
We seek a student with a strong conservation background (with a degree in a relevant discipline, e.g. conservation, ecology, zoology, environmental sciences), but ideally with experience in remote sensing datasets and machine learning. They should at least be able to evidence the capacity to develop such skills. They should also be willing to work collaboratively with our machine learning partners as well as partners in conservation NGOs and government agencies.
Key Information
Institute scholars will receive the following:
- Annual stipend at UKRI rates (£17,668 in 2022/23);
- Annual tuition fees at Home rates (£4,596 in 2022/23)
2023/24 rates to be announced.
Home and International candidates are eligible to apply but international candidates must provide evidence on how they would cover the difference between home and international fee rates.
Candidates should apply by 23:59 GMT on 17th February 2023 using the online form here.
Shortlisted candidates will be invited for an interview taking place the week commencing 6 March 2023.
References
1) Kalamandeen, M., Gloor, E., Johnson, I., Agard, S., Katow, M., Vanbrooke, A., Ashley, D., Batterman, S.A., Ziv, G., Holder‐Collins, K., Phillips, O.L., Brondizio, E.S., Vieira, I., Galbraith, D., 2020. Limited biomass recovery from gold mining in Amazonian forests. Journal of Applied Ecology 57, 1730–1740.
2) Bicknell, J.E., Collins, M.B., Pickles, R.S.A., McCann, N.P., Bernard, C.R., Fernandes, D.J., Miller, M.G.R., James, S.M., Williams, A.U., Struebig, M.J., Davies, Z.G., Smith, R.J., 2017. Designing protected area networks that translate international conservation commitments into national action. Biological Conservation 214, 168–175.
3) Pfeifer, M., Lefebvre, V., Peres, C.A., Banks-Leite, C., Wearn, O.R., Marsh, C.J., Butchart, S.H.M., Arroyo-Rodríguez, V., Barlow, J., Cerezo, A., Cisneros, L., D’Cruze, N., Faria, D., Hadley, A., Harris, S.M., Klingbeil, B.T., Kormann, U., Lens, L., Medina-Rangel, G.F., Morante-Filho, J.C., Olivier, P., Peters, S.L., Pidgeon, A., Ribeiro, D.B., Scherber, C., Schneider-Maunoury, L., Struebig, M., Urbina-Cardona, N., Watling, J.I., Willig, M.R., Wood, E.M., Ewers, R.M., 2017. Creation of forest edges has a global impact on forest vertebrates. Nature 551, 187.
4) Rodriguez, D. R. O., Rother, D. C., de Figueiredo, F. D. M., Marcelo-Peña, J. L., Pollito, P. A. Z., Valdeiglesias, J. P., & Chambi-Legoas, R. (2021). Natural Regeneration After Gold Mining in the Peruvian Amazon: Implications for Restoration of Tropical Forests. Frontiers in Forests and Global Change.
5) Hao, S., Zhou, Y., & Guo, Y. (2020). A brief survey on semantic segmentation with deep learning. Neurocomputing, 406, 302–321.
Continue with Facebook
