Funded PhD Opportunity Monitoring glacier change in the Arctic using multi-spectral satellite imagery and machine learning techniques

This opportunity is now closed.

Subject: Geography and Environmental Studies

Summary

Ice sheets and glaciers are an integral part of the global climate system that are sensitive to climate change and are known to influence global sea levels. Currently, there is concern regarding the future stability of the cryosphere in a warming world and if ice sheets and glaciers continue to melt, global sea levels will potentially rise by over 90m. Given that most of the world’s population lives within several kilometres of the coast, it is critical that we develop a better understanding of how ice sheets and glaciers are responding to climatic warming so that governments can develop mitigation strategies for future generations.  Currently, glaciers in the Arctic are showing increasing signs of instability due to Arctic amplification of the climate system, making it a key region for monitoring changes.  Future projections of glacier change in the Arctic rely on observations of recent glacier change and an understanding of the processes that are driving these changes.

To date, such observations have been undertaken in a piecemeal fashion, with most research focussed on particular regions and with observations over variable time periods. Furthermore, there is a large imbalance across the Arctic, with most studies focussing on the Greenland Ice Sheet and Canadian Arctic Archipelago and with less research for example on Svalbard, and only a handful of studies in the Russian High Arctic. This means that our understanding of past changes is incomplete and that there are large uncertainties for some regions when attempting to predict future change.  To understand what is happening we need better monitoring tools that can be applied Pan Arctic and using new satellite remote sensing datasets such as the Copernicus Earth observation (EO) programme provides an unprecedented opportunity to develop new methods for mapping changes to glacial systems in the Arctic.

This interdisciplinary PhD project involving the Geography and Environmental Sciences Research Institute and Intelligent Systems Research Centre will build on the expertise developed in both research centres to focus on change detection of Arctic glaciers using remote sensing data and novel machine learning techniques. The rapid increase in computing power has enabled the use of powerful machine learning algorithms on large datasets. In particular, recent breakthroughs in computer vision methods and deep learning models for image classification and object detection now make it possible to automatically obtain a much more accurate representation of the composition of the environment than could be previously achieved. It is only recently that advances in deep learning methods have permitted the analysis of satellite imagery and as such, land use classification is still at an early stage but this approach shows great potential for identifying landscape changes. Much work has focussed on the application of convolutional neural networks for land use classification. Deep learning architectures which are pre-trained on natural images from the ImageNet dataset are often used due to a lack of appropriately labelled land-use image data.

In this approach, called transfer learning, a deep network is first trained using the ImageNet dataset and then the network weights are refined using a smaller set of labelled land use images. In this project, low-level (pixel) image processing approaches and object-based approaches will be used in conjunction with deep learning for temporal analysis of image satellite image data for change detection in glaciated landscapes. Algorithms will be developed to automatically learn region representations using a deep neural network in a data-driven fashion. Based on these highly discriminative representations, changes will be determined and predicted using low-level and object based labelling the candidate images.

Essential Criteria

  • Upper Second Class Honours (2:1) Degree from a UK institution (or overseas award deemed equivalent via UK NARIC)
  • A comprehensive and articulate personal statement

Desirable Criteria

If the University receives a large number of applicants for the project, the following desirable criteria may be applied to shortlist applicants for interview.

  • First Class Honours (1st) Degree
  • Masters at 65%
  • Research project completion within taught Masters degree or MRES
  • Experience using research methods or other approaches relevant to the subject domain
  • Work experience relevant to the proposed project
  • Experience of presentation of research findings

Funding

    Vice Chancellors Research Scholarships (VCRS)

    The scholarships will cover tuition fees and a maintenance award of £14,777 per annum for three years (subject to satisfactory academic performance). Applications are invited from UK, European Union and overseas students.

    DFE

    The scholarship will cover tuition fees at the Home rate and a maintenance allowance of £ 14,777 per annum for three years. EU applicants will only be eligible for the fees component of the studentship (no maintenance award is provided).  For Non EU nationals the candidate must be "settled" in the UK.

Other information

The Doctoral College at Ulster University

Launch of the Doctoral College

Current PhD researchers and an alumnus shared their experiences, career development and the social impact of their work at the launch of the Doctoral College at Ulster University.

Watch Video

Key Dates

Submission Deadline
Monday 19 February 2018
Interview Date
March 2018

Contact Supervisor

Dr Paul Dunlop

Other Supervisors

Apply online

Visit https://www.ulster.ac.uk/applyonline and quote reference number #237345 when applying for this PhD opportunity