PhD scholarship: Machine learning for computationally efficient flood modelling
The rapid expansion and growing density of our cities, together with climate change, have dramatically increased the risks of flooding. Two-dimensional (2D) flood models are increasingly used to accurately simulate flood inundations and assess the impact of mitigation solutions. Such models solve the governing physics equations (Shallow Water Equations - SWEs) using finite element or finite volume numerical methods. However, they are computationally demanding and require long run times. Therefore, these models are of limited use for exploratory modelling or real-time simulations. This interdisciplinary PhD will discover computationally fast flood simulation algorithms using state-of-the-art Machine Learning (ML) approaches. It will test whether existing numerical methods for solving the governing equations could be advanced/replaced using ML techniques, e.g., deep learning and physics-informed neural networks.
The successful candidate will receive a living stipend of $28,597 per annum (2021 rate) funded by Australian Research Council and a Tuition Fee Scholarship (TFS) from UNSW.
The candidate should:
- Preferably be a citizen/permanent resident of Australia or currently live in Australia on a temporary visa (but we will also consider exceptional overseas students)
- Have excelled in previous academic studies (Applicants must be Honours 1 or equivalent to be competitive).
- Hold a degree in civil engineering (hydraulics) / mechanical engineering / computational mathematics / applied mathematics or computer science
- Have excellent computer programming skills and experience in scientific computing, machine learning or numerical methods
- Meet UNSW English Requirements for PhD in engineering
The full-time PhD will be enrolled within the School of Computer Science and Engineering at UNSW (Kensington Campus) and will be guided by a group of supervisors from UNSW, Eawag (Switzerland), the University of British Colombia (Canada) and the University of Colorado Boulder (USA).
How to apply:
Please contact Dr Behzad Jamali (firstname.lastname@example.org) or A/Prof. Aleksandar Ignjatovic (email@example.com) with a subject heading “ARC Fast Flood Modelling PhD scholarship” with a CV and summarised transcripts. Please submit your CV and transcripts by the 10 December 2020.