(Belated!) Congratulations to Dr Tomas Beuzen

Dr Tomas Beuzen completed his PhD at WRL titled “Modelling coastal storm erosion using Bayesian networks” in July 2019. Tomas is now a Teaching Fellow with the Master of Data Science Program at the University of British Columbia in Vancouver. 

During his PhD, under the supervision of Professor Ian Turner, Dr Kristen Splinter, Dr Mitchell Harley and A/Professor Lucy Marshall, Tom investigated the use of data-driven Bayesian networks (and other data mining and machine learning techniques) to assist coastal scientists, engineers and managers to better understand and predict sandy coastline storm erosion using large observational datasets acquired by coastal remote sensing technologies.

While studying, Tomas also assisted as demonstrator for several undergraduate and postgraduate courses within the School of Civil and Environmental Engineering, worked as a coastal engineer for Royal HaskoningDHV and engaged in several international research collaborations.

You can keep up with Tomas’ work on Google Scholar (https://scholar.google.com.au/citations?user=v8Di958AAAAJ&hl=en&oi=ao), LinkedIn (https://au.linkedin.com/in/tomas-beuzen-2b432a118), or his personal website (https://tomasbeuzen.github.io/).

The main contributions of Tomas' PhD research are:

Clarifying the implementation of Bayesian networks for modelling coastal data: A comprehensive review of Bayesian network theory and implementations in coastal modelling was conducted followed by a methodological study using a 10-year dataset of coastal storm events extracted from Argus coastal imaging technology. It was found that Bayesian networks have key advantages for modelling storm erosion including the illumination of causality, uncertainty quantification, and low computational cost. However, these advantages can be limited by the data requirements of the approach. It was determined that, based on data availability and the modelling objective, Bayesian networks can be applied to descriptive or predictive applications. Descriptive networks can be useful for exploring the physical relationships between variables within a specific dataset. In contrast, predictive networks identify general relationships in a dataset that can be used to predict unseen data.

Using a Bayesian network to provide insights into controls of beach and dune storm erosion: A descriptive Bayesian network was used to investigate controls of spatial variability in storm erosion of the berm and dune based on detailed storm erosion observations spanning a 400 km region of the southeast Australian coastline. It was found that spatial variability in storm erosion was driven by both the antecedent morphology of the coastline and hydrodynamic forcing conditions of the storm event, and that controls of erosion of the berm and dune were different.

The development of a generalised predictive model of coastal storm erosion: A predictive Bayesian network was developed using multiple large storm erosion datasets from three different global locations. Results showed that while Bayesian networks developed at one specific location poorly predicted the storm erosion response at differing locations, a Bayesian network developed on all three datasets generalised storm erosion response well, predicting changes to the shoreline, dune toe and dune crest with accuracies between 61 – 76% when tested on unseen data. Importantly, similar variables were important in the Bayesian network for predicting storm erosion at the different study regions and included descriptors of the antecedent beach morphology and hydrodynamics forcing of the storm event.

Journal articles:

1. Beuzen, T., Chickadel, C. C., & Horner-Devine, A. R. (2016). Influence of Subsurface Stratification on Turbulence and Aeration in a Tidal River. IEEE Geoscience and Remote Sensing Letters, 13(12), 1975-1978.
2. Beuzen, T., Splinter, K. D., Marshall, L. A., Turner, I. L., Harley, M. D., & Palmsten, M. L. (2018). Bayesian Networks in coastal engineering: Distinguishing descriptive and predictive applications. Coastal Engineering, 135, 16-30. 10.1016/j.coastaleng.2018.01.005
3. Beuzen, T., Marshall, L., & Splinter, K. D. (2018). A comparison of methods for discretizing continuous variables in Bayesian Networks. Environmental modelling & software, 108, 61-66. 10.1016/j.envsoft.2018.07.007
4. Beuzen, T., & Simmons, J. (2019). A variable selection package driving Netica with Python. Environmental Modelling & Software, 115, 1-5. 10.1016/j.envsoft.2019.01.018 
5. Beuzen, T., Turner, I. L., Blenkinsopp, C. E., Atkinson, A., Flocard, F., & Baldock, T. E. (2018). Physical model study of beach profile evolution by sea level rise in the presence of seawalls. Coastal Engineering, 136, 172-182.
6. Beuzen, T., Harley, M. D., Splinter, K. D., & Turner, I. L. (2019). Controls of variability in berm and dune storm erosion. Journal of Geophysical Research: Earth Surface, 124(11), 2647-2665. 10.1029/2019JF005184
7. Beuzen, T., Goldstein, E. B., & Splinter, K. D. (2019). Ensemble models from machine learning: an example of wave runup and coastal dune erosion. Natural Hazards & Earth System Sciences, 19(10). 10.5194/nhess-19-2295-2019
8. Beuzen, T. (2019). pybeach: A Python package for extracting the location of dune toes on beach profile transects. Journal of Open Source Software, 4(44), 1890. 10.21105/joss.01890

Conference proceedings:

1. Beuzen, T., Splinter, K. D., Turner, I. L., Harley, M. D., & Marshall, L. (2017). Predicting storm erosion on sandy coastlines using a Bayesian network. Australasian Coasts & Ports 2017: Working with Nature, 102.
2. Beuzen, T., Splinter, K. D., Turner, I. L., Harley, M. D., Marshall, L. A., Palmsten, M. L., Stockdon, H. F. & Plant, N. G. (2018). A Probabilistic Model of Regional-scale Response to Extreme Storm Events. Coastal Engineering Proceedings, 1(36), 46.
3. Beuzen, T., Goldstein, E. B., Leaman, C., Simmons, J., Vos, K., Splinter, K. D. Harley, M. D. & Turner, I. L. (2018). A new parameterization for wave runup within a dune impact model. AGU 2018 Fall Meeting Abstracts.
4. Beuzen, T., Splinter, K. D., Harley, M. D., Turner, I. L., Marshall, L. A., Kinsela, M. A. & Middleton, J. H., (2017). A machine learning approach for the prediction of coastal storm erosion at the regional scale. AGU 2017 Fall Meeting Abstracts.
5. Beuzen, T., Simmons, J., Harley, M., Plant, N. G., & Stockdon, H. F. (2019). A machine learning approach for identifying dune toes on beach profile transects. In AGU Fall Meeting 2019, AGU.