Quantified Assessment of Shoreline Change Analysis Using Machine Learning Techniques

点击翻译按钮获取中文摘要

The shoreline is likely to be severely impacted by climatic changes, with particular concerns over the impacts of sea level rise and increased storm activity. This can lead to the rates of beach erosion increasing, impacting both coastal ecology and socio-economics alike. Developing an understanding of the processes that change shorelines and being able to predict what changes are likely to occur are imperative for coastal management and is a challenging task. The growing availability of shoreline morphology data offers the ability to explore newer 'data driven' approaches to coastal modelling. Machine Learning is a data driven approach that helps to identify complex relationships between variables, offering greater accuracies in prediction. Recently, shoreline change studies have started to implement these methods, however, there are limited studies that quantify the performance of different ML approaches, and the quantities of data needed to effectively model shoreline change. This project will aim to perform shoreline change analysis on a section of the UK coastline using ML models, assessing the performance of different ML models. The project will then aim to identify the data quantity limits to best perform SCA, and provide a framework for ML SCA studies on less 'data rich' regions.