Automated Statewide Coastal Erosion Assessment of South Australia – Balancing Innovation and Judgement

Abstract

Coastal erosion poses a significant challenge to South Australia. In response, the Department for Environment and Water, in collaboration with the Local Government Association of South Australia, has commissioned a
statewide erosion hazard assessment under the SA Climate Ready Coasts Data Workstream Program. This study explores the integration of automation in numerical coastal modelling while highlighting the necessity of expert judgement in refining outputs and model settings.

Advancements in national and international data sources, coupled with increasingly sophisticated numerical models, offer opportunities for rapid large-scale coastal assessments. However, the accuracy and applicability
of these automated tools to South Australia's variable and dynamic coastal environment remains uncertain. This study examines the efficacy of an automated XBeach model generation tool, which incorporates modelled
wave conditions, surveyed beach profiles, archived sediment size data, the Smartline coastal classification system, and CSIRO’s CANUTE extreme water level dataset. The case study demonstrates the workflow, challenges, and practical implications of automation in coastal erosion modelling of sandy beaches.

Findings indicate that automation allows for more efficient utilisation of site-specific data in large-scale assessments by integrating detailed datasets and capturing key physical processes. However, automated
modelling alone is insufficient; sense checks remain essential for refining simulations and ensuring reliability. The study also highlights the role of automated workflows in improving quality assurance and quality control
(QA/QC), particularly when managing extensive 1D model datasets (7000+ simulations). Looking ahead, the question remains, can artificial intelligence evolve to bridge the gap between automation and expert intuition?