Coastal Risk Assessment and Hazard Forecast Analysis via a Bayesian Network
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
In this study, a Bayesian network (BN) module is proposed for coastal hazard risk assessment and forecasting on the basis of the statistical prior and conditional probabilities, which indicates the mechanism of transitivity and causality. The BN module is applicable for quantifying the probabilities of various risk possibilities, which materialize and interpret the phenomenon of natural science. BN-based risk assessment is quantified by the causality of factors in this study instead of individual concerns. Using long-term coastal, environmental, and humanities data from Taitung city, which is situated on along the coast of Taiwan, causality and conditional probability are proposed to forecast hazard risk in the 2°C global warming scenario. The causality verification indicates a strong correlation (R-squared ≈ 0.81) between the wave energy and shoreline change rate, which reflects the performance of the BN module in this study. In the 2°C global warming scenario, the hazard risk degree of the wave energy-propagated shoreline change rate factor will increase in the estuary sector and the sediment supply termination sector; therefore, the maximum probability of a high hazard risk degree will increase to 60% in the estuary sector of the Bainen River. The degree of risk to the shoreline change rate has been revealed to be more polarized under climate change. The estuary and sediment supply termination sectors are identified as erosion hotspots under the climate change scenario. Hazard forecast analysis via a BN is an informative and well-founded process for sustainable coastal management.