National Evaluation of Chinese Coastal Erosion to Sea Level Rise Using a Bayesian Approach

Q. Zhan,X. Fan,X. Du,J. Zhu
DOI: https://doi.org/10.1088/1755-1315/18/1/012136
2014-01-01
IOP Conference Series Earth and Environmental Science
Abstract:In this paper a Causal Bayesian network is developed to predict decadal-scale shoreline evolution of China to sea-level rise. The Bayesian model defines relationships between 6 factors of Chinese coastal system such as coastal geomorphology, mean tide range, mean wave height, coastal slope, relative sea-level rise rate and shoreline erosion rate. Using the Bayesian probabilistic model, we make quantitative assessment of china's shoreline evolution in response to different future sea level rise rates. Results indicate that the probability of coastal erosion with high and very high rates increases from 28% to 32.3% when relative sea-level rise rates is 4 similar to 6mm/a, and to 44.9% when relative sea-level rise rates is more than 6mm/a. A hindcast evaluation of the Bayesian model shows that the model correctly predicts 79.3% of the cases. Model test indicates that the Bayesian model shows higher predictive capabilities for stable coasts and very highly eroding coasts than moderately and highly eroding coasts. This study demonstrates that the Bayesian model is adapted to predicting decadal-scale Chinese coastal erosion associated with sea-level rise.
What problem does this paper attempt to address?