Gaining insight into regional coastal changes on La Réunion island through a Bayesian data mining approach

Abstract

International audienceRecent works have highlighted the interest in coastal geographical databases - collected for coastal management purposes - for obtaining insight into current shoreline changes. On La Réunion, a tropical volcanic high island located in the Southern Indian Ocean, a dataset is available which describes shoreline changes, the coastal geomorphology and the presence of anthropic structures. This database is first supplemented with information on the exposure of each coastal segment to energetic waves and to estuarine sediment inputs. To incorporate relative sea-level changes along the coast in the database, levelling data are analysed in combination with GPS, satellite altimetry and sea-level reconstructions. Finally, a method based on Bayesian networks is used to assess the probabilistic relationships between the variables in the database. The results highlight the high degree of dependency between variables: a retrospective model is able to reproduce 81% of the observations of shoreline mobility. Importantly, we report coastal ground motions for La Réunion island of the order of 1 to 2 mm/year along the coast. However, the resulting differing rates of relative sea-level rise do not significantly impact on shoreline changes. Instead, the results suggest a major control of geological processes and local coastal geomorphic settings on shoreline evolution. While any exploration of a coastal database needs to be complemented with human reasoning to interpret the results in terms of physical processes, this study highlights the significance of revisiting other datasets to gain insight into coastal processes and factors causing shoreline changes, including sea-level changes

Preliminary Results of an Analysis Using Association Rules to Find Relations between Medical Opinions About the non-Realization of Autopsies in a Mexican Hospital

“…In the study [10] authors offered a general overview of the recent works undertaking the predictive and descriptive tasks in the building field, with the aim of improving building performance. On the other hand, the works [11,12] confirmed the considerable potential of Bayesian networks to explore coastal databases. The software industry has also benefited from DM techniques as seen in [13], where Ahmed H. Yousef proposed a solution architecture that enhances software development based on data in software repositories and presented a benchmark that provides an ensemble of DM models in the defective modules prediction problem and compares the results.…”

Determining the Shoreline Retreat Rate of Australia Using Discrete and Hybrid Bayesian Networks

“…Bulteau et al. (2015) included the presence of estuaries as a new variable in the BN and ranked this variable as the second significant next to the geomorphology. Sahin et al.…”

“…A similar spatial BN as Gutierrez et al (2011) was constructed but trained with a Eurasian data set by Yates and Le Cozannet (2012), which found that SLR was of secondary importance on shoreline change after geomorphology. Bulteau et al (2015) included the presence of estuaries as a new variable in the BN and ranked this variable as the second significant next to the geomorphology. Sahin et al (2019) applied a spatial BN to a reef island in Vanuatu by extracting data for different variables either from globally available data sets or by manual inspection.…”

Detecting Changes in European Shoreline Evolution Trends Using Markov Chains and the Eurosion Database

“…Here, we explore how different groups of coastal sites may have contrasting coastal behaviors, following the clustering approach of Gutierrez et al (2011). For example, the coastal database of the eastern United States shoreline was used to detect similarities between the spatial patterns of sea-level rise and shoreline changes, which provide insight into the role of sea-level rise in shoreline erosion (e.g., Gutierrez et al, 2011;Bulteau et al, 2015). Less attention has been given to the temporal evolution of shoreline changes.…”