Statistical and geostatistical modelling for mapping surficial sediments of the continental shelf of Algiers bay, Algeria
摘要
Detailed seabed mapping is increasingly necessary. Besides bathymetry, sedimentology is the most crucial variable because it governs sediment transport processes, and the determination of benthic habitats. This paper deals with the characterization of the surficial sediment of the continental shelf of Algiers. Statistical analysis has been applied to the collected data with principal component analysis (PCA) and Hierarchical ascending classification (HAC). Additionally, a geostatistical analysis has been used to predict and map the fractions of mud, silt, clay, illite, kaolinite, chlorite, carbonates and the median grainsize along the continental shelf of Algiers. Both geostatistical approaches employed: ordinary kriging (OK) and ordinary co-kriging (OCK), have been compared by computing a cross-validation. In this study, the bathymetry has been used as a secondary variable to compute OCK. PCA and HAC revealed that Algiers continental shelf can be divided into four sectors: continental slope sector, center of the bay, coastal sector and Cap Matifou sector. OK seemed to be the best method for mapping silt, illite and chlorite while OCK was the best interpolation method for the rest of the parameters. The sediment types have been mapped using the Folk (1954) sediment classification and highlighted the distribution of sediments in an offshore direction along the shelf. In general, combining statistical and geostatistical methods revealed very satisfactory results. Finally, the results of this study can enhance the knowledge of the spatial distribution of the surficial sediments of the Algiers continental shelf.