Discrimination between coastal subenvironments using textural characteristics



The objective of this study was to discriminate between modern beach subenvironments based on textural characteristics obtained using the graphical (percentile) method, the moment method, and the log-hyperbolic distribution (LHD). A total of 126 surface sedimentation units were sampled at the nodes of a 21 x 6 rectangular grid (1000 m2) on a carbonate sand beach, Oahu, Hawaii. Sampling was conducted at low energy conditions from the lower foreshore to the backshore. Non-parametric discriminant analysis was used as an objective tool in defining distinct subenvironments. Confidence bands around the canonical variates derived from the graphic mean, sorting, skewness and kurtosis indicated four separate subenvironments (lower foreshore, mid-foreshore, upper foreshore and backshore). Three distinct subenvironments were identified using the mean, sorting (standard deviation) and skewness measures derived by the method of moments. A similar subenvironment distinction was obtained using five statistics of the LHD (gamma, γ; nu ν; delta, δ; tau, τ; and xi, ξ). No significant difference was noted in textural characteristics between the upper foreshore and backshore zones, and these zones were grouped into one subenvironment. These results indicate that different process scenarios would be needed to explain different subenvironment partitioning based simply on the approach adopted. Discriminant analysis indicated that fewer subenvironment samples were misclassified and separation distances between subenvironments in bivariate canonical plots were greater for the standard moment measures compared with the statistics derived from fitting the computationally intensive LHD model. Examination of the mass frequency grain size distributions indicated that the LHD was generally the most appropriate model. These observations were confirmed by the hyperbolic shape triangle which indicated that the LHD rather than the more commonly used log-normal distribution was generally optimal in describing sediments. These results support the use of the LHD statistical measures in subenvironment discrimination over the graphic-inclusive measures.