Space–time series modelling of beach and shoreline data

Abstract

This paper focuses on utilizing Box–Jenkins modelling procedures to identify models which best describe a time series (1978–1994) of beach and shoreline data collected from the Northeast Beach, Point Pelee, Lake Erie, Canada. The results highlight the influence of localized stochastic processes on beach sediment flux levels and on shoreline change. A spatial autoregressive model best describes the variation for beach net sediment flux. The model reinforces the hypothesis that beach net sediment flux in adjacent sites are interdependent and represent the cumulative spatial effects of short-term sediment flow processes such as longshore drift and persistent localized sediment circulation cells. Additional findings also suggest that a space–time autoregressive model best fits the data on spatial–temporal variations of shoreline retreat. The implication of this model is that the shoreline will retreat quite rapidly, but readjusts to an aggradational state at a much slower rate.