Empirical orthogonal function analysis of shoreline changes behind two different designs of detached breakwaters

This paper compares the shoreline responses immediately shoreward of two adjacent schemes of segmented shore parallel rubble mound breakwaters undergoing the same forcing by waves and tides. Scheme one consists of four longer, emergent breakwaters that have produced tidal tombolos in their lee. Scheme two consists of five shorter, lower breakwaters that are submerged at higher tides with salients in the breakwater's lee. Empirical orthogonal function analysis was used to decompose a video derived shoreline dataset into the dominant modes of shoreline change for both schemes. The two schemes showed similar modes of change. The primary mode of change for both schemes was the cross-shore growth and shrinking of the salients/tombolos. The secondary mode of change was the longshore movement of the salients/tombolos. For all modes of change, the dominant length scale was dictated by the breakwater dimensions and locations. A new manifestation of hydrodynamic parameters is introduced: the cumulative integral of the de-meaned parameters. This parameterisation allowed for meaningful correlation of the temporal EOF components with forcing parameters and identification of the important influence of the tide on observed morphodynamic change. Clear differences were observed between the shoreline responses measured in schemes one and two; including differences in bay erosion/accretion; and the longshore translation of salients/tombolos. The beaches in scheme two showed behavioural patterns similar to unprotected beaches which were not observable in scheme one. It is postulated that these differences are caused both by the different breakwater designs and by variation in longshore sediment supply.