Modeling dune response to an East Coast Low

Coastal dunes can act as a method of soft coastal protection against inundation and direct impact of waves during storms if they are substantially large enough in volume to withstand erosion without breaching. However, the time evolution of sand dunes under direct wave impact is not well understood and many available models require site specific calibration and have had limited verification at field scales. Here we test three models of varying complexity in their ability to predict both dry beach erosion volumes and dune to a retreat for an East Coast Low storm event that occurred on the Gold Coast, Australia. The process-based model, XBeach, which models the entire profile was able to reproduce both dune toe retreat and dry beach volume, however, was sensitive to calibration parameters. The two parametric models that only modeled erosion above the initial dune toe position were capable of accurately predicting dune toe retreat, however, under-estimated dry beach erosion volumes. With no calibration, the parametric model proposed by Palmsten and Holman (2012) produced the smallest errors of dune toe retreat with mean error in final dune position of 6.6 m, or 18% of the total measured dune retreat. With minimal calibration estimated absolute error in average dune toe retreat was less than 13% of observed retreat for all three models.

► Field-scale testing of 3 dune erosion models during a storm event.
► The parametric model PH12 accurately predicts dune toe retreat with no calibration.
► XBeach predicts dry beach erosion volumes and dune toe retreat when calibrated.
► XBeach results are sensitive to calibration parameters.
► PH12 can be used as a first estimate of dune toe retreat when no bathy is available.