Laboratory investigation of dune erosion using stereo video

Simple parameterizations of dune erosion are necessary for forecasting erosion potential prior to an oncoming storm. Dune erosion may be parameterized in terms of the elevation of the total water level (composed of surge, tide, and wave runup) above the dune base and period of exposure of the dune to waves. In this work, we test several versions of this model using observations from a large wave tank experiment designed to model a storm hydrograph, and we develop a new method for acquiring the appropriate data with confidence intervals using stereo video techniques. The stereo method results in observations of dune morphology at higher spatial and temporal resolutions than traditional survey methods allow. Resolution of the stereo technique was 0.1 m in the horizontal and 0.04 m in the vertical, and errors in stereo observations were on the order of 0.02 to 0.08 m (1 to 2 pixels) when compared with surveys. A new method was developed to estimate confidence intervals on stereo observations. When the unchanging dune top was repeatedly sampled, the new confidence intervals encompassed 2 standard deviations of scatter about the mean dune surface 98% of the time. Observations from the stereo method were used to quantify wave runup and dune erosion. We tested a variety of runup statistics based on a Gaussian distribution of swash properties, and found that the most predictive statistic for dune erosion was the 16% exceedance elevation above the dune base, lower than the often used 2% exceedance value. We found that the parameterization of runup was sensitive to the definition of beach slope and that the most accurate beach slope for predicting runup was through the region of the beach profile defined by the mean water level plus one standard deviation of swash. The dune base retreated along a relatively constant trajectory that was a half of the initial beach slope. Finally, a simple model for dune erosion was tested and found to reproduce 64% of the observed variance in dune erosion rate given known forcing at the dune and 49% of the observed variance in dune erosion rate given parameterized forcing. Integrating the simple model over time, 93% of the observed dune retreat distance was reproduced given offshore forcing. (C) 2011 Elsevier B.V. All rights reserved.