Automated determination of wave run up from time-variance video images

Thesis submitted to Plymouth University, Devon.A novel remote sensing method is presented that accurately predicts 2% run up and run down thresholds on a gravel beach under calm (Hs < 2 m) conditions. This overcomes the common problem of ascertaining accurate field measurements in the energetic swash zone of a gravel beach where damage to equipment is commonplace. The optical image intensity from time-exposure and time-variance Argus images is interrogated in order to extract the swash parameters of interest. Predictions are validated against field observation and result in a vertical RMS error of 17 cm for run up and 18 cm for back wash. The method alleviates the need for manual digitisation of swash events as has previously been commonplace, enabling swift creation of large datasets for validation of empirical formulae. The use of time-variance images was also seen to increase the number of useable images collected from Argus stations in adverse conditions when compared to time-exposure imagery. This paper outlines a solid proof of concept for the method but acknowledges that extensive further field validation is required, specifically under energetic conditions and at other gravel beaches