American Geophysical Union, Fall Meeting 1996:

**********************************************

Local Generation of Infragravity Energy in the Swash Zone? Model Evaluation

A B Murray and B T Werner (Institute of Geophysics and Planetary Physics, University of California, San Diego, La Jolla, CA 92093-0225); R Ramamoorthi (Division of Physics, Mathematics and Astronomy 200-36, California Institute of Technology, Pasadena, CA 91125)

The energy of fluid motions in the swash zone is commonly dominated by infragravity waves, at frequencies well below the sea and swell-dominated incident band. It has been hypothesized that this low-frequency energy originates outside the swash zone, for example with bound waves that are tied to wave groups until they are released at the break point. Alternatively, we propose that low-frequency motions could be generated by local interactions in the swash zone. The interaction between a collapsing bore and runout can regulate how far up the beach the swash travels and how long it takes to return, which then affects the next interaction. Interactive systems of this type, such as a ball bouncing on a vibrating plate, can easily be destabilized, resulting in chaotic motions rich in low frequencies.

A low-dimensional model of the surf and swash zones treating local interactions generates time series of swash front positions that have energy spectra similar to those of natural series. To more accurately assess the level and importance of local interations on natural beaches, we compare model series with field data by comparing the properties of delay-coordinate-embedding graphs, in which each maximum swash height is plotted versus the previous height(s). Using delay-coordinate embedding, common linear and nonlinear forecasting schemes can be used to make short term preditions of successive vaules in a series given the current and previous values. We compare the accuracy of these predictions for model time series with those for natural series. In addition, we use a new technique for directly comparing the distribution of the data in such graphs, to asses whether the frequency of occurence of different sequences of maximum heights in the model is realistic.

Supported by the Andrew W. Mellon Foundation and the Office of Naval Research, Coastal Dynamics.