Development of a wave classification scheme to examine climate variability and nearshore response

Abstract

Nearshore response to changing wave conditions, both at the short-term (storm) and long-term (climate) time scales, can be observed through temporal variations in shoreline position and orientation. Approximately 500,000 m3/yr of sand is estimated to travel northwards past the Gold Coast, however, several studies (Delft Hydraulics Laboratory 1992, Patterson 2007) have shown this can vary considerably along the coast and from year to year. Gradients in longshore transport (both spatially and temporally) can play a significant role in temporal variations of the shoreline, thus understanding the driving mechanisms (breaking wave characteristics) and their variability are key to predicting future shoreline change. The roughly east-facing coast is exposed to energetic wave conditions throughout the year. South-East ground swell is the predominant wave signature, however, isolated events such as East-Coast Lows (ECL) and tropical cyclones (TC) also contribute to the large longshore drift. Wave models, such as NOAA's Wave Watch III (WWIII) and ECMWF's ERA reanalysis provide offshore wave data at roughly 6-hr intervals and can be used as offshore boundary conditions in nearshore spectral models to estimate wave breaking conditions. However, running spectral wave models in near real time requires large computational costs and in most cases is redundant given that offshore wave conditions are often repeatable and can be grouped based on similarities in wave properties, thus reducing the number of individual model runs drastically. Here we first develop a method to classify the yearly offshore wave climate into distinct bins (classes) describing the various forcing mechanisms. Compared to joint pdf methods, the classification scheme considers the three parameters (wave height, Hs, wave period, Tp, and mean wave direction, ?) as a single entity and provides a succinct way of describing a yearly wave climate (composed of 100s of observations) into a much more manageable number of approximately 10 classes.