Consideration of finite amplitude bar response greatly complicates modeling. While several, good, time- domain models exist [e.g. De Vriend, 1987; Roelvink and Stive, 1989], the feedback between the sand bar response and the wave forcing provides the basic conditions associated with nonlinear dynamical systems. Thus, the overall behavior of the nearshore system may be complex in ways that we cannot yet imagine. One approach, as with small scale processes, is the use of cellular automaton techniques, applying very simplified rules to the wave-beach system to study its inherent behavior. Werner and Fink [1993] use this approach to show that beach cusps (rhythmic undulations in the foreshore slope) can result from self-organization of swash on an erodeable slope, in contrast to the popular alternate model of generation by subharmonic (frequency half that of the incident waves), standing edge waves [ Guza and Inman, 1975]. Greater understanding and use of this approach seems important to future progress.
Because long-term behavior of a feedback system can be unexpected (and therefore hard to model), a second approach to understanding the beach system behavior is to collect and examine long time series data from natural systems. Through study of many realizations of beach response, simplicities may emerge, for example, discovery of an appropriate phase plane. At a minimum, the nature of the phenomenology that requires modeling should become apparent. The value of long-term study should not be understated, as initial work has revealed surprises. For example, based on a six-year study of sand bar variability at Duck, NC, Lippmann and Holman [1993] unexpectedly found that the statistics of the inner bar depended on the presence or absence or a second, outer bar. Also, in confirmation of previous work by Birkemeier [1985] and by the Dutch [ Vroeg, 1987], they found a long term tendency for sand bars to form at the shore and to propagate and disappear offshore over roughly a five year time scale (Figure 3). These types of behavior are not intuitive and are not easily modeled. Studies of the long duration and large space scale behavior of the nearshore is part of a rapidly emerging research area called Large Scale Coastal Behavior, examining the predictability of the nearshore system over time scales that are relevant to management decision-makers.