The spatial and temporal variability of nearshore sand bar morphology is quantified using a unique data set spanning 2 years. The data consist of daily time exposure images of incident wave breaking on an open coast sandy beach which may be used to infer bar morphology (Lippmann and Holman, 1989). The morphology in each image is classified into an eight state morphologic scheme in which bars are uniquely defined by four independent criteria. The most frequently observed morphologies are the longshore-periodic (rhythmic) bars, observed in 68% of the data. Linear bars occur under highest wave conditions H-s=1.78 m) and are unstable (mean residence time≈2 days). Shore-attached rhythmic bars are the most stable (mean residence time≈11 days) and generally form 5--16 days following peak wave events. Non-rhythmic, three-dimensional bars are very transient (mean residence time ≈3 days). Eighty-seven percent of transitions to lower bar types (defined in text) occurred one state at a time, supporting our selection of the ordering of states, and suggesting the suitability of a sequential morphology model. Transitions to higher states occurred under rising wave energy and were evenly spread among the possible higher states, with more substantial changes in morphology resulting from larger wave increases.

This suggests that up-state, erosional transitions (based on offshore bar migration) are better described by an equilibrium model where response is better correlated with incident wave energy than with preceding morphological state. Time exposure images were also digitized to yield quantitative estimates of bar crest location as a function of longshore distance. Principal component analysis was used to decompose bar position into two-dimensional (linear) and three-dimensional (longshore variable) components. Cross-shore (linear) bar position ranges ¿50 m about the 2-year mean (27 m standard deviation) and dominates bar variability (74.6%). Three-dimensional bar structure accounts for ~14% of the variance (12 m standard deviation). Changes in incident wave height precede cross-shore bar migration by less than 1 day. Changes in longshore variability are inversely correlated to changing wave conditions, with bar morphology becoming linear rapidly during storms (on time scales of less than 1 day). Evolution to significantly three-dimensional structure typically occurs over 5--7 days following peak wave events. ¿ American Geophysical Union 1990