(a) On an overall basis, the results of the present study agree 

 reasonably well with some of the previous studies. The dimensionless 

 constant K in the relationship I = KPjj^g was found to range from 0.19 

 to 4.18 with an average value K = 0.98 for the Channel Islands data. 

 In view of the large scatter within the data set, it is recommended 

 that no change be made in the SPM (U.S. Army, Corps of Engineers, 

 Coastal Engineering Research Center, 1977) relationship for sand 

 transport. 



(b) For the phase I survey periods, the K values obtained here 

 are substantially lower than those based on visual wave estimates 

 obtained by Bruno and Gable (1976). Their average K value is 1.40 

 as compared to 0.98 obtained here for the Channel Islands data. 



(c) A pair of wave gages appears to perform reasonably well to 

 determine overall directional characteristics such as P^s! however, 

 the quality of the detailed directional results could be improved 

 substantially by a greater number of gages. 



(d) In calculating wave direction from an array of gages, it is 

 extremely important to accurately determine the relative locations 

 of the gages and the shoreline orientation. For the present data, 

 an angular misalinement of +0.5° can cause an associated change in 

 P^g, ranging from 8 to 18 percent. 



(e) LEO wave height observations appear to be subjective with 

 strong observer bias. As a result, LEO measurements should be cor- 

 related with wave gage or other instrumentation measurements before 

 their acceptance. 



(f) There are variations in the K values that cannot be explained 

 through any evident mechanisms or processes other than poor data 

 quality in specific data sets. 



(g) The environmental and economic significance of an improved 

 quantitative predictor for longshore sediment transport justifies 

 substantial future field programs to discern the effects of individ- 

 ual variables (e.g., grain size, beach slope, etc.) and in general, 

 to develop a more reliable predictor. 



39 



