Part IV) . Subsequent calculations used in the analysis were carried out from 

 a point on the shoreward end of the profile where no change occurred during 

 the run to a seaward point where there was no movement of material (typically, 

 to the horizontal part of the tank beyond the toe of the beach) . 



305. In Equation 18, sand porosity has been incorporated in q , the 

 cross -shore transport rate, implying that the porosity is independent of time 

 and space. Qualitatively, it was noted during the CE experiments that the 

 foreshore sand tended to compact, whereas sand at the flanks of the breakpoint 

 bar tended to be looser (Kraus and Larson 1988a) . The error introduced by 

 assuming constant porosity is believed to be negligible. 



306. Errors may be introduced through limitations in accuracy of the 

 profile surveys or due to long time interval between surveys. A small 

 systematic error in profile depth measurements may give a finite contribution 

 if summed over the profile length and could give rise to an apparent transport 

 at the seaward boundary of the profile where no transport actually occurred. 

 In particular, the CE surveys, having a 1.2-m (4-ft) spatial interval, were in 

 some cases not taken frequently enough to indicate negligible transport at the 

 seaward boundary. Therefore, in these cases, to proceed with the analysis, 

 one of the profiles was displaced vertically to achieve the condition of zero 

 transport at the boundary. The vertical displacement was in general small 

 (less than 1 cm) and less than the measurement accuracy of the profile survey 

 (±1.5 cm). A small displacement of one of the profiles exerts some influence 

 on the magnitudes of maximum and minimum transport rates but only slightly 

 changes the shape of the transport rate distribution. 



General Features of Cross -Shore Transport 



307. Distributions of the net cross-shore transport rate were deter- 

 mined for both the CE experiment (18 cases) and the CRIEPI experiment (24 

 cases). However, in most of the analyses a subset comprising 33 cases was 

 used which encompassed those cases starting from an initial plane slope. 

 Profile behavior was thereby more readily isolated, and the added complexity 

 of an arbitrary initial profile shape avoided. Between 5 and 10 transport 

 rate distributions were calculated for each case, depending on the number of 



119 



