Table 5 Is constructed like Table 4 but for sediment samples collected 

 in May and December 1968 at Newport Beach, California. The squares of the 

 standard error of the mean ratios indicate that the variation in mean grain 

 sizes is about three to six times greater along profiles than between pro- 

 files, and there is a smaller difference between seasonal means than for 

 the Pt. Mugu example (Table 4). Data obtained either in OctobeT or November 

 at Newport Beach would have probably been more representative of the 

 "summer" condition, and would show greater differences with May than do 

 the December, data. The variation between profiles is much greater at Newport 

 Beach than at Pt. Mugu. 



Tables 4 and 5 indicate that more samples are needed to determine the 

 across-profile component of variability than needed for the along-beach 

 component but how many more? Using squared standard error ratios, the 

 answer to this need appears to be from 3 to 35 times as many samples. The 

 lower values could be applied to beaches of fairly regular topography with 

 perhaps eight samples collected along each of four profiles twice a year 

 (64 samples] ; a greater numbeT of profiles would be required for beaches 

 of more irregular topography. Until the expected magnitudes of all four 

 components of variability have been analyzed., a conservative approach would 

 be to use the coarser (usually winter) composite for beach- fill calculations. 



3. Borrow Site Sampling . 



Sediments suitable for beach nourishment have been traditionally obtained 

 from land and estuarine sources. Offshore sources have increased in impor- 

 tance as land-derived materials become scarcer and because lagoonal mate- 

 rials are often too fine grained or unavailable due to environmental 

 constraints. Regardless of type, the suitability of a potential borrow 

 deposit must be determined by sampling according to a plan that reflects 

 both geometric and genetic aspects of the deposit. These constraints often 

 result in sampling schemes as varied as the borrow sites. Random or regular 

 core sampling on a grid might be appropriate for characterizing a sand 

 deposit of unknown geometry whereas a grid-sampling plan would be inade- 

 quate for evaluating a sinuous offshore deposit. The points to be made are 

 that the composite characteristics of the borrow material must be as repre- 

 sentative as the beach composite to predict a fill's performance and, that 

 care must be taken to design an adequate sampling plan. Also, any addi- 

 tional information such as seismic reflection surveys or topographic 

 expression of the sea floor should be used in evaluating a potential borrow 

 deposit. 



Figure 5 shows kinds of sites requiring different sampling procedures. 

 A potential offshore region of interest (Fig. 5, a) is defined by general, 

 shore -parallel boundaries. The position of the seaward boundary is deter- 

 mined by the maximum water depth (D 2 ) in which a modern dredge can 

 operate, or the farthest distance offshore that pipeline pumping is feasi- 

 ble or sand-hauling economic. Environmental considerations or physical 

 constraints such as wave climate or the maximum depth to which beach 

 processes operate might determine the location of the nearshore boundary 

 {D}). This region of interest between Dj and D 2 could be sampled according 



