each of the selected profile lines. Previous studies of MSL shoreline and 
beach volume changes had concluded that a well-defined seasonal response was 
not found on Long Beach Island (Everts and Czerniak, 1977). Goldsmith, Farrell, 
and Goldsmith (1974) also concluded that during a l-year study (1973-74) of 
biweekly profile measurements at the southern end of Long Beach Island there 
was no overall seasonal trend. These data were not analyzed using empirical 
eigenfunctions. The fact that the seasonal signal can be shown here by a 
quantitative, objective method emphasizes the value of the technique applied 
to well-planned data sets. 
c. Long-Term Changes. One of the goals of the BEP was to determine the 
long-term trend of beach development at the various evaluation sites. This is 
important from the standpoint of planning the beach preservation strategy as 
well as assessing the effectiveness of existing beach protection measures. A 
study of long-term shoreline changes along the mid-Atlantic coast, from Beach 
Haven Inlet to Shackleford Bank, North Carolina, was recently completed by 
Dolan, et al. (1979). Air photos, some dating back to 1930, were used to determine 
the rate of shoreline change of the barrier islands and headlands along the 
630-kilometer section of coastline. Erosion rates averaged 1.5 meters per year, 
but were extremely variable with the greatest erosion rates occurring on 
unstructured, small barrier islands. Accretion was observed in developed areas 
near the north end of the study area where beaches are maintained by groins, 
jetties, and beach nourishment. The results of Dolan,et al. (1979) should not 
be extrapolated to Long Beach Island, but they do indicate that long-term 
erosion is not a necessary and general condition in the adjacent area to the 
south. 
Trends in the change of MSL shoreline position and volume over the period 
of the study are apparent from the figures in Appendixes B and E. A qualita- 
tive indication of the trend in the above parameters as well as that apparent 
from the first beach eigenfunction of the total and demeaned data set is shown 
in Table 4. Linear regression was not used because the resulting slope implies 
a degree of precision and predictability that is unwarranted by these data. The 
volume change shows that only one profile line (21) showed a decrease over the 
study period. All of the other profiles indicated a volume increasing with 
time or no change. The first eigenfunction of the demeaned data is particularly 
well correlated with the rate of volume change. Comparison of the trends shows 
a one to one correspondence in most cases. Profile lines 17 and 18 indicate a 
negative correlation between the first demeaned beach eigenfunction and rate 
of volume change. This is due to a sign ambiguity which exists in the 
numerical solution for the eigenfunctions and eigenvalues. It may be resolved 
by integrating the product of the first temporal eigenfunction and the first 
spatial eigenfunction over the length of the profile. No long-term trends 
were obtainable from the l-year record of profiles within the groin field 
(profile lines 21 to 30). Profile line 7, in that region, showed no trend 
in the rate of volume change. This analysis indicates that, in spite of large 
variability in above MSL volume, most of the profiles along Long Beach Island 
are stable and many are accreting over the term of the study. 
Profile measurements were extended offshore to a depth of approximately 
10 meters in 1937, 1955, 1963, and 1965. Detailed soundings were made in 
Barnegat and Beach Haven Inlets and, during the latter 3 years, were rela- 
tively evenly spaced along the beaches as well (Fig. 4). Many of the offshore 
45 
