62

F.S. Buonaiuto, N.C. Kraus / Coastal Engineering 48 (2003) 5165

annual deepwater significant wave height (m). The

Because the limiting depth over an ebb shoal is

factor of 3.6 in front of the wave height in Eq. (3) for

controlled by both wave height and tidal prism, a

ebb shoals is much larger than that for the depth over

combination of those two quantities might have

crest for depth-limited waves breaking on longshore

greater predictive power. For this purpose, the param-

eter (HSP)1/4 is introduced that has units of length (m).

bars (which is 0.66; see Larson and Kraus, 1989)

because of the erosion by the ebb jet. Although Eq. (3)

Regression with an assumed linear dependence yields:

predicts depth over the crest of the main ebb shoal

hC 0:066 0:046HSP1=4

MLLW

5

six data points fall outside of the 95% confidence

intervals. Also, the relatively large constant or inter-

which has an intercept of 0.066 m that is consid-

cept, 0.27 m, suggests that a pertinent governing

erably less than the uncertainty in individual depth

process is omitted, such as the tidal prism.

soundings, and Eq. (5) is convenient in having homo-

The seaward extent of the ebb shoal is known to

genous units. Six data points remain outside of the

depend on the tidal prism P (Floyd, 1968), so the

confidence limits (Fig. 10), and the bands are tighter

depth over ebb-shoal crest was regressed with a

to the regression line (R2 = 0.87) than in Figs. 8 and 9.

power-law form similar to those listed in Table 1,

The three regression equations account for the

namely hC = CPn, giving:

variance in the data set reasonably well. The depth

to the ebb-shoal crest was marginally better predicted

hC 0:0063P0:35

MLLW

4

by the combination of the incident wave height and

tidal prism (Fig. 10), based upon R2 values and cluster

with R2 = 0.83, but with eight data points lying outside

within the 95% confidence limits. Homogeneous units

the 95% confidence limits (Fig. 9). Eq. (4) is not

and scaled magnitude (reasonable magnitude of val-

ues) of the parameter (HSP)1/4 make it convenient for

convenient because of the awkward dimensional

empirical coefficient.

calculations.

Fig. 9. Power law regression between the depth over crest of an ebb shoal and tidal prism.

Integrated Publishing, Inc. |