Figure 13 shows the calculation results for Test R2, which encompassed

blocking of a majority of the waves before incipient breaking. (However,

because random waves were employed, breaking would occur for some

waves.) The shoaling phase is well described and in agreement with trends

obtained for the monochromatic waves. Also, it seems like the recorded

typical blocking point is shoreward of the simulated one. The maximum rms

wave height is overestimated by about 20 percent, which was the case for

some monochromatic tests as well (compare Figure 9). The simulation

results for Test R4 display similar characteristics as the results for Test R2

(see Figure 14).

Test R15 shows some breaking before blocking is calculated to occur, but

otherwise displays the same tendencies as Tests R2 and R4 (Figure 15).

However, the conditions for Test R19 were such that wave blocking was not

predicted to occur (linear dispersion theory). Figure 16 shows the results of

the simulations together with the measurements. Again, the maximum rms

wave height is overestimated, and the discrepancy in the shoreward portion

of the profile is marked. The poor description in the shoreward area owes to

the fact that a majority of the waves is not calculated to break on the current,

implying that they propagate to the area of maximum current without losing

energy. Some waves break, but dissipate their energy down to the stable

wave height fairly quickly, giving a constant wave height in the shoreward

part of the profile. The Monte-Carlo simulation technique with constant

wave period is one reason for the discrepancy between calculations and

measurements. In shallow water (e.g., the CHL-I data), the influence of wave

period is not as pronounced as in deep water, making it more reasonable to

only employ one period in the simulations for the CHL-I data.

To represent the dependence of the momentum transport of the roller on

the cross-shore wave height distribution, the wave energy balance equation

(Equation 31) was included in NMLong-CW. Thus, Equation 32 was solved

after the wave action equation to yield the momentum fluxes associated with

the roller in the cross-shore and longshore directions. Relative wave

quantities were employed in the roller equation to account for the wave-

current interaction. However, before applying the enhanced model to the

CHL-I and C&K data sets, sensitivity tests were carried out, including both

monochromatic and random waves. Monochromatic waves occasionally

caused numerical instability, because the break point represents a

discontinuity in the forcing. This problem and how it was circumvented is

discussed in the chapter dealing with modeling the longshore current. For

random waves, the forcing constitutes a smooth function across the profile,

and no difficulties were encountered in such simulations.

In the CHL-I and C&K experiments, the waves propagated across the

profile (i.e., perpendicular to the shoreline), and the only manifestation of

including the roller would be on the mean water level. In test simulations for

these two data sets, the mean water level was only marginally changed by

including the roller momentum, which in turn did not noticeably change the

wave height distribution across the profile. In comparing, calculated mean

water level from NMLong-CW simulations with and without roller, the

simulations with the roller displayed the expected shoreward shift in the

water level shape.

35

Chapter 4 Verification of Wave Model

Integrated Publishing, Inc. |