Edge waves are shallow water waves which are trapped along the
shoreline waveguide by reflection and refraction. The cross-shore
pattern of nodes and antinodes in edge waves has long been
hypothesized to have the right lengthscale to explain longshore bars.
However, different frequencies and modes have antinodes in different
locations, and there is little existing field data to suggest
frequency and mode selection. A bar can also
provide a local minimum in shallow water wave speed,
C+(ghbar)1/2
(where h is depth), needed to produce a separate waveguide, which can
trap and amplify edge waves relative to the shoreline. These trapped
solutions have similar shapes over the bar, regardless of frequency or
wavenumber. Calculations of drift velocity, in the absence of phase
locking, show drift convergence near the top of the bar, at the top of
the bottom boundary layer; these bar-trapped waves may cause bar
growth. Edge waves react to a longshore current as though it were a
change in the bottom topography; the profile will be deeper for edge
waves traveling with the current and shallower for edge waves
traveling against the current, with the magnitude of the change
dependent on the direction and the strength of the current shear
relative to the bottom slope. For strong shears, the bottom can be
changed to the degree of creating a virtual bar, on which edge waves
can also be trapped and amplified. This could be a mechanism for
moving a bar, or creating a bar on a plane beach.
These theories were tested using frequency-wavenumber spectra of the longshore component or orbital velocity from observations taken during the DELILAH experiment, October 1990, Duck, N.C. Continuous, unexplained, diagonal lines of variance have been observed in this data, and similar data from other experiments. Here, these diagonal lines are shown to be evidence of bar-trapped edge waves. These lines not only have the same frequency-wavenumber coordinates as bar-trapped edge waves, but also vary in a predictable manner with changes in the longshore current and depth profile. For example, when the effect of the current was strong enough to remove the effect of the bar in theoretical predictions, the diagonal line of variance disappeared. (The diagonal line reappeared, when this strong current shear moved shoreward into the trough, at high tide). On such days, when the edge wave shape is strongly controlled by current (or for example on a plane beach), the expected affect on topography is unclear because the location of edge wave trapping moves, when the longshore current profile changes with changes the tide. However, on days when the edge wave shape is strongly influenced by the bar, calculations of cross-shore drift, using the DELILAH data to obtain realistic magnitudes, show the drift should allow the bar to grow or maintain itself against gravity and other destructive forces.