A high-resolution primitive equation model simulation is used to form an energy budget for the principal semidiurnal tide (M 2 ) over a region of the Hawaiian Ridge from Niihau to Maui. This region includes the Kaena Ridge, one of the three main internal tide generation sites along the Hawaiian Ridge and the main study site of the Hawaii Ocean Mixing Experiment. The 0.01°-horizontal resolution simulation has a high level of skill when compared to satellite and in situ sea level observations, moored ADCP currents, and notably reasonable agreement with microstructure data. Barotropic and baroclinic energy equations are derived from the model's sigma coordinate governing equations and are evaluated from the model simulation to form an energy budget. The M 2 barotropic tide loses 2.7 GW of energy over the study region. Of this, 163 MW (6%) is dissipated by bottom friction and 2.3 GW (85%) is converted into internal tides. Internal tide generation primarily occurs along the flanks of the Kaena Ridge and south of Niihau and Kauai. The majority of the baroclinic energy (1.7 GW) is radiated out of the model domain, while 0.45 GW is dissipated close to the generation regions. The modeled baroclinic dissipation within the 1000-m isobath for the Kaena Ridge agrees to within a factor of 2 with the area-weighted dissipation from 313 microstructure profiles. Topographic resolution is important, with the present 0.01°resolution model resulting in 20% more barotropic-to-baroclinic conversion compared to when the same analysis is performed on a 4-km resolution simulation. A simple extrapolation of these results to the entire Hawaiian Ridge is in qualitative agreement with recent estimates based on satellite altimetry data.

The effects of water level variations on breaking wave setup over fringing reefs are assessed using field measurements obtained at three study sites in the Republic of the Marshall Islands and the Mariana Islands in the western tropical Pacific Ocean. At each site, reef flat setup varies over the tidal range with weaker setup at high tide and stronger setup at low tide for a given incident wave height. The observed water level dependence is interpreted in the context of radiation stress gradients specified by an idealized point break model generalized for nonnormally incident waves. The tidally varying setup is due in part to depth-limited wave heights on the reef flat, as anticipated from previous reef studies, but also to tidally dependent breaking on the reef face. The tidal dependence of the breaking is interpreted in the context of the point break model in terms of a tidally varying wave height to water depth ratio at breaking. Implications for predictions of wave-driven setup at reef-fringed island shorelines are discussed.

[1] Measurements obtained across a shore-attached, fringing reef on the southeast coast of the island of Guam are examined to determine the relationship between incident waves and wave-driven setup during storm and nonstorm conditions. Wave setup on the reef flat correlates well (r > 0.95) and scales near the shore as approximately 35% of the incident root mean square wave height in 8 m water depth. Waves generated by tropical storm Man-Yi result in a 1.3 m setup during the peak of the storm. Predictions based on traditional setup theory (steady state, inviscid cross-shore momentum and depth-limited wave breaking) and an idealized model of localized wave breaking at the fore reef are in agreement with the observations. The reef flat setup is used to estimate a similarity parameter at breaking that is in agreement with observations from a steeply sloping sandy beach. A weak (∼10%) increase in setup is observed across the reef flat during wave events. The inclusion of bottom stress in the cross-shore momentum balance may account for a portion of this signal, but this assessment is inconclusive as the reef flat currents in some cases are in the wrong direction to account for the increase. An independent check of fringing reef setup dynamics is carried out for measurements at the neighboring island of Saipan with good agreement.

[1] A GPS station in Manaus, near the center of the Amazon basin, manifests an annual cycle of vertical displacement with a peak-to-peak amplitude of 50-75 mm. This is by far the largest crustal oscillation observed to date, and nearly 2 -3 times larger than the amplitude predicted for this region. Vertical ground displacement is strongly anticorrelated with the local stage height of the Amazon river, with no detectable time lag between the two time series. This suggests that we are observing, for the first time, a purely elastic response to changes in the weight of a flowing river system. We use a simple hydrological model to relate stage height to the regional pattern of flooding, and argue that the elastic oscillations observed in Manaus are dominated by changes in water loading developed within $200 km of the GPS station. Citation:

[1] Open basin resonant modes have been observed at tidal frequencies on coastal shelves, but their excitation on coral reef platforms has only been suggested. The topography of and water depth over most fringing reefs correspond to resonant periods that are the order of tens of minutes and fall outside of the energetic part of the wave spectrum. During tropical storm Man-Yi, low frequency near resonant oscillations dominated the variance of the sea surface elevation at the shoreline of Ipan reef, Guam. The excitation of the resonant modes resulted both from a large increase in water level over the reef due to wave setup, which increased the resonant frequencies, and the occurrence of wave group forcing with time scales commensurate with these increased resonant frequencies. Citation: Péquignet, A. C. N., J. M.Becker, M.