OS34A-01 INVITED
Modeling Coastal Morphodynamics using Local Estimates of Alongshore Sediment Transport: Limits and Alternatives
Two- and three-dimensional hydrodynamics-based models for simulating wave-driven nearshore flows, sediment transport, and sea-bed evolution have reached a level of development where their application to shoreline change prediction can now be considered. Presently their utility for this purpose is far from established, and far simpler and much less computationally expensive methods for estimating alongshore sediment transport gradients and subsequent shoreline change (e.g., one-line models) continue to be widely used for both scientific investigations and coastal engineering applications. These methods include empirical relations for the total alongshore sediment transport, such as the CERC equation, and 1D quasi processes- based integrations of alongshore transport on cross-shore transects, such as employed by the UNIBEST and LITPACK systems. These approaches can be classified together as "local" models, in that the transport at each location along the coast is found independently of other locations with the assumption of a quasi- uniform coast. Recent model comparisons have demonstrated that these local approaches often predict alongshore variations in transport that differ greatly from similar estimates made by 2D and 3D hydrodynamics-based models. This is especially true when offshore irregularities in bathymetry, such as shoals or depressions, modify the incoming wave field. In such cases the hydrodynamics-based models demonstrate that alongshore-variable wave forcing results in alongshore advective flow accelerations which in turn govern the pattern of transport gradients and the resulting locations of shoreline erosion and accretion. The local approaches used by one-line models assume that such accelerations are small enough that they can be neglected. Using 2D and 3D hydrodynamics-based models (such as Delft3D and ROMS), we examine a range of test cases to determine the conditions under which advective flow accelerations are significant and should prevent the application of the local approaches. These include idealized cases with simulated bathymetric irregularities, an engineered field case of a deep borrow pit at Delray Beach Florida, and a natural case with a large ebb-tidal delta offshore from Ocean Beach, California. It is hoped that this work will accelerate the testing and improvement of hydrodynamics-based models for application to long-term shoreline change prediction.
OS34A-02 INVITED
Wave-driven Littoral Subcells in the Southern California Bight
An extensive network of directional wave buoys has been used to reconstruct a detailed, 5-year nearshore wave climate for the mainland coast of California from San Diego to Pt Conception. Resulting cumulative annual statistics of potential alongshore and cross-shore transport rates suggest that numerous wave-driven subcells exist within the larger submarine canyon-bounded littoral cell systems. Interestingly, the wave-driven subcells are typically bounded by well-known surfing sites (aka, potential erosion hotspots) and an adjacent "updrift" section of coastline where the net annual cumulative alongshore transport rate is near zero. These adjacent pairs of weak net alongshore transport and strong net offshore transport create wave-driven sinks that have significant implications for regional sediment management. The shelf bathymetry which creates the wave-driven subcells will be discussed, along with ongoing efforts to corroborate the existence of the subcells by monitoring nearshore sand elevations in San Diego County.
OS34A-03
The Spatial and Temporal Variability of a High-Energy Beach: Insight Gained From Over 50 High-Resolution Sub-aerial Surveys
Since April 2004 a monitoring program of 7 km-long Ocean Beach, San Francisco, CA, has led to the completion of 55 Global Positioning System topographic surveys of the sub-aerial beach. The four-year timeseries contains over 1 million beach elevation measurements and documents detailed changes of the beach over a variety of spatial, temporal, and physical forcing scales. The goal of this ongoing data collection is to understand the variability in beach response as a function of wave forcing and offshore morphology which will ultimately aid in sediment management and erosion mitigation efforts. Several statistical methods are used to describe and account for the observed beach change, including empirical orthogonal functions (EOFs) and linear regression. Results from the EOF analysis show that the first mode, and approximately 50% of the observed variance of either the mean high water (MHW) or mean sea level (MSL) position, is explained by the seasonal movement of sediment on and offshore. The second mode, and approximately 15% of the variance, is dominated by alongshore variability, possibly corresponding to the position of cusps and embayments. Higher level modes become increasingly variable in the alongshore direction and each explain little of the observed variance. In both cases the first temporal mode is well correlated (R2~=0.7) with offshore significant wave height averaged over the previous 80 to 110 days, suggesting that seasonal wave height variations are the primary driver of intra-annual shoreline position. No other modes exhibit good correlation with offshore wave parameters regardless of the averaging time. The observed seasonal change is superimposed on a longer term trend of net annual accretion at the north end of Ocean Beach and erosion at the south end. Areas at the northern end have seen as much as 60 m of cumulative shoreline progradation since 2004, while some areas of the southern portion have retrograded nearly as much. This pattern shows an overall rotation in the shoreline position hinged on a nodal point in the middle portion of the beach. The nodal point roughly corresponds to the location where the southern lobe of a large ebb tidal delta welds to the coast, suggesting that wave conditions coupled with offshore morphologic variability is a key control on short and long-term beach evolution.
OS34A-04
Optimum wave height and runup parameterization for use during hurricane conditions
Barrier-island response to hurricanes is driven not only by storm surge, but also by wave-induced runup. During extreme storms, the 2% exceedence elevation of wave runup can be of equal or greater magnitude than surge levels, making runup a major factor in storm-induced coastal topographic and bathymetric change. Empirical formulations of runup show a dependence on locally measured, shallow-water wave heights that have been reverse-shoaled to deep water. This transformation supported intercomparisons of runup observations from a variety of scenarios. In order to improve the predictive skill of parameterized models and apply the results to predictions of coastal morphologic evolution, we use hydrodynamic models that provide a detailed description of wave conditions in the nearshore as input for the runup parameterizations. Comparisons of parameterized runup predictions to video-based observations of runup during Hurricane Isabel (September 2003) were used to examine the optimum wave height for use in empirical formulations. Nearshore wave and surge conditions were modeled using Delft3D and verified through comparison to observations from local buoys and wave gauges. Several measures of local wave height were extracted from the model and used to drive the runup parameterization, including (1) pre-breaking wave heights obtained just outside the surf zone, (2) a spectrally weighted wave height that is related to runup, and (3) a reverse- shoaled version of each of these. Clarification of wave-height input for runup, setup, and swash parameterizations will allow greater applicability of the equations, particularly during storms when wave heights evolve significantly across the shelf and in the nearshore.
OS34A-05
Modeling Wave Overtopping on the Chandeleur Islands during Hurricane Katrina using XBeach
Tropical cyclones that enter or form in the Gulf of Mexico generate storm surge and large waves that impact low-lying coastlines of along the Gulf Coast. Much of the Gulf Coast is ringed with barrier islands that provide inland marshes and the mainland some protection from storm events. The Chandeleur Islands, are located 161 km east of New Orleans, Louisiana and are oriented from north to south, and act to dissipate some of this energy. After a series of major storm events between 2001 and 2005, Hurricane Katrina's devastation in the fall of 2005 was particularly violent, destroying two-thirds of the area associated with the island chain. We would like to evaluate the predictability of hurricane-induced barrier island erosion and accretion. We test the ability of a time-dependent hydrodynamic and morphodynamic model, XBeach, to predict the impact of Hurricane Katrina on portions of Chandeleur Islands. Pre-storm LIDAR-derived bathymetry/topography and surge and wave data were used to drive a number of XBeach simulations. Model-predicted morphology was compared to post-storm LIDAR data. The accuracy of these predictions, including model sensitivity tests with varying grid size and temporal resolutions, are presented.
OS34A-06
Testable Predictions for Large-Scale Coastline-Shape Change in Response to Changing Storm Climate
Recent modeling (Ashton et al. 2001; Ashton and Murray, 2006a) and observations (Ashton and Murray 2006b) suggest that sandy coastlines self-organize into large-scale, plan-view shapes that depend sensitively on the regional wave climate—the distribution of influences on alongshore sediment transport from different deep-water wave-approach angles. Subsequent modeling (Slott et al., 2007) shows that even moderate changes in wave climate, as are likely to arise as storm behaviors shift in the coming century, will cause coastlines to change shape rapidly, compared to a steady-wave-climate scenario. Such large-scale shape changes involve greatly accentuated rates of local erosion, and highly variable erosion/accretion rates. A recent analysis of wave records from the Southeastern US (Komar and Allen, 2007) indicates that wave climates have already been changing over the past three decades; the heights of waves attributable to tropical storms have been increasing, changing the angular distribution of wave influences. Modeling based on these observations leads to predictions of how coastlines in this region should already be changing shape (McNamara et al., in prep.). As a case study, we are examining historical shorelines for the Carolina coastline, to test whether the predicted alongshore patterns of shoreline change can already be detected.
OS34A-07
Observations of bottom boundary-layer dynamics at the edge of a sorted grain-size feature on the inner shelf
Sorted grain-size features (SGFs) are common on sandy inner shelves, and they may influence circulation through their affect on topography, ripple distribution, and bottom roughness. We made measurements from two tripods at the edge of a SGF to better understand the mechanisms that maintain these features and to examine their affect on ripple geometry and hydraulic bottom roughness. The measurements were made at 12-m depth on the inner shelf near the Martha's Vineyard Coastal Observatory, Massachusetts. There, SGFs are characterized by bathymetric undulations with amplitude of ~0.5 m, with alternating patches of coarse (0.5 mm) sand with large ripples (heights of 0.10-0.15 cm, wavelengths of 0.6 to 0.8 m) and fine (0.125 mm) sand with hummocky topography and small ripples (heights of 0.01 m, wavelengths of 0.1 m). The features extend from the shallowest region surveyed (~6 m) to depths of ~17 m about 3 km offshore, with a maximum alongshore width of less than 1 km. Our tripods were deployed about 10 m apart along the north-south trending boundary between a coarse patch (east) and a fine patch (west) for 10 weeks. At this location, tidal currents and episodic wind-driven flows are predominantly east-west, and waves approach from the south. Instruments on the northern tripod provided sonar images of ripple geometry on both sides of the coarse/fine boundary, and optical measurements of the size and concentration of suspended sediments. Flow measurements were from the southern tripod with an upward-looking 1200 KHz acoustic Doppler profiler, three acoustic Doppler velocimeters, and a downward-looking 1500 KHz pulse- coherent acoustic Doppler profiler. We determined shear (using the log-profile method), dissipation rate (using the spectral method), and Reynolds stress (using the covariance method) at ~0.4 meters above the bottom. To the best of our knowledge, these are the first bottom boundary-layer measurements that allow direct comparison between flow properties over coarse sand with large ripples and fine sand with small ripples using the same instruments. Surprisingly, there was no discernable difference in bottom drag between eastward and westward flow when waves were small, and average drag coefficient C100 was the same (0.0026) in both directions. However, when waves were larger (near-bottom orbital velocities >10 cm/s), the drag coefficient increased significantly (0.005) when flow was westward from the coarse patch, but remained virtually unchanged (0.003) when flow was eastward from the fine patch. This supports the hypothesis that flow is more turbulent over the rough patch, which would tend to remove fine sand. In addition, these data confirm earlier observations that roughness tends to decrease at higher flow speeds.
OS34A-08
A combined observation-modeling approach for estimating water and suspended- sediment flux through a large tidal inlet: the Golden Gate, San Francisco, USA
Knowledge of water and sediment flux at the interface between estuaries and the ocean is needed to develop sediment budgets in support of estuarine and coastal zone morphology, ecology, and water quality studies. Humans have dramatically altered the delivery rate of sediment from watersheds, which, combined with ongoing climate change and sea-level rise, will likely continue to affect the morphology of estuaries and the coastal zone. Monitoring these fluxes, however, is a challenge owing to tidal forcing and freshwater pulses that occur over short time scales, and, in some cases, because of the sheer size of the channels. For the Golden Gate, which links San Francisco Bay to the Pacific Ocean, the size and depth of the channel as well as the tidal prism of approximately 2 billion cubic meters precludes the use of standard sampling techniques for measuring the suspended-sediment flux. That is, the amount of time that it would take to collect representative samples across the channel is equal to or greater than the time scale over which conditions are changing due to the tides. To overcome this, we employed an approach that combines observations with computational modeling. We made relatively rapid measurements of water and sediment flux using boat- mounted acoustic Doppler current profilers (ADCP), utilizing empirical calibrations between backscatter intensity and sediment concentration. Each ADCP transect, which results in concurrent water and suspended-sediment flux measurements, took about 15-20 minutes depending on transect location. In contrast, the conventional approach would comprise an ADCP transect for water flux plus depth-integrated suspended-sediment samples at multiple locations across the channel. When we conducted instrumentation drops at 7 locations across the channel, these transects typically took between 1-1.5 hours. The periodic ADCP-based flux measurements were then used to calibrate and test a three-dimensional computational model of San Francisco Bay and the coastal ocean. The model was subsequently used to fill in gaps between the measurements, thus providing flux estimates at the temporal resolution of the model. Another major advantage of the modeling approach is that it allows for estimating synoptic total fluxes during past and future time periods, provided that boundary conditions are known or can be estimated. This type of combined observation-modeling approach allows us to overcome the challenges posed by the tidal time scale and spatial scale of the Golden Gate, and similar approaches should prove useful at other tidal inlets.