Nearshore Buoy Research Articles

An assessment of the impact of boundary conditions in dynamical downscaling techniques for fetch-limited waves

ABSTRACT The handling of boundary conditions (BCs) is a key issue for the development of dynamical downscaling techniques in which nearshore wave models require the definition of the sea conditions at the offshore boundaries. This study addresses the implications of using simplified BCs represented by bulk or partitioned wave parameters with respect to using the complete information from directional wave spectra. As a large-scale model, we used WAVEWATCH III (WW3) running over the Mediterranean Sea providing the wave BCs for the small-scale SWAN model. One year of measured directional wave spectra from three deepwater buoys allowed the assessment of the WW3 model accuracy. The nested SWAN model simulations propagated the wave field over three nearshore areas. The accuracy of SWAN simulations was assessed through the comparison of computed results with wave measurements from nearshore buoys. Being the wave conditions dominated by unimodal seas, modest differences of error metrics are found between the datasets forced by different types of BCs. Conversely, focusing the attention on multimodal seas, an accurate definition of wave BCs achieved by the use of the directional wave spectra or, in alternative, by the reconstruction of wave spectra from partitioned wave parameters becomes crucial for nearshore wave predictions.

Kratkoročna prognoza značajnih valnih visina u zaštićenim akvatorijima koristeći se strojnim učenjem i Copernicus bazom podataka

Long-term time series of wave parameters play a critical role in coastal structure design and maritime activities. At sites with limited buoy measurements, methods are used to extend the available time series data. To date, wave hindcasting research using machine learning methods has mainly focused on filling in missing buoy measurements or finding a mapping function between two nearshore buoy locations. This work aims to implement machine learning methods for hindcasting wave parameters using only publicly available Copernicus data. Ensemble regression and artificial neural networks were used as machine learning methods and the optimal hyperparameters were determined by the Bayesian optimization algorithm. As inputs, data from the MEDSEA reanalysis wave model were used for the wave parameters and data from the ERA5 atmospheric reanalysis model were used for the wind parameters. The results of this study show that the normalized RMSE of the test data improved by 29% for Rijeka and 12% for Split compared to the original MEDSEA wave hindcast at buoy locations. The proposed method was extremely efficient in removing bias in the original MEDSEA hindcasts (e.g., NBIAS = -0.35 for Rijeka) to negligible values for both Split and Rijeka (NBIAS < 0.03).

CLASI: Coordinating Innovative Observations and Modeling to Improve Coastal Environmental Prediction Systems

Abstract The Coastal Land–Air–Sea Interaction (CLASI) project aims to develop new “coast-aware” atmospheric boundary and surface layer parameterizations that represent the complex land–sea transition region through innovative observational and numerical modeling studies. The CLASI field effort involves an extensive array of more than 40 land- and ocean-based moorings and towers deployed within varying coastal domains, including sandy, rocky, urban, and mountainous shorelines. Eight Air–Sea Interaction Spar (ASIS) buoys are positioned within the coastal and nearshore zone, the largest and most concentrated deployment of this unique, established measurement platform. Additionally, an array of novel nearshore buoys and a network of land-based surface flux towers are complemented by spatial sampling from aircraft, shore-based radars, drones, and satellites. CLASI also incorporates unique electromagnetic wave (EM) propagation measurements using a coherent array, drone receiver, and a marine radar to understand evaporation duct variability in the coastal zone. The goal of CLASI is to provide a rich dataset for validation of coupled, data assimilating large-eddy simulations (LES) and the Navy’s Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS). CLASI observes four distinct coastal regimes within Monterey Bay, California (MB). By coordinating observations with COAMPS and LES simulations, the CLASI efforts will result in enhanced understanding of coastal physical processes and their representation in numerical weather prediction (NWP) models tailored to the coastal transition region. CLASI will also render a rich dataset for model evaluation and testing in support of future improvements to operational forecast models.

Estimating Nearshore Waves by Assimilating Buoy Directional Spectrum Data in SWAN

AbstractThis paper describes a variational data assimilation algorithm based on the Simulating WAves Nearshore (SWAN) wave spectrum model. The approach allows single-point wave spectrum observations to be used to estimate the wave field for a nearshore region under stationary conditions, assuming a spatially uniform incident wave spectrum at the offshore boundary. The assimilated data are in the form of Fourier directional coefficients, the standard output from operational wave buoys, and are used directly by incorporating the relationship between directional spectrum and the Fourier coefficients into the formulation. The algorithm was tested on data from nearshore buoys deployed off the coast of North Carolina in May 2012, and the estimated wave field is compared to both the input data and independent observation data. The results compare favorably to the independent data with overall RMS errors of 10%–20% for significant wave height, about half a second for mean wave period, and as much as three to four SWAN spectral grid cells for mean direction. Overall, the results show that the algorithm can be effectively used to estimate the offshore boundary spectrum and accurately reproduce wave conditions in the domain.

The impact of ERA-Interim winds on wave generation model performance in the Southern Caspian Sea region

The aim of the present work is to evaluate wind dataset in generation of the wave characteristic along the southern basin of the Caspian Sea (SBC). To achieve this purpose, satellite altimetry data, numerical model results and field measurements were considered for 2011. The mixed re-analysis/forecast wind data from European Center for Medium Range Weather Forecast, ERAI_MIXED, are analyzed. Comparisons are made using altimetry data from CERSAT, mixed re-analysis/forecast GFS datasets (GFS_MIXED) and field observation from two buoys and one ADCP located in SBC. The modeling was carried out using the SWAN model over a composite-overlap nested grid system. The results showed that the ERAI_MIXED underestimated wind data in SBC, similar to other datasets from ECMWF and GFS. However, the simulation results implied that using the same configuration for the wave model, ERAI_MIXED led to superior results compared to GFS_MIXED and ERAI_MIXED-COR. Comparison with limited observation showed that ERAI_MIXED data slightly overestimated the wave height in the offshore station and underestimated it at the nearshore buoy. Finally, model results showed that moderate to low wind speed over the SBC might generate the waves up to 4 m wave height.

Modeling Long-Period Swell in Southern California: Practical Boundary Conditions from Buoy Observations and Global Wave Model Predictions

AbstractAccurate, unbiased, high-resolution (in space and time) nearshore wave predictions are needed to drive models of beach erosion; coastal flooding; and alongshore transport of sediment, biota, and pollutants. On sheltered shorelines, wave predictions are sensitive to the directions of onshore propagating waves, and nearshore model prediction error is often dominated by directional uncertainty offshore. Here, regional wave model skill in highly sheltered Southern California is compared for different offshore boundary conditions created from offshore buoy observations and global wave model hindcasts [NOAA WaveWatch III (WW3)]. Spectral ray-tracing methods are used to transform incident offshore swell (0.04–0.09 Hz) energy at high directional resolution (1°). Model skill is assessed for predictions (wave height, direction, directional spread, and alongshore radiation stress) at 16 nearshore buoy sites between 2000 and 2009. Buoy-derived boundary conditions using various estimators (maximum entropy, maximum smoothness) have similar skill and all outperform WW3-derived boundary conditions. A new method for estimating offshore boundary conditions, CMB-ADJ, combines buoy observations with WW3 predictions. Although CMB-ADJ skill is comparable to buoy-only methods, it may be more robust in varying regions and wave climatologies, and will benefit from future improvements in global wave model (GWM) predictions. A case study at Oceanside Harbor shows strong sensitivity of alongshore sediment transport estimates to the boundary condition method. However, patterns in alongshore gradients of transport (e.g., the location of model accretion and erosion zones) are similar across methods. Weak, tidally modulated coastal reflection is evident in both shallow and deep buoy observations, and significantly increases the observed directional spread.

The California coastal wave monitoring and prediction system

A decade-long effort to estimate nearshore (20m depth) wave conditions based on offshore buoy observations along the California coast is described. Offshore, deep water directional wave buoys are used to initialize a non-stationary, linear, spectral refraction wave model. Model hindcasts of spectral parameters commonly used in nearshore process studies and engineering design are validated against nearshore buoy observations seaward of the surfzone. The buoy-driven wave model shows significant skill at most validation sites, but prediction errors for individual swell or sea events can be large. Model skill is high in north San Diego County, and low in the Santa Barbara Channel and along the southern Monterey Bay coast. Overall, the buoy-driven model hindcasts have relatively low bias and therefore are best suited for quantifying mean (e.g. monthly or annual) nearshore wave climate conditions rather than extreme or individual wave events. Model error correlation with the incident offshore wave energy, and between neighboring validation sites, may be useful in identifying sources of regional modeling errors.

WindSat satellite comparisons with nearshore buoy wind data near the U.S. west and east coasts

Nearshore wind speeds retrieved by WindSat are validated by a comparison with the moored buoy observations near the U.S. west and east coasts. A 30 min and 25 km collection window is used for the WindSat wind data and buoy measurements from January 2004 to December 2014. Comparisons show that the overall root-mean-square error is better than 1.44 m/s near the U.S. coasts, and the result for the east coast is better than that for the west coast. The retrieval accuracy of the descending portions is slightly better than that of the ascending portions. Most buoy-to-buoy variations are not significantly correlated with the coastal topography, the longitude and the distance from the shore or satellite-buoy separation distance. In addition, comparisons between a polarimetric microwave radiometer and a microwave scatterometer are accomplished with the nearshore buoy observations from 2007 to 2008. The WindSat-derived winds tend to be lower than the buoy observations near the U.S. coasts. In contrast, the QuikSCAT-derived winds tend to be higher than the buoy observations. Overall, the retrieval accuracy of WindSat is slightly better than that of QuikSCAT, and these satellite-derived winds are sufficiently accurate for scientific studies.

Wave hindcast studies using SWAN nested in WAVEWATCH III - comparison with measured nearshore buoy data off Karwar, eastern Arabian Sea

Waves in the nearshore waters (~15m water-depth) of eastern Arabian Sea were simulated using SWAN nested in WAVEWATCH III (WW3) for the year 2014. The sensitivity of the numerical wave model WW3 towards different source term (ST) packages was tested. The performance of WW3 was evaluated with the measured deep water buoy data at 15.0000° N, 69.0000° E. Overall, the WW3 wave hindcast results using ST4 physics in deep water show a reasonable match (r=0.97 and SI=0.16) with the measured significant wave height (Hs). In deep water, a large overestimation (~23.7%) of mean wave period was observed using ST2 compared to ~8.2% using ST4. Komen et al. (1984) whitecapping formulation was found suitable based on the sensitivity studies of SWAN. Simulated Hs using nested SWAN model shows good correlation (r=0.96) with the nearshore measured data. During high sea state (Hs>2m), an overestimation (~19.8% for ST2 and ~21.4% for ST4) and during low sea state (Hs<1m), an underestimation (~16.1% for ST2 and ~31.5% for ST4) in mean wave period was observed in the nearshore. WW3 model forced by modified wind field (speed increased by 5%) during June and July resulted in a reduction in underestimation of Hs from 8.6% to 2.2%.

Thirty-four years of Hawaii wave hindcast from downscaling of climate forecast system reanalysis

The complex wave climate of Hawaii includes a mix of seasonal swells and wind waves from all directions across the Pacific. Numerical hindcasting from surface winds provides essential space-time information to complement buoy and satellite observations for studies of the marine environment. We utilize WAVEWATCH III and SWAN (Simulating WAves Nearshore) in a nested grid system to model basin-wide processes as well as high-resolution wave conditions around the Hawaiian Islands from 1979 to 2013. The wind forcing includes the Climate Forecast System Reanalysis (CFSR) for the globe and downscaled regional winds from the Weather Research and Forecasting (WRF) model. Long-term in-situ buoy measurements and remotely-sensed wind speeds and wave heights allow thorough assessment of the modeling approach and data products for practical application. The high-resolution WRF winds, which include orographic and land-surface effects, are validated with QuickSCAT observations from 2000 to 2009. The wave hindcast reproduces the spatial patterns of swell and wind wave events detected by altimeters on multiple platforms between 1991 and 2009 as well as the seasonal variations recorded at 16 offshore and nearshore buoys around the Hawaiian Islands from 1979 to 2013. The hindcast captures heightened seas in interisland channels and around prominent headlands, but tends to overestimate the heights of approaching northwest swells and give lower estimates in sheltered areas. The validated high-resolution hindcast sets a baseline for future improvement of spectral wave models.

The Use of a Statistical Model of Storm Surge as a Bias Correction for Dynamical Surge Models and its Applicability along the U.S. East Coast

The present study extends the applicability of a statistical model for prediction of storm surge originally developed for The Battery, NY in two ways: I. the statistical model is used as a biascorrection for operationally produced dynamical surge forecasts, and II. the statistical model is applied to the region of the east coast of the U.S. susceptible to winter extratropical storms. The statistical prediction is based on a regression relation between the “storm maximum” storm surge and the storm composite significant wave height predicted ata nearby location. The use of the statistical surge prediction as an alternative bias correction for the National Oceanic and Atmospheric Administration (NOAA) operational storm surge forecasts is shownhere to be statistically equivalent to the existing bias correctiontechnique and potentially applicable for much longer forecast lead times as well as for storm surge climate prediction. Applying the statistical model to locations along the east coast shows that the regression relation can be “trained” with data from tide gauge measurements and near-shore buoys along the coast from North Carolina to Maine, and that it provides accurate estimates of storm surge.

EVALUATION OF THE ECMWF ERA-INTERIM WIND DATA FOR NUMERICAL WAVE HIND-CASTING IN THE CASPIAN SEA

One of the most important factors in design of coastal and marine structures is the wind-induced wave characteristics. Hence, an accurate estimation of wave parameters is considerably important. In this paper, SWAN third generation spectral models with unstructured mesh have been used for the prediction of wave parameters in the Caspian Sea. The new reanalysis wind datasets provided by ECMWF which is known as ERA-Interim have been used as wind forcing. Significant wave height (Hs), peak spectral period (Tp) and mean wave direction were hindcasted in the study. The field dataset consist of buoy measurements in two offshore and one nearshore stations has been used for evaluating the performance of the model. In addition, the predicted significant wave height has been compared with satellite observations in some paths. The results show that the simulations follow the wave climate trend quite well. All of the error measures for Hs are in the perfect range for the offshore buoy locations. For nearshore buoy, the model is slightly underestimates wave peaks, but still performed well for the peak spectral period. The comparison of the results with satellite paths also shows that the model can predict wave height accurately in the whole area of the Caspian Sea.

Did a submarine landslide contribute to the 2011 Tohoku tsunami?

Many studies have modeled the Tohoku tsunami of March 11, 2011 as being due entirely to slip on an earthquake fault, but the following discrepancies suggest that further research is warranted. (1) Published models of tsunami propagation and coastal impact underpredict the observed runup heights of up to 40 m measured along the coast of the Sanriku district in the northeast part of Honshu Island. (2) Published models cannot reproduce the timing and high-frequency content of tsunami waves recorded at three nearshore buoys off Sanriku, nor the timing and dispersion properties of the waveforms at offshore DART buoy #21418. (3) The rupture centroids obtained by tsunami inversions are biased about 60 km NNE of that obtained by the Global CMT Project. Based on an analysis of seismic and geodetic data, together with recorded tsunami waveforms, we propose that, while the primary source of the tsunami was the vertical displacement of the seafloor due to the earthquake, an additional tsunami source is also required. We infer the location of the proposed additional source based on an analysis of the travel times of higher-frequency tsunami waves observed at nearshore buoys. We further propose that the most likely additional tsunami source was a submarine mass failure (SMF—i.e., a submarine landslide). A comparison of pre- and post-tsunami bathymetric surveys reveals tens of meters of vertical seafloor movement at the proposed SMF location, and a slope stability analysis confirms that the horizontal acceleration from the earthquake was sufficient to trigger an SMF. Forward modeling of the tsunami generated by a combination of the earthquake and the SMF reproduces the recorded on-, near- and offshore tsunami observations well, particularly the high-frequency component of the tsunami waves off Sanriku, which were not well simulated by previous models. The conclusion that a significant part of the 2011 Tohoku tsunami was generated by an SMF source has important implications for estimates of tsunami hazard in the Tohoku region as well as in other tectonically similar regions.

Comparison of gridded multi-mission and along-track mono-mission satellite altimetry wave heights with in situ near-shore buoy data

The applicability of altimeter data for the coastal region is examined by comparing the gridded multi-mission and along-track mono-mission significant wave height (SWH) data with the in situ buoy measurements at four stations off the east and west coasts of India. Among all of the satellites, a greater number of collocated points and better correlation were found for Jason-2. A comparison of the SWH data from multi-mission products (Jason-1, Jason-2, Envisat, and ERS-2) obtained from AVISO with the measured buoy data shows that all of the satellites overestimated the SWH. Applying a correction factor of 0.75, 0.70 and 0.69 to the altimeter data for stations 1, 2 and 4, respectively, resulted in close agreement with the buoy data (correlation coefficient ~0.9–0.96). Monthly averaged buoy and altimeter SWH values show convincing correlation ranging from 0.97 to 0.98 for three stations. The smallest correlation coefficient (0.6) was found at station 3, where the selected grid points are located close to the coast and are influenced by land contamination and wave decay effects. For waves with a SWH<0.5m, the correlation coefficient is sparingly low for all stations, which underlines the fact that the altimeter observations for SWH<0.5m are unreliable.

Performance and validation of a coupled parallel ADCIRC–SWAN model for THANE cyclone in the Bay of Bengal

An accurate prediction of near-shore sea-state is imperative during extreme events such as cyclones required in an operational centre. The mutual interaction between physical processessuchastides,wavesandcurrentsdeterminethephysicalenvironmentforanycoastal region, and hence the need of a parallelized coupled wave and hydrodynamic model. The presentstudyisanapplicationofvariousstate-of-artmodelssuchasWRF,WAM,SWANand ADCIRCusedtocoupleandsimulateaseverecyclonicstorm ThanethatdevelopedintheBay ofBengalduringDecember2011.Thecoupledmodel(ADCIRC-SWAN)wasruninaparallel mode on a flexible unstructured mesh. Thane had its landfall on 30 December, 2011 between Cuddalore and Pondicherry where in-situ observations were available to validate model performance.Comprehensiveexperimentontheimpactofmeteorologicalforcingparameters with two forecasted tracks derived from WRF model, and JTWC best track on the overall performance of coupled model was assessed. Further an extensive validation experiment was performed for significant wave heights and surface currents during Thane event. The significant wave heights measured along satellite tracks by three satellites viz; ENVISAT, JASON-1 and JASON-2, as well in-situ near-shore buoy observation off Pondicherry was used for comparison with model results. In addition, qualitative validation was performed for modelcomputedcurrentswithHFRadarObservationoffCuddaloreduring Thaneevent. The importance of WRF atmospheric model during cyclones and its robustness in the coupled model performance is highlighted. This study signifies the importance of coupled parallel ADCIRC-SWAN model for operational needs during extreme events in the North Indian Ocean.

QuikSCAT Satellite Comparisons with Nearshore Buoy Wind Data off the U.S. West Coast

To determine the accuracy of nearshore winds from the QuikSCAT satellite, winds from three satellite datasets (scientifically processed swath, gridded near-real-time, and gridded science datasets) were compared to those from 12 nearshore and 3 offshore U.S. West Coast buoys. Satellite observations from August 1999 to December 2000 that were within 25 km and 30 min of each buoy were used. Comparisons showed that satellite–buoy wind differences near shore were larger than those offshore. Editing the satellite data by discarding observations recorded in rain and those recorded in light winds improved the accuracy of all three datasets. After removing rain-flagged data and wind speeds less than 3 m s−1, root-mean-squared differences (satellite minus buoy) for swath data, the best of the three datasets, were 1.4 m s−1 and 37° based on 5741 nearshore comparisons. By removing winds less than 6 m s−1, these differences were reduced to 1.3 m s−1 and 26°. At the three offshore buoys, the root-mean-squared differences for the swath data, with both rain and winds less than 6 m s−1 removed, were 1.0 m s−1 and 15° based on 1920 comparisons. Although the satellite's scientifically processed swath data near shore do not match buoy observations as closely as those offshore, they are sufficiently accurate for many coastal studies.

The near‐coastal microseism spectrum: Spatial and temporal wave climate relationships

Comparison of the ambient noise data recorded at near‐coastal ocean bottom and inland seismic stations at the Oregon coast with both offshore and nearshore buoy data shows that the near‐coastal microseism spectrum results primarily from nearshore gravity wave activity. Low double‐frequency (DF), microseism energy is observed at near‐coastal locations when seas nearby are calm, even when very energetic seas are present at buoys 500 km offshore. At wave periods &gt;8 s, shore reflection is the dominant source of opposing wave components for near‐coastal DF microseism generation, with the variation of DF microseism levels poorly correlated with local wind speed. Near‐coastal ocean bottom DF levels are consistently ∼20 dB higher than nearby DF levels on land, suggesting that Rayleigh/Stoneley waves with much of the mode energy propagating in the water column dominate the near‐coastal ocean bottom microseism spectrum. Monitoring the southward propagation of swell from an extreme storm concentrated at the Oregon coast shows that near‐coastal DF microseism levels are dominated by wave activity at the shoreline closest to the seismic station. Microseism attenuation estimates between on‐land near‐coastal stations and seismic stations ∼150 km inland indicate a zone of higher attenuation along the California coast between San Francisco and the Oregon border.