Datawell Waverider Research Articles

Error analysis of wave intercomparison based on sampling variability

Quantitative analysis of error sources in wave measurement intercomparison under real-world conditions remains challenging due to various factors such as spatial and temporal offsets in the field, diverse configurations of wave sensors and platforms, and distinct measurement algorithms. This study introduces a modified error analysis method based on sampling variability. Measurements from two co-deployed Datawell Waverider buoys at the National Marine Test Site (Weihai) in China offer an opportunity to assess the spatial variability error. The results reveal that for non-directional wave parameters, like significant wave height and mean zero-crossing period, the wave conditions in the field can be considered stationary, with random errors primarily arising from inherent sampling variability. Spatial offsets significantly impact directional wave parameters at peak frequencies, with large deviations (greater than 20°) in peak wave direction primarily attributed to bimodal distribution. For data with significant wave heights exceeding 0.5 m, random errors in the mean wave direction at the same frequency can be mainly attributed to sampling variability. The bulk directional wave parameters weighted by the energy spectrum demonstrate lower sensitivity to spatial offsets, and there is excellent agreement between the bulk mean directions when considering only the unimodal distribution data.

Performance of moored GPS wave buoys

ABSTRACT Small, lightweight, and easy-to-deploy Global Positioning Satellite (GPS)-based miniature wave buoys (mini-buoys) are increasing in popularity. Most are deployed as Lagrangian drifters, so their utility for engineering applications, requiring a nearshore mooring, has remained in question. We evaluated the performance of five moored mini-buoys off the coast of Duck, North Carolina at Field Research Facility (FRF). We collected 8 months of data over two deployments; first near the FRF’s 8-m array and then near a Datawell Waverider in 17-m depth. We deployed an Acoustic Doppler Current Profiler (ADCP) that measured waves and currents, and compared it to the nearby operational system as a performance benchmark for the mini-buoys. The mini-buoys performed well, with statistics similar to the benchmark for wave frequency spectra and spectral parameters, and they outperformed the benchmark on directional parameters. However, mini-buoy communication was sometimes interrupted by tilting or submersion, which manifested as low-frequency noise or drop-outs. If a mooring was taut, forced by winds or currents and exacerbated by tangled lines or shallower depths, then mini-buoys were more susceptible to communication issues. Overall, these issues were rare; thus, mini-buoys were found to be appropriate for many applications.

Analysis of data from the full-scale prototype testing of the WASP – A novel wave measuring buoy.

To assess the viability of locations for wave energy farms, and design effective coastal protection measures, knowledge of local wave regimes is required. Current regime-measuring devices are expensive, and the aim of the WASP Project is to develop a low-cost, self-powering wave-measuring device. The Wave-Activated Sensor Power Buoy (WASP) comprises a floating body with a moonpool. The relative motion of the water level in the moonpool to the buoy will pressurise and depressurise the air above the water column, which will be used to drive a bidirectional turbine in the manner of an oscillating water column, which will operate in conjunction with a generator to recharge an on-board battery pack. It has previously been demonstrated through tank testing of scale-models that a sea-state may be estimated from measurements of the pressure within the air above the water column under controlled conditions [1]. Important statistical parameters relating to the sea state, such as the significant wave height, zero-cross period et cetera. may then be estimated from the spectral moments of the wave spectrum [2].
 This paper seeks to validate the use of measurements taken of the pressure in a volume of air contained in a chamber above an oscillating water column (OWC) (which is in communication with the ocean) to estimate the sea state which induces motion of the water column.
 The WASP was deployed from March to July 2019 at the Smart Bay test facility in Galway Bay, Ireland. The ¼ Scale test site was previously characterised by B Cahill UCC 2011 [3] using data captured by the Marine Institute Datawell WaveRider buoy from 2006 to 2011. Pressure signal data recorded by the WASP at a frequency of 8Hz was processed in to 30-minutes blocks to reflect the presented data from the WaveRider buoy, from which the WASP was moored a distance of 400m from during deployment. Power Density Spectra for 6,500 thirty-minute sea states were generated and an inverse transfer function between the WASP and the WaveRider produced. This allowed for the calculation of the associated significant wave height, Hs, and zero-crossing period, Tz, from the WASP pressure data for each thirty-minute block of data. At this time, outliers and bimodal spectra were removed to simplify the analysis, and the data was binned based on Hs and Tz ranges. For each bin, average inverse transfer functions were generated. A Bretschneider Spectrum [4] was produced using the Hs and Tz values from the WaveRider data for the respective binned WASP data. An average transfer function was then generated between the Bretschneider models and the corresponding sea states for the WASP. Using this final transfer function, a scatter plot of Hs and Tz for the test site was generated and compared against the previously produced scatter plot from the initial resource characterisation of the Galway Bay Test Site in 2011 in order to determine if the WASP is capable of measuring sea-states through measurement of time series pressure signals in the volume of air above the water column in the OWC.

DISPLACEMENT BASED COMPARISON OF ACCELEROMETER AND LOW-COST GNSS WAVE BUOYS

Wave observations are critical in both coastal and deep water and feed into applications including improving ocean forecasting, industry activities, and marine safety. Wave buoys are typically the most robust method to collect in situ wave observations, particularly when real-time data is required, however they have historically been expensive to purchase and operate. With the advent of low-cost and small-format Global Navigation Satellite System (GNSS) based wave buoys (e.g. Sofar Spotter), the ability to collect wave observations across a wider range of locations is feasible at a much lower cost. However, understanding the performance and behaviors of small-format GNSS buoys (e.g. spectral response and mooring effects), against more traditional accelerometer based platforms, is critical to inform their use across research and industry applications. Here, we focus on comparing the Sofar Spotter to Datawell Waverider (Mark 3 and 4) buoys which are commonly deployed in operational networks globally. However, instead of comparing processed products (e.g. spectral statistics) produced by the buoy’s onboard software, we begin our analysis with the buoy displacement time series directly to ensure consistency. We also assess the detailed spectral properties/moments, the GNSS buoy performance in different conditions (locations), and with different mooring configurations.

Comparative Performance of Radar, Laser, and Waverider Buoy Measurements of Ocean Waves. Part I: Frequency Domain Analysis

Abstract Accurate ocean wave measurements are needed for the safe design and operation of offshore facilities, but despite many ocean wave measurements, the accuracy of wave measurement systems remains an ongoing issue. Of paramount importance are measurements during extreme sea states. This paper examines wave measurements made with an Optech Laser (Laser), a Rosemount WaveRadar (Radar), and a Datawell Waverider buoy at North Rankin A (NRA) platform, Australia; Ekofisk, North Sea; and several South China Sea locations. We evaluate the relative performance of these instruments based upon various frequency domain comparisons, including comparisons of their 1D frequency spectra using spectrograms, spectral moments, high-frequency tail slopes, and significant wave heights derived from their wave spectra. A spectral relationship (transfer function) in terms of mean spectral ratio of the instruments is developed, which can be used for spectral calibration. On average, Laser and Waverider spectral estimates agree well at all sea states. However, at low wind speeds, the higher-frequency spectral levels of the Laser are relatively high and noisy compared with the other two instruments. Radar higher-frequency spectral estimates are relatively low compared to the other two instruments, particularly at lower sea states. In addition, the higher-frequency tail slopes of all three instruments vary between f−4 and f−5. However, at higher sea states, the Waverider tail slopes become steeper than f−5. The Radar produces the lowest significant wave heights (Hm0) compared to the Laser and Waverider, but its second-moment period (Tm02) estimates are longer than the Laser and Waverider.

Spectral shapes and parameters from three different wave sensors

The quality of wave measurements is of primary importance for the validation of wave forecasting models, satellite wave calibration and validation, wave physics, offshore operations and design and climate monitoring. Validation of global wave forecasts revealed significant regional differences, which were linked to the different wave buoy systems used by different countries. To fully understand the differences between the wave measurement systems, it is necessary to go beyond investigations of the integral wave parameters height, period and direction, into the frequency spectra and the four directional Fourier parameters that are used to estimate the directional distribution. We here analyse wave data measured from three different sensors (non-directional Datawell Waverider buoy, WaveRadar Rex, Optech laser) operating at the Ekofisk oil production platform located in the central North Sea over a period of several months, with significant wave height ranging from 1 to 10 m. In general, all three sensors provide similar measurements of the integral wave properties and frequency spectra, although there are some significant differences which could impact design and operations, forecast verification and climate monitoring. For example, the radar underestimates energy in frequency bands higher than 8 s by 3–5%, swell (12.5–16 s) by 5–13%, while the laser has 1–2% more energy than the Waverider in the most energetic bands. Lee effects of structures are also estimated. Lower energy at the frequency tail with the radar has an effect on wave periods (they are higher); wave steepness is seen to be reduced by 10% in the wind seas. Goda peakedness and the unidirectional Benjamin-Feir index are also examined for the three sensors.

Analysis of UHF-band sea clutter reflectivity at low grazing angles in offshore waters of the Yellow Sea

ABSTRACT Reflectivity is an important characteristic of sea clutter that depends on the parameters of the radar and the environment. This paper presents reflectivity measurements of UHF-band horizontal polarized sea clutter based upon data collected in the offshore waters of the Yellow Sea of China at low grazing angles. During the experiment, the oceanic and meteorological conditions were recorded by the Datawell Waverider 4 and Lufft WS700-UMB on the sea surface. The factors that affect the reflectivity acquisition are analysed and the selective median filtering method in average Doppler power spectrum is presented to remove the effect of ground clutter from the sidelobes of the antenna beam pattern and the normalized reflectivity of the sea surface is computed. The applicability of the four classical reflectivity empirical models to the datasets is analysed. The results show that the Hybrid (HYB) and Technology Service Corporation (TSC) models are suitable while the Georgia Institute of Technology (GIT) and Naval Research Laboratory (NRL) models are unsuitable for the UHF-band horizontal polarized sea clutter in the offshore waters of the Yellow Sea of China.

Measurements of Directional Wave Spectra and Wind Stress from a Wave Glider Autonomous Surface Vehicle

AbstractMethods for measuring waves and winds from a Wave Glider autonomous surface vehicle (ASV) are described and evaluated. The wave method utilizes the frequency spectra of orbital velocities measured by GPS, and the wind stress method utilizes the frequency spectra of turbulent wind fluctuations measured by an ultrasonic anemometer. Both methods evaluate contaminations from vehicle motion. The methods were evaluated with 68 days of data over a full range of open ocean conditions, in which wave heights varied from 1 to 8 m and wind speeds varied from 1 to 17 m s−1. Reference data were collected using additional sensors on board the vehicle. For the waves method, several additional datasets are included that use independently moored Datawell Waverider buoys as reference data. Bulk wave parameters are determined within 5% error with biases of less than 5%. Wind stress is determined within 4% error with 1% bias. Wave directional spectra also compare well, although the Wave Glider results have more spread at low frequencies.

Development of a high resolution wave climate modelling methodology for offshore, nearshore and onshore locations of interest

The aim of this investigation was to develop a methodology to characterise the wave climate for offshore, nearshore and onshore locations of interest. The proposed methodology was applied to a domain located off the west coast of Ireland, adjacent to Loop Head, Co. Clare, and this was an area of interest for ocean energy development. A 3rd generation spectral wave model for the domain was developed, calibrated and validated using DHI’s MIKE 21 spectral wave modelling software along with sea state, met-ocean and bathymetric data obtained from a wide range of sources.The model yielded information on a range of indices including annual mean wave power, exploitable wave power, maximum wave height, percentage occurrence of a range of sea states and wave directionality at off-shore, near-shore, and on-shore locations.The models predictions were compared with local wave measurements for calibration and validation purposes obtained from a non-directional Datawell Waverider wave measurement buoy. The wave climate model predicted significant wave height and energy period at this location exhibiting a scatter index of less than 0.25 and a correlation coefficient greater than 0.85. The inclusion of a high resolution bathymetry dataset in the on-shore and coastal regions of the domain was the main driver in achieving this level of agreement between the data.The spectral wave model developed in this research programme will be useful for future wave energy research. Results from this model indicate that the domain off County Clare possesses an exploitable resource with promising levels of energy for potential wave energy projects. They also show that the near-shore and on-shore environment in this domain maintains good levels of power with relative filtering of extreme waves.

Assessment of sea wave spectra using a surfaced glider

The determination of non-directional and directional sea wave spectra is attempted by analyzing the dynamical response of a surfaced Slocum glider. The method makes use of the glider heave motion to infer non-directional properties of the wave spectra. In addition, surge and sway motions are considered to derive wave directionality. The transfer functions for heave, surge and sway for a surfaced Slocum glider has been computed to determine the impact of the body geometry/inertia and the angle of incidence on the response of the surfaced platform when excited by regular waves. Numerical results show that for wave periods longer than 6s, motions measured at the glider platform can be assumed equivalent to motions experimented by the water parcels at the sea surface. Spectral information about sea state conditions can then be directly derived from the measurement of glider responses. Results also reveal a natural period of the surfaced glider at around 5s. A series of field experiments were conducted during March 7th, May 17th and May 27th of 2013 in a marine region off-shore La Spezia, to validate the methodology. Specifically, the wave spectra derived from a Slocum glider equipped with a set of accelerometers was compared against the values provided by a Datawell Waverider Mk3 moored in the vicinity of the glider deployment. Agreement has been found between the non-directional estimates from the glider and the values measured by the waverider. Regarding wave directionality, glider estimates of peak direction agree well with those reported by the waverider. More significant differences are found between the directional spread estimates.

Drifter observations of the effects of shoals and tidal-currents on wave evolution in San Francisco Bight

Detailed observations of wave evolution and wave–current interaction in tidal inlets and river mouths are practically non-existent. This is in part due to the practical difficulty of installing and maintaining fixed instruments in this harsh environment with large waves, strong currents, dynamic seabed morphology, and often busy ship traffic, but also due to the fact that it is difficult to resolve the spatial variability and evolution of the wave and current field from an array of point measurements. This work explores the use of newly developed small, free-drifting buoys to collect wave and current measurements in a coastal inlet. The instruments, referred to as wave-resolving drifters (or WRD), are small and lightweight enough so that they can be deployed and retrieved from small vessels, and relatively inexpensive so that large numbers can be used. The surface-following drifters resolve the three-dimensional wave orbital motion and surface current field by combining GPS and accelerometer measurements. We validate the WRD platform and its sensor package in open ocean conditions in Monterey Bay by comparing the WRD observations to observations made by a collocated 40cm-diameter Datawell Waverider buoy. To study wave evolution in the San Francisco Bight, 30 WRDs are deployed near the San Francisco Bay entrance (Golden Gate) during peak ebb tide so that the drifters flow out of the bay, and into the incident wave field. Wave statistics are estimated through local ensemble averaging of drifter observations, and ensemble-averaged wave spectra are used to capture the wave evolution through the inlet area. Comparisons with numerical simulations of the simulating waves near shore (SWAN) model help identify the various processes acting on different frequency ranges of the wave field, and ray diagrams show the distinct effects of refraction by variable depth on the lower-frequency swells and refraction by currents on the higher-frequency wind waves. This combined analysis demonstrates the potential of using relatively inexpensive surface-following drifters to investigate surface dynamics in a complicated and energetic coastal inlet.

Marine radar ocean wave retrieval’s dependency on range and azimuth

The strength of the surface wave signal in marine X-band radar (MR) images strongly depends on range and azimuth (i.e., the angle between antenna look and peak wave direction). Traditionally, MR wave analysis is carried out in a set of rectangular windows covering the radar field of view (FOV). The FOV is typically partially obstructed, e.g., due to the coastline or ship superstructures. Especially for ships that are subject to regular course changes, this results in an increased variability or error associated with wave parameters. Using MR measurements from R/P FLIP, acquired off California during the 2010 US Office of Naval Research (ONR) high resolution air–sea interaction (Hi-Res) experiment, this study quantifies the dependency of the radar-based 2D wave spectrum and parameters on range and azimuth. With the help of reference data from a nearby Datawell Waverider buoy, we propose empirical methods to remove the dependency and we illustrate their efficacy.

Properties of the Umov vector in shallow water and its dependence on sea surface conditions

In this work, the effects of a rough sea surface on shallow water acoustic propagation are examined using experimental data collected during the ONR sponsored Target and Reverberation Experiment (TREX) off of the coast of Panama City, Florida, in May 2013. The acoustic data were collected from a bottom deployed recording tower that coherently recorded data on a horizontal line array (HLA), a vertical line array (VLA), and on an accelerometer-based vector sensor which was combined with a co-located hydrophone to formulate the Umov vector, or instantaneous intensity vector. The source was lowered from the stern of a research vessel to a depth one-third and two-thirds of the water depth (18 m), and transmitted a multi-frequency pulse from 1 to 4 kHz. These measurements were repeated at positions approximately 10, 20, and 40 water depths away from the tower, along a bearing perpendicular and parallel to the surface wave-crests. During the experiment, the sea surface directional-wave spectrum was measured by a Datawell Waverider buoy moored at the experimental site. Properties of the Umov vector are shown to relate to roughness and directional characteristics of the sea surface. The Umov vector is also studied in relation to the HLA and VLA measurements.

Sea state characterisation of the test site of an offshore wave energy plant

The Archimedes Wave Swing (AWS) pilot plant, rated at 2 MW, was tested offshore at Póvoa de Varzim in northern Portugal in 2004, gathering a large team of engineers and scientists. The Mechanical Engineering Department of Instituto Superior Técnico was involved, following previous work done for the project. The present paper describes one of the studies conducted. It aims to characterise the sea state at the actual location of the pilot plant using the available pressure sensors. Two approaches were performed: a first one purely based in linear wave theory, neglecting the presence of the device, and a second one, based on the results from a Boundary Element Method (BEM) code named AQUADYN, originally developed at École Centrale de Nantes (France), which allowed a detailed quantification of the effects of the presence of the plant on the wave profile directly above it. Comparisons with a Datawell Waverider buoy located at a certain distance from the plant also played an important role in the current study.

Waves at Holderness from X-band radar

During the Holderness Experiment of 1994–1995 and 1995–1996, data were collected by a coastal deployment of an X-band radar with simultaneous deployments of various moored and bottom-mounted wave measurement devices. The radar data consist of sequences of images of the sea surface at intervals of 1.5 h over several months. Three-dimensional wave number-frequency spectra were derived from the sequences of captured images using an analysis method developed by Young et al. [J. Geophys. Res. 90(C1) (1985) 1049]. Some aspects of this analysis were refined. An empirical transfer function was derived relating the radar image spectra to the wave spectra measured by other instruments. A new calibration procedure for obtaining wave height from the radar spectra was also derived. The radar spectra were compared with directional spectra collected by a bottom-mounted InterOcean S4DW directional wave-current meter and with Datawell Waverider data. The capabilities of X-band radar in a coastal situation are discussed including the derivation of water depth required to optimise the analysis.

Deep Ocean Wave Measurements Using a Vertically Oriented Sonar

A vertically oriented 200-kHz sonar is used to make estimates of (one-dimensional) surface wave spectra in the deep ocean. Two independent approaches are used to determine wave spectra from this system: a direct measurement of range to the surface and the vertical velocity estimate of the surface. Estimated wave spectra are compared with observations from a Datawell Waverider buoy positioned directly above the acoustic instrument. Comparisons are provided for relatively calm conditions and also during a period of mixed swell and wind waves forced by a 13 m s−1 wind. Wave spectra from the three estimates appear qualitatively similar and agree to within 10% in total wave energy density.

A SOUTH AFRICAN WAVE CLIMATE STUDY

A study was made, using all the available sources of deepsea wave data around the South African coast, to establish a statistical representation of the wave climate in this region. Data obtained from clinometer observations, the Datawell Waverider and voluntary observing ships were found to be the most useful. The most accurate source of height, period and spectral information came from the Waverider which has been in extensive use in the region since 1969. To optimise the information value of these data a sophisticated data qualification technique was developed for digitised Waverider data in order to exclude poor data. This technique was also found to be useful in optimising the operational efficiency of a Waverider station. A clear pattern was obtained for the general wave climate around the South African coast; this is in accordance with the pattern that may be expected from the weather climate affecting this region. Of interest are the SE'ly winds generating increased wave heights in proportion to the distance offshore from the west coast as well as the effect of the Agulhas current, opposing the SW'ly waves along the east coast, also creating high-wave conditions.