1. Introduction
The benthic boundary layer plays an important role in aquatic environments, and it is essential to have information about the turbulent flow structure in this zone (Smith 1975). Sediment particle transport is directly influenced by the hydrodynamic conditions (e.g., Kawanisi and Yokosi 1993; Trowbridge and Agrawal 1995; Cheng et al. 1997). In addition, solute fluxes, such as oxygen across the sediment–water interface, are strongly affected by flow conditions immediately above the sediment (Pamatmat 1971; Belanger 1981; Boynton et al. 1981; Mackenthum and Stefan 1998). This is due to the hydrodynamic control of mass transfer in the benthic boundary layer (Jørgensen and Revsbech 1985; Dade 1993; Inoue et al. 2000). Therefore, precise measurements of the flow structure are indispensable when considering material cycling at and around the sediment–water interface.
For field observation of the velocity field, propeller current meters have been used (Soulsby and Dyer 1981; Heathershaw and Langhorne 1988), but they cannot detect turbulent motions. To measure turbulent flow, electromagnetic current meters are commonly employed. Kawanisi and Yokosi (1993) measured three-dimensional velocity using electromagnetic current meters fixed at 15, 25, and 55 cm above the riverbed to investigate the turbulence characteristics. Le Couturier et al. (2000) measured horizontal current velocity and direction at 35, 78, and 120 cm above the seabed using electromagnetic current meters to analyze the characteristics and effects of large-scale flow structures developed in the benthic boundary layer. Another useful instrument is the acoustic current meter (Williams et al. 1987). Grant et al. (1984) obtained near-bottom velocity profiles using four acoustic-travel-time current meters, which measured velocity averaged over a 15-cm pathlength, mounted on tripods at 30, 55, 105, and 205 cm above the bottom. However, with these methods, the spatial resolution of velocity measurements depends on the size of the sensors and the pathlength, and is usually on an order of 10 cm (Voulgaris and Trowbridge 1998). Therefore, it is difficult to conduct direct measurements of the benthic boundary layer (Cacchione et al. 1987). Moreover, multiple velocimeters are needed for vertical profile measurement (Huntley 1988), and their vertical positions and mutual intervals must be decided beforehand.
In recent years, the acoustic Doppler current profiler (ADCP) has been employed to measure three-dimensional velocity profiles. Cheng et al. (1997) measured velocity profiles every 2 min covering the water column from near the bed to a distance about 175 cm from the bed in 5-cm increments by using a broadband acoustic Doppler current profiler, and reported the results of a continuing field investigation of sediment dynamics in a shallow estuarine environment. Kawanisi (2004) also measured three-dimensional velocity at 3-s intervals and with a vertical resolution of as fine as 3 cm in the range of 2–98 cm above the bottom by using a high-resolution current profiler. In these measurements, the temporal resolution was fine enough to measure turbulent characteristics; however, the spatial resolution was a little rough (3–5 cm).
For more precise measurements, a particle image velocimetry (PIV) system and laser Doppler velocimetry (LDV) have been used. Doron et al. (2001) measured turbulence characteristics in the coastal ocean bottom boundary layer using a submersible PIV system. Its temporal and spatial resolutions are fine enough to measure turbulent characteristics. However, PIV provides distributions of only two velocity components unless stereo photography is employed (for details of PIV, see Bertuccioli et al. 1999). Using a two-axis submerged LDV, Agrawal and Aubrey (1992) measured velocity profiles above the rippled bed of a beach. The most successful measurements using profiling LDV were done by Trowbridge and Agrawal (1995) and Trowbridge (1998). They revealed a vertical structure of two horizontal components of velocity at heights of 0.5, 1.0, 2.0, 4.0, 8.0, and 16.0 cm above a nominal bottom using a microprocessor controlling the LDV profiler. The LDV method offers the finest temporal and spatial resolution, but is strongly affected by transmissivity.
As an alternative device, much attention has been paid to the acoustic Doppler velocimeter (ADV). ADV was first introduced by Kraus et al. (1994), and its performance was evaluated by Lohrmann et al. (1994), Voulgaris and Trowbridge (1998), and Snyder and Castro (1999). Recently, ADV has become popular because of its accuracy and handiness (Trowbridge and Elgar 2001; Elgar et al. 2005) and has been employed for field measurements in the benthic boundary layer. However, in most studies, the ADV is mounted on a frame fixed on the bed, and the measurement height cannot be changed (e.g., Howarth and Souza 2005; Uchiyama et al. 2005). Nikora et al. (2002) also conducted ADV measurements in the benthic boundary layer. In their study, the average elevations of ADV sampling volume were 2, 6, 12, 29, and 61 cm above the bed, and the height of the sensor was changed by divers. However, they could not determine the exact height of the boundary between the roughness sublayer and the logarithmic layer, because of the rough vertical intervals of the sensor. They also pointed out the inaccuracy of using logarithmic velocity profiles to estimate the shear velocity because of the limited number of measurement points, that is, only 2–4 points were available for fitting of the logarithmic profiles.
In this paper, we introduce an ADV system for precise measurement of vertical profiles of three velocity components and the turbulent flow structure immediately above the sediment–water interface. We also present the results of field measurements performed in Hiroshima Bay, Japan, and discuss them with respect to turbulence behavior.
2. Materials and methods
a. Instrumentation
For the flow measurements, the ADV (Nortec, ADV field) technique was employed. The ADV field is used as a bistatic acoustic Doppler system and consists of a transmitter and three receivers (Voulgaris and Trowbridge 1998). This setup allows simultaneous collection of real-time data of eastward, northward, and upward components of the velocity, using compass and slope measurement instruments to convert 3D flow components. The velocimeter conducts measurements by recording the Doppler effect caused by suspended particles in the water on 10-MHz pulses of sound. The accuracy of the ADV field employed in this study is ±0.5% of the measured value, or ±0.1 cm s−1. The control volume is 0.03 cm3 at a distance of 10 cm from the transmitter situated at the center of the probes. The size of the control volume is a function of the length of the transmitting pulse, the width of the receiving window, and the beam pattern of the receiving and transmitting pulses (Voulgaris and Trowbridge 1998). The maximum sampling frequency is 25 Hz. The detection of reflected pulses of sound enables measurement of the distance from the sampling point to the boundary with a precision of 0.01 cm. Prior to data collection, the operator must input the water temperature and salinity in the software to correct the raw data, because they affect sound speed. The operator must also choose the software-selectable range of the flow velocity (Snyder and Castro 1999). Further details of the ADV technique can be found in Kraus et al. (1994).
In this study, the ADV field was attached to an elevation system (Alec Electronics Co., Ltd.; Fig. 1), which makes it possible to elevate and fix the ADV arbitrarily within a range of 0.1–27 cm from the sediment surface. The elevation system consisted of an onboard controller, a 60-m cable, a stepping motor system as an elevation device, and a pedestal. Figure 2 shows the simplified circuit chart. The onboard controller for the stepping motor has an up-and-down changeover switch and a stepwise switch to elevate the cylinder of the stepping motor and the ADV in 0.1- or 1-cm steps. The relative vertical position of the ADV is displayed on the liquid crystal monitor. The stepping motor can vertically traverse the ADV field over a range of 27 cm in 0.1-cm steps with an accuracy of 0.02 cm. A pedestal of dimensions 120 cm in width and 85 cm in height is fixed on the sediment by using four sinkers and stakes. As the sole of each leg of the pedestal has a sufficiently large area, the pedestal does not sink into the sediment even on silty sediment. The ADV and the elevation system are operated by an onboard operator via a cabled control. In this system, the compass and slope measurement instruments are separated from the ADV, and the operator must install each device carefully. An ac 100-V power supply is also required. See Table 1 for more details of the system.
b. Observation site and measurement details
The field observation was conducted in Hiroshima Bay, an enclosed bay in Japan (946 km2 in surface area, 25.6 m in mean depth, see Fig. 3). The bay mouths are located to the south. The main inflow river, the Ohta River, discharges into the northern side of Hiroshima Bay at 41.4 m3 s−1. The observational point (34°20′N, 132°25′E) was in front of the river mouth of the Ohta River, and the mean depth was about 18 m. The surface sediment mainly consisted of silt and was visibly flat. The tidal conditions of the day were a spring tide with a tidal range of about 3.5 m.
The precise profiles of the three-component velocities were measured in the range from 0.5 to 20 cm above the sediment with a sampling frequency of 25 Hz. Spatial intervals were about 1 mm for within 0–1 cm above the sediment, 0.5–1 cm within 1–3 cm, and 2–5 cm within 3–20 cm. Six vertical profiles of three-component velocities were measured, and each profile consisted of 10–15 measurement points. The measurement time for each point was 1 or 5 min, and it took about 20 min to obtain one vertical profile for the 1-min measurement and about 1½ h for the 5-min measurement. Results from 1-min measurements were used to examine the vertical profiles of mean velocity, and those from 5-min measurements were used to calculate the turbulence properties as discussed below. Scuba divers visually confirmed the situation of the device such as the inclination and settlement of the pedestal before and after the observation.
c. Data analysis
3. Results
In Hiroshima Bay, the bay mouths are located to the south; therefore, the main tidal current is northward at the flood tide and southward at ebb tide. The wind was blowing from the northeast throughout the measurement period, and this wind was inferred to induce a northeastward current (7–12 cm s−1) in the bottom layer to compensate for the southwestward current in the surface layer. Moreover, it was supposed that the direction of the gravitational circulation flow in the bottom layer due to the freshwater inflow from the north was in the northward direction. Therefore, a relatively high northeastward velocity at the flood tide and low southwestward velocity (1–5 cm s−1) at the ebb tide were observed with the measurement terms, and the observed flow was inferred to be a combination of the tidal currents, the gravitational circulation current, and the wind-induced current (Fig. 4).
The horizontal mean velocity profile during the ebb period showed good agreement with the logarithmic profile (r 2 = 0.932–0.990). The calculated shear velocity by Eq. (1) varied from 0.11 (at high tide) to 0.94 cm s−1 (at flood tide). Because of the nonsteady condition of the current velocity, only limited datasets were available to calculate the shear velocity using the logarithmic method, resulting from 0.11 (at high tide) to 0.37 cm s−1 (at ebb tide). Shear velocities calculated in both ways had the same order values and were also comparable with those from 0.41 to 0.48 cm s−1, obtained by Grant et al. (1984). In addition, the roughness height z0, which is an important factor in governing the turbulent viscosity, shear stress, and the mass flux at the sediment–water interface, were calculated as 0.000 01–0.0639 cm by using the logarithmic method. From Eqs. (2) and (3), we obtained error estimates from ±14.2% to ±33.8% for shear velocities and from ±5.2% to ±14.4% for roughness heights by using six to nine measurement points. Gross and Nowell (1983) mentioned that the often-cited value of R2 = 0.95 gives errors in u*, for six measurement points of ±32% and for three measurements points of ±292%. Therefore, multipoint measurement using this method is significant for improving accuracy.
The maximum Reynolds stress value during the flood and ebb period was found at around 1.0–1.5 cm above the sediment (Fig. 5a). Reynolds stress values (0.02–0.94 g cm−1 s−2) obtained in this study are comparable with those measured by electromagnetic current meters by Kawanisi and Yokosi (1993), though Lohrmann et al. (1994) pointed out that Reynolds stress values were overestimated by the ADV at low shear stresses. Figures 5b and 5c show examples of calculated profiles of k and ɛ at flood tide when the current velocity varied between 5 and 15 cm s−1. Note that the observational time differs for each height. The k values, which were smaller than those obtained by Le Couturier et al. (2000), varied between 12 and 43 cm2 s−2 when the current velocity varied between 30 and 65 cm s−1. However, the ɛ values, which varied between 0.05 and 0.5 cm2 s−3, are typical of near-bed flows (Le Couturier et al. 2000).
4. Discussion
a. Advantages and disadvantages of the measurement system
Our ADV system offers several advantages. First, ADV has sufficiently high spatial and temporal resolutions to allow discussion of the turbulent flow because its control volume is sufficiently small (0.03 cm3) and the maximum sampling frequency is 25 Hz. In addition, it is not influenced by turbid water. The second advantage is that our ADV system can detect the distance from the sediment surface to the control volume with a precision of 0.01 cm by reflected pulses of sound. Therefore, the onboard operator can easily know the measurement height. Third, the elevation device enables us to move the ADV vertically to any point as desired, and to measure velocity fields at multiple heights, for example, 15–20 heights, while many other measurements have been done at only 3–6 heights. Such an increase of measurement points can improve the accuracy of the estimation of shear velocities and roughness heights. Thus, this system allows measurement of the precise profile of the three-dimensional mean velocity and the turbulent flow structure, and calculation of the flow and turbulence properties within the benthic boundary layer. The fourth advantage is that an operator can monitor measurement signals (e.g., height, velocity, acoustic amplitude, etc.) via the connecting cable in real time.
Of course, there are also some disadvantages. The system cannot be employed under stormy conditions because of its cable system. Also, the operator must remain at the observation point. Another disadvantage is the fact that because it takes time to obtain a statistically sound representation of the turbulent flow structure, the whole vertical measurements require a longer time (e.g., 1.5 h in this study). Thus, the question arises of whether a steady flow field can be maintained during the measurements. To resolve this issue, we must employ other velocimeter(s) at fixed point(s) to conduct simultaneous multipoint measurements and confirm the steadiness.
b. Flow field distortion by the instrument package
c. Analytical methods
In this paper, shear velocity was calculated in two ways: the correlation method and the logarithmic profile method. These methods have been used generally, but some difficulties have been pointed out (Gross et al. 1994), such as the modification of shear stress by a nonlinear interaction between current and surface-wave velocity components (Grant and Madsen 1986). Direct measurements of shear velocity are not possible without detailed measurements within 15 cm above the bed (Gross et al. 1992), but almost all measurements in the bottom boundary layer have not included this data. The system introduced here makes this possible. As an alternative method, indirect measurement via turbulent kinetic energy dissipation has been used (Gross and Nowell 1985; Huntley 1988; Green 1992). This method was not adopted in the present study because of the small Reynolds number, Re < 500 (Gross et al. 1994).
The power-law scaling of the energy spectrum and the estimate of ɛ is valid only if the ratio of the spectral strain rate s(k) to the mean rate of strain S is much greater than 1 (Gross and Nowell 1985). As the calculated values of s(k)/S were less than 0.4 in this study, ɛ was estimated by turbulent velocity scale and shear scale [Eq. (5)]. The shape of the vertical profile of ɛ is very similar to that of k. These results show the same properties of flow as those shown by laboratory experiments and literature values. Therefore, we concluded that the analytical methods developed by laboratory experiments on a small scale may be useful for in situ analysis of small-scale phenomena in the benthic boundary layer whose representative length scale is a few centimeters.
5. Conclusions
We developed a new flow velocity measurement system, which consists of ADV, an elevation system, and an onboard controller. It can measure the precise vertical profile of three-dimensional flow velocity immediately above the sediment surface in situ. This system allows direct and flexible measurements of a vertical flow structure profile within a range of 0.1–27 cm from the sediment surface, at 0.1-cm intervals, which can provide essential data to calculate turbulent properties. The system can be used to study hydromechanics immediately above the sediment. This system was employed in field observations performed in Hiroshima Bay. The observed results enabled us to obtain vertical profiles of mean velocity and turbulent properties. The observed results of the flow structure in the benthic boundary layer showed good agreement with those obtained from laboratory experiments. Consequently, the proposed system can be a promising tool to elucidate the physical processes around the sediment surface, for example, sediment transport or material balance across the sediment–water interface.
Acknowledgments
The authors wish to thank Dr. Yoshio Takasugi of the Institute for Marine Resources and Environment, AIST, for his technical help and discussions. The authors also want to thank Dr. Yoshiaki Kuriyama, Mr. Yasuyuki Nakagawa, Dr. Tomohiro Kuwae, Dr. Ken-ichi Uzaki, Dr. Yusuke Uchiyama, and Mr. Shinya Hosokawa of Port and Airport Research Institute for their helpful comments. The manuscript was greatly improved by the valuable comments made by the anonymous reviewers. This study was partly supported by a grant from the Environmental Agency, Japan.
REFERENCES
Agrawal, Y. C., and Aubrey D. G. , 1992: Velocity observations above a rippled bed using laser Doppler velocimetry. J. Geophys. Res., 97 , 20249–20259.
Belanger, T. V., 1981: Benthic oxygen demand in Lake Apopka, Florida. Water Res., 15 , 267–274.
Bertuccioli, L., Roth G. I. , Katz J. , and Osborn T. R. , 1999: A submersible particle image velocimetry system for turbulence measurements in the bottom boundary layer. J. Atmos. Oceanic Technol., 16 , 1635–1646.
Boynton, W. R., Kemp W. M. , Osborne C. G. , Kaumeyer K. R. , and Jenkins M. C. , 1981: Influence of water circulation rate on in situ measurements of benthic community respiration. Mar. Biol., 65 , 185–190.
Cacchione, D. A., Grant W. D. , Drake D. E. , and Glenn S. M. , 1987: Storm-dominated bottom boundary layer dynamics on the northern California continental shelf: Measurements and predictions. J. Geophys. Res., 92 , 1817–1827.
Cheng, R. T., Gartner J. W. , and Smith R. E. , 1997: Bottom boundary layer in South San Francisco Bay. J. Coastal Res., 25 , Special Issue. 49–62.
Dade, W. B., 1993: Near-bed turbulence and hydrodynamic control of diffusional mass transfer at the sea floor. Limnol. Oceanogr., 38 , 52–69.
Doron, P., Bertuccioli L. , Katz J. , and Osborn T. R. , 2001: Turbulence characteristics and dissipation estimates in the coastal ocean bottom boundary layer from PIV data. J. Phys. Oceanogr., 31 , 2108–2134.
Elgar, S., Raubenheimer B. , and Guza R. T. , 2005: Quality control of acoustic Doppler velocimeter data in the surfzone. Meas. Sci. Technol., 16 , 1889–1893.
Grant, W. D., and Madsen O. S. , 1986: The continental-shelf bottom boundary layer. Annu. Rev. Fluid Mech., 18 , 265–305.
Grant, W. D., Williams A. J. III, and Glenn S. M. , 1984: Bottom stress estimates and their prediction on the northern California continental shelf during CODE-1: The importance of wave–current interaction. J. Phys. Oceanogr., 14 , 506–527.
Green, M. O., 1992: Spectral estimates of bed shear stress at subcritical Reynolds numbers in a tidal boundary layer. J. Phys. Oceanogr., 22 , 903–917.
Gross, T. F., and Nowell A. R. M. , 1983: Mean flow and turbulence scaling in a tidal boundary layer. Cont. Shelf Res., 2 , 109–126.
Gross, T. F., and Nowell A. R. M. , 1985: Spectral scaling in a tidal boundary layer. J. Phys. Oceanogr., 15 , 496–508.
Gross, T. F., Isley A. E. , and Sherwood C. R. , 1992: Estimation of stress and bed roughness during storms on the northern California shelf. Cont. Shelf Res., 12 , 389–413.
Gross, T. F., Williams A. J. , and Terray E. A. , 1994: Bottom boundary layer spectral dissipation estimates in the presence of wave motions. Cont. Shelf Res., 14 , 1239–1256.
Heathershaw, A. D., and Langhorne D. N. , 1988: Observations of near-bed velocity profiles and seabed roughness in tidal currents flowing over sandy gravels. Estuarine Coastal Shelf Sci., 26 , 459–482.
Howarth, M. J., and Souza A. J. , 2005: Reynolds stress observations in continental shelf seas. Deep-Sea Res. II, 52 , 1075–1086.
Huntley, D. A., 1988: A modified inertial dissipation method for estimating seabed stresses at low Reynolds numbers, with application to wave/current boundary layer measurements. J. Phys. Oceanogr., 18 , 339–346.
Ikeda, S., 1999: Hydraulics. Gihodo Shuppan, 435 pp.
Inoue, T., Nakamura Y. , and Adachi Y. , 2000: Non-steady variations of SOD and phosphate release rate due to changes in water quality of the overlying water. Water Sci. Technol., 42 , 265–272.
Jørgensen, B. B., and Revsbech N. P. , 1985: Diffusive boundary layers and the oxygen uptake of sediments and detritus. Limnol. Oceanogr., 30 , 111–122.
Kawanisi, K., 2004: Structure of turbulent flow in a shallow tidal estuary. J. Hydraul. Eng., 130 , 360–370.
Kawanisi, K., and Yokosi S. , 1993: Measurements of turbulence and suspended sediment in tidal river. J. Hydraul. Eng., 119 , 704–724.
Kraus, N. C., Lohrmann A. , and Cabrera R. , 1994: New acoustic meter for measuring 3D laboratory flows. J. Hydraul. Eng., 120 , 406–412.
Le Couturier, M. N., Grochowski N. T. , Heathershaw A. , Oikonomou E. , and Collins M. B. , 2000: Turbulent and macro-turbulent structures developed in the benthic boundary layer downstream of topographic features. Estuarine Coastal Shelf Sci., 50 , 817–833.
Lohrmann, A., Cabrera R. , and Kraus N. C. , 1994: Acoustic-Doppler velocimeter (ADV) for laboratory use. Fundamentals and Advancements in Hydraulic Measurements and Experimentation, C. A. Pugh, Ed., American Society of Civil Engineers, 351–365.
Mackenthum, A. A., and Stefan H. G. , 1998: Effect of flow velocity on sediment oxygen demand: Experiments. J. Environ. Eng., 124 , 222–229.
Nikora, V., Green M. O. , Thrush S. F. , Hume T. M. , and Goring D. , 2002: Structure of the internal boundary layer over a patch of pinnid bivalves (Atrina zelandica) in an estuary. J. Mar. Res., 60 , 121–150.
Nishioka, M., and Sato H. , 1974: Measurements of velocity distributions in the wake of a circular cylinder at low Reynolds numbers. J. Fluid Mech., 65 , 97–112.
Pamatmat, M. M., 1971: Oxygen consumption by the seabed. IV. Shipboard and laboratory experiments. Limnol. Oceanogr., 16 , 536–550.
Roshko, A., 1953: On the development of turbulent wakes. NACA Tech. Note 2913, 77 pp.
Roshko, A., 1954: On the drag and shedding frequency of two-dimensional bluff bodies. NACA Tech. Note 3169, 29 pp.
Smith, I. R., 1975: Turbulence in Lakes and Rivers. Freshwater Biological Association, 79 pp.
Snyder, W. H., and Castro I. P. , 1999: Acoustic Doppler velocimeter evaluation in stratified towing tank. J. Hydraul. Eng., 125 , 595–603.
Soulsby, R. L., and Dyer K. R. , 1981: The form of the near-bed velocity profile in a tidally accelerating flow. J. Geophys. Res., 86 , 8067–8074.
Trowbridge, J. H., 1998: On a technique for measurement of turbulent shear stress in the presence of surface waves. J. Atmos. Oceanic Technol., 15 , 290–298.
Trowbridge, J. H., and Agrawal Y. C. , 1995: Glimpses of a wave boundary layer. J. Geophys. Res., 100 , 20729–20744.
Trowbridge, J. H., and Elgar S. , 2001: Turbulence measurements in the surf zone. J. Phys. Oceanogr., 31 , 2403–2417.
Uchiyama, Y., Nakano H. , Kurosaka M. , Yamawaki S. , Yanagishima S. , and Kuriyama Y. , 2005: Near-bed velocity measurement on an estuarine intertidal sandflat. Annu. J. Coastal Eng., 52 , 81–85.
Voulgaris, G., and Trowbridge J. H. , 1998: Evaluation of the acoustic Doppler velocimeter (ADV) for turbulence measurements. J. Atmos. Oceanic Technol., 15 , 272–289.
Williams A. J. III, , Tochko J. S. , Koehler R. L. , Grant W. D. , Gross T. F. , and Dunn C. V. R. , 1987: Measurement of turbulence in the oceanic bottom boundary layer with an acoustic current meter array. J. Atmos. Oceanic Technol., 4 , 312–327.
A schematic view of the flow measurement system.
Citation: Journal of Atmospheric and Oceanic Technology 25, 5; 10.1175/2007JTECHO531.1
Outline chart of the system setup. Stepwise switch A sends 500 pulses per push for 1-mm elevation of the cylinder. Stepwise switch B sends 5000 pulses per push for 10-mm elevation of the cylinder.
Citation: Journal of Atmospheric and Oceanic Technology 25, 5; 10.1175/2007JTECHO531.1
Location of Hiroshima Bay and the observation site.
Citation: Journal of Atmospheric and Oceanic Technology 25, 5; 10.1175/2007JTECHO531.1
Mean velocity profiles at the flood and ebb tide.
Citation: Journal of Atmospheric and Oceanic Technology 25, 5; 10.1175/2007JTECHO531.1
Examples of the vertical profile of (a) Reynolds stress, (b) turbulent kinetic energy, and (c) dissipation rate of turbulent energy during the flood tide.
Citation: Journal of Atmospheric and Oceanic Technology 25, 5; 10.1175/2007JTECHO531.1
Composition and specifications of the elevation system.