1. Introduction
The daily cycles of heating, cooling, winds, rain, and variations in temperature and/or humidity produce a hierarchy of physical processes that act and interact to stir the upper ocean (Moum and Smyth 2001). The mixed layer is a result of the manifestation of those active mixing processes, and it plays an important role in studies of climate, biological productivity, and marine pollution (Yamazaki et al. 2002).
The observations of the upper-ocean processes, particularly those involving turbulence measurements, are performed by instruments categorized as vertical or horizontal microstructure profilers. According to Oakey and Elliott (1982), vertical profiling instruments had demonstrated the spatial and temporal variability of observed turbulence structures. However, vertical profilers provide sparse horizontal sampling due to their deployment logistics, particularly in the upper ocean, where the horizontal inhomogeneity and the effect of phenomena such as Langmuir cells inside the mixed layer can be significant (Garrett 1996).
According to Lueck et al. (2002), horizontal turbulence measurements began in the 1950s and revived in the 1980s with a variety of instruments, such as towed vehicles, autonomous underwater vehicles (AUVs), submarines, and free-fall gliders. The EPSONDE-Glider, introduced by Greenan and Oakey (1999), is a quasi-horizontal tethered glider that measures ocean turbulence, and it had demonstrated advantages over vertical profilers in exploring the ocean mixed layer under various forcing conditions.
We have developed a new tethered quasi-horizontal microstructure glider: Turbulence Ocean Microstructure Acquisition Profiler-Glider [TurboMAP-Glider (TMG)]. The TMG prototype (Arima et al. 2010) was based on a vertical laser-based profiler, TurboMAP-L (TM; Wolk et al. 2002). The TMG measures ocean microstructure temperature and turbulent velocity shear. Also, it carries high-resolution bio-optical sensors that simultaneously measure chlorophyll and turbidity. This capability makes the TMG a unique instrument capable of studying the effect of turbulence upon biological signals through a quasi-horizontal perspective. In this paper, we present the design and outcomes of the TMG during field experiments near Joga-shima. Japan, as well as its performance, and compare it against its predecessor, the TM.
2. The TMG
a. Design
The TMG is 2.70 m long and has a mass of 30.24 kg, a density of 1045.4 kg m−1, and a volume of 28.93 × 10−3 m3 (Fig. 1). The main wing has a symmetric National Advisor Committee for Aeronautics 0009 (NACA0009) profile, a chord of 0.1675 m, and a span of 0.5 m (each side). This wing has a maximum ratio of lift to drag, CL/CD = 27, occurring at an angle of attack (AOA) of 4° and at low Reynolds number (~0.5 × 105), and the wing stalls at an angle of 16°.
The TMG basic configuration: (a) acetal copolymer pressure case—head, containing all the electronics and sensors; (b) aluminum pressure case—tail; (c) main wings; and (d) tail wings.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
The fuselage of the TMG is composed of two separated pressure cases. The fore half (1.19 m) is made of acetal copolymer and contains all sensors and electronics. The aft case (1.20 m) is made of aluminum. The tail section (0.31 m) contains the vertical and horizontal stabilizers, and a wet compartment that holds steel plates for adjusting the centers of mass and volume of the glider (Fig. 2).
(a) Tail wings, (b) adjustable weight compartment with its enclosure, and (c) weight plates inside the compartment.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
The tethered cable is made of high molecular weight polyethylene fiber with a density of 970 kg m−3 and a breaking strength of 9000 N. Data are recorded internally and downloaded after the TMG is recovered using its tether.
The TMG carries two shear probes, an FP07 thermistor and a three-axis accelerometer, which are sampled at a rate of 512 Hz (Fig. 3; Table 1). The light-emitting diode (LED) fluorescence/turbidity probe and the laser fluorescence probe (Doubell et al. 2009) are sampled at a rate of 256 Hz. Electrical conductivity, temperature, and depth are sampled at a rate of 32 Hz, as are the single-axis electromagnetic flowmeter and the compass.
(a) FP07 thermistor, (b) two velocity shear probes measuring in z direction, (c) guards that protect sensors from mechanical impact, (d) laser fluorometer, (e) LED fluorometer, (f) conductivity and temperature sensors, (g) electromagnetic flowmeter, and (h) pressure sensor.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
Parameters and sensors from the TMG.
b. Deployment
The launcher for the TMG consists of a steel platform and two buoys to keep it above the water surface. The TMG is mounted on the launcher and lowered to the water by the ship’s A-frame or a crane (Fig. 4a). Thereafter, the TMG is released (Fig. 4b), and the launcher is retrieved. The TMG descends immediately after its release and is promptly pulled back to the surface (Fig. 4c). When the TMG nose starts to sink, the tethered cable is paid out manually at a rate to maintain several meters of slack. This decouples the glider from the ship and allows it to glide freely. The flight is terminated when the tether is exhausted or when the instrument has reached its target depth. The TMG is pulled back to the surface, either manually or using a Yoing Ocean Data Acquisition (YODA) Profiler winch (Masunaga and Yamazaki 2014), and brought aboard (Fig. 4d).
Deployment of TMG from R/V Seiyo Maru: (a) the launcher coupled with the TMG is lowered to the water surface, (b) the TMG is released and the launcher is retrieved, (c) the TMG is pulled back to the surface in order to start its glide, and (d) the TMG is retrieved using the ship’s A-frame or crane.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
c. Steady forces and moments
The steady forces on TMG: U is the TMG velocity, θ is the pitch angle, α is the angle of attack, (θ + α) is the path angle, and M is the moment due to drag, FD, and lift, FL. Buoyancy is denoted by B and W is the weight. The glider-based coordinates are x (along axis), y (to port), and z (up and orthogonal), while the earth-based coordinates are X, Y (both horizontal), and Z (up).
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
The center of mass position along the axis of the TMG xm is found by weighing the glider in air from two points of suspension. When there are no trim plates in the tail, xm = −1.065 m with respect to the front nose cone. The z coordinate zm is found by measuring the torque about the x axis when the y axis is vertical. The center of mass is at 4 × 10−3 m below the centerline because a stainless steel drop weight is located in the bottom half of the glider near the middle of the fuselage, which provides roll stability. Internal trim weights bring TMG’s y coordinate to zero, so that the center of mass is in the x–z plane. The axial position of the center of buoyancy xB is determined by weighing the glider in water from two points of suspension. It is at xB = −1.115 m when there are no trim plates in the tail. The z coordinate zB has not been determined, but it must be very close to the centerline because the fuselage is rotationally symmetric. The only off-center objects are the tail wing and a small lifting eye near the center of the fuselage, both of which raise the center of buoyancy above the axis of the glider. We assume zB = 0.
The torque on the glider due to lift and drag cannot be accurately predicted because the point of effect of these forces is nebulous. Both forces are generated by flow over the entire glider—fuselage, wings, sensors, and all other appendages. The glider will sink, nose downward, if the center of mass is forward of the center of buoyancy. If the center of lift is forward of the center of mass, the lift will raise the nose, and if the configuration is stable, the TMG will glide according to Eqs. (1)–(3). We have placed the center of the wing at x = −1.110 m, so that the center of mass can be brought aft of this position using the trim plates in the tail.
d. Pitch and path angles
Glider path geometry: C, D, and E are right-triangle legs; θ is the pitch angle; Δz is the vertical distance moved by TMG in 1 s; U is the x axis component of the velocity measured by the flowmeter; and V is the along-path component of velocity.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
e. Trimming the TMG for flight
We used an empirical approach to trim the glider for flight. In February 2011, we carried out a test in Lake Biwa, Japan, where we progressively added stainless steel weights to the tail (Fig. 2c) to move the centers of mass and buoyancy rearward. Each kilogram of added mass moves the center of mass by 5.5 × 10−2 m and the center of buoyancy by 0.7 × 10−2 m. These centers coincide at x = −1.123 m when the added mass is 2.2 kg, a position that is 0.013 m aft of the center of the wings.
We set the angle of the wings to 0° and tested the path angle, θ + α, for added masses of 0.6–2.2 kg (Table 2). The TMG was stable but flew with unsatisfactory path angles for all masses smaller than 2.2 kg. Vibrations were large for 1.6 and 1.9 kg, indicating flow separation from the wings. A wing angle of 2° performed well for masses of 1.9 and 2.0 kg. A wing angle of 4° performed well with a mass of 1.9 kg, but the glider was marginally stable at 2.0 kg. With a mass of 1.6 kg, a wing angle of 8° was stable and even had a slightly negative AOA. However, the glider was unstable when the mass was increased to 1.9 kg. Thus, there is a range of added mass (1.6–2.0 kg) and wing angles (2°–8°) that gives a stable flight, which means path angles between 13° and 18°, small vibrations, and no abrupt changes of direction or angles.
Flight tests of TMG with different wing angles and tail weights in February 2011 in Lake Biwa.
f. Joga-shima experiments
From 18 to 20 June 2011, from 5 to 6 September 2012, and on 24 June 2013, we carried out TMG and TM deployments near Joga-shima from the Research Vessel (R/V) Seiyo Maru (Fig. 7). We deployed the TMG first, and as soon as we retrieved the TMG, we deployed the TM. The time interval between two successive deployments was less than 30 min. We carried out 34 TMG profiles in 2011, 12 in 2012, and 24 in 2013. For TM, these numbers are 32, 13, and 21, respectively.
Location of the deployments from the TMG and the TM (black star).
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
The TMG’s deepest deployment reached 100 m, during which it had also traveled a total of 300 m horizontally away from the ship, using a combination of 1.9 kg of weight inside the tail compartment and a wing angle of 4° (Table 2). We used the deepest deployment to show the performance of the flight in the longest time series (Fig. 8). We also discuss data from different deployments in the next sections.
The TMG flight performance parameters on 18 Jun 2011: (a) depth, (b) path angle and angle of attack (c) heading, (d) horizontal path, and (e) TMG speed.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
The TMG started to increase speed, reaching a maximum of 1 m s−1 around 8-m depth (10 s after its launch). At 15-m depth, the speed settled to 0.6 m s−1 (50 s after launch; Fig. 8e). The horizontal path of the TMG depends not only on the background flow but also on even the smallest athwartship asymmetry, which can cause the TMG to deviate from its path. However, we did not observe abrupt direction changes, and the rate of deviation of 0.34° s−1 was approximately constant (Figs. 8c,d). The vertical path angle started around 40° and quickly decreased to 20° in the first 8 m (10 s after launch). Then, it reached 13° around 10 m (30 s after launch) and remained approximately constant until 20 m (100 s after launch). Finally, it increased and reached 22° (at 100 m) at the end of the flight (Fig. 8b). The TMG never reaches a truly steady state. With increasing travel distance, more cable is deployed, which increases the drag on the glider and, consequently, its path angle [Eq. (6)]. However, the glide was smooth right to the end.
3. Microstructure
a. Data processing
b. Minimizing vibration effects
The TMG normally reports strong acceleration and shear during the first 20 s of its flight (e.g., Fig. 9), and vibration severely contaminates the shear probe signals. The glider is still settling into its quasi-steady orientation. The vibration affects mainly the y axis (Ay) and the z axis (Az) and is usually weak in the x axis (Ax) as shown in Figs. 9c–e. After the starting transient, the vibrations are very small and the shear varies intermittently, which is a common characteristic of near-surface stably stratified waters.
(a) Velocity shear; (b) temperature gradient; (c),(d) and (e) are x, y, and z accelerations, respectively; and (f) depth, path angle, and velocity. The gray rectangle denotes the initial transient phase of the glide. Data from 20 Jun 2011.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
(top) Shear and acceleration spectra for the strong vibration region identified by (bottom) the black rectangle in the shear probe signal time series. The vertical gray line represents the half of the Kolmogorov wavenumber (kn). The acceleration spectra are offset in the vertical direction by a factor of 10−2 for clarity. The speed was obtained from the averaged profiling speed in the identified segment. Data from 20 Jun 2011.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
(top) Shear and acceleration spectra for the quiescent shear region identified by (bottom) the black rectangle in the shear probe signal time series. The vertical gray line represents the half of the Kolmogorov wavenumber (kn). The acceleration spectra are offset in the vertical direction by a factor of 10−2 for clarity. The speed was obtained from the averaged profiling speed in the identified segment. Data from 20 Jun 2011.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
(top) Shear and acceleration spectra for the strong velocity shear with low vibration noise region identified by (bottom) the black rectangle in the shear probe signal time series. The vertical gray line represents the half of the Kolmogorov wavenumber (kn). The speed was obtained from the averaged profiling speed in the identified segment. Data from 6 Sep 2012.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
Empirical CDF of the Ozmidov length scale (LO) estimated by TMG. The dashed line represents the TMG length L. Data from all the profiles obtained in the years of 2011–13. The total number of samples is equal to 1239.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
c. Estimation of kinetic energy dissipation rate
Figure 14 shows a quiescent velocity shear spectrum, collected on 24 June 2012 between 20.4- and 21.4-m depth. At these low levels of turbulent energy, it is still possible to infer the correct dissipation rate from the measured spectra up to half of kn, which represents about 87% of shear variance (Wolk et al. 2002). The TMG was capable of measuring ε as low as 0.5 × 10−10 W kg−1, which is in the order of the lowest dissipation rate of turbulent motion and comparable to the best performance of most vertical free-fall profilers (Lueck et al. 2002).
Dissipation spectrum of a quiescent section of velocity shear between 20.4- and 21.4-m depth near Joga-shima on 6 Sep 2012. The vertical gray line represents the half of the Kolmogorov wavenumber (kn).
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
4. Chlorophyll-a
Fluorescence measurements from the LED sensor were first low-pass filtered at 50 Hz to suppress sensor noise while retaining the spatial scales resolved by the sensor (Wolk et al. 2006). No correction was required for fluorescence measured by the laser sensor (Doubell et al. 2009). The arbitrary units used to measure fluorescence are calibrated using sodium fluorescein, so that the output units are approximately equivalent to μg L−1 of chlorophyll-a. The LED sensor samples a volume of 4 mL and has an approximated spatial resolution of 2 cm (Wolk et al. 2001). The reduced sample volume of the laser probe (32 μL) in comparison to the LED probe (4 mL) allows for measurements of chlorophyll-a with increased spatial resolution and gives independent measures of the fluorescence field approximately every 2–3 mm at typical profiling speeds, between 0.50 and 0.80 m s−1 (Doubell et al. 2009). Therefore, the variance of the signal from the laser probe is much larger than that from the LED unit and shows that the phytoplankton spatial variability becomes increasingly intermittent when measured with increased resolution (Fig. 15a). Still, according to Doubell et al. (2009), it is likely that the peak structures identified by the laser probe constitute patches of increased biomass, which may include individual phytoplankton cells as well as chains and aggregates. We defined phytoplankton-related patches as laser fluorescence peaks 2 times bigger than the background value, which in turn is the laser signal low passed at 1 Hz (Fig. 15b).
(a) Chlorophyll-a measured with the laser (black line) and the LED (gray line) sensors. The gray circles represent phytoplankton patches measured using laser sensor. (b) Zoomed area from the segment identified by the rectangle in (a). The dashed line represents the laser signal low passed at 1 Hz (background value). Data from 24 Jun 2013.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
We analyzed the log-transformed histograms of 4-m-depth segments of fluorescence data from both fluorescence sensors. The histograms of log-normalized fluorescence values measured by the LED and laser sensors were best fit by normal and Gumbel (also known as type I extreme value distribution) distributions, respectively (Fig. 16). Our findings corroborate with the results obtained from vertically measured microstructure fluorescence by Doubell et al. (2014). These authors suggest that the clear shift from a lognormal to a skewed extreme value distribution with a reduction in sample volume demonstrates the existence of a critical scale at which the underlying nature of the fluorescence field diverges. We have also observed this trend quasi horizontally, though further investigation is needed to infer about possible differences between our results and those observed vertically.
Histogram of log-normalized fluorescence values measured by the TMG (a) LED and (b) laser sensors between 21- and 25-m depth. The LED sensor histogram was fitted with a lognormal distribution (dashed line). The histogram from the laser sensor was best fitted with a Gumbel extreme value distribution (dashed line). Corresponding quantile–quantile (Q–Q) plots show the comparison of the distribution of the (c) LED and the (d) laser fluorescence values to theoretical lognormal and Gumbel extreme value distributions, respectively. Data from deployments conducted on 24 Jun 2013. The number of samples is equal to 1024 for each histogram.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
(a) Mean (m) vs standard deviation (s) from laser (black stars) and LED (gray circles). The areas identified by the ellipses represent different averaged coefficients of variation, CV1 and CV2 from the laser sensor. (b) Averaged (1 m) chlorophyll-a measured from laser and LED sensors. The dashed line represents equality between the two sensors. Data from deployments conducted during ebb and flood tide on 24 Jun 2013.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
We have calculated the averaged laser CV for the areas identified by the ellipses in Fig. 17. These areas are labeled CV1 and CV2 and are equal to 2.13 and 1.01, respectively. We used the profiles related to CV1 and CV2 to build the temperature–salinity (T–S) diagram shown in Fig. 18, where the plot of T–S shows that CV1 and CV2 are also associated with two water masses near Joga-shima. Area CV1 corresponds to the flood tide and to the phytoplankton community from Sagami Bay, where the water has oceanic characteristics. In contrast, CV2 is related to the water from Tokyo Bay that reaches Joga-shima during the ebb tide.
The T–S diagram from profiles related to CV1 and CV2. Data refers to deployments during ebb and flood tide on 24 Jun 2013.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
5. Comparisons between TMG and TM
Histograms (%) of log10ε from (top) TMG and (middle) TM with N samples. (bottom) The cumulative distribution functions (CDF) from TMG and TM from the years of 2011–13.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
CDF of I (ε/νN2) from TMG and TM. The number of samples for the TMG is equal to 620 in 2011, 256 in 2012, and 363 in 2013. For TM, these numbers are 735, 361, and 416, respectively.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
The along-path measurement of density from the TMG shows inversions, which could be either true local overturns or just vertical undulations of the isopycnals by internal waves transected by the path of the glider. A theoretical study from Thorpe (2012) shows that gliders can show false overturns while crossing internal waves or Kelvin–Helmholtz instabilities (for a simplified illustration, see Fig. 21). We used the Ozmidov and the Thorpe (Thorpe 1977) length scales to try to distinguish these two possible explanations for the along-path inversions. We are relying on the linear relation between these scales when there are true overturns (see Finnigan et al. 2002 for a review).
TMG path crossing a sinusoid internal wave traveling through the interface between two layers with densities equal to ρ1 and ρ2.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
More than 50% of the estimates of the ratio a measured by the glider are smaller than 0.25 (Fig. 22), while more than 73% of the TM-based estimates fall into the range of 0.25–4. The consistently higher value of LT compared to LO between depths of 6 and 15 m, where apparent inversions are common, indicates that most of these are false overturns of the isopycnals (Fig. 23); that is, the kinetic energy of the motions (as indicated by LO) is insufficient for overturns of scale LT.
Scatterplot of LT and LO from (left) the TMG and the (right) TM, where the top and bottom straight lines represent a = 4 and a = 0.25, respectively. Data from all the profiles obtained in the years of 2011–13.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
(left) Terms LT and LO distribution from a representative TMG profile. Large values of LT are accompanied by apparent density inversions. Data from 19 Jun 2011.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
TM profiles do not show big and frequent inversions such as those found in TMG profiles, as exemplified in Fig. 24. The conductivity (C) and temperature (T) are measured by a combined C–T sensor consisting of a platinum wire thermometer and an inductive conductivity cell. A different time response between these two sensors (the temperature sensor is slower than the conductivity sensor) can lead to salinity spiking and can cause false inversions in the density profiles. We corrected this problem by computing the lagged cross correlation between T and C and calculating the averaged lag, which is used to advance the T signal. Because T and C data were processed identically on the TMG and TM, and because they profile at similar speeds, the inversions are not an artifact of sensor response. Rather, they are caused by the difference in the profiling direction by these two platforms.
Distribution of potential density anomaly (σθ) of two deployments of TMG and TM on (left) 6 Sep 2012 and (right) 24 Jun 2013. The time interval between the TMG and TM profiles was less than 30 min in both examples.
Citation: Journal of Atmospheric and Oceanic Technology 31, 10; 10.1175/JTECH-D-13-00240.1
6. Summary
We have developed an instrument that glides smoothly through the upper ocean, over a horizontal range of 300 m, starting from a depth of a few meters to as much as 100 m, and measures oceanic biophysical microstructure. Its shallow path angle (~13°) provides a very high vertical resolution for a range of wing angles and trim weights, and it is easily adjusted for variations in the density of water, since the path angle is independent of the weight and buoyancy of the glider. The TMG measures ε as low as 0.5 × 10−10 W kg−1, which is on the order of the lowest dissipation rate of turbulent motion. This paper presents some new and relevant empirical results about the quasi-horizontal application of a high-resolution fluorescence sensor to understand the spatial structure of phytoplankton. The log-normalized fluorescence distributions from LED (normal distribution) and laser (Gumbel distribution) sensors are consistent with results obtained by a vertical profiler (Doubell et al. 2014), but further investigation is needed to understand any possible difference between vertical and quasi-horizontal sampling. The a ratio results indicate that the false density overturns observed by the TMG are observational evidence of internal waves or Kelvin–Helmholtz instabilities crossed by the glider, which was demonstrated theoretically by Thorpe (2012).
Acknowledgments
We thank the crew of the R/V Seiyo Maru of the Tokyo University of Marine Science and Technology for its assistance during the field experiment. We also acknowledge Takeyoshi Nagai, Hua Li,Hikaru Honma, Eiji Masunaga, Mamoru Tanaka, and Lynn L. Allmon for their cooperation during the various steps of our work. H. F.-N. is financially supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan. This research was funded by Grant-in-Aid for Science Research (B2)20340127 from the Japan Society for the Promotion of Science. Also, a partial fund was made available by JST, CREST, and Tohoku Ecosystem-Associated Marine Sciences.
REFERENCES
Ara, K., and Hiromi J. , 2007: Temporal variability in primary and copepod production in Sagami Bay, Japan. J. Plankton Res., 29, i85–i96, doi:10.1093/plankt/fbl069.
Arima, M., Nakamura H. , and Yamazaki H. , 2010: Development of the glider-type turbulence ocean microstructures acquisition profiler, TurboMap-G. Proceedings of the Twentieth (2010) International Offshore and Polar Engineering Conference, J. S. Chung et al., Eds., Vol. 1, ISOPE, 309–314.
Dillon, T. M., 1982: Vertical overturns: A comparison of Thorpe and Ozmidov length scales. J. Geophys. Res., 87, 9601–9613, doi:10.1029/JC087iC12p09601.
Doubell, M. J., Yamazaki H. , Li H. , and Kokubu Y. , 2009: An advanced laser-based fluorescence microstructure profiler (TurboMap-L) for measuring bio-physical coupling in aquatic systems. J. Plankton Res., 31, 1441–1452, doi:10.1093/plankt/fbp092.
Doubell, M. J., Prairie J. C. , and Yamazaki H. , 2014: Millimeter scale profiles of chlorophyll fluorescence: Deciphering the microscale spatial structure of phytoplankton. Deep-Sea Res. II, 101, 207–215, doi:10.1016/j.dsr2.2012.12.009.
Finnigan, T. D., Luther D. S. , and Lukas R. , 2002: Observations of enhanced diapycnal mixing near the Hawaiian ridge. J. Phys. Oceanogr., 32, 2988–3002, doi:10.1175/1520-0485(2002)032<2988:OOEDMN>2.0.CO;2.
Gargett, A. E., Osborn T. R. , and Nasmyth P. W. , 1984: Local isotropy and the decay of turbulence in a stratified fluid. J. Fluid Mech., 144, 231–280, doi:10.1017/S0022112084001592.
Garrett, C., 1996: Processes in the surface mixed layer of the ocean. Dyn. Atmos. Oceans, 23, 19–34, doi:10.1016/0377-0265(95)00421-1.
Goodman, L., Levine E. R. , and Lueck R. G. , 2006: On measuring the terms of the turbulent kinetic energy budget from an AUV. J. Atmos. Oceanic Technol., 23, 977–990, doi:10.1175/JTECH1889.1.
Greenan, B. J. W., and Oakey N. S. , 1999: A tethered free-fall glider to measure ocean turbulence. J. Atmos. Oceanic Technol., 16, 1545–1555, doi:10.1175/1520-0426(1999)016<1545:ATFFGT>2.0.CO;2.
Han, M., and Furuya K. , 2000: Size and species-specific primary productivity and community structure of phytoplankton in Tokyo Bay. J. Plankton Res., 22, 1221–1235, doi:10.1093/plankt/22.7.1221.
He, M., Williams C. D. , and Bachmayer R. , 2009: Additional modelling for the prediction of the performance of ocean gliders. Sixth International Symposium on Underwater Technology 2009, W. Cui, M. Zhang, and B.-L. Xang, Eds., China Ship Scientific Research Center, 158–164.
Levine, M. D., and Boyd T. J. , 2006: Tidally forced internal waves and overturns observed on a slope: Results from the HOME. J. Phys. Oceanogr., 36, 1184–1201, doi:10.1175/JPO2887.1.
Lueck, R. G., Wolk F. , and Yamazaki H. , 2002: Oceanic velocity microstructure measurements in the 20th century. J. Oceanogr., 58, 153–174, doi:10.1023/A:1015837020019.
Macoun, P., and Lueck R. , 2004: Modeling the spatial response of the airfoil shear probe using different sized probes. J. Atmos. Oceanic Technol., 21, 284–297, doi:10.1175/1520-0426(2004)021<0284:MTSROT>2.0.CO;2.
Masunaga, E., and Yamazaki H. , 2014: A new tow-yo instrument to observe high-resolution coastal phenomena. J. Mar. Syst., 129, 425–436, doi:10.1016/j.jmarsys.2013.09.005.
Moum, J. N., and Smyth W. D. , 2001: Upper ocean mixing processes. Encyclopedia of Ocean Sciences, J. H. Steele, S. A. Thorpe, and K. K. Turekian, Eds., Vol. 6, Academic Press, 3093–3100.
Nasmyth, P. W., 1970: Oceanic turbulence. Ph.D. thesis, University of British Columbia, 69 pp.
Oakey, N. S., 1982: Determination of the rate of dissipation of turbulent energy from simultaneous temperature and velocity shear microstructure measurements. J. Phys. Oceanogr., 12, 256–271, doi:10.1175/1520-0485(1982)012<0256:DOTROD>2.0.CO;2.
Oakey, N. S., and Elliott J. A. , 1982: Dissipation within the surface mixed layer. J. Phys. Oceanogr., 12, 171–185, doi:10.1175/1520-0485(1982)012<0171:DWTSML>2.0.CO;2.
Ozmidov, R. V., 1965: On the turbulent exchange in a stably stratified ocean. Izv., Acad. Sci., USSR, Atmos. Oceanic Phys., 1, 853–860.
Thorpe, S. A., 1977: Turbulence and mixing in a Scottish loch. Philos. Trans. Roy. Soc. London, A286, 125–181, doi:10.1098/rsta.1977.0112.
Thorpe, S. A., 2012: Measuring overturns with gliders. J. Mar. Res., 70, 93–117, doi:10.1357/002224012800502417.
Wesson, J. C., and Gregg M. C. , 1994: Mixing at Camarinal Sill in the Strait of Gibraltar. J. Geophys. Res., 99, 9847–9878, doi:10.1029/94JC00256.
Wolk, F., Seuront L. , and Yamazaki H. , 2001: Spatial resolution of a new micro-optical probe for chlorophyll and turbidity. J. Tokyo Univ. Fish., 87, 13–21.
Wolk, F., Yamazaki H. , Seuront L. , and Lueck R. G. , 2002: A new free-fall profiler for measuring biophysical microstructure. J. Atmos. Oceanic Technol., 19, 780–793, doi:10.1175/1520-0426(2002)019<0780:ANFFPF>2.0.CO;2.
Wolk, F., Yamazaki H. , Li H. , and Lueck R. G. , 2006: Calibrating the spatial response of bio-optical sensors. J. Atmos. Oceanic Technol., 23, 511–516, doi:10.1175/JTECH1863.1.
Yamazaki, H., Mackas D. L. , and Denman K. L. , 2002: Coupling small-scale physical processes with biology. Biological–Physical Interactions in the Sea: Emergent Findings and New Directions, A. R. Robinson, J. J. McCarthy, and B. J. Rothschild, Eds., The Sea: Ideas and Observations on Progress in the Study of the Seas, Vol. 12, Harvard University Press, 51–112.
