Article (refereed) Published version

Transcription

1 Article (referee) Publishe version Sen, Uwe; Fowler, George; Siall, Greg; Beanlans, Brian; Pittman, Merle; Walmann, Christoph; Karstensen, Johannes; Lampitt, Richar SeaCycler: A Moore Open-Ocean Profiling System for the Upper Ocean in Extene Self- Containe Deployments. Journal of Atmospheric an Oceanic Technology, 30 (7) /JTECH-D This version available at NERC has evelope NORA to enable users to access research outputs wholly or partially fune by NERC. Copyright an other rights for material on this site are retaine by the rights owners. Users shoul rea the terms an conitions of use of this material at Copyright 2013 American Meteorological Society (AMS). Permission to use figures, tables, an brief excerpts from this work in scientific an eucational works is hereby grante provie that the source is acknowlege. Any use of material in this work that is etermine to be fair use uner Section 107 of the U.S. Copyright Act September 2010 Page 2 or that satisfies the conitions specifie in Section 108 of the U.S. Copyright Act (17 USC 108, as revise by P.L ) oes not require the AMS s permission. Republication, systematic reprouction, posting in electronic form, such as on a web site or in a searchable atabase, or other uses of this material, except as exempte by the above statement, requires written permission or a license from the AMS. Aitional etails are provie in the AMS Copyright Policy, available on the AMS Web site locate at ( or from the AMS at or copyrights@ametsoc.org. Contact NOC NORA team at publications@noc.soton.ac.uk The NERC an NOC traemarks an logos ( the Traemarks ) are registere traemarks of NERC in the UK an other countries, an may not be use without the prior written consent of the Traemark owner.

2 JULY 2013 S E N D E T A L SeaCycler: A Moore Open-Ocean Profiling System for the Upper Ocean in Extene Self-Containe Deployments UWE SEND Scripps Institution of Oceanography, La Jolla, California GEORGE FOWLER, GREG SIDDALL, BRIAN BEANLANDS, AND MERLE PITTMAN Befor Institute of Oceanography, Dartmouth, Nova Scotia, Canaa CHRISTOPH WALDMANN Center for Marine Environmental Sciences, University of Bremen, Bremen, Germany JOHANNES KARSTENSEN GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Germany RICHARD LAMPITT National Oceanographic Center, Southampton, Unite Kingom (Manuscript receive 19 September 2011, in final form 1 August 2012) ABSTRACT The upper ocean, incluing the biologically prouctive euphotic zone an the mixe layer, has great relevance for stuies of physical, biogeochemical, an ecosystem processes an their interaction. Observing this layer with a continuous presence, sampling many of the relevant variables, an with sufficient vertical resolution, has remaine a challenge. Here a system is presente that can be eploye on the top of eep-ocean moorings, with a rive mechanism at epths of m, which mechanically winches a large sensor float an smaller communications float tethere above it to the surface an back own again, typically twice per ay for perios up to 1 year. The sensor float can carry several sizeable sensors, an it has enough buoyancy to reach the near surface an for the communications float to pierce the surface even in the presence of strong currents. The system can survive mooring blowover to 1000-m epth. The battery-powere esign is mae possible by using a balance energy-conserving principle. Reliability is enhance with a rive assembly that employs a single rotating part that has no slip rings or rotating seals. The profiling boies can break the surface to sample the near-surface layer an to establish satellite communication for ata relay or reception of new commans. An inuctive pass-through moe allows communication with other mooring components throughout the water column beneath the system. A number of successful emonstration eployments have been complete. 1. Introuction The upper layer of the ocean, from the surface to a epth of approximately m, is a very ynamic component of the oceanic water column. It contains important physical, biogeochemical, an biological Corresponing author aress: Uwe Sen, Scripps Institution of Oceanography, UCSD 0230, 9500 Gilman Drive, La Jolla, CA usen@ucs.eu processes, which nee to be observe with goo temporal an vertical resolution while maintaining a long presence in orer to unravel their interconnection or even just to gain information about the short-term variability, climate-riven responses, or long-term evolutions in this layer. For a wie variety of quantities it is necessary to know the vertical structure (graients or maximum/minimum layers) an/or the vertical integral or the vertical movement of layers. Prominent examples are the mixe-layer structure (ensity graients, heat istribution), phytoplankton (which usually have a DOI: /JTECH-D Ó 2013 American Meteorological Society

3 1556 J O U R N A L O F A T M O S P H E R I C A N D O C E A N I C T E C H N O L O G Y VOLUME 30 subsurface maximum), nutrients, or pco 2 (whose vertical istribution is neee for carbon bugets an fluxes). For these reasons, time series collecte with fixe-point sensors often eliver insufficient information. Some variables can now be observe with small an power-efficient sensors, such that they can be mounte on unerwater gliers or profiling floats, to obtain vertical profile information. Other variables require larger or more power-hungry sensors, for example, imaging flow-through systems like Laser Optical Plankton Counters (LOPCs) an wet chemical sensors for carbon variables or nutrients. Also, time series may be neee in locations where gliers cannot hol station well enough (strong current systems or in ey fiels). This requires a profiling technology that can be mounte on moorings to transport sensors through the surface layer. Moorings with a surface buoy are ifficult to use for profiling systems since the mooring wire can move violently uner the action of surface waves. The amage potential of the surface or near surface is also well recognize, an thus minimizing the time spent there is a common feature share by many profiling systems operating on subsurface moorings. Various profiling esigns have successfully employe variable buoyancy to rive a near-neutrally buoyant element up an own a taut mooring wire (Van Leer et al. 1974; Eriksen et al. 1982; Provost an u Chauffaut 1996; Walmann 1999; Buéus 2009). The near-neutral buoyancy requirement, which minimizes energy input, tens to restrict the instrumentation suite that may be carrie, an, since the force evelope by buoyancy change is quite small, ambient currents can negatively affect the system s ability to move vertically. The concept of operating on a taut wire has been ramatically extene by Doherty et al. (1999) using a motor/pinch wheel system running on a taut-wire subsurface mooring cable. This system has been eploye operationally on wie-ranging oceanographic stuies (Morrison et al. 2000; Krishfiel et al. 2008; Nikoloupoulos et al. 2009; Toole et al. 2011). Like the profilers that rely on buoyancy change, the riving force is low so ambient conitions can have an influence on performance an the near-neutral buoyancy requirement can impose instrument suite/power capacity challenges. But, epening on mooring configuration, ambient conitions, an water epth, these systems can operate from near the bottom an approach the near surface. Because all these esigns operate on subsurface moorings, there is the implication that, without some parallel structure or operating system, ata nee to be store internally. The current approach iscusse in this paper also employs a subsurface mooring but one that ens approximately 150 m below the surface an incorporates a winchlike system at this epth. This arrangement is more tolerant of extreme weather conitions since it can stay well below the surface when waves an win are too severe an is less likely to be amage by ships or vanalism. A winche system also avois the reef effect, that is, marine life that gathers aroun an attaches to near-surface moorings, an thus can observe a more unisturbe marine ecosystem since sensors parke at a epth of 150 m are less affecte by biofouling. But there are a number of challenges associate with a moore unerwater winche system that nee to be overcome. A main factor is the energy efficiency, assuming the entire mooring is self-containe an thus battery powere. Unerwater winch systems have been evelope, however, that can operate from the bottom or from a miwater platform. An innovative profiler that carries the winching component on boar a buoyant profiling element has been evelope (Barnar et al. 2010) an has been operationally eploye. This system is esigne to pierce the surface to permit ata transmission. But because the magnitue of the force, exerte by buoyancy, that is require to raise the system to the surface is highly epenent on the ambient current, the size of the profiling package, an operational epth, a potentially restrictive balance exists between the power available an the uration an number of applications of that force. Just the same, it has been emonstrate that many profiles are possible in weeklong eployments in shallow water (Sullivan et al. 2010; Babin et al. 2005). A compact variant of the onboar winch system has been use to obtain temperature ata from the upper water layer beneath Arctic ice by Pickart (2007). Plain winching requires significant energy to pull own a boy that has enough buoyancy to overcome the blowover ue to horizontal rag in typical ocean surface currents. Drag is especially serious when large an heavy sensors are to be eploye on the profiling boy. This ifficulty has been aresse by Fowler et al. (1997). Here wave energy is use to rive a buoyant profiling element own a mooring line, which is then permitte to rise uner its own buoyancy. The energy available permits the use of a substantial sensor package an makes the system insensitive to ambient current but also makes it virtually impossible to stop the profiling element in miprofile if a sensor might require it. Collecte ata are store internally an transmitte inuctively to the surface where a two-way communication system can transmit the ata to shore or receive an relay commans from shore to the profiling package. The ownsie to this approach is that keeping a permanent surface expression in place an functioning properly in all weather conitions is ifficult. This rive system has

4 JULY 2013 S E N D E T A L been later uplicate by others (Rainville an Pinkel 2001; Pinkel et al. 2011). A secon challenge is the operation of rotating mechanical parts an electric motors unerwater over long urations. Typically, this requires rotating seals an unerwater electrical slip rings, which increase the risk of failure when eploye for time perios in the orer of 1 year. A thir complication is the fact that subsurface moorings in the eep ocean (5000-m epth) may be blown over by strong current events such as eies. At high latitues these currents can be eep reaching an may cause the components that are normally at a epth of 150 m to be pushe own to epths of m. Thus the entire winch assembly nees to be pressure resistant to such epths. Finally, to establish communication to shore, it is necessary to break the surface an remain there while transmitting ata or receiving commans. This is hazarous an challenging because a large float with ample buoyancy will be subject to snap loaing in the wave fiel, while a small float may be continually swampe by waves or may not even reach the surface in the presence of currents. This paper presents an approach that tries to respon to all the challenges resulting from the above requirements an represents consierable collaboration between engineering an science teams over the course of 6 years to prouce the system now calle SeaCycler. Several ieas an principles are erive from an earlier system calle ICYCLER (Fowler et al. 2004), which was evelope for making aily measurements uner ice for a perio of a year. The solution an implementation presente here combine the following features: an energy-conserving principle, to increase power efficiency by an orer of magnitue over conventional systems (Fowler 2002); a totally enclose rive system (no rotating seals or slip rings) to increase reliability; a large instrument payloa (60 kg in air) permitting flexible scientific stuies; a Sensor Float buoyancy of 110 kg to allow surfacing in strong currents; a pressure rating of 1000 m, to allow eployment on eep open-ocean moorings; extra cable storage (total of 373-m net) to compensate for blowover in currents; a parking epth of 150 m, to avoi storm waves, reef effect, an vanalism an reuce biofouling; ambient wave sensing capability to avoi surfacing when conitions are too severe; a separate Communication Float to establish shore telemetry even when the Sensor Float remains below the surface; remote retasking; simple straight-through cable routing, anchor to surface, allowing inuctive moem coupling to eeperwater instrumentation; an enurance of approximately two or four 150-m profiles per ay in a yearlong eployment, for alkaline or lithium batteries, respectively; 550-kg buoyancy to help maintain a taut mooring; an an ability to surface the Sensor Float for maintenance without recovering the mooring. This system has been eploye for engineering an emonstration purposes on multiple occasions an uring the most recent eployment carrie out 644 roun-trip profiles from a epth of 150 m using an alkaline battery pack. 2. Technical implementation a. Mechanical esign As shown in Fig. 1, the SeaCycler system comprises three floats connecte by electromechanical cable. At the top is a Communication Float (short Comm Float, 5-kg net buoyancy), followe by a Sensor Float (105-kg net buoyancy incluing an extensive sensor suite). Both floats travel in tanem through the water column uner the action of the lower Mechanism Float (440-kg buoyancy), which also provies flotation for the mooring that connects it to the ocean bottom. The Mechanism Float contains a winch rum/motor assembly, shown in the etail in Fig. 1, which is not only highly efficient but also mechanically simple. The smaller iameter section of the rum stores 6-mm-iameter steel galvanize plastic-jackete mooring wire (1800-kg breaking strength), an the larger-iameter section carries a near-neutrally buoyant plastic-jackete, spectra strength member, three conuctor, profiling cable leaing to the Sensor Float. Rotation of the ouble rum prouces ifferential movement of the two cables in the ratio of the rum iameters, here set at 5:1. Since the cables are woun in opposite irections, rum rotation causes the profiling floats an the Mechanism Float to move vertically in opposite irections. Because the various buoyancies are carefully esigne to prouce tensions in the cables that are in the inverse ratio, that is, 1:5, the rum is in static balance an can therefore be rotate with very little torque an resultant power. Put another way, rotation of the rum changes the potential energy of the Sensor an Comm Floats but this is offset by an equal an opposite change in potential energy of the Mechanism Float. This energy-conserving principle has been patente. The balance of the system is critical for energy conservation, but minor variations are

5 1558 J O U R N A L O F A T M O S P H E R I C A N D O C E A N I C T E C H N O L O G Y VOLUME 30 FIG. 2. Cutaway view of the neutrally buoyant winch rum assembly showing how the torus motor, winch electronics, an battery packs are mounte. FIG. 1. Schematic showing the overall esign an configuration of the SeaCycler system. The Mechanism Float (MF) is typically parke at 165-m epth with the Sensor Float (SF) pulle in close. The Communication Float (CF) is connecte via 23 m of fixelength cable. During profiling, the MF moves ownwar while the SF ascens, in a 5:1 ratio. If there are no water currents an associate blowover of either the mooring with the MF an/or of the SF, the MF winches itself own to 195 m while the SF reaches the surface. To allow for mooring blowover, the total cable store allows for spooling out 466 m of cable for the SF, an this requires 93 m of cable capacity for the MF. At maximum payout the MF may thus be at a epth of 258 m, resulting in a net SF cable length (relative to 150 m) of 373 m, or 223 m of spare profiling capacity allowing for mooring blowover. Dimensions of the floats are as follows: for MF, length is 4.0 m, maximum iameter is 1.8 m, air weight is 1850 kg, an buoyancy is 440 kg; for SF, length is 2.5 m, maximum iameter is 0.6 m, air weight is 230 kg, an buoyancy is 105 kg; an for CF, length is 1.4 m, maximum iameter is 0.1 m, air weight is 18 kg, an buoyancy is 0.2 kg. Arrows on the left in the rum etail inicate biirectional rotation an the associate translation force by the axially mounte lea screw on the right. tolerable. The cables, which are alternately spoole on an off the rum, can contribute to an imbalance but are chosen, particularly the profiling cable, to have minimum in-water weight. As a result, only minor variations are etecte in the rive motor power consumption throughout a profile. Several challenges were met in the esign an integration of the winch rum s rive motor assembly. Primary among these was the nee to overcome the projecte cyclical unbalancing torques cause by wave forcing when profiling elements approach the surface. These forces, in combination with the large torque arm offere by the rum, force a new approach to unerwater motor esign. Instea of mounting the rive motor on the centerline, where immense output torque woul be require, it was connecte near the outsie iameter of the rum to a large internal gear. To resist anticipate high ambient pressures, this assembly was house in a torus-shape pressure case (1.1-m outer iameter) (Fig. 2). This geometry offers a substantial iameter to create torque while keeping the wall thickness of the pressure case thin (11 mm) to generate a lightweight assembly. The rive mechanism insie the torus consists of the large internal gear integral with a substantial steel ring that is supporte on five bearings mounte on the torus enclosure wall. These bearings isconnect the ring, an gear, rotationally, from the torus. The ring is eccentrically weighte to create a penulum. A small c motor (40 mm 3 70 mm, 150 W) mounte on the torus is engage with the internal gear on the ring so that when the motor rotates it causes the torus, with attache winch rum, to rotate aroun the centerline of the penulum ring. Since the batteries an control electronics are also locate insie the rum an thus rotate with the entire assembly incluing the motor, no slip rings are require to transmit the power nor is there a nee for any rotary seals. Significantly, gravity, working on the penulum ring, acts as an elastic vertical reference, or foot on the groun from which to create torque. When the torus rotates uner no-loa conitions the penulum ring remains comparatively stationary, but uner loa the penulum rotates to create torque so that the whole assembly is rotationally compliant an

6 JULY 2013 S E N D E T A L absolutely critical feature for a structure that operates in the wave zone. Notably, all the gearing an relative motion require to prouce rum rotation occur in air, within the torus itself, enhancing efficiency. At low profiling spees the major part of the energy require to move the Sensor an Comm Floats in the water column is prouce by the frictional forces within the mechanism itself. To limit frictional losses, the neutrally buoyant rum is riven horizontally by a fixe, axial lea screw uner stationary fairleas while cable, lai in a single wrap, is pulle in or pai out (Fig. 1). This eliminates the nee for power-consuming an mechanically complex spooling mechanisms. Since friction reuction is so critical for power conservation, great care was taken with the esign of all rotating elements. All fairleas an the main winch shaft are supporte on ball bearings that are enclose in oil-fille, pressurecompensate housings that isolate them from seawater an have proven to be highly efficient. Drum translation is supporte on simple low-friction bushings that consume little power at the extremely low translation spees involve. At first glance the winch rum may seem ungainly but its large size actually serves multiple purposes. The larger section is 1.15 m in iameter an 1 m long an is capable of storing 466 m of profiling cable in a single layer. It is also large enough to house the electronics an all the batteries (576 alkaline D-cells in four packs) neee to power rum rotation. Although the mechanism is in static balance ue to the buoyancies an cable wrapping, external forces such as hyroynamic wave loaing on the profiling floats as they approach the surface can impose significant torsional forces on the rum. These forces are resiste by the motor assembly with the torus-shape pressure housing having almost the same major iameter as the larger section of the rum itself. The motor is thus capable of substantial output torque. Finally, sufficient space is available to inclue enough syntactic foam to rener the whole rum assembly neutrally buoyant, which is essential to maintain level trim as the rum translates. The motor assembly, winch batteries, an control electronics all rotate with the rum proviing a seamless cable routing right through the entire SeaCycler assembly from the ocean floor to the surface. (Fig. 1, rum etail) b. Power bugets It is essential that aequate float buoyancy be provie to ensure that oceanographic sensors an communication elements reach the surface when high water currents are encountere. Further, both the ascent an escent must be accomplishe uner controlle conitions to ensure proper instrument function. For the operational parameters efine in this project where the parking epth is set at 150 m, an with a substantial sensor suite that can a to float size, moels preict that a combine buoyancy on the Sensor Float an Comm Float of 110 kg is require to lift the profiling elements to the surface when near-surface currents reach as high as 0.8 m s 21 (assuming no lower mooring knockover). Of course, this is an oversimplification. The mooring is affecte by currents throughout the entire water column, an when the system is moore in eeper water aitional buoyancy will be require beneath the assembly to keep the mechanism float within the profiling range. Actual fiel experience inicates that the SeaCycler operates with an average overall power consumption of 60.7 W, an this inclues power for mechanism control an monitoring electronics. Comparisons with a conventional winch system, where the profiling buoyancy must be pulle own by brute force but is allowe to free ascen uner control to the surface, are ifficult because of assumptions that must be mae about efficiencies an low loa power requirements. Nonetheless, calculations show that the SeaCycler shoul be on the orer of times more efficient. For equal onboar power that means times more profiles. The Mechanism Float carries 600 A h of energy at 24 V in alkaline batteries for profiling an to power the electronics. In the current configuration, power is aequate to complete 650 profiles: 150-m roun-trip profiles or 195-km profiler travel. The Sensor Float carries a 14-V lithium battery pack with 320 A h of energy that powers the main system control electronics, all the sensors [at present: conuctivity temperature epth (CTD) an issolve oxygen] plus the Comm Float electronics an transceivers. Replacing the Mechanism Float batteries with lithium cells woul permit more than oubling the number of profiles, or alternatively, completing up to two profiles per ay for a year in areas that experience much higher water currents that will nee more cable payout to reach the surface. To o this woul also require oubling the Sensor Float power since instruments will be on for longer perios of time. Within the 0.6-m circular envelope of the Sensor Float, with two 0.6-m bays, there is sufficient space an, by removing the current 20 kg of lea ballast neee to achieve balance, aequate buoyancy to accommoate this change as well as increase sensor payloa to eight instruments. c. Electronic interfacing, communication During profiling, main functional control, instrument management, winch control, an communication resie on the Sensor Float along with CompactFlash rive ata storage. During ata telemetry to shore, the Comm

7 1560 J O U R N A L O F A T M O S P H E R I C A N D O C E A N I C T E C H N O L O G Y VOLUME 30 Float becomes the master an the Sensor Float respons to its commans either locally from the Comm Float or remotely from shore via the Comm Float. Ancillary an backup CompactFlash rive ata storage is site on both the Mechanism an Comm Floats. Intercomponent communication among the three floats is accomplishe through a irect, full-uplex serial link using three conuctors on the interconnecting electromechanical cables. The Sensor Float manages the mission planning as well as ata file transfers between all floats. Functional control inclues parameters such as the profiling interval, profiling spee, an the minimum epth to which the Sensor Float is profile, or stop epth, on the way up. On the way own, stops can be orere to accommoate sensor equilibration. Depth control is effecte by the pressure signal from the onboar CTD ata. All of these parameters can be moifie by the shore operator uring any of the regular telemetry sessions. Provisions have been mae for the Sensor Float to wake up an/or reset any of the SeaCycler subsystems as require. The profiling sequence is governe entirely by Sensor Float commans, which can be isperse to all instruments an subsystems. In aition, an acoustic moem is inclue on the Sensor Float to provie control an status uring perios where the Comm Float is submerge. Currently, it is configure to act solely as a Full System Reset Mechanism to bring the Sensor Float to the surface in the case of a catastrophic electronic communication failure. Provisions have been mae, though, for auxiliary instrument ata transfer, system control, an status reporting. The Mechanism Float contains its own control system that respons to both simple an complex commans from the Sensor Float. Simple commans inclue functions such as turning the brake on or off, while more complicate commans can effect a complete surfacing profile base solely on the Mechanism Float s internally establishe criteria. The Mechanism Float electronics incorporates sensors that allow it to control an monitor all of its internal functions. Operating parameters, such as winch rum spee, maximum allowable torque, an motor current are accesse locally but can be overrien by commans irectly from the Sensor Float or from the shore operator via the Comm Float to the Sensor Float. Two-way communication over the Internet between a shore computer an the SeaCycler is accomplishe via an Iriium Communications, Inc., transceiver locate on the Comm Float, which also inclues a GPS engine. Local communication with the surface Comm Float, that is, to a ship in the vicinity, can also be accomplishe via a FreeWave Technologies, Inc., transceiver. The Comm Float activates a sniffer session at the beginning of each telemetry session. During this sniffer phase, a user in the area can ownloa ata or gain control of the mooring via FreeWave. If there is no FreeWave signal that is sense to talk to the Comm Float, it will follow up with an Iriium session attempt to shore. If the FreeWave attempt is successful, the Iriium session is abanone for that profile. The Comm Float is a completely self-containe communications subsystem. All of the Iriium, FreeWave, an GPS communications are controlle by the Comm Float electronics. Files estine for shore are typically transferre from the Sensor Float to the Comm Float uring the surfacing phase of the profile, where they are store in the Comm Float s internal file system. The Comm Float ata storage provies full reunancy for all files throughout a eployment. A comman from the Sensor Float then relinquishes control to the Comm Float where it will establish the connection, transfer files, an receive new commans from shore. All new files are automatically transferre to shore, but any of the archive files may be retransmitte at the request of the shore operator. Time upates from the GPS an commans from the shore operator are transferre to the Sensor Float to be later isperse throughout the system. The uninterrupte nature of the cable routing from the Comm Float through the Sensor Float through the Mechanism Float winch rum to the mooring line below means that irect communication is possible from shore to the ocean bottom. Currently, communication with instrumentation locate on the mooring line beneath the Mechanism Float has been accomplishe using an inuctive moem. Iriium/GPS emergency recovery beacons are locate on both the Sensor Float an the Mechanism Float. With a planne stan-alone power aition on the Comm Float, it will be able to act as an emergency recovery beacon as well.. Performance aspects There are four separate functional features that affect the ability of a system to approach the surface, pierce it to sen an receive ata, an then submerge. The first is the nee for extra profiling cable beyon the absolute epth of the system. SeaCycler carries 466 m of profiling cable that, when it is all eploye, results in a net upwar movement of 373 m by the Sensor Float to reach the surface. This in effect accommoates a 223-m mooring knockover. It must be note that the Sensor Float parks itself approximately 3 m above the Mechanism Float, an as such imparts a small profiling gap (or blin spot) between the top of the Mechanism Float an the Sensor Float. The part of the mooring between the

8 JULY 2013 S E N D E T A L FIG. 3. Communication Float in its operating position at the surface. Tank an fiel stuies have shown remarkable stability with waves ranging from capillary to win waves an swell, always keeping the antennas out of the water. parking epth an lowest possible epth of the Mechanism Float is also a section where the mooring can carry no sensors an where the Sensor Float oes not reach. In the current configuration this epth range is 93 m long an woul be a blin spot unless sensors esire for this interval are mounte on the Mechanism Float. The secon is the effect that varying wave forces have on any structure or boy at or near the surface. These forces can have a very negative effect on the longevity of systems that are unyieling an have the potential of imposing exaggerate snap-loas on fixe cable structures. The esign of the SeaCycler motor, however, has built-in an automatic compliance that raically reuces potential stress on the system an can, uner certain circumstances even give up cable if forces become excessive. The thir aspect, piercing the surface, is accomplishe by SeaCycler s Comm Float. This relatively small component, about 1.5 m long, floats near vertical when submerge at the top of a 23-m-long ouble-armore steel cable that is renere neutrally buoyant by the aition of iscrete syntactic foam buoyancy elements. When it pierces the surface it flips to an almost level attitue because of off-center ballasting. In this state it projects a three-element antenna above the surface; see Fig. 3. The combination of neutrally buoyant cable leain, ballast placement, an the very large water-plane area create by its near-horizontal attitue ramatically enhances stability, allowing the Comm Float to transmit an receive messages in significant waves. Data have been successfully transferre in 4.1-m waves. The fourth function, submerging in heavy weather, however, represente a significant challenge in early sea trials. It was foun that when the weather got rough, in seas of over 4 m, the Comm Float coul sometimes be left on the surface for extene perios after an Iriium communication session. Wave rag force on the profiling elements exceee maximum motor torque so that they coul not be haule own. This was eventually overcome with a stratagem that took avantage of Sea- Cyler s unique motor/energy balance principle. As note, the three buoyancies that compose the assembly are organize to maintain balance. When this balance is upset, for instance, when transient wave forces are encountere, the system attempts to restore this balance automatically an autonomously in a very useful way. In the normal stoppe position, for example, when on the surface an transmitting, the system is locke with an internal brake. We foun, however, that if the brake was isengage, the system s preisposition to maintain balance combine with the Mechanism Float s large buoyancy took over an the profiling elements were ratchete own by passing waves as the Mechanism Float, momentarily out of balance with applie cable tensions, rose in the water column to restore balance. This technique has become stanar proceure, an the system has been programme to remove the brake for 2 min after each surfacing session. Even in relatively calm, 1-m seas, the profiling elements are often haule own to a epth of 10 m. But as wave height increases, the ratcheting effect becomes more intense so that, instea of expening consierable energy to submerge, the waves provie a free rie own to 20 m or more in larger waves. This is particularly avantageous in helping the SeaCycler escape from rough sea conitions. The more severe the threat is from waves, the eeper the waves rive the profiling floats own away from the challenging wave environment, thus protecting the system from potential amage. It shoul be note that surfacing the Sensor Float on comman allows it to be accesse, for example, to service or replace sensors, while keeping the remaining mooring incluing the Mechanism Float in place an operational. Figure 4 shows the Comm an Sensor Floats on the surface. 3. Demonstration Between March 2010 an May 2011, seven eployments have been accomplishe: three in shallow local waters, two in ;150-m water epth 32 km off Halifax, an two at the ege of the Scotian shelf in ;1100-m epth 250 km offshore. These fiel tests were combine with countless laboratory an jetty tests. The five inshore an near-shore test eployments were of short uration, typically 3 ays, with the offshore eployments lasting 74 an 41 ays, respectively. As woul be expecte for a evelopment this ambitious, early eployments ientifie minor shortcomings. These were correcte with aitional innovations or aitions to

9 1562 J O U R N A L O F A T M O S P H E R I C A N D O C E A N I C T E C H N O L O G Y VOLUME 30 FIG. 4. Photographs showing the Comm an Sensor Floats on the surface uring the mile of eployment. The CTD sensor is out of the water, an the remainer of the Sensor Float remaine submerge. culminate in the last eployment, which was highly successful both from a performance perspective but also from the stanpoint of operational evelopment. Chief among these was the implementation an refinement of the autonomous wave-riven submergence. Over the uration of the last eployment the power savings realize through this technique represente twenty-six 300-m roun-trip profiles, or 4% of the 644 profiles complete, expening no rotational power at all. In the local waters tests, the Mechanism Float was towe to the eployment site an the other two floating components were streame aft before eploying mooring line an ropping the anchor. The offshore eployments were accomplishe using Coast Guar vessels of various types, but in all cases operations were conucte from the foreeck or waist rather than from the stern. This cumbersome metho was only mae possible with the ai of a seconary small boat to tow the floating components away from the ship an keep them organize in a straight line as the ship move away crabwise, eploying mooring line, before ropping the anchor. Operational plans call for working from an oceanographic vessel where components can be eploye sequentially, mooring top first an anchor last, from the stern, which is our normal practice. It goes without saying that, whatever platform is use to eploy the large Mechanism Float, care an proper rigging are essential to combat its potentially large inertial forces. Figure 5 shows all three float boies on a Coast Guar vessel prior to eployment. A significant portion of the testing process was concerne with the evaluation of communication capability. Initial satellite communication ifficulties were ientifie as a possible compatibility issue with the TCP/IP stack in Microsoft Corporation Winows XP an the shoresie server software. After migrating to Winows 7 (which has a more current TCP/IP stack esign), the FIG. 5. View of all three float boies on a Coast Guar vessel prior to eployment. The Sensor Float is seen to have ample spare capacity for aitional sensors, batteries, or electronics. problem isappeare. Further investigation is ongoing with Microsoft an Iriium to fully unerstan the matter, but for now it is not viewe as a serious issue. This malfunction resulte in many roppe calls but, for these, ata were recovere uring shore-requeste retransmission. The system s operating characteristics such as profile scheule, profile stop epth, an torus-motor penulum angle, which efines available motor torque output, were varie from shore. This was one primarily to test functionality but at the beginning of the eployment we were actually learning how to best run the system an garner some iea of what the operational limits might be. In fact, the team is still learning about how the system respons to its environment an what is the best way to set parameters to maximize operational efficiency. At the beginning of the eployment, maximum motor current was varie to assess its impact on SeaCycler s ability to approach the surface in varying win an wave conitions, an this is easily seen in the early part of the oxygen recor of Fig. 7 as stops occurre as eep as 30 m. Typically, we start to see or feel the effects of the surface as eep as 45 m. Once we ha gaine some information on performance, normal Sensor Float stop epth was set at 5 m. But on 82 occasions it was brought to within 1 m of the surface, an on 23 profiles the CTD water inlet was surface into air. Inee, on comman, the top en of the Sensor Float itself was actually brought above the surface. The graph in Fig. 6 shows, with respect to wave height, the occasions when the system i not achieve its instructe stop epth. After the initial experimentation with surface approach, only eight profiles faile to reach esire epth, which represents only about 1.3% of the total number of profiles. Even though the Sensor Float i not reach requeste epth, the 23 m of cable above it meant that the Comm Float was at least able to make an attempt at communicating with the satellite with routine success. The otte tren line shows the anticipate upper limit of profiles with respect to wave height an confirms the

10 JULY 2013 S E N D E T A L FIG. 6. Profiles that i not reach the requeste stop epth are shown for various wave heights. These represent a very small number relative to the number of profiles complete. Even though these profiles stoppe early, there was still an excellent chance that the Comm Float woul pierce the surface to relay ata ue to the 23-m cable separation between the Comm Float an the Sensor Float where these epth measurements were actually mae. original esign stuy results. One faile profile is not shown on the graph since profiling was terminate after only 7 m of travel because of an unexplaine motor shaft encoer error. On the other en of the spectrum, successful two-way communication was emonstrate in wave heights over 4 m. The instrumentation carrie on all the eployments worke flawlessly with 100% ata recovery rate. There were occasions when ata file transmissions were terminate prematurely, a few for no apparent reason, but invariably these were recovere on comman in a later transmission. Some instrument ata are shown in Fig. 7 for the 644 profiles of the most recent eployment. Power consumption was foun to be very close to original estimates with an average winch power expeniture of 60.7 W while profiling or 15.1 W h per 300-m roun-trip profile, which inclues aitional power emans of surfacing an submerging. The total number of profiles complete is commensurate with the original esign objective. Although project planning calle for only 365 profiles, supplementary battery power was provie to eploy an recover an aitional amount of cable to surface the profiling floats in higher water currents. The site chosen for the eployment has been extensively stuie over past years, an, although currents were juge to be low, it was anticipate that occasional higher-current events coul be expecte. In the event, water currents at the site prove to be consistently very low so that only 2 4 m of aitional cable was require to reach the surface. The extra energy conserve by reuce profiling istance was use to complete 644 m m roun-trip profiles instea. 4. Outlook an future applications At the time of writing, SeaCyler is in the water for a test eployment as part of a National Science Founation FIG. 7. Time series isplay of the real-time recovere ata for all 644 profiles from the eployment in 1100-m water epth in the open ocean off Halifax (April May 2011), together with wave conitions from a near-by National Data Buoy Center buoy. The lowest two panels show ata that were retrieve from a Sea-Bir Electronics, Inc., MicroCAT farther own in the mooring, using the inuctive communication capability mae possible by the single connecte cable routing from the Comm Float to the mooring wire below SeaCycler. (NSF)-fune Ocean Observatories Initiative (OOI) effort. For this it carries a pco 2 sensor an an acoustic current meter in aition to the CTD an issolve oxygen sensor. The plan is to migrate the technology from the Befor Institute of Oceanography (BIO) Ocean Physics group to the commercial manufacturer/venor, Rolls-Royce Canaa Limite Naval Marine. We feel confient that the SeaCycler principle provies a very robust an energy-efficient metho of obtaining profiling ata in the upper ocean. Sensors that len themselves to integration range from CTD an current meters, fluorometers an backscatter sensors, an incoming raiation sensors to acoustic zooplankton sonars, wet chemical systems for carbon an nutrient measurements, anlopcsystems.itispossibletomovetosteelwire for all cables, proviing more fishbite resistance. In this case aitional electronic cable communication complexity will be necessary to permit operation using a single

11 1564 J O U R N A L O F A T M O S P H E R I C A N D O C E A N I C T E C H N O L O G Y VOLUME 30 conuctor rather than the multiconuctor system currently employe. The aitional weight of the wire spoole out can be compensate by tapere rums to keep the system balance. Experience nees to be gathere with proceures an possibly harware for safe eployment an recovery of the large an heavy SeaCycler system. Recovery may be simplifie by first etaching SeaCycler from the subsurface mooring with an acoustic release this will be explore uring the OOI test eployment. An aitional moification for future applications may be possible by proviing power to the mechanism float from below (in case a seafloor cable is available to provie power). Also, this version of SeaCycler is a most ambitious esign, allowing for blowover to 1000-m epth. It may be possible to buil moifie versions for coastal applications that only nee to operate to epths of 200 m. Overall, SeaCycler is an unerwater moore winch system that is esigne for applications in emaning situations, that is highly flexible an robust, an that has proven its reainess for extene fiel eployments in research applications. Acknowlegments. We acknowlege funing from the European Commission integrate project CARBOOCEAN, Contract an from the NSF OCE Technology Grant OCE Many people gave generous support to the esign team at BIO. Jim Hamilton provie, an inee continues to provie, invaluable insights into mooring performance, which allowe us to properly establish the system s operational parameters. In many aspects of mechanical esign an execution, Neil Mackinnon provie us with critical practical insights an execution as i Rany King, Dan Moffatt, an Scott Young, who came out of retirement to help us. Mechanically, much of the Sea- Cycler was constructe within BIO, an this woul not have been possible without the eication an skill emonstrate in the machine an weling shops uner the guiance of John Conro. On the electronics sie we were fortunate to have Mike Vining, Jeremy Lai, George States, an, in particular, Don Belliveau who, besies acting as an oft-consulte intellectual resource, was our meiator with management an the Coast Guar. Much of the project s invaluable testing was accomplishe with ships generously provie by the Canaian Coast Guar with Dave Morse as our constant avocate in procuring ship time. Deployment an recovery woul not have been possible without the active participation of our Technical Operations group with Rick Boyce, Jason Burtch, an Jay Barthelotte. Aministratively, an one cannot iscount this important contribution, we were supporte by Val Pattenen, Sany Burtch, an Helen Dussault, with the ivision s manager Tim Milligan campaigning within the science community an upper management on our behalf. Special thanks are ue to Simon Prinsenberg who has acte as our unremitting science champion right from the first conceptual esign iea. Many others at Scripps Institution of Oceanography, the University of Bremen Center for Marine Environmental Sciences, an GEOMAR Helmholtz Centre for Ocean Research Kiel helpe the project to success. The machine shops at Scripps Institution of Oceanography uner supervision by Ken Duff took on the major challenge of manufacturing the torus, with the help of Eric Slater, as well as many other components from rawings prouce a continent away. Early engineering insights an guiance were provie by Lloy Green. In this eneavor, invaluable coorination an assistance were provie, an continue to be reaily given, by Matt Molovan of Scripps. Other components such as the Comm Float an the Sensor Float were constructe entirely in Germany in Kiel an Bremen by Anreas Pinck an Markus Bergenthal an were successfully integrate with the assembly an ocean away. Finally, the esign team thanks the science principal investigators for their unfailing an patient support an encouragement throughout the life of the project. Even as we explore ea ens an encountere technical roablocks they never wavere. It has been a rewaring experience to have worke with them. REFERENCES Babin, M., an Coauthors, 2005: New approaches an technologies for observing harmful algal blooms. Oceanography, 18 (2), Barnar, A. H., an Coauthors, 2010: The coastal autonomous profiler an bounary layer system (CAPABLE), Proc. Oceans 10 MTS/IEEE Conf., Seattle, WA, IEEE, 1 7. Buéus, G. Th., 2009: Autonomous aily CTD profiles between 3700 meters an ocean surface. Sea Technol., 10, Doherty, K. W., D. E. Frye, S. P. Liberatore, an J. M. Toole, 1999: A moore profiling instrument. J. Atmos. Oceanic Technol., 16, Eriksen, C. C., J. M. Dahlen, an J. T. Shillingfor Jr., 1982: An upper ocean moore current an ensity profiler applie to winter conitions near Bermua. J. Geophys. Res., 87 (C10), Fowler, G. A., 2002: A moore energy conserving oceanographic profiler. Proc. Oceanology Int. 2002, Lonon, Unite Kingom, Spearhea Exhibitions, 11 pp. [Available online at gc.ca/science/research-recherche/ocean/ice-glace/ocuments/ fowler01.pf.], J. M. Hamilton, B. D. Beanlans, D. J. Belliveau, an A. R. Furlong, 1997: A wave powere profiler for long-term monitoring. Proc. Oceans 97 MTS/IEEE Conf., Halifax, NS, Canaa, IEEE,

12 JULY 2013 S E N D E T A L. 1565, G. R. Siall, an S. Prinsenberg, 2004: An energy conserving oceanographic profiler for use uner mobile ice cover: ICYLER. Int. J. Offshore Polar Eng., 14, Krishfiel, R., J. Toole, A. Proshutinsky, an M.-L. Timmerman, 2008: Automate ice-tethere profilers for seawater observations uner pack ice in all seasons. J. Atmos. Oceanic Technol., 25, Morrison, A. T., III, J. D. Billings, J. D. Doherty, an K. W. Toole, 2000: The McLane moore profiler: A platform for physical, biological an chemical oceanographic measurements. Proc. Oceanology Int. 2000, Lonon, Unite Kingom, Spearhea Exhibitions, Nikolopoulos, A., R. S. Pickart, P. S. Fratantoni, K. Shimaa, D. J. Torres, an E. P. Jones, 2009: The western Arctic bounary current at 1528W: Structure, variability, an transport. Deep-Sea Res. II, 56, Pickart, R. S., 2007: Reaching up into perilous, icy waters. Oceanus Magazine. [Available online at viewarticle.o?i ] Pinkel, R., A. Golin, J. A. Smith, O. M. Sun, A. A. Aja, M. N. Bui, an T. Hughen, 2011: The Wirewalker: A vertically profiling instrument carrier powere by ocean waves. J. Atmos. Oceanic Technol., 28, Provost, C., an M. u Chauffaut, 1996: Yoyo Profiler : An autonomous multisensory. Sea Technol., 10, Rainville, L., an R. Pinkel, 2001: Wirewalker: An autonomous wave-powere vertical profiler. J. Atmos. Oceanic Technol., 18, Sullivan, J. M., P. L. Donaghay, an J. E. B. Rines, 2010: Coastal thin layer ynamics: Consequences to biology an optics. Cont. Shelf Res., 30, Toole, J. M., R. G. Curry, T. M. Joyce, M. McCartney, an B. Peña- Molino, 2011: Transport of the North Atlantic Deep Western Bounary Current about 398N, 708W. Deep-Sea Res. II, 58 (17 18), Van Leer, J., W. Düing, R. Erath, E. Kennelly, an A. Speiel, 1974: The cyclesone: An unattene vertical profiler for scalar an vector quantities in the upper ocean. Deep-Sea Res., 21, Walmann, C., 1999: Performance ata of a buoyancy-riven eepsea YoYo profiler for long term moore eployment. Proc. Oceans 99 MTS/IEEE Conf., Seattle WA, IEEE,