The small AUV - a tool for detecting marine life

Remote sensing of the oceans by satellite sensors is perhaps the only tool by which one can study large-scale oceanic processes both in space and time. For example, basin scale maps of chlorophyll distributions due to phytoplankton can be derived from ocean color spectro-radiometers fitted on sun synchronous satellite platforms like the OCM monitor on IRS-P4, the MOS sensor ( Modular Opto-electronic scanner ) on IRS-P3, and the well known SeaWiFs sensor on the Seastar satellite. However, satellites provide integrated information of the surface layer of the ocean only. In contrast, conventional ships provide the opportunity to study the vertical structure of the water column, but offer only point sampling of the sea surface. The ship approach suffers from the problem of extreme under-sampling and inadequate spatial resolution.
 
 

An extended arm of the ship - the small AUV

Between these two sampling approaches, and one that has not been sufficiently exploited, is the possibility of sampling the environmental "volume" around a research vessel in great detail. These notes address the notion of using small Autonomous Underwater Vehicles [AUV] as extended arms of the ship to support research at sea. The AUV approach overcomes the inherent drawbacks of satellite and ship platforms, as they could be used to explore the surface properties of a wide area ( several kms. ) around a ship, and by enabling repeated dives of the vehicle, detailed knowledge on water column properties can be obtained.
 
 

A small AUV for use in the coastal zone would offer significant advantages-

• There is a scarcity of data in the coastal areas of most countries. Variability of ocean processes over short spatial and time scales is also largest along the coastal belts. AUV's deployed in the surf zone will be important.
• Ships are expensive machines to run, and hence ship time must be properly utilised. The small AUV must complete it's mission within the allocated ship halt, while oceanographers carry out normal investigations on the ship.
• Ease of deploying and retrieving it without a special crane or winch on a ship. Retrieval could happen by driving it into a net floating on the sea surface.
• No exacting technological demands on range and endurance. A typical endurance of ~ 1 hour to match average station halts during a research cruise. At an average speed of 4 kts ( ~ 2m / s), the AUV range could be as large as 3.5 kms around the ship.
• Vertical dives of the small AUV would be at best to 100 m, but more typically 50 m. At a dive speed of 1m/s to 70 meters, the AUV would be up at the surface within 3 mins. We can therefore achieve in principle, 20 repeated dives in 1 hour of ship halt. However, a carefully crafted mission scenario would have to worked out so that a judicious balance between areal coverage and vertical descents is made so as to optimally utilise allocated ship halt times.

Past developments in AUV's have tended to confront underwater technological challenges of long mission ranges of several tens of kms, and deep depth dives to 600 meters. The application area has been in bathymetric surveys and pipe line inspection. Consequently, these developments have resulted in a class of large heavy AUV's with buoyancy of 1300 kgs , body lengths of ~ 5 meters or more, and high costs ~ $200,000 thereby rendering them beyond the budgets of most oceanographic applications. Present developments of small AUV's are being seen in developments of REMUS ( WHOI), Fetch ( Silas Patterson Inc.), Ocean Explorer ( FAU), Cetus ( LMC ). Most of these have been tested as "prototypes' from several research institutions and universities in the US , Europe and Japan.

In the small AUV approach, we focus on science problems that will shape the specifications of the AUV. In particular we chose the following marine life problems -

• Arabian Sea Paradox
• Myctophid mystery
• Respiration rates in the oxygen minimum layer
• Satellite validation of ocean color
• Trichodesmium blooms

1. The Arabian Sea Paradox refers to the apparent invariance in space and time of the mesozooplankton standing stock, despite large seasonal oscillations in primary production over short spatial and time scales. This constancy in zooplankton biomass is even present in the less productive intermonsoon season between April and May ( Madhupratap et al 1996 )
2. Myctophids are mesopelagic fish that are strong diel migrators, which ascend to the sea surface layer at night and descend to depths of 200 to 400 m during the day. Myctophids occur in huge amounts everywhere in the Arabian Sea, even in oligotrophic waters, and are especially adapted to survive in low ambient oxygen layers (< 0.1 ml/l). There have been few direct measurements of the abundance of myctophids estimated at 80g.m^-2 (Gjosaeter 1984). A rough calculation by (Naqvi personal communication) shows that primary produced carbon at a rate of ~ 2g.m^-2/ day through the year would have to be produced so as to support huge biomass abundance of myctophids. So how do myctophids survive in such vast numbers in the Arabian Sea, and where are the primary producers located ?
3. A third riddle referred to as the 'respiration riddle' by Naqvi (1993a,b ) is the large bacterial production and high enzymatic respiration rates (inferred from electron transport system [ETS] data) in the oxygen minimum zone (150 to 500 m) in the Arabian Sea. This results in carbon consumption rates that are an order of magnitude higher that what be produced by sinking particulate carbon from surface layer primary production. This requires additional pathways of carbon to suboxic waters. How does this happen?

The resolution of these three inter-linked problems require further detailed investigations with a suitably equipped AUV having -

• A scientific echosounder mounted on the AUV to monitor changes in the migrating myctophid layer whist exploring the spatial extent and biomass abundance of the layer beneath the ship.
• The night time shoaling of the myctophid layer ( otherwise known as the deep scattering layer) to the sub-surface chlorophyll maximum (SCM ) could be studied in detail by positioning the AUV within the SCM to record particle properties with a beam attenuation meter and chlorophyll levels with a fluorometer, and to make periodic dives to the SCM layer ( ~ 70 meters).

Other suitable sensors ( oxygen, temperature) could be added. A well planned mission programme would have to be formulated for the AUV to execute.
 

An optics equipped AUV can be made to roam around the ship sampling the water column over several satellite pixels around the ship. The Indian satellite OCM which was launched recently has a pixel size of 360 meters compared to a pixel size of 1100 meters for the SeaWiFs sensor. If the range of the AUV is ~ 3kms then optical data averaged over 16 pixels of OCM sensor would be statistically more meaningful that having only 1 pixel coverage from the ship as would be obtained from conventional ship cruises.

The present way of validating satellite ocean color sensors is by taking the ship along a defined cruise track. At each station, the light fields of irradiance [Ed] and radiance [Lu] are measured at the surface and vertically along the water column upto a depth of ~ 60 meters. At each meter, the spectral signatures of Ed and Lu are measured in real time using a cable connected to the instrument. Soon after an optical cast, the biologist collects samples of sea water at fixed depths which then are analyzed to give the 'chlorophyll ' profile.
 
 

The optical AUV will have the following characteristics

• Shallow water dives to 100 meters
• It has Ed and Lu sensors that can be mounted on a pair of retractable or fixed fins. This is important as the AUV body must not shadow the sensors, or the optical data will be invalid.
• AUV dives to say 100 meters, and then ascends with no thruster power so as to measure an undisturbed light field.
• The chlorophyll field can be measured with an active fluorometer.
• Ability to roam upto a distance of 3 kms from the ship with programmed dives, and endurance of ~ 1hr..
• AUV is used only during daylight hours, and in calm weather

Trichodesmium blooms
 

Trichodesmium blooms are of great scientific interest now in the Satellite Remote Sensing Community. The phytoplankton in this bloom fixes nitrogen gas (N₂ ) under fully aerobic conditions while photosynthetically evolving oxygen It is now known to occur throughout the oligotrophic and sub tropical oceans. The (N₂ ) fixation property of Trichodesmium is likely to be a major input to the marine and global nitrogen cycle.

Trichodesmium Paradox

Although Trichodesmium is seen easily in the tropical seas, it has been difficult to detect it remotely by satellite. A robust detection protocol needs to be developed, but this is hampered by the fact that the spectral signature of Trichodesmium is strikingly identical to that of sediment plumes. High resolution optical spectra at sea using hyper-spectral radiometers would resolve this problem. Present day commercial radiometers used routinely by oceanographers are equipped with a few wavelengths that correspond to the satellite wavebands. These low-resolution radiometers miss the spectral structure from the phycoerythrobilin [PEB] and phycourobilin [PUB] pigments in Trichodesmium. With a hyperspectral sensor, the diel variations in the ratio of [PUB/PEB] can be studied in detail throughout the day, and used in a new protocol.

There is a further logistic problem of measuring Trichodesmium spectra from a ship, as the act of lowering the radiometer into a Trichodesmium patch tends to separate and break the patches. This is because Trichodesmium is highly buoyant, and tends to be driven apart by the smallest disturbance on the sea surface

Sallow water AUV in a process study of Trichodesmium

An AUV could solve the problem of recording high-resolution Trichodesmium spectra, and the way would be to:

• To make the AUV ride below a Trichodesmium patch.
• In this mode the optical sensors of radiance and irradiance could be deployed outwards from the AUV body for spectral collection.
• Other parameters of interest are temperature, chlorophyll, nitrates, and particle backscatter coefficients.

A small AUV would require small sensor payloads to be accommodated within the limited space permitted. Fortunately, off-the-shelf sensors of small size are now available for most oceanographic parameters. They have built in data loggers with serial outputs, low power requirements, and low costs. Some examples are:

• Miniature high resolution multi-spectral radiometers that measure the complete spectrum from 280-720 nm ( see www.trios.de )
• Portable scientific echosounders for use in shallow water environments ( see www.simrad.com )
• Multi parameter sensor packs combining CTD and fluorescence in one unit ( see www.chelsea.co.uk or www.falmouth.com )
• ADCP's from various commercial firms ( Sontek, RDI, and Nortek)

Other sensors rate gyros, doppler logs , GPS and radio modems for control and navigation purposes are also available in small dimensions.

The target audience for the small AUV application is the oceanographic community, and in particular its use to study and monitor marine life in the coastal zone, as these present the most challenging problems to solve. Some general specifications can be drawn -

• Each problem will require specific payloads on a sensor nose-cone to be fitted on the vehicle for a specific mission namely optics for satellite validation of water leaving radiance, acoustics and optics to study myctophids or blooms
• Single thruster and at least two fins for diving in shallow depths ~100 m
• The AUV should be able to 'sniff' a sub-surface chlorophyll maximum, or an oxygen minimum layer, or detect the extent of the mixed layer depth. This would need oceanographic variables to play a critical role in the control algorithms for the AUV.
• The small AUV will use only GPS position fixes and a high accuracy attitude unit to navigate in the water column without acoustic transponders. The AUV would need to re-surface periodically so as to update its dead reckoned coordinates.
• A lightweight add-on aluminum framework with small thrusters attached to the basic hull, would make it possible to examine coral reefs with a video camera or to hover in shallow water seabed regimes.
• Low endurance times ( ~ 2 hrs in the worst case ), and low range capability upto 3 kms.
• A user friendly mission programming interface using graphics that would require no special purpose computer skills from the oceanographer.
• High energy density batteries ( lithium ion , lithium polymer ) that would not require frequent battery pack changes.

The era of the small AUV is clearly at hand. There have been major advances in the technology of miniaturizing scientific sensors, and in the control and navigation algorithms of underwater vehicles over the last decade. Battery technology on high energy density cells is emerging, and could be usefully employed to power the small AUV for long periods. Similar advances in thruster motors with built in motor controllers, rate gyros, fin controllers. doppler logs, and miniature Ultra Short Baseline (USBL) systems are in the market, and can be used in the small AUV. Miniature GPS systems, and high-speed radio modems are other breakthroughs that benefit the small AUV development. On the software side, embedded micro-controller cards with real time operating systems running high level languages make it sufficiently easy to program the AUV in a well structured modular approach that is quickly amenable to change, and to fault detection analysis.

Research cruise ships carrying a variety of mission specific small AUV's could address a wide variety of investigations that would encompass the major disciplines in oceanography - biology, chemistry and the physical oceanography of the coastal zone.

Finally, we may consider the possibility of networking several AUV's in coastal areas to search and coordinate an intelligently coordinated data collection exercise at sea by reporting to the mother ship.

Gjosaeter J.(1984) Mesopelagic fish, a large potential resource in the Arabian Sea . Deep Sea Res., 31 , 1019-1035

Madhupratap, M., T.C. Gopalalkrishnan, P. Haridas, K.K.C. Nair, P.N. Arvindakshan, G. Padmavati and S. Paul ( 1996) . Lack of seasonal and geographic variation in mesozooplankton biomass in the Arabian Sea and its structure in the mixed layer . Curr. Sci., 71, 863-868

Naqvi, S.W.A., M. Dileep Kumar, P.V. Narvekar, S.N. DeSousa, M.D. George and C. D'Silva ( 1993a) . An intermediate nepheloid layer associated with high microbial metabolic rates and denitrification in the northwest Indian Ocean . J. Geophysics Res., 98, 16469-16479.

This page is maintained and was last updated 22-Nov-00 by Pip Sumsion (sumsionp@dfo-mpo.gc.ca).