91av

THROUGH the gloom you can just pick out the shape of the science module, its
titanium exterior bristling with sensors. From ports in its side, communications
cables snake out across the alien terrain. Occasionally, a robot vehicle
trundles out of the blackness, docks and downloads the instructions for its next
job—an assignment planned by scientists watching the scene on flickering
TV monitors.

But this isn’t Mars—it’s the next step in a very Earth-bound branch of
science. Tired of having to settle for mere snapshots of life beneath the waves,
marine scientists are stashing their wellies and oilskins at the back of the
darkest locker in favour of space-age technology that will give them a permanent
presence in the deepest, darkest recesses of our planet—the sea floor.

It’s a difficult world to study. The most interesting things happening in the
ocean tend to be sudden and short-lived, such as volcanic eruptions or blooms of
plankton. If you’re not in the right place at the right time, you miss all the
fun.

Not any more though. Two American marine scientists have hatched a plan to
keep a permanent eye on vast swathes of the ocean. They are about to start
building a sophisticated network of sensors, robot subs and autonomous labs
across the seabed, all linked to the Internet by fibre-optic cables. With this
array, aptly named Neptune, scientists will survey over half a million square
kilometres of the north-eastern Pacific without ever leaving their cosy,
comfortable labs. At their disposal will be a fleet of robot subs, ready to
scuttle off at a moment’s notice to check out a new eruption. Without stirring
from their computers, they’ll be able to collect delicate samples, deposit them
in deep-sea freezers or even analyse them immediately, right there on the
seabed.

The plan is nothing if not ambitious. First conceived ten years ago by John
Delaney, a marine geologist at the University of Washington in Seattle, and Alan
Chave from the Woods Hole Oceanographic Institution in Massachusetts, it will
unite dozens of researchers from institutes across the US and Canada, and comes
with a hefty $200-million price tag. Neptune will require dozens of lab
modules, robot subs and a vast power and communications grid that can survive
for up to 30 years in corrosive sea water at an average depth of two thousand
metres. Yet its designers aren’t jumping straight in at the deep end.

Neptune is a natural extension of marine scientists’ growing penchant for
building observatories in the ocean. In the mid-80s, researchers began to deploy
arrays of buoys that recorded weather conditions, currents and water temperature
from mid-ocean and beamed the data back to researchers via satellite. Then, in
1996, marine scientists built the long-term ecosystem observatory (LEO-15, their
first permanent lab on the sea floor) to study the shallow waters off New
Jersey.

LEO-15 has two instrument-laden modules set on the seabed 15 metres below the
waves, connected to shore by seven kilometres of fibre-optic cables. Because it
is hard-wired to shore, LEO-15 can gather and record a lot more data than
free-floating buoys. And as well as giving ocean scientists a long-term picture
of the interactions between the sea and the shore, LEO-15 is also the first
network of its kind to employ a fleet of autonomous underwater vehicles
(AUVs)—torpedo-shaped, propeller-driven subs that are loaded with
instruments and follow a pre-programmed course when launched into the sea.

A lot further from the safety of shore is the Hawaii-2 Observatory, the first
cabled outpost in the deep ocean, which started up last year. A titanium frame
sitting in more than four thousand metres of water, H2O is essentially a
junction box with six sockets into which a variety of experiments can be
plugged. It is linked to Hawaii, more than two thousand kilometres away, via a
disused trans-Pacific telephone cable. Since it is also equipped with a
seismometer, H2O gives geophysicists the first-ever live seismic data from the
deep ocean floor.

Neptune will be a purpose-built deep-water observing system. It will be sited
in an area of immense scientific interest—the Juan de Fuca plate, just off
the Pacific coast of North America
(see Diagram). The wide variety of geological
processes happening at the edges of this tectonic plate make it a magnet for
marine geophysicists. “It embodies most of the basic processes that operate on
our planet,” says Delaney.

Juan de Fuca’s western margin is a 500-kilometre-long submarine ridge where
new ocean crust is forming. Its top is scarred by a valley a few kilometres
wide, dotted with underwater volcanoes, fault lines and hydrothermal vents.
Meanwhile the plate’s eastern boundary dives under the North American plate
where it creates frequent earthquakes and feeds the volcanoes that surround
Seattle and Vancouver.

The ocean above the plate is also fascinating to scientists. The California
Current, a northward-flowing body of cool, nutrient-rich water, meets Arctic
water flowing south in the vicinity of the plate and rises up to the surface,
providing a prime feeding ground for fish and sea mammals such as blue
whales.

To allow Neptune to survey as much of this region as possible, the
observatory will have 30 corrosion-proof titanium lab modules or “nodes”
encircling most of the Juan de Fuca plate. Robot subs will wander about across
the network, communicating with the nodes and each other using acoustic signals.
The nodes will be linked to labs on shore by a network of combined power and
communications cables. The 3000 kilometres of cable will carry a total of 100
kilowatts of energy, and provide two-way data links to every node.

Much of the real-time data on the physics of the deep ocean will come from a
suite of sensors mounted on each science module. Vibration sensors and
magnetometers will measure the sea floor’s seismic activity while other
detectors record temperature, salinity, pressure and currents in the water
column above. Still and video cameras will record the flora and fauna around
each node.

Sound waves too will play an integral part in Neptune. “Acoustics really are
our vision under water,” says Bruce Howe, an acoustics expert at the University
of Washington. Each of Neptune’s nodes will house an acoustic transceiver, a
device that can send out and receive high frequency sound waves. These signals
will image the otherwise transparent water column using a technique called
acoustic tomography.

By measuring the time it takes for sound to travel between transmitters and
receivers at different locations, you can measure the speed of sound in the
water—and use that to calculate water temperature and the movement of
currents. Because Neptune will have so many transceivers, it will be possible to
gather high-resolution data from the ocean above the array. These same
transceivers are also acoustic modems, converting instructions sent out by
shore-based scientists into bursts of high-frequency sound that AUVs can detect
while they are out foraging. A craft might be told to collect samples, for
example, or return to base.

Neptune engineers are planning a total of 15 AUVs, some that trundle across
the seabed on caterpillar tracks and others that work in the open ocean much
like those already going through their paces at LEO-15. Here, they have been
operating since 1998, running back and forth over the study area, observing
changes in sediment deposits on the sea floor with side-scanning sonar, and
monitoring levels of nutrients and phytoplankton in the water column. Other,
more manoeuvrable models such as the autonomous benthic explorer developed by
Dana Yoeger, a marine engineer at Woods Hole, can work closer to the seabed,
using magnetometers and cameras to survey hydrothermal vents.

However, these craft are not yet completely autonomous. Since their batteries
only provide power for journeys of up to 100 kilometres, boats have to recover
the AUVs for charging and to retrieve the data they have collected. And subs
like those used at LEO-15 are not yet robust enough to handle the rigours of the
deep sea for long. “Things with moving parts that have to stay working in the
ocean for a long time are a big challenge,” says Yoeger.

But AUV technology is maturing fast. In 1998, for example, Jim Bellingham and
his colleagues at Monterey Bay Aquarium Research Institute in California
successfully tested an AUV docking system, which lets a sub locate a submerged
docking station and clamp onto it. By the time Neptune is built in 2006 a new
generation of autonomous vehicles should be ready to go to work. They will be
able to dock at special moorings and download mission schedules and recharge
their batteries without any human involvement. “The AUVs will move from mooring
to mooring much like a flying squirrel moves from tree to tree,” says
Delaney.

These docking stations will rise above the nodes on cables suspended beneath
floating buoys. The buoys will carry instruments to measure the physics and
chemistry of the water column while “profilers”—spherical, motorised
robots which can scuttle up and down the cables —will measure changes in
the sea’s temperature and nutrient content at different depths. These profilers
have already been tested at Woods Hole.

One of Neptune’s main purposes is to maintain a fleet of AUVs ready to
scramble at the first sign of fresh eruptions from volcanoes or vents along the
Juan de Fuca Ridge. So far scientists have missed the chance to watch these
eruptions. “We’d have an interactive capability that would allow us, when an
event did happen, to fire up all the AUVs and send them after it,” says Yoeger.
This is vital, says Chave. “If you’re not there when it happens, you’re not
going to understand it.”

During quieter periods, AUVs will map the seabed and study the chemistry of
the ocean. Mapping will give marine geophysicists important information about
how the sea floor changes over time, especially after an eruption or earth
tremor. Yoeger has recently adapted a technique called bathymetric sonar to
allow AUVs to map the seafloor on a scale of centimetres—a one
hundred-fold improvement in resolution over older, ship-based sonar.

The eastern edge of Neptune’s realm—where the Juan de Fuca plate
crumples under the North American plate—is an important site for studying
methane hydrate formation and breakdown. Hydrates could be a useful fuel source
and may have played a role in past climate change, but there’s still plenty to
learn about the way they form (91av, 30 May 1998, p 38).
Tremors in the sea floor, for example, could release large amounts of gas that
become trapped as hydrates. AUVs would visit the sites of tremors to hunt for
the fresh release of gas or new hydrate deposits.

The flood of data collected from the nodes and AUVs will be sent via the
communications cables to onshore control centres in Vancouver and Seattle. From
here the data will be pooled and posted on the Web for the benefit of far-flung
researchers. The fibre-optic cables will be able to transfer more than 100
gigabytes of data each second—easily enough to send high-definition colour
video pictures from cameras on the ocean floor in real time. “Imagine being able
to have monitors in your laboratory and share that information with other people
on the Internet,” says Delaney.

This high-speed exchange of information is vital for one of Neptune’s most
ambitious goals: real-time sample analysis on the sea floor. Researchers at the
Monterey Bay Aquarium Research Institute, for example, have already tested a
sensor attached to a mooring in Monterey Bay. Built by Chris Scholin at MBARI,
it is designed to identify species of phytoplankton by their RNA. The device
sucks in a small amount of sea water, filters out the cells, ruptures them with
enzymes and injects the cell contents into a chamber containing molecular probes
that bind to RNA. When they stick to a particular RNA sequence, the probes
change colour, a change that’s picked up by a camera. “The device can even
communicate via radio to tell you what it’s doing,” says Ed de Long, a
microbiologist at MBARI.

Marine microbiologists working on Neptune hope to include this analytical
technology at the nodes. Each will be fitted with a junction box so the network
can listen in on and control “lab-on-a-chip” experiments such as an RNA sensor.
These seabed labs would be particularly useful for studying the bizarre and
ancient bacteria that live in the deep ocean. Attempts to retrieve bacteria from
the deep and cultivate them in the lab often fail, so De Long is looking forward
to devices that can characterise microbes at the node, using DNA microarrays or
molecular probes, in much the same way as MBARI’s phytoplankton sensor seeks out
toxic algae. “Many of the organisms out there are difficult to grow, or we
haven’t even characterised them,” he says.

De Long also hopes to see miniature versions of the PCR devices labs use to
amplify small samples of DNA, but adapted for the pressures of the deep. He is
already using PCR to survey bacterial populations in surface waters off
California and would love to do the same in the deep ocean.

This powerful army of sensors and subs isn’t only attracting the attention of
marine researchers. NASA’s Jet Propulsion Laboratory in Pasadena, California,
became a partner in Neptune to help them develop tools for exploring Jupiter’s
moon Europa. Planetary scientists suspect that liquid water lies beneath
Europa’s frozen surface, making it a prime candidate in the search for life
beyond Earth. “When we go to Europa and if we get underneath the ice, we’ll
certainly have to have to do some sensing,” says Pat Beauchamp, Director of
JPL’s Center of Excellence for In-Situ Exploration and Sample Return. “We’re
also interested in how you communicate under water.” She hopes that working with
Neptune scientists will help NASA mission planners decide how to go about
exploring Europa, as well as how to get the data they collect back to Earth.

While Neptune will be the first large-scale ocean observatory, our
exploration of “inner space” is moving on rapidly. A European, American and
Japanese initiative called the Deep Earth Observing System aims to network the
world’s oceans with cabled or moored observatories. There are systems under
development for the Atlantic, the waters around Japan as well as the
Mediterranean. “My guess,” says Delaney, “is that in 50 to 100 years the entire
ocean will be wired.”