What's worse than looking for a needle in a haystack? Trying to find a whale in a sea shrouded in fog or roiled by wind.
It's a daily ordeal for government agencies that use aerial surveys to find North Atlantic right whales and then warn ships steaming toward the endangered animals to slow down. It's also a headache for scientists who can roam the sea for weeks without seeing a whale.
"I've been involved with a lot of projects where we can't even study the whales because we can't find them," said Mark Baumgartner, a biologist at Woods Hole Oceanographic Institution. "I wanted a way we could ... minimize the time searching for them and maximize the time actually studying them."
So Baumgartner and his colleagues at Woods Hole are now letting a robot scout for them -- using ears instead of eyes.
The underwater scout -- known to technocrats as an autonomous underwater vehicle, or AUV -- is equipped with a recorder that listens for whale sounds and transmits data every two hours to researchers via satellite phone.
Last month, scientists deployed two AUVs in the Gulf of Maine. It was the first time whale sounds of four baleen whale species -- right, humpback, fin and sei -- were recorded and transmitted in near real time from a roving vehicle.
The 23-day deployment began on a high note as the robots picked up whale calls within two days of sliding into the water. Baumgartner and his colleagues found the right whales, confirmed the data's accuracy, then phoned the National Oceanic and Atmospheric Administration, which directed mariners to reduce their speed in that area for the next 15 days.
Ship strikes are the leading cause of right whale mortality.
"Right whales possess all the characteristics needed to make them susceptible to ship strikes," said Michael Asaro, a marine mammal specialist with NOAA's Northeast Regional Office.
They spend a lot of time at the water's surface, move slowly and don't react as vessels approach. With between 300 and 500 individuals remaining, preventing strikes is seen as key to helping the species survive.
AUVs could boost the whales' odds, scientists say, especially as agencies are reducing aerial surveys to save money in a severe federal budget crunch.
How it works
The robots look like 6-foot yellow torpedoes. They're called gliders because they have wings that let them move forward at an angle as they slowly sink or float.
The 100-pound gliders move up and down by getting either heavier or lighter than the surrounding seawater by sucking in or expelling about 12 ounces of water, said David Fratantoni, a physical oceanographer and engineer at Woods Hole.
An on-board computer, GPS system and satellite phone allow a computer in Fratantoni's lab to send coordinates to the glider so it knows where to go, typically every two days. The onboard systems calculate a course, taking into account currents and other factors.
Moving at about 1 nautical mile per hour -- the pace of a leisurely walk, Fratantoni said -- one glider on the Gulf of Maine deployment traveled an area half the size of Connecticut twice, while the other stayed in one area. An AUV can roam until its batteries run out of juice, about 60 days.
Along the way, the acoustic recorders ran continuously. (Gliders are ideal for acoustic recordings because they are silent.) However, that's a huge, "almost insurmountable" volume of data to sift through, not to mention transmit over satellite, Baumgartner said.
Baumgartner developed software that did the heavy lifting, constantly checking the noises coming in against a library of known whale calls. When there was a match, the program translated the sounds into a "pitch graph," colorful lines that depicted the length, frequency and volume of the call, much like a sheet of music.
The graphs were relayed to the researchers when the AUVs surfaced about every two hours, enabling them to connect to the satellite. The researchers looked at the pictures to identify the whale species.
For the first time, researchers knew what whales were out there before leaving the dock, and where to find them.
Baumgartner notes that acoustic recorders on stationary buoys in shipping lanes near Boston already alert managers in real time when right whales are present. And the devices have gone on lengthy dives atop autonomous vehicles, but researchers couldn't retrieve any data until the end of the mission.
Getting those data every two hours, for multiple species, from a mobile platform is a big improvement, said Baumgartner, who has been collaborating with Fratantoni to take that next step since 2005.
The concept and most of the mechanics for doing autonomous data collection have been around since the 1980s, Fratantoni said. But it wasn't until a robust and reliable two-way satellite phone network was established about 10 years ago that it was possible to advance to near-real-time observations.
"It's great to have vehicles that can stay out in the ocean for weeks and months at a time, but if you can't talk to them, it's not very helpful," Fratantoni said.
The AUVs were not just listening for whales. Other sensors on board took measurements like temperature, salinity and plankton abundance, which researchers hope will reveal clues about why whales are congregating in certain locales.
Now that the gliders have passed the test run, managers need to determine how they might be used routinely to support NOAA's monitoring efforts.
NOAA and the Navy helped fund the technology development -- so there is interest, but, Baumgartner said, "There has to be the will and the resources to pour into a new technology like this."
The AUVs are attractive because, while not cheap, they are far less expensive than aerial surveys or research cruises, which require significantly more manpower, equipment and fuel.
They also are able to listen 24 hours a day, providing new insight for managers.
"Right whales are around pretty much all the time ... always calling," said Sofie Van Parijs, who leads NOAA's passive acoustics research group. "We never understood how pervasive it was."
AUVs are also opening up new times of year and survey areas. Very few data on whale distribution exist for winter months because of the difficulty of getting out there.
The middle of the Atlantic is also a mystery, Baumgartner said. Managers could send the gliders to shipping lanes between London and Boston in place of a research cruise that would be prohibitively expensive.
Robots won't replace humans completely, Van Parijs said. People are still needed to count and identify individuals once they are located.
The data gathered are sure to inform management protocols.
Currently, seasonal management areas are in place that require vessels longer than 65 feet (think large tankers) to reduce their speed to 10 knots, which is about half of normal. Compliance is mandatory during the months whales migrate between southeastern Florida and Nova Scotia.
Temporary, voluntary "dynamic management areas" (DMAs) like the one implemented recently are issued whenever a group of right whales is spotted by NOAA or other sources elsewhere.
More than 90 DMAs have been issued since the program began in 2008, Asaro said. The regulation authorizing the management zones is set to expire at the end of this year. Analyses of how effective they have been in reducing collisions, and the impacts of slowing down commerce, are being prepared to inform discussions on whether to continue the zones, Asaro said.
"This year will be a big year for these regulations," Asaro said.
The need for such research and monitoring for right whales is clear to Baumgartner and other scientists who see the scars left by fishing gear entanglements or ship propellers down the whales' backs.
But because right whales aren't in the right areas for whale watching trips, tourists rarely encounter a right whale.
"You have a highly endangered species that is clearly impacted by our activities, but the public is largely unaware," Baumgartner said. "They are spectacular, but it's sad, I wish other people could see them."