Newswise — Far below the steep, whitewashed villages of Greece’s famed Santorini Island lies an ancient submarine volcano with a violent past.
Kolumbo volcano—which sits 500 meters below the surface within the fault-heavy Hellenic Volcanic Arc just off Santorini—is the Aegean Sea’s most active and potentially dangerous volcano. It’s been quiet over the past few centuries. But during its last eruption in 1650 CE, things got ugly. The volcano blasted pumice and ash as far as neighboring Turkey and triggered a tsunami that inundated the flat coastal areas surrounding the island.
“Beyond the natural devastation it caused, Kolumbo burped up CO2 and other gasses that asphyxiated people and animals on Santorini,” says Rich Camilli, an associate scientist at Woods Hole Oceanographic Institution (WHOI).
The poisonous cloud hung heavy in the air from September through December of that year, a four-month period of Greek history known as the “Time of Evil.”
Now, Camilli and a team of scientists and engineers are headed to the active volcano site as part of a NASA-funded program that will attempt to answer a number of key questions: What can the organisms living in the extremes of this dark and chemical-laden underworld tell us about life on Earth and beyond? Are there signs of geohazards down there that may help predict the next eruption? And to what extent can we hand over the decision-making to ocean robots and let them explore without human control?
The Mediterranean is dead calm as our makeshift research vessel—a cable-laying workhorse named Ocean Link—floats above Kolumbo volcano. The view from the aft deck is idyllic: Santorini’s soaring, multicolored cliffs rise directly to our left, while lower-lying isles are scattered off in the distance.
But despite the postcard view, there’s trouble in paradise. A month ago, scientists from GEOMAR and the University of Athens detected the restless rumble of earthquakes right below the placid surface.
“From a geological point of view, this is the most active volcano in the Aegean Sea,” says Evi Nomikou, a marine geologist from the National and Kapodistrian University of Athens who used ocean bottom seismometers to take the volcano’s pulse. “The cliffs are very steep, so if there’s a large enough earthquake, it could cause landslides and trigger a tsunami.”
But for now, the conditions are just right for sending Nereid Under Ice—or NUI—into the caldera.
The WHOI-developed robotic vehicle is a hybrid—it can operate as an autonomous underwater vehicle (AUV) following a pre-programmed mission, or as a remotely operated vehicle (ROV) connected to the surface by a wispy optical fiber tether no thicker than a human hair.
As its name implies, NUI was designed to explore the underside of Artic sea ice, but it’s also well-suited for other types of tough marine environments. Like an active submarine volcano with the shakes.
“There are a lot of obstacles down there that NUI will have to contend with, including walls that can run hundreds of feet high and other geological features,” says Camilli. “It’s not far from a suicide mission for most ocean robots.”
Teetering on the winch, NUI looks like a red Smart Car being lowered into the Aegean Sea. Within minutes, it plunges to its target depth of 500 meters. There, the vehicle’s lights pierce the ancient darkness and an otherworldly terrain pops up on the control room monitor.
Casey Machado, a mechanical engineer and ROV pilot at WHOI, moves the vehicle around with an Xbox controller and quickly closes in on a hydrothermal vent jutting out of the lunar-like landscape. Its 10-foot chimney gushes plumes of CO2 and other chemicals from below the seafloor. The infusion of carbon makes the seawater so acidic, it could dissolve a hard-shell clam.
But hard-shell clams don’t live here. In fact, the chemical soup—which includes traces of hydrogen sulfide and methane—isn’t a viable food source for most living things.
There are, however, some takers. Thick carpets of chemical-craving microbes blanket the benthos in red, orange and white. Maria Pachiadaki, a marine biologist at WHOI, perks up at her first glimpse of the microbial mats.
"These bacteria are oxidizing the inorganic compounds found in the hydrothermal fluids, a process that creates energy that they can use to turn CO2 into biomass," says Pachiadaki.
Camilli says the ability of marine organisms to take CO2 and convert it to food without photosynthesis is a phenomenon that has caught the attention of planetary scientists.
“If organisms possess that capability here on Earth, it may be possible to find similar lifeforms on ocean worlds beyond our planet like Jupiter’s moon Europa or Saturn’s moon Enceladus,” he says.
With self-driving cars, handing the wheel over to a computer algorithm can be unsettling. The same goes for ocean robots, especially when they need to work in tricky and hazardous environments. But that’s precisely one of the main goals of the expedition: to get NUI to work—and think—autonomously in an unforgiving place that typically requires an experienced ROV pilot at the controls.
“For this cruise, we’re using Artificial Intelligence (AI)-based systems for automated planners, interpreters, and controllers in conjunction with NUI, including one named ‘Spock’ that will allow the robot to do things autonomously like decide where to explore," says Camilli. The technology is being developed as part of NASA’s Planetary Science and Technology from Analog Research (PSTAR) interdisciplinary research program. "We have another automated planner named 'SMIRC' that intelligently uses NUI's robotic arms and tools to collect samples without pilot control.”
“One of our first goals is to toss out the joystick.”
Gideon Billings, a guest student from the University of Michigan who hasn’t stopped banging out AI code since we arrived, is ready to give it a shot. Pachiadaki sees a small patch of microbial mat she’d like the robot to sample from Kolumbo’s mineral-rich seafloor. Machado surrenders her Xbox controller as Billings punches a few commands and the computer brain takes control of the arm.
Moments later…presto! The hose extends down to the precise sample location and sucks up the filaments like a steam cleaner.
Applause rings out in the small but crowded control room.
“This is a huge step forward,” says Camilli, who says its the first known instance of autonomous sample collection using a robotic arm in the ocean.
Billings puts the successful maneuver in a broader context. “If we have this grand vision of sending robots to places like Europa to explore, they will ultimately need to work independently like this without the assistance of a pilot,” he says.
Machado is back to piloting NUI. She navigates the vehicle into an unexplored part of the volcano. But this time, she didn’t choose the site: Spock did.
The Spock planner, created by Ben Ayton, a doctoral candidate in the MIT-WHOI Joint Program in Oceanography, is part of an autonomous control system developed by MIT’s Model-based Embedded & Robotic Systems (MERS) group. “It uses logic to choose targets that are likely scientifically valuable, or have novel characteristics to investigate,” says Ayton. “It also allows us to find interesting locations we would otherwise miss—in this case new evidence of seafloor vents and unique organisms.”
Additional sheaths of microbes extend along the seafloor. But biologically, this part of the volcano is different. New forms of life start to appear. A handful of lollipop sponges stick out of a lava mound like a party centerpiece, while a spiny sea urchin clings to its peak.
Suddenly, a baseball-sized, purple-and-white anemone enters the scene.
Pachiadaki inches closer to the screen. “Based on the pattern, it looks like it could be a new species!” she exclaims.
The creature’s identity won’t be confirmed until biological analysis is done, but Pachiadaki doesn’t recognize it, nor does Voula Polymenakou, a microbiologist from the Institute of Marine Biology, Biotechnology and Aquaculture (IMBBC) on the Greek island of Crete.
A thought occurs to Camilli as he takes in the biological wonders: Do organisms here move in and out of certain areas as vent activity waxes and wanes? What are the dynamics at play between hydrothermalism and life in Kolumbo?
As NUI moves toward the main slope of the volcano—another one of Spock’s target sites—tiny hydrothermal vent chimneys no bigger than tree stumps come into view.
Nomikou hasn’t seen these miniature vents before during prior field visits to Kolumbo and doesn’t jump to conclusions about the discovery. But she says hydrothermal vents, in general, represent “weak points” in the volcano that could make it more vulnerable to future eruptions.
“This is something that will require more investigation,” says Nomikou. “The more we know about vents in Kolumbo and how they’re changing over time, the better we’ll be able to conduct our risk analysis,” she says.
The observations provide a potentially important snapshot in time, but Nomikou is thinking beyond snapshots. She’s been after the Greek government for a permanent, real-time seafloor monitoring observatory in Kolumbo that can act as an advanced warning system for those living on Santorini and neighboring islands.
“If Kolumbo was sitting in a remote area of the Mediterranean Sea, it wouldn’t matter so much, but its proximity to Santorini is a very important factor,” she says.
The expedition carries on for several more days without a hitch. In all, NUI makes five dives, most of which were planned and carried out by Spock.
Camilli says that surrendering pilot control of NUI was nerve-wracking at times and it created blips of tension among scientists who would have opted for the vehicle to focus on sites that were already known.
“I had to draw a line in the sand and told them we’re going to let the automated planner handle things,” says Camilli. “And sure enough, each of the sites Spock picked were not only astounding, but scientifically-relevant as well. It became pretty much unanimous that the planner was doing things that would not have been possible if we had simply gone with what we already knew.”
For the biologists on board, the expedition proved fruitful. Pachiadaki and Polymenakou now have milkcrates full of mat samples, hydrothermal fluids, and sea sponges. Biological and genetic analysis of the samples will help the scientists learn more about the biodiversity of Kolumbo, and if there are species that can be useful in biomedical applications.
Nomikou will use her vent observation data to expand the growing knowledgebase of hydrothermal activity in Kolumbo and continue building the case for a dedicated volcano monitoring system to prepare for the possibility of another eruption.
And Camilli will continue working with researchers and students from NASA, MIT, the University of Michigan, the University of Sydney's Australian Centre for Field Robotics, and the Toyota Technology Institute of Chicago to push the automation technology forward. The work will include training ocean robots to "see" the way an ROV pilot sees using “gaze tracking” technology, and building a robust human language-based interface so scientists can talk directly to robots without a pilot or a complicated computer code go-between.
His ultimate vision hints of “Skynet”—the fictional artificial intelligence network that made some high-stakes decisions of its own in James Cameron’s 1984 classic movie, “The Terminator.”
“We can eventually see having a network of AI-based ocean robots where there’s a shared intelligence spanning an entire fleet, with each vehicle working cooperatively like bees in a hive,” Camilli says. “It will go well beyond losing the joystick.”
The futuristic talk provides an interesting contrast to the ancient backdrop of Santorini’s clustered villages we see from the ship. Nestled in one of them is a blue-domed structure known as the Church of the Panagia Tou Kalou (the Holy Virgin of the Good) that was built at the end of Santorini’s Time of Evil. The wrath of the eruption had finally subsided and the island could breathe again. The church, presumably, was a sign of better days ahead.
Only time will tell what the future holds for Kolumbo volcano and Santorini island. But Camilli feels that pushing the limits of science and technology—particularly in this type of highly collaborative and interdisciplinary fashion—is crucial to really understanding what’s going on down there.
“We all knew it would be a high-risk proposition to send these new technologies down into such a dangerous environment,” he says. “But given Kolumbo’s active and awake state, it seemed a lot riskier not to.”
Funding for this project was provided by a NASA PSTAR Grant #NNX16AL08 and a National Science Foundation National Robotics Initiative grant #IIS-1830500.