The night of February 13, 2014, Indonesia's Kelud volcano burst to life with the ground-rattling energy of some 250 megatons of TNT. Its billowing ash plume shot 16 miles high, sprinkling tiny shards of rock for hundreds of miles.
But that wasn't the only thing the ferocious display had in store: Hundreds of volcanic lightning strokes crackled overhead, spreading their spidery tendrils across darkened skies. Now, scientists say such lightning may be just as useful as it is beautiful. A new study, published in the Journal of Volcanology and Geothermal Research, takes another step toward the development of lightning as a monitoring tool to track the ever-shifting dangers of a volcanic eruption.
“It sort of fills a niche that no other volcanic eruption monitoring tool can cover,” says study author Alexa Van Eaton, a volcanologist with the U.S. Geological Survey's Cascades Volcano Observatory.
The latest work relies on data from the World Wide Lightning Location Network (WWLLN), which is run through a collaboration of over 50 universities and institutions. Using this network and satellite imagery, Van Eaton and her team showed that lightning rates tend to peak at certain times during an eruption, offering clues to how a volcano is behaving even before eyewitness reports are available. (These are the most dangerous volcanoes in the U.S., according to a recent government report.)
“We're starting to use those global networks, and that gives a much sweeter perspective on these volcanoes that we would never otherwise be able to access or collect data for,” she says.
One volcano is not like the others
A large part of scientists' clues to pending volcanic eruptions comes from a network of seismometers that measures Earth's many grumbles. But these instruments are expensive to install and upkeep, making it impossible to wire up the more than 1,500 potentially active volcanos on Earth.
Instead, monitoring largely focuses on volcanoes with populations nestled nearby, but that doesn't mean the other geologic giants are hazard-free. People are taking to the skies more than ever in human history, which means remote volcanos are ever increasing risks. Flying through swollen plumes of volcanic ash can cause clogged air filters and even complete engine failure. In 1982, volcanic ash caused all four engines on a Boeing 747 to shut down mid-flight.
Satellite imagery can help fill in some monitoring gaps, but no system is perfect. Clouds or nighttime darkness easily obscure satellite imagery. In recent years, infrasound has also become a promising monitoring tool to pinpoint explosive eruptions. But the paths these sound waves take as they travel hundreds of miles can affect what the mechanical “ears” pick up.
These challenges were on display during a 2016 eruption of Alaska's Bogoslof volcano, which is cradled in the arm of the Aleutian archipelago. While the tiny island lies under a well-worn flight path, its shores change too quickly for seismometers there to last for long and clouds often obscure the skies overhead. When it started erupting in mid-December, it took more than a week for anyone to notice.
By contrast, lightning detection is fast and can work from hundreds of miles away through haze or pitch-black night. And unlike sound, light doesn't suffer so-called path effects, Van Eaton says, making it a valuable tool in the volcano-monitoring arsenal. (See photos of people who live near active volcanoes.)
Grassland birds of the Great Plains wade by the water's edge as a storm begins to take shape in the background.
Lightning's usefulness is rooted in why these sparklers happen in the first place. Lightning in general comes from the development of what's known as a charge imbalance, or the stripping of electrons from particles. One important way this happens is friction—the same force at work when you rub a balloon against your head to stand your hair on end.
“A volcanic plume is a perfect environment for friction,” says Corrado Cimarelli, an experimental volcanologist at Ludwig-Maximilian's University of Munich who was not involved in the new study. “You have a lot of turbulence, you have a lot of particles, [and] these particles collide with each other, and they gain charge.”
Material in ash plumes can also fracture and eject electrons, generating a charge imbalance. Or, if the steaming plumes expand up toward the chilly high altitudes, ice formation can contribute to the situation. Thus, the ashier—and more dangerous—the eruption, the brighter the light show.
The flashy show at Kelud
In 2010, the WWLLN started separately tracking lightning activity around 1,563 active volcanoes to help with ash plume detection. It was this data that Van Eaton and her colleagues harnessed for the new analysis. During Kelud's 2014 eruption, the ash plume produced nearly 500 strikes of lightning, according to the WWLLN data. That's not even every crackle of light, Van Eaton notes, since WWLLN detectors only pick up the most energetic of flashes.
“We’re just taking the cream off of the milk,” she says.
The team combined this data with satellite imagery to monitor the plume's expansion from above, using these images to calculate the rate at which the volcano spewed molten rock and ash.
“The more stuff coming out at a time, the faster that radius expands,” explains Van Eaton. What they found was that lighting peaked at six strokes a minute during the early intensification of the eruption and then tapered off as the plume reached a steady expansion.
Van Eaton and colleagues observed a similar spike in lightning during a 2015 eruption of the Calbuco volcano in Chile. In that instance, the lightning peaked at the onset of a pyroclastic flow—a dangerous avalanche of hot rocks and ash. Together, the results hint that these spikes in lightning rates reflect some key change early in the eruption.
Detecting lightning's limits
“It's an interesting paper,” says Cimarelli, who calls the results encouraging but kind of simplistic. As Van Eaton and her colleagues note in the study, there's still a large amount of uncertainty associated with general estimates of mass spewing from a volcano based on satellite imagery. There's also the issue of WWLLN sensors detecting just a fraction of the lightning.
“That doesn't mean that it's not without value,” says atmospheric scientist Sonja Behnke of the Los Alamos National Laboratory. “It's just, we see something, but there's really a whole lot more going on.”
Cimarelli, a National Geographic grantee, replicates volcanic lightning in miniature eruptions to meticulously control each factor. Though the scales vastly differ, lightning in these lab-made eruptions similarly peaks in their initial stages. Then, the flashes become less frequent as the plume expands, likely because the distance between particles—and their ability to charge and discharge—grows ever longer, he says.
“Even with these, let's say, general observations, they already seem to get something which we also see in the [lab] experiments,” Cimarelli says. Still, there's much more to be studied. For one, it's tough to tell if WWLLN-detected flashes are due to storms or volcanoes. The new study hints there may be some difference, but it seems to be small.
“Further work needs to be done to fully distinguish the two types, but there is great potential here,” says volcanologist Rebecca Williams of the University of Hull.
Van Eaton agrees, adding that there's a long way to go before this method can be put into popular use: “What we really have with this paper is some juicy observations. I hope that this will trigger a lot of interesting modeling work, and people who can take these observations and take them to the next level.”