There are more volcanoes in Alaska than in any other state, and the Alaska Volcano Observatory keeps an eye on which ones show signs of unrest.
The organization measures seismic and infrasound activity on more than 20 volcanoes, and they watch for visual signs like ash clouds via satellite and webcam.
Now, a group of scientists from Colorado is using a new technology to measure volcanic activity: drones. By attaching gas detectors to the unmanned aircraft and flying them over volcanic vents, scientists can measure output to “smell” if a volcano’s activity is changing.
In August, KUCB’s Andy Lusk sat down with Christoph Kern from the Alaska Volcano Observatory. Kern worked with the drones alongside Black Swift Technologies CEO and co-founder Jack Elston and CTO Maciej Stachura.
This conversation has been edited slightly for length and clarity.
TRANSCRIPTION
Andy Lusk: Why are you here? What brings you to Unalaska Island?
Christoph Kern: We are out here to take some measurements of volcanic gasses and photogrammetry measurements of the summit of Makushin Volcano. The new thing about what we're doing this time around — we did it once two years ago — is that we're trying to do this without the use of a helicopter or a fixed-wing aircraft. We are doing it solely reliant on a drone.
Lusk: In the simplest terms, what kind of tech are you using?
Maciej Stachura: The drones we're using for this mission, we originally developed in partnership with NASA as a new aircraft for their Earth science missions.
Originally, we were doing some calibration of satellites that they have. But as time went on, we got more requests to do missions.
I wouldn't call this the pinnacle in terms of challenges, but there’s a lot of good engineering improvements because the goal is to fly in these kinds of environments. This specific aircraft design has flown in Greenland, it's flown in really hot weather, it's flown over volcanoes in the jungles of Costa Rica, it's flown at really high altitudes.
Lusk: Because you're launching right from the airstrip on the airport, right?
Kern: Yeah, we've been launching directly from the airport here in Dutch Harbor. The idea is to develop a technology that can be used from a safe distance in the event of an erupting volcano. That means at least 10 miles.
Jack Elston: One of the reasons why we picked Makushin is we didn't have to take a boat to get to it.
Lusk: When you're mapping something like Makushin, how long does that entire process take?
Kern: For photogrammetry, you break it down into two parts. One part is the acquisition of the data, basically just taking photos. For that, you have to ensure that your photos have enough overlap. You're flying back and forth — kind of like mowing the lawn over some of the volcano — taking hundreds of photographs.
The computer does a lot of heavy lifting on the analysis side. In fact, the photogrammetric analyses are some of the most computationally expensive, we say. You feed the images into a software that then can take anywhere from minutes to several hours to process those and convert them into what we're after, which is the digital elevation model.
Those models, depending on the resolution of the cameras and the altitude at which the aircraft flies above the ground, tend to have precision on the order of centimeters. They're very, very accurate.
Lusk: How do you deal with the issue of snow given that we're in Alaska? That seems like a problem.
Kern: We're focusing on areas that are snow-free, and there are actually quite a few. Those are there because there are hot gas emissions, so some of the craters don't have snow in them. We're seeing whether those craters are changing their shape. That actually could be indicative of changing activity of the volcano itself.
In terms of the gas measurements that we're also making, we're trying to measure two things. We're measuring the composition of the gasses that are emitted, and the emission rate — that’s how much gas per hour or day. Both of those are indicative of processes that are occurring within the volcano itself.
Lusk: Very last question. When will the results of your study be published and when can the public know what you found out here?
Kern: It'll be another couple of weeks before these results are made public, but they will most likely be a USGS data release, and there will be a little bit of interpretation there as well. They’ll be posted online at the USGS data portal, ScienceBase.
