Volcanoes are known to be unpredictable. It’s often very difficult to know when an eruption might happen, and forecasting the time of an eruption is currently not possible. This is why GE is cooperating with a team of engineers and scientists, that is led by Sam Cossman of Qwake, to try and forecast the time of an eruption given enough data on a particular volcano. The team went on an expedition to the Masaya Volcano in Nicaragua.
Masaya is a caldera volcano, which means that it is mostly just a crater in the ground. It is found a few miles outside Managua, the capital of Nicaragua. Since Managua has more than a million residents, it is important to understand this particular volcano. The GE team spent six weeks around the volcano, studying and recording all the necessary data they could get. The team was particularly interested in the active lava lake, since it is dangerous. The Masaya volcano is one of the few volcanoes in the world with an active lava lake, and its thousand-degree temperatures and fragile earth make it difficult to study.
To be able to get close enough to install the team’s equipment, they used drones and build a system of zip lines that were hundreds of feet tall. The scientists wore special heat-resistant suits and placed sensors inside the lake rim and around the entire volcano. The team installed a total of ten sensors which recorded temperatures, air pressures, gas emissions, and humidity.
With sensors positioned, the team was able to study a rare event which was the opening of a volcanic vent. The sensors were able to record swings in temperature, humidity, carbon dioxide levels, etc. Within 11 days, the sensors recorded these changes until the vent closed again.
By using machine-learning algorithms to study the collected data, the team believes that they can predict future volcanic activity by studying similar changes in temperature, CO2 levels, as well as humidity. The team is still testing their hypothesis, but if they prove this to be correct, then this could have a huge impact on everyone inhabiting the area near Masaya.
This could be applied to different volcanoes as well. There are more than a thousand active volcanoes on the planet, and have around 800 million people living close by. The GE research team’s idea of collecting data and using artificial intelligence to analyze the data could change the way we study volcanoes. This could turn an unpredictable natural disaster into a weather phenomenon everyone could foresee.