In early May, the Kilauea volcano erupted in Hawaii, causing serious damage to the island. Lava flow, clouds of gas and ash, acidic vapors and, more recently, lava rain turned the island into a hellish place. And not a single scientist was able to predict this natural disaster more than two weeks before it occurred.

Measuring volcanic activity

According to the French volcanologist Jean-Marie Bardintzeff, there are many tools to predict volcanic eruptions. On the website Futura Planète, he explains that any forecast must first be based on a precise observation of the volcano. With the help of seismographs, laser sensors and satellite images, scientists are able to measure the seismic activity of a volcano. In Kilauea’s case, an irregular seismic activity allowed scientists to predict the eruption two weeks in advance, according to Nature magazine. Scientists also rely on temperature readings near volcanoes, taken from fumes escaping surfaces and craters. Finally, they analyze the gas emissions generated by each volcano. All of these observations make it possible to follow the volcano’s activity and, in the short term, predict an eruption.

Unpredictable explosions

However, these observations can’t always be done. “Effusive” eruptions – characterized by the emission of lava at a temperature of 1,200 ° C – are easier to predict in the short term. Sensors placed on the volcano’s surface and satellite images make it possible to follow and map the magma flow that will eventually generate lava flow. Conversely, “explosive” eruptions – which begin far below the Earth’s crust – release little lava and can generate gigantic explosions. Current technology and warning signals are limited in this regard and offer little chance of predicting explosive eruptions, as geophysicist Robert Scarpa explained in an article published in the North American magazine Science in 2001.

Geochronology, key to predicting eruptions

In the short term, some eruptions – especially effusive ones – can be predicted. But “instant” data provided by various volcano observations don’t allow us to make long-term forecasts. The United States Geological Survey – the body responsible for monitoring volcanoes in the United States – explains on its website that “long-term forecasts are based on the geological past of each volcano.” We can learn about this past by analyzing ash and lava deposits from previous eruptions. These allow us to know the nature of the volcano – if it is explosive or effusive – as well as understand the warning signs of an eruption.

Each volcano is unique

However, the monitoring and geochronological analysis of volcanoes remain imperfect. The current state of technology, but also the scale of the task, impose some limits: how can we possibly monitor all the volcanoes in the world? Moreover, if an eruption can be detected more or less in the long term, it remains difficult to predict its nature and effects, as volcanologist Simon Carn explained in Popular Science. He recalls that the unique qualities of each volcano prevent us from predicting when and how an eruption will occur, even if we’re comparing similar volcanoes.

In other words, volcanology is not an exact science. Despite the increased monitoring of volcanoes and steadfast geochronology work, scientists don’t have access to all the necessary data.