The Horseshoe Abyssal Plain, a geological formation in the Atlantic Ocean not far from the Gorringe Bank underwater mountain, on the border between the Eurasian and African tectonic plates, is the geographic origin of the earthquake measuring nearly 8 on the Richter scale that shook Lisbon and other regions of the country in 1969.

The fact that it is a flat geological formation, with no known major seismic faults, fuelled "a puzzle" in the scientific community about how a region with these characteristics could have caused earthquakes of such magnitude. However, a study by the Faculty of Sciences of the University of Lisbon, published today in the journal Nature Geosciences, offers a new possible explanation.

In that area, "no obvious fault" had ever been found to explain an earthquake of the magnitude of the 1969 earthquake, said João Duarte, one of the study's co-authors, a geologist, professor at the Faculty of Sciences of the University of Lisbon, and researcher at the Dom Luiz Institute.

Adding to this puzzle as a motivation for the research is the fact that "a fault with sufficient dimensions to generate an earthquake like the one of 1755" has never been found, which is believed to have had an intensity close to 9 on the Richter scale.

According to João Duarte, what has now been revealed and which may explain the origin of both earthquakes is that "a portion of the tectonic plate is separating," in a process called delamination.

This delamination implies that the plate is undergoing a horizontal fracture, as if the rock were being separated by a sheet, opening a fissure that causes the lower part to sink, having already reached a depth of 200 kilometres toward the Earth's mantle, when the normal depth is around 100 kilometres.

The upper part of the plate remains in an unaltered horizontal position, making it impossible to detect any geological changes in that location by observing the seabed, the researcher explained to Lusa.

This horizontal plate separation process, which has been occurring slowly for five to ten million years, was identified using a kind of "Earth ultrasound," the professor explained, referring to the use of seismic tomography and the sound captured—the sound of earthquakes themselves—to understand what is happening beneath the seafloor.

"We conducted a study that placed seismometers on the seafloor for eight months, recording small earthquakes. We noticed that in that area there was a 'cluster,' a set of small earthquakes at great depth, about 30 to 40 kilometres deep, which is somewhat abnormal. Therefore, there is a combination of several observations here that indicate that a process is occurring there that is generating seismicity," the researcher explained.

"We know that the African plate is also moving very slowly, converging with the Eurasian plate. It's as if we imagined we had two books and one started to slide into the other. The plates are converging, and in reality, the area where the plate begins to separate behaves a bit like a leaf, because it starts to slide into the other plate. There's always contact, but it's a more horizontal contact. In other words, there's no hole left, no space left. That space is then occupied by other rock," explained João Duarte.

Future investigations

João Duarte hopes that this study will lead to more detailed future investigations into that area. With what we already know, he says it's inevitable that this delamination process will be taken into account "in the characterization of seismic hazard and risk" in the country, already considered a high-risk area due to the confluence of two tectonic plates in this region.

The researcher sees an opportunity in the installation of a new generation of submarine cables—communications cables connecting both sides of the Atlantic, passing through the Azores and Madeira archipelagos and also through the Horseshoe Abyssal Plain.

"They'll have seismic sensors, so the cables will run through that area, making it possible to better monitor and characterize that seismicity. And we'll probably also have more data, more records," he said.

The visibility of earthquakes remains a pipe dream, but João Duarte believes that artificial intelligence could provide a step forward, even if a learning model based on previous events and data presents challenges in this case, because large earthquakes are rare phenomena.

The key may lie in a more systematic study of smaller earthquakes, which occur every day, including in Portugal, "with some hope" of understanding the seismicity process and "making some inferences and using statistics to understand larger earthquakes."