Since spring 2018, many earthquakes have occurred at sea off Mayotte to the east of the island. Some were felt strongly by the island’s inhabitants. Observation and research operations conducted by several French research institutions and laboratories (IPGP/CNRS/BRGM/IFREMER/IPGS) as part of a CNRS-INSU programme funded by the French Ministry of Research and the French Ministry of Ecological and Solidarity Transition shed light on the mechanisms behind these phenomena.
Nathalie Feuillet, a researcher at the Institut de physique du globe de Paris, head of the INSU TELLUS SISMAYOTTE project and co-head of the MAYOBS 1 mission, answered three questions about on the operations that led to the discovery of a new active volcano east of Mayotte.
What did we know about volcanism around Mayotte before these missions?
“Mayotte is an island where very few earthquakes have been felt during the historical period. As a result, scientists had rather incomplete studies and little data. We know that it is an island of volcanic origin which began to form about 20 million years ago. The most recent traces of volcanic activity have been found in sediment cores collected in the Mayotte lagoon. These are ash deposits that have been dated to around 7,000 years ago. The very well preserved morphology of volcanic cones and craters in the northeast of the island suggests more recent volcanic activity in this area, particularly on Petite-Terre and Mamoudzou. Petite-Terre’s Lake Dziani is known for the degassing that occurs there, and studies on it have already been by scientists from the BRGM and the IPGP.
Despite these indications of relatively recent activity on a geological time scale, Mayotte was not considered to be a volcanically or seismically active island. Before May 2010, only two earthquakes had been felt by the population in recent decades, and there was no volcanic activity during the historical period. There was therefore no systematic observation of seismic and volcanic activity. The network of seismological stations installed and maintained by BRGM scientists was very useful for monitoring the crisis, but the type of stations (accelerometers) and their geographical distribution were not suitable to accurately locate the earthquakes and understand their origin. The nearest seismological stations, located in particular around the active volcano of Kartala (Grande Comore) and in Madagascar, were too far away to significantly improve the ability to locate earthquakes.
To complicate matters, the seismic crisis is occurring at sea. Although high-resolution bathymetry* data had been acquired in this area (BATHYMAY campaign, BRGM and SHOM transits), the coverage was too partial to be able get an idea of the structures that could have caused the earthquakes.”
How did you discover Mayotte’s new volcano?
“All of the observations accumulated since 2018 tended to show that the earthquakes recorded off Mayotte were caused by an underwater volcanic phenomenon. The earthquakes were distributed in swarms, which may be typical of volcanic activity. Our colleagues from the ENS had determined that the surface deformations recorded by the GNSS stations could be modelled by the emptying of a magma reservoir located deep below the seismic swarm zone (more information available on the ENS website). In the fall of 2018, with colleagues from the IPGP-CNRS (including those from the Piton de la Fournaise volcanological observatory), BRGM, IFREMER, IPGS-EOST and several other CNRS laboratories, we set up and proposed a large-scale project (called SISMAYOTTE) at the sea-land interface as part of the INSU-CNRS “TELLUS Mayotte” call for projects. Our project was selected and funded at the beginning of 2019 and it was immediately launched.
Our colleagues from IPGS-EOST and the BRGM installed three seismometers from the INSU fleet coupled with GPS in Mayotte. Scientists from the IPGP and the Piton de la Fournaise volcanological observatory installed a seismological station and a GPS on Grande Glorieuse Island in March. At the end of February, colleagues from the INSU's OBS** fleet deployed six OBS offshore to the east of Mayotte over a wide area surrounding the earthquake swarms.
The MAYOBS 1 campaign was the last component of the SISMAYOTTE project. It took place from 2 to 19 May 2019 aboard the Marion Dufresne oceanographic vessel. With Stéphan Jorry, the co-director of IFREMER, we had several objectives: to recover the 6 INSU OBS deployed in February and the seismic data they had recorded and then redeploy them with 8 other OBS from IFREMER, and to map in detail the area where the earthquakes were occurring from the ship's multibeam and sediment sounders to try to detect traces of recent volcanic activity. We also planned to image the water column to detect possible fluid outlets. If recent traces of volcanic activity were detected, we had also considered dredging the floor to recover volcanic rocks.
On board, we put in place a protocol to analyse the seismic signals recorded by the OBS. The teams operated around the clock, broken down into shifts, and we were able to precisely locate, in less than 2 weeks, the nearly 800 largest earthquakes (of magnitudes between 3.5 and 4.9). We discovered that these earthquakes were, for the most part, located in an area quite close to the island (10 km from the east coast of the island) but were deep (between 20 and 50 km deep).
Bathymetry and reflectivity data*** acquired during a systematic survey showed the existence of a volcanic ridge oriented N110°E off Petite-Terre, on the island slope. This ridge is made up of dozens of volcanic landforms and has many lava flows. The morphology of these landforms suggests activity which is very recent on a geological time scale. The earthquakes we located occurred along this ridge.
We have therefore focused our efforts in this area. One evening, the team on shift detected significant plumes in the water column. These plumes came out of the ocean floor about 50 km from Mayotte, from depth of about 3000 m and coming up two kilometres high. This was the first time we had detected such large plumes. In hydrothermal zones, plumes were also detected at the ridges, but they were much smaller and often difficult to recognise. But what we were observing here was simply extraordinary.
We were fortunate to have the old data acquired with the Beautemps-Beaupré vessel by the French Marines Hydrography and Oceanography Service (SHOM) in 2014 in this area. We immediately compared the new bathymetric data from the MAYOBS 1 campaign with those from 2014. We then discovered a new relief 5 km in diameter and more than 800 m high, which is unprecedented. The plume came out from the top of this relief. So we were witnessing a volcanic eruption and the birth of a volcano on French territory!
We then mapped the area of the new volcano very precisely with very close profiles to increase the resolution of the bathymetry and reflectivity maps. We took samples at different levels of the water column to obtain information on the nature of the fluids. We also dredged the northeast side of the new volcano. We were able to sample several hundred kilograms of freshly placed rocks. When these rocks were brought to the surface at the stern of the ship, they exploded, throwing pieces several metres high. These very particular rocks are called “Popping Rocks”. They are full of gas that decompresses during the ascent that causes the rock to explode, and they have only been sampled very rarely on the planet.”
What are the next steps in the project?
“At the end of the mission, scientists from the institutions involved worked to compile the data as quickly as possible. And there is still a lot of data to process.
The second mission, MAYOBS 2 (still CNRS/IFREMER/IPGP/BRGM, as part of an inter-ministerial project), took place in June, so the OBS had their data read were and deployed again. The mapping of the seabed was refined and showed that volcanic activity was still ongoing and had produced a new relief, south of the new volcano discovered during MAYOBS 1. The mapping and study of the earthquake swarm area, 10 km east of Mayotte, was also completed. The fluid plumes identified during the MAYOBS 1 campaign are still present above the area of this earthquake swarm.
The study of the samples taken during these two missions is also in progress and should provide a better understanding of the formation of this volcanic zone.
We are leaving in July for two new missions (MAYOBS 3 and 4) to continue to collect data, monitor the growth of volcanic landforms, analyse and sample fluid plumes in the water column, obtain very high resolution imagery and bathymetry, and better locate the earthquakes.”
* Bathymetry data are seabed topography data
** OBS (Ocean Bottom Seismometers) are seabed seismometers
*** Reflectivity data give an idea of the nature of the soil. For example, sediments are very poorly reflective while volcanic rocks, such as lava flows, are very reflective.
Nathalie Feuillet, lithosphere tectonics and mechanics team