Latest News ...


August 2015: First MERMAID statistics published in Nature Communications

Note: This page is still evolving - we shall publish a link to the Nature paper and show some more figures after the publication date

In a paper, to appear August 20 in Nature Communications (*), we report the successful recording of a large number of small and big earthquakes by robots floating in the Mediterranean, the Indian Ocean and the Pacific Ocean. Some of these earthquakes were too small to be observed by land stations but one of our robots (which are named MERMAIDS) was close enough to transmit no less than 235 seismograms in four weeks.


A MERMAID on its way from a drifting depth (normally 1500 m) to the surface for data transmission via the Iridium satellite network. Figure source: Geoazur.



A MERMAID transmitting data to the satellite, here photographed during a test in the bay of Villefranche-sur-Mer. The Geoazur scientists receive the recorded seismograms per email. Figure source: Geoazur.



However, it is not the observation of small earthquakes that formed the main motivation for this development. The question that has been at the heart of the GLOBALSEIS project was: why is the Earth not a cold, dead planet like Mars? In fact, the Earth has cooled very little since it formed as a hot, probably molten ball about 4.5 billion years ago. The constant temperature has allowed the surface to become habitable, thus providing an environment that allowed life to develop.

But geophysicists do not quite understand why the convective processes that bring heat from the Earth's mantle to the surface are not more drastic in cooling the interior.


One can use seismic waves to make a scan of temperature differences in the Earth in an attempt to understand what exactly is happening. One important feature of the cooling process are the mantle plumes, narrow upwellings of hot rock beneath volcanic islands such as Hawaii and La Reunion that bring heat from depth to the surface. Because these plumes are almost all located in oceanic areas where seismic stations are absent, the scanned images of such plumes are very blurred, such as the image of Hawaii shown in this figure - it is like making a hospital scan of a tumor with most sensors broken down. With sharp images absent, some scientists have even questioned the existence of plumes in the Earth's mantle. Figure source: Nolet et al., 2006.



From the statistics of what we can observe with MERMAIDs under different noise conditions we have determined that a network of about a thousand such robots can produce enough seismograms in five years to be able to image mantle plumes with high precision.

Such a network would not only serve the imaging of our planet's interior. Among the surprising discoveries was an earthquake swarm in the Indian Ocean that led to the transmission of 235 small earthquake signals of which only one was also observed on land stations. A new prototype of the robot (see below) will also be able to record whale signals and even estimate rainfall, thereby making a global monitoring system for the oceanic environment a practical possibility.

(*) Alexey Sukhovich, Sébastien Bonnieux, Yann Hello, Jean-Olivier Irisson, Frederik J. Simons & Guust Nolet, Seismic monitoring in the oceans by autonomous floats, Nat. Commun. 6:8027 doi: 10.1038/ncomms9027 (2015).
More information: scroll down on this page and check
Contact: Guust Nolet, +
The MERMAIDs were developed as part of the GLOBALSEIS project, funded by the European Research Council



This succesful publicaton comes at a crucial time, since the new version of the MERMAID, which will soon become available for purchase, will undergo its final tests in Villefranche as soon as the tourist boat activity (which renders tests a little risky) dies down in September.


The new Mermaid (here shown during a test in Villefranche-sur-Mer) will be able to dive to greater depth, and have a longevity of at least six years. In addition to being able to record earthquakes, it measures ocean temperature and salinity. It can handle up to eight different instruments to monitor the oceanic environment. In the future we hope to develop very user friendly software that would allow a the robot MERMAID to be programmed directly by the scientist, who then remains in full control about what to observe, when to surface, and what to transmit. Figure source: Geoazur.


For information about availability of the new MERMAID, please contact our industrial partner OSEAN directly: email, phone +


February 2015: ERC projects ends: What next?

With the ERC Globalseis project coming to an end we are proud to say it has spawned a large number of continuing activities:

- an array of 9 MERMAIDs is still operational around the Galapagos islands and will continue to transmit data into 2016, after which we hope to be able to get very much improved tomographic images of the mantle plume under this complex,

- we continue to use ships of opportunity to launch MERMAIDs in the Ligurian Sea, which we retrieve when the float too far west, in an effort to image the upper mantle under this failed ocean rift,

- with OSEAN, we are testing the next generation MERMAID, which has a very long lifespan (6-8 years) and is able to hold up to eight different instruments. An acoustic version, able to observe low- and high frequency acoustics as well as temperature and salinity, will come on the market by the end of 2015. For more information send an email to OSEAN (

- we are investigating errors in existing surface wave and delay time data using the method described in the paper by Voronin et al. (2014) , before doing a full global tomography that is compatible with finite-frequency theory.


August 2014: Galapagos deployment highly succesful

Our first - still preliminary results - show that the noise level in the Pacific Ocean is very low. The first six weeks of deployment have already yielded many useful seismograms. These will eventually be used to obtain images of the Galapagos plume system to large depth.

The crucial question we attempt to answer is whether the basalt rises up from the lower mantle, or from much shallower depth. This, in turn, may help us understand how Earth gets rid of her calories. Geological evidence shows that the Earth cooled very little in the past 4 billion years. This modest cooling prevented Earth to become a dead planet, like Mars, but maintain tectonic activity. Even though earthquakes and volcanoes are dangerous for man, it is this activity that shaped our landscape and determined the chemistry of our atmosphere, and it was essential for the development of life. We do not quite know how Earth managed to stay as warm as she is. Radioactive elements in mantle rocks contribute to fight off the cooling, but not enough, and some of the heat must be stored heat from the time of formation of our planet. How it was able to stay inside is the ultimate riddle we try to solve. Our hypothesis is that plumes like Galapagos play an important role - perhaps these are the only places where heat is able to rise to the surface and cool the planet?


July 2014: Mermaid 17 recovered

A rare opportunity offered itself when the Spanish scientific vessel Garcia del Cid took to sea at Malaga, heading for Barcelona. We noticed that its trajectory would bring it ver close to Mermaid 17, which was launched in December 2012 and was nearing the end of its battery life.
Thanks to the cooperation of Miguel A. Ojeda of the Unidad de Tecnologia Marina - CSIC, and of the crew of the Garcia del Cid, Yann Hello was able to sail on this ship and recuperate the Mermaid.


May 2014: Mermaid goes West!

We have started to cover our third - and last - oceanic domain, with the most ambitious project so far. Ten Mermaids are being launched in the Pacific Ocean near the Galapagos Islands, an experiment conducted in close cooperation with scientists from INOCAR, the oceanographic institution of Ecuador.



Right: Sebastien Bonnieux in Pacific waters....


Left: the first Mermaids have already started to return seismograms


We expect to finish the launch of these Mermaids by the end of May. At the same time, we install three land stations, as well as three OBS's (ocean bottom seismometers) closer to the islands. The combination of these with the Mermaids should give us an unequaled resolution to lower mantle depths of the deep structure of this hot spot, which may very well be a plume arising from as deep as the lower mantle.


We are very grateful to the people in Ecuador who are making this experiment possible, in particular Mario Ruiz, director of the IG EPN, Quito, CPNV-EM Humerto Gomez, Asesor ministro de Defensa, CPFG-EM Juan Carlos Proano, Director de Inocar, CPFG-EM Pablo Pazmino, Commandant del BDE Orion, CDPF-EM Mac Mera, Director de Intereses Maritimos, CPFG-EM Freddy Vergara, Director del Centro de Datos Ocenograficos, and CPCB-EM Andres Pazmino, Director de Investigation del Inocar.

25 November 2013: New Mermaid launched in Indian Ocean records a seismic swarm!

Barely in deep water for 24 hours, a new Mermaid just launched in the Indian Ocean surprised us by giving more than 200 'triggers'. It is drifting close to the Indian Ocean triple junction (a spot were three plates meet: the Australian/Indian, the African and the Antarctic plate). And it happened to catch an important seismic swarm


A collection of seismograms sent via the Iridium satellite to our server. The first one (labeled 24th) represents a precursor, recorded during the very first dive. The mermaid was instructed to come up to the surface after 24 hours and send us a state of health message, which is why it missed the two main shocks (a magnitude 5.1 and 4.5). The small earthquakes it did record are of much smaller magnitude and were not recorded by stations on land or islands.

November 2013: First Multimermaid tests in deep water

A new version of our Mermaids, based on the same technology that Geoazur uses for its Ocean Bottom Seismometers (OBS) is being developed in close cooperation with OSEAN in Le Pradet (Var). The name reflects its ability to carry multiple scientific instruments and serve seismology, marine geophysics, biology and meteorology at the same time. In contrast to an OBS, the Multimermaid is designed to float and come to the surface rather than rest on the ocean bottom. In contrast to the current Mermaid design, the Multimermaid is multi-disciplinary and lasts up to 5 years on a single battery charge. For a short movie, click here.


To conduct the tests we use a common fishing vessel.


The MultiMermaid is ready for its test dive. For security reasons, it is kept on a leash.


A view from below.

October 2013: MERMAID recovered close to Spanish coast


Mermaid 18 was launched just before Christmas 2012 in the Ligurian Sea. While floating westwards, it has been recording a large number of teleseismic P waves, but by October it was entering an area were fishing is done to a depth of 800m (the seabottom), near Cabo Creus Canyon north of Barcelona. We therefore decided to recover it.


Yann Hello and Sébastien Bonnieux leased a small fishing vessel. At its last surfacing, we used the two-way communication via Iridium satellite to command the Mermaid to come to the surface on Monday October 21 and transmit its position.


Despite a rather rough sea, the Mermaid was quickly spotted and recuperated. To the surprise of everyone, it was still very clean and free of any shells.


The operation has given us hope that, in the future, Mermaids can be recovered when nearing coastal areas. In fact, every Mermaid has been labeled in multiple languages following a UNESCO standard. If it is inadvertently picked up by a fishing vessel, we hope the fishermen will contact JCOMMOPS, the UNESCO office in charge of monitoring float use in the world's oceans.

March 2013: MERMAIDS in the Indian Ocean


This figure shows the launch sites of the first three Mermaids in an open ocean. For more details, see our BLOG.

December 2012: Michel Foundotos discovers new seismic waves!

In a poster presented at the Fall Meeting of the American Geophysical Union, we have presented a world record: the first observations ever of P waves reflected five and six times off the surface of the Earth. Such waves are very difficult to observe since at every reflection they lose some of their energy to converted shear waves that then disappear in the lower mantle. The so-called 6P and 7P waves sample the upper mantle much like a surface wave does this with shear waves. The high station density of US Array allowed Michel to apply stacking techniques that can zoom in onto a particular wave and amplify its amplitude.


From left to right: Observations of 5P, 6P and 7P waves for two events in the west Pacific. The coloured diagrams are Vespagrams showing the energy in the seismogram as function of time (vertical axis) and incidence angle (horizontal). By summing over a local subarray we then obtain the amplified waveform shown at the right of each Vespagram.

Paper ranks among ten best papers in Nature Geociences

To celebrate its fifth anniversary, the editors of Nature Geociences have selected ten articles, among which the paper by Sigloch, McQuarrie and Nolet (2008), as the ten favourite papers in the first five years of the journal. This paper was at the basis of the Globalseis project.

August 2012: First Mermaid fleet has arrived

After working hard to complete the final design of the Mermaid robots, we ordered early this year a fleet of 17 instruments, which arrived in August and are currently being tested one for one. Early October we plan to launch two of them in the Mediterranean, early 2013 six of them will be shipped to La Reunion and launched in the Indian Ocean. Plans for the further future include a deployment of a small fleet near the Galapagos islands as well.


Mermaids on the quay of Villefranche being tested for GPS and Iridium communication (photo courtesy Yann Hello).

October 2011: Mermaids record small magnitude earthquakes

The analysis of the recordings made by the two Mermaids in September gave us further insight into the magnitude range of earthquakes that rise above the noise level. Among the P waves that are clearly recognizable in the data are a PKP wave that has crossed the Earth's core, coming from an earthquake of magnitude 6.4 in Tonga, at a distance of 152 degrees, and a P wave from a magnitude 5.5 earthquake in Turkey (distance 24 degrees).

Below we show some examples from Mermaid III:





June 2011: Mermaids come of age!

On June 28 the Mermaid prototypes completed their longest journey to date (two weeks) in the Ligurian Sea, cruising at different depths down to 2 km. A first scan of the data showed this amazing recording of a P wave from a magnitude 7.4 earthquake in the Aleutians, at a distance that is almost a quarter of the circumference of the globe. The signal-to-noise ratio exceeds 50 in the first float (top). Because the second float had already started a pre-programmed ascent to the surface, its recording is a little noisier. These records confirm the earlier estimate by Frederik Simons (JGR 2009, doi:10.1029/2008JB006088) that earthquakes with magnitude higher than 6 should generate strong enough P wave signals to rise above the noise level in Mermaid seismograms.


April 2011: Christophe Zaroli receives prestigious prize

The Société des Amis des Universités de l'Académie de Strasbourg has awarded its Prix de Thèse to Christophe Zaroli for the originality of the research reported in his recent PhD thesis. He was selected among a very competitive field of candidates by the Conseil Scientifique de l'Université de Strasbourg. He receives the prize, which is accompanied by a monetary award of 1500€, during a ceremony planned for June 10 in the salle Pasteur of the Palais Universitaire in Strasbourg.


October 2010: Mermaid records PP wave

Our two Mermaid prototypes made their maiden voyage in the Mediterranean between Sep 26 and Oct 1. Our primary goals was to test the technology and to sample noise at different depths.

The wave at the center (near time mark 5150) is the PP wave from a magnitude 7 earthquake in Irian Jaya on September 29, at an epicentral distance of 119 degree (more than 13,000 km away). Under the seismogram we plot a 'spectrogram', which allows us to analyze the frequencies present in the signal. The vertical axis of the spectogram runs from 0.1 (bottom) to 1 Hz (top), and the time axis is the same as in the seismogram.

This is the first time that a PP wave has been observed in a freely floating hydrophone!

July 2010: First test with Mermaid succesful!

The first tests with the Mermaid prototype, done early July in the vicinity of the port of Villefranche, have been a complete succes. As programmed by Antony Ogé, the float descended to a depth of 20 m, followed by divers Yann Hello and Alain Anglade, who took the beautiful photographs you see here. We are currently waiting to check the integrity of the first recordings of underwater acoustic noise (the divers, mostly!) to be recuperated from the solid state memory and analysed by Alexey Sukhovich. The next brief sortie of the prototype will be outside of the harbour and reach a depth of about 300 m. A sortie of five days in which the prototype will collect data at several depth levels down to 2000 m is programmed for the end of September when we have access to the CNRS vessel Tethys II.

Figure 1. The Mermaid has received instructions to descend. It pauses at the surface to take its final position with the GPS antenna (yellow). The red housing protects the hydrophone that records acoustic noise (and earthquake signals!).

Figure 2. A pump empties a bladder, which decreases the floatability of the Mermaid. It starts its descent. For this first test, we programmed it to stop near 20 m depth, since we wished to follow it closely.

Figure 3. Once arrived at its floating depth, the instrument is inspected by the diver (Yann Hello).

Figure 4. After an hour the Mermaid prototype again works the pump, but now to feed the bladder - this increases the volume of the float, which starts its ascent. The bubbles are not caused by the Mermaid, but by the divers who cannot come to the surface as quickly as the Mermaid....