How scientists are studying the deep sea floor

Insti­tut Poly­tech­nique de Par­is (IP Par­is) is in the pro­cess of cre­at­ing a new Inter­dis­cip­lin­ary Centre for the Study of Seas and Oceans (Centre Inter­dis­cip­lin­aire pour l’Etude des Mers et Océans, CIMO). This pro­ject is the res­ult of the forth­com­ing mer­ger of ENSTA Bretagne and ENSTA Par­is, which gives the IP Par­is an ocean cam­pus in Brest and sig­ni­fic­ant poten­tial for mar­ine and mari­time edu­ca­tion and research. Ocean obser­va­tion is one of CIMO’s key areas of research. Giv­en the cli­mate and biod­iversity crises and the sus­tain­able devel­op­ment object­ives, it is now of vital import­ance to observe the oceans.

The United Nations has launched the “Dec­ade of Ocean Sci­ences for Sus­tain­able Devel­op­ment” (2021–2030), led by UNESCO. And the United Nations is organ­ising the 3rd United Nations Ocean’s Con­fer­ence (UNOC), which will take place in Nice next year. As engin­eers, IP Par­is sci­ent­ists can bring a fresh per­spect­ive to research into the mar­ine envir­on­ment and mari­time activ­it­ies, and the CIMO will be a melt­ing pot for this.

Ocean obser­va­tion tech­niques have pro­gressed con­sid­er­ably in recent dec­ades. While obser­va­tions have long been made from research, com­mer­cial or even pleas­ure and racing ves­sels, it was satel­lite obser­va­tions in the 1970s that revolu­tion­ised many aspects of land and ocean obser­va­tion. Today, in the age of robot­ics, sur­face obser­va­tions by satel­lites can travel com­pletely autonom­ously from the sur­face to the seabed. Autonom­ous sys­tems, and gliders in par­tic­u­lar, have revolu­tion­ised obser­va­tion of the mar­ine envir­on­ment. They are inex­pens­ive and can carry mini­atur­ised sci­entif­ic sensors to depths of nearly 6,000 metres. And they are spark­ing a host of innov­a­tions in many fields.

To make the most of these small robots, which are deployed in large num­bers – there are cur­rently 4,000 “Argo” pro­filers (the simplest of these robots) – spe­cial­ised infra­struc­tures are needed.

In Europe, there are a num­ber of major Research Infra­struc­tures (RIs) ded­ic­ated to dif­fer­ent sci­ences or major soci­et­al issues, organ­ised and largely fun­ded at a European Uni­on level. One of these, which every­one has heard of, is CERN [Editor’s note: European Organ­iz­a­tion for Nuc­le­ar Research]. Anoth­er is the European South­ern Obser­vat­ory (ESO) in Chile, a col­lec­tion of very large tele­scopes. In the con­text of ocean obser­va­tion, the Hori­zon 2020 GROOM II pro­ject (Gliders for Research, Ocean Obser­va­tion and Man­age­ment Infra­struc­ture and Innov­a­tion) is devel­op­ing a coordin­ated European RI to sup­port research and Ocean Obser­va­tion Sys­tems (OOSs) with autonom­ous sys­tems cap­able of remain­ing autonom­ous in the ocean for months or even years.

Over the past 20 years, Laurent Mor­ti­er, from ENSTA Par­is, has devoted his career to set­ting up these types of RIs and OOSs. He is cur­rently coordin­at­or of the Hori­zon Europe Advanced Mar­ine Research Infra­struc­ture Togeth­er (AMRIT) pro­ject, after hav­ing coordin­ated GROOM II, which has just come to an end. Europe is increas­ingly encour­aging the integ­ra­tion of RIs and innov­a­tion, and in this respect autonom­ous mar­ine sys­tems and the GROOM II pro­pos­als will play a corner­stone role in the future edi­fice of mar­ine RIs. In par­tic­u­lar, AMRIT will devel­op stand­ards, best prac­tices and tools to ensure that obser­va­tion data can be optim­ally integ­rated into exist­ing and future cli­mate pre­dic­tion mod­els, serving the needs of research and the wider blue eco­nomy and society.

“One of the aims of AMRIT is to improve the ocean com­pon­ent of the Coper­ni­cus pro­gramme [Editor’s note: an EU pro­gramme that col­lects and reports data on the state of the Earth on a con­tinu­ous basis],” he explains. By observing the ocean, which influ­ences the cli­mate, mod­els will be able to bet­ter pre­dict ocean dynam­ics, as well as weath­er and cli­mate. “This is obvi­ously essen­tial for under­stand­ing cli­mate change, but above all for pro­pos­ing mit­ig­a­tion and adapt­a­tion meas­ures,” adds Laurent Mor­ti­er. Today, ocean fore­cast­ing and inform­a­tion ser­vices are mainly provided by the Coper­ni­cus Mar­ine Ser­vice, man­aged by Mer­cat­or Ocean Inter­na­tion­al in Toulouse. This entity was largely cre­ated by poly­tech­ni­cians from the French Navy’s Hydro­graph­ic and Ocean­o­graph­ic Ser­vice.

To this end, research­ers have already turned to arti­fi­cial intel­li­gence (AI) tech­niques. Des­tin­a­tion Earth, a major pro­ject of the European Com­mis­sion and the European Space Agency, is devel­op­ing a digit­al twin of the Earth, with its ocean branch EDITO. These digit­al twins are based on advanced mod­els of the Earth sys­tem, which are then used to integ­rate more applic­at­ive digit­al twins. But to func­tion, these mod­els and digit­al twins need a reg­u­lar flow of obser­va­tions and data, cov­er­ing all phys­ic­al and liv­ing ele­ments, for example in extreme envir­on­ments, at great depths or under the Arc­tic ice. “It’s an almost impossible task, unless you go in with autonom­ous under­wa­ter sys­tems,” explains Laurent Mor­ti­er. “Robot­ics is one solu­tion, but it’s not easy to send robots under the ice and the instru­ments they carry can get lost. These twins will prove use­ful in design­ing the obser­va­tion sys­tems of the 21^st Century.”

“France has often been a pion­eer, and AI has been very use­ful in improv­ing the design of the SWOT satel­lite mis­sion to map ocean cur­rents from space at high res­ol­u­tion” he adds. IP Par­is could pos­i­tion itself in this field, as it has a num­ber of labor­at­or­ies cap­able of tak­ing on this type of work on much more com­plex prob­lems involving a large num­ber of parameters.

“In addi­tion to Argo, the revolu­tion­ary obser­va­tion pro­gramme launched in the 1990s, which is the corner­stone of the Glob­al Ocean Observing Sys­tem (GOOS), we now need to integ­rate all the obser­va­tion sys­tems to ensure that these digit­al twins are really use­ful,” explains Laurent Mor­ti­er. “And GOOS wouldn’t really exist without fund­ing from the US Nation­al Ocean­ic and Atmo­spher­ic Admin­is­tra­tion (NOAA). Europe doesn’t have the equi­val­ent of NOAA for the oceans. Agen­cies such as Ifre­mer, research bod­ies and uni­ver­sit­ies are try­ing hard to coordin­ate the European com­pon­ent of GOOS, the European Ocean Observing Sys­tem (EOOS), but neither the Com­mis­sion nor the Mem­ber States have yet found a way to make it work and, above all, to fund it. The European Com­mis­sion con­tac­ted me recently because it sees AMRIT as a pro­ject that could change all that.”

He adds that a new reg­u­la­tion pro­posed by the Com­mis­sion entitled “Ocean Obser­va­tion – Shar­ing Respons­ib­il­ity” could be a decis­ive step for­ward. If adop­ted by the next Com­mis­sion, it will oblige EU Mem­ber States to observe the oceans in an oper­a­tion­al way. “Obser­va­tion of the oceans includes many ele­ments: tem­per­at­ure, salin­ity, of course car­bon, but also fish and para­met­ers more related to mari­time activ­it­ies, such as noise – and of course pol­lu­tion. Car­bon is the para­met­er that we all want to try and meas­ure much more sys­tem­at­ic­ally, because the ocean is a car­bon pump, and this pump is being dan­ger­ously weakened by cli­mate change. Bet­ter mon­it­or­ing of the ocean’s capa­city to absorb car­bon is now essen­tial, and it’s a glob­al issue.” The Glob­al Green House Gases Watch (G3W), an ongo­ing pro­gramme of the World Met­eor­o­lo­gic­al Organ­isa­tion (WMO), is work­ing in this dir­ec­tion, and meas­ur­ing car­bon diox­ide exchanges could become compulsory.

“This will be the focus of my work over the com­ing months. With my col­leagues from the European mar­ine RIs, we have every inten­tion of influ­en­cing EOOS and pro­pos­ing solu­tions. And with its excep­tion­al research poten­tial, IP Par­is must be part of this col­lect­ive effort” says Laurent Mortier.

Interview by Isabelle Dumé

Ref­er­ence:

Ocean­Gliders: A Com­pon­ent of the Integ­rated GOOS DOI:10.3389/fmars.2019.00422