Oceanographers and mathematicians, working together for the climate
Julie Deshayes, member of the Pierre Simon Laplace Institute, works in the Oceanography and Climate Laboratory (LOCEAN).
This physicist oceanographer is interested in modeling ocean dynamics. For several years, she has helped produce the climate change scenarios that feed into the reports of the Intergovernmental Panel on Climate Change (IPCC). In 2019, she launched the interdisciplinary Andiamo project, supported by the Institute of Computational and Data Sciences (ISCD), to bring together mathematicians and oceanographers.
Your research in physical oceanography at LOCEAN* is used in particular to better model climate change. Could you explain the role that the ocean plays in global warming?
Julie Deshayes: The ocean, which covers 71% of the Earth’s surface, is the main source of heat for the climate. It plays a thermoregulatory role by transporting the surplus of solar energy received at the equator towards the poles, thus making it possible to broaden the habitable zone on Earth. It also limits global warming by storing part of the greenhouse gases produced by human activities.
The ocean plays a key role in climate change. If it released the greenhouse gases and/or excess heat it has stored into the atmosphere, the climate would warm up even faster. If, on the contrary, it continues to absorb this excess energy, it will limit the intensity of the warming. It is therefore necessary to make models simulating the future of the oceans and the climate in order to better anticipate the impact on societies, biodiversity, and so on. This is the focus of my research in physical oceanography.
What is the objective of the Andiamo project that you are carrying out with the ISCD?
JD: The models that produce future climate scenarios can be seen as virtual planets representing all of the physical characteristics of the climate actors (including temperature, salinity, currents in the oceans, winds, pressure, humidity in the atmosphere). These numerical climate models have reached such a level of complexity in terms of quantity of information, interactions, finesse of the representation of processes, that it is no longer possible to control their mathematical properties. In addition, they require massive computing power, which is expensive and also has a strong impact on the environment.
The Andiamo project was born out of my desire to improve these models to make more precise climate forecasts that are more useful for society, with a reduced carbon footprint. Instead of systematically increasing their resolution and their level of complexity, my objective is to question the relevance of the mathematical tools used by oceanographers to develop models of ocean dynamics adapted to the scientific questions in which they are interested.
Why do you have mathematicians and oceanographers working together?
JD: Oceanography is based on physical equations that describe fluids and environments. In order for calculators to be able to make a numerical representation of them, they must be discretized, that is to say, transposed into a binary language made of 0 and 1. This is where mathematicians come in.
The whole problem comes from the fact that the mathematical tools available to discretize these equations are not specific to my discipline. We must therefore adapt them to our constraints and to our questions from oceanographers. It is at this stage that we can make bad choices and introduce errors into our models. In this context, it seems essential to me that mathematicians and oceanographers take the time to work together.
What does this interdisciplinary collaboration look like on a daily basis?
JD: Oceanography and mathematics are two disciplines that are closely linked in theory. But in reality, few oceanographers and mathematicians rub shoulders because we are each caught up in the race for publications and the search for funding in our specialty.
In the Andiamo project, with the mathematician Nina Aguillon and the engineer Sibylle Téchené, we wanted to take the time to recreate this interconnection. Sibylle was recruited to work with Nina and me, and she is trained in both modeling and oceanography. With this double competence, it helps us to find a common language. The same terms do not always have the same meaning depending on our disciplines, which is why we spend a lot of time exchanging and sharing our visions of certain concepts. We read and comment together on articles in oceanography and mathematics. Using maps, photos, or measurements, I describe the physical objects I study. My colleagues explain to me in return the properties, advantages and disadvantages of the mathematical tools that I use and introduce me to the latest numerical developments that I could take advantage of.
Beyond the Andiamo project funding, what support does the ISCD give you?
JD: The ISCD brings us enormous added value. At a time when government policy requires us to design three-year scientific projects, with very tight deadlines and strong pressure to publish, the ISCD gives me carte blanche to take the time to reflect and speak with my colleagues in mathematics. It offers me institutional recognition of this time spent building together a real field of interdisciplinary collaboration.
With us, the institute is betting on the future. They are aware that we may not produce immediate results. But they are confident that if we succeed in creating a common language between oceanographers and mathematicians, we will then be able to offer an innovative vision for research in our respective disciplines.
From a logistical point of view, the ISCD allows us to extract ourselves from the context of our specific communities by providing us with neutral premises where we meet every week. We will also organize a multi-day retreat between scientists from different disciplines. This is an opportunity for us to get out of our jargon, to interact with other communities and to take a step back from our research.
Finally, the ISCD offers us the opportunity to offer training modules to students and young researchers from our two disciplines, in order to teach them our common language as soon as possible. We hope to facilitate the realization of truly interdisciplinary research projects.
You launched the Andiamo project with two other women. Is there a particular impact on the fact that this project is carried by three women?
JD: This project is the fruit of ten years of research where I wondered if what I had been doing until then was really the best way to contribute to my scientific community. This questioning, my mathematician and engineer colleagues, Nina and Sibylle, share it. Taking a step aside and taking the risk of not publishing is a choice. A difficult choice, but one that is important to make in order to be proactive in our field. We made this bet. Would we have done it if we had been men? I do not know.
The Institute of Computational and Data Sciences
This center for research, expertise and training in scientific computing, the Institute for Computing and Data Sciences using high performance computing and visualization means.
Dedicated to research activities around numerical simulation, it offers a welcoming structure to multidisciplinary teams who engage in a collaborative project in scientific computing.
Its mission and objectives are to accompany, support, structure and stimulate the projects of the community of researchers, engineers and students concerned with scientific computing and simulation.
To contribute to the challenges that face our contemporary society but which do not belong to any science in particular, the Sorbonne University Alliance has set up several institutes and initiatives. which bring an interdisciplinary approach, offering opportunities to compare knowledge and experience, both in research and in education.
* LOCEAN (Sorbonne University/CNRS/IRD/MNHN/IPSL/Ecce Terra). Within this laboratory, Julie Deshayes is attached to the National Institute of Universe Sciences.