Annonce

Les commentaires sont clos.

IMT-Atlantique M2 internship: High-resolution reconstruction of ocean global oxygen

11 Janvier 2024


Catégorie : Stagiaire


IMT-Atlantique M2 internship: High-Resolution Reconstruction of Ocean Global Oxygen

 

IMT-Atlantique M2 internship:

High-resolution reconstruction of ocean global oxygen

 

Introduction

There is growing evidence indicating a decline of the global ocean's oxygen content over recent decades. The distribution of dissolved oxygen in the ocean reflects the interplay between ocean circulation, biological processes, and air–sea gas exchange. The long-term loss of O2 is primarily caused by the weakening of the ventilation of the intermediate ocean, mostly driven by surface warming. To investigate this phenomenon of ocean deoxygenation, recent three-dimensional reconstructions of global O2 concentration have been developed. These reconstructions rely on optimal interpolation of O2 profiles extracted from the World Ocean Database (WOD) dataset [1]. A recent study also utilized machine learning to combine WOD data with ARGO climatology, resulting in a new interpolated O2 product [2].

The objective of this study is to go beyond low resolution oxygen products as in [1,2] and to generate a high-resolution three-dimensional reconstruction of oxygen concentration in the global ocean. This is achieved by constructing a neural network emulator of oxygen concentration, which is trained on both remote sensing measurements and in-situ observations of physical and biogeochemical data (e.g., temperature, salinity, ocean color, etc.). The model will be employed to generate artificial O2 profiles, which will be combined with WOD measurements to create a high-resolution space-time reconstruction.

Workplan

  • Study state-of-the-art AI methods for reconstruction of geophysical fields and oxygen reconstruction.
  • Familiarisation with WOD data, training a neural network emulator for O2 prediction.
  • Emulation of high resolution O2 profiles.
  • Evaluation, tuning and validation.
  • 3D reconstruction.
  • Analysis of long-term O2 trends based on the 3D reconstruction.

Profile of the candidate

  • Enrolled at Master 2 or in the last year of engineering school.
  • Knowledge in Machine Learning, Deep Learning and generative models.
  • Good understanding of the use of Deep Learning Frameworks (Pytorch,Tensorflow, JAX).
  • Prior experience in research is highly advantageous.

Documents to be submitted

Please send following documents in a single pdf page with the title of Application-Internship-Odyssey-IA4O2' for evaluations to: said.ouala@imt-atlantique.frronan.fablet@imt-atlantique.fr and zouhair.lachkar@nyu.edu

  • Curriculum Vitae.
  • Motivation letter (max 1 page).
  • Current diploma and transcripts.

Application deadlines, selection process and start of work

  • Application deadline is 31 January 2024.
  • Selection to be planned from 2nd to 5th february 2024.
  • Start of work flexible, between March/April 2024.

Further questions

Further questions can be addressed to:

About the hosting organisation

IMT Atlantique is internationally recognized for the quality of its research, is a leading French technological university under the supervision of the Ministry of Industry and Digital Technology. IMT Atlantique maintains privileged relationships with major national and international industrial partners, as well as with a dense network of SMEs, start-ups, and innovation networks. With 290 permanent staff, 2,200 students, including 300 doctoral students, IMT Atlantique produces 1,000 publications each year and raises 18€ million in research funds.

References

[1] Ito, Takamitsu. "Optimal interpolation of global dissolved oxygen: 1965–2015." (2022): 167-176.

[2] Sharp, Jonathan D., et al. "GOBAI-O 2: temporally and spatially resolved fields of ocean interior dissolved oxygen over nearly two decades." Earth System Science Data Discussions (2022): 1-46.