Research Internship at LSCE - Spring 2024
Impact of environment and management on the mortality of the Chantilly forest
Context: For nearly a decade, the Institut de France has been monitoring the gradual deterioration of its trees, with the oak trees, which represent the predominant species in the 6,300-hectare forest, being particularly affected. Currently, approximately 40% of the oaks are damaged. A collaborative science initiative amassed a vast quantity of data, including stand descriptions (species, height, volume, decay, humus, flora), soil horizon descriptions (providing valuable information on water reserves and carbon stocks), and high-resolution airborne LiDAR measurements.
Goal: The team created maps showing the height and evolution of the forest using deep learning algorithms. The goal is to use these unprecedented data to pinpoint the factors driving mortality using explainable machine learning methods. The team already used successfully Shapley indexes. The intern will work closely with members of the research group and IGN, and will be able to use existing tools and infrastructure.
Working environment:
A group of fifteen researchers from ENS, LSCE, IGN (French National Institute of Geographic and Forest Information), collaborations with institutions such as ONF (French National Forests Office), INRAE, German and Danish research groups.
Roadmap:
- Review of literature
- Data preparation
- Model training
- Assessment of the features impact
- Writing a report/paper.
Profile:
- You are currently doing a masters in AI/Machine learning/Datascience/Computer vision, and looking for a research internship in an academic lab.
- You enjoy coding, especially in python, and already have a bit of experience from previous internships and/or university projects.
- You are already familiar with forest ecology or you are willing to learn.
- You know how to train and test a machine learning model.
- You like problem solving, you are autonomous, but want to work in a collaborative environment.
- Working with big and heterogeneous data is not a problem for you.
- You are curious, enjoy learning, and want to develop skills in remote sensing.
- You are motivated to work on a problem that can have an impact on forests.
Location: Laboratoire des Science du Climat et de l’Environnement (https://www.lsce.ipsl.fr) located about 20 km from the heart of Paris in the Orme des Merisiers green area. LSCE is a world-class research laboratory established and a collaboration between CEA, CNRS and the University of Versailles Saint-Quentin (UVSQ). It is part of the Institute Pierre Simon Laplace (IPSL). LSCE hosts approximately 300 researchers, engineers and administrative staff including many PhD and master’s students.
Start: Spring 2024
Duration: 4-6 months
Supervision: Agnès Pellissier-Tanon, post-doc researcher. Team of Philippe Ciais. In collaboration with Laurent Saint André, INRAE
Compensation: 700€ to 750€ per month at the CEA, depending on qualifications.
Contact: agnes.pellissier-tanon@lsce.ipsl.fr and Laurent Saint-André laurent.saint-andre@inrae.fr
Applicants should submit a CV by email.
References:
Schwartz, Martin, et al. "FORMS: Forest Multiple Source height, wood volume, and biomass maps in France at 10 to 30 m resolution based on Sentinel-1, Sentinel-2, and Global Ecosystem Dynamics Investigation (GEDI) data with a deep learning approach." Earth System Science Data 15.11 (2023): 4927-4945.
PHoCEA DRF - Tous droits réservés - Mentions légales - Protection des données personnelles - Dernière mise à jour : 25-04-2024
