En poursuivant votre navigation sur ce site, vous acceptez le dépôt de cookies dans votre navigateur. (En savoir plus)

M/F PhD in coastal bathymetry and sea state prediction using deep learning : satellite observation and short-term prediction

This offer is available in the following languages:
Français - Anglais

Date Limite Candidature : vendredi 30 octobre 2020

Assurez-vous que votre profil candidat soit correctement renseigné avant de postuler. Les informations de votre profil complètent celles associées à chaque candidature. Afin d’augmenter votre visibilité sur notre Portail Emploi et ainsi permettre aux recruteurs de consulter votre profil candidat, vous avez la possibilité de déposer votre CV dans notre CVThèque en un clic !

Faites connaître cette offre !

General information

Reference : UMR5566-RAFALM-001
Date of publication : Friday, October 09, 2020
Scientific Responsible name : Rafael Almar/Dennis Wilson/Jean-Marc Delvit
Type of Contract : PhD Student contract / Thesis offer
Contract Period : 36 months
Start date of the thesis : 16 November 2020
Proportion of work : Full time
Remuneration : 2 135,00 € gross monthly

Description of the thesis topic

Coastal regions are currently facing environmental and resource problems
aggravated by population pressure and overexploitation. The environmental context or
extreme events (floods, coastal erosion) combined with demographic pressure are a
limiting factor for coastal development. The general objective of this internship is to
improve the representation of sea state and subsequent coastal dynamics at the event scale.
In practice, direct measurements (wave buoys, CANDHIS network) remain costly and
difficult. On the other hand, models have been developed and implemented for both coastal
circulation and wave representation but are computationally costly and subject to large
uncertainties in coastal areas. The internship will investigate the possibilities to 1) derivate
sea states (waves) using optical images from regular basis satellites with global coverage
such as Sentinel-2 and 2) replace costly wave models to propagate from offshore deep
waters to the coast (incl. extreme sea level/setup). Using deep learning would represent an
efficient way to solve computationally costly wave observation and modelling in coastal
zones. The training will be conducted on a synthetic dataset of more than 12000 numerical
simulations of waves with random conditions and the application on Sentinel-2 images.

Work Context


Constraints and risks


Additional Information

1/2 grant CNES

We talk about it on Twitter!