Description du sujet de thèse

Numerical Modeling on a Small Scale of the Transport of Sargassum Algae in Coastal Waters Near the Shore

Contexte de travail

1. **Context**
Since 2011, significant strandings of Sargassum algae have affected the Caribbean islands, with major health, economic, and environmental impacts. Detecting these algae in the sea and predicting their trajectory is crucial for limiting these effects, especially through the deployment of floating barriers. Current research largely relies on satellite imagery to detect Sargassum, and on oceanographic and atmospheric data to model its movement. Recent models, such as those based on HYCOM, NEMO-Sarg1.0, or Lagrangian and Eulerian approaches, have been developed. Some integrate artificial intelligence methods (machine and deep learning) to improve the accuracy of predictions. However, these models face limitations on a small scale (a few kilometers), mainly due to satellite image accuracy and disturbances (clouds, artifacts). Alternative approaches (coastal imaging, videos) are still experimental. To improve prediction accuracy, it is necessary to better integrate oceanic and atmospheric dynamics, as well as local factors such as bathymetry, shoreline configuration, and tides.

2. **Objectives**
The main objective of this PhD is to provide a small-scale model to simulate the transport of Sargassum adapted to a coastal area near Caribbean islands. The proposed approach is deterministic, but it is important that the fluid-structure interface model is not too computationally expensive, as it is desirable for it to provide forecasts of stranding areas. Such a tool is currently highly awaited by stakeholders and territorial managers in Guadeloupe.

3. **Description**
Based on preliminary work, the aim is to minimize the impact of Sargassum on seawater and consider a multilayer Saint-Venant-type model, in which the Sargassum transport at the surface layer speed is coupled with the system. To faithfully simulate the algae drift, the numerical scheme must be as non-diffusive as possible, at least for the component modeling the algae. However, it is also certain that the model's surface layer is affected by surface wind forcing. Classic test cases will be reproduced, with or without constant bathymetry (dam break, etc.), and salinity will then be added. After studying the 1D model, a finite volume solver will be developed with several versions of numerical flux (Central-Upwind scheme, HLL, staggered grids, etc.) for a comparative analysis.
The model will then be extended in a 2D configuration in a coastal bay, in order to validate it on known and observed scenarios. The model is expected to reproduce dynamics and morphogenesis of Sargassum mats and rafts. Furthermore, considering floating barrier devices seems appropriate, as they would retain the algae while allowing marine current and waves to pass, providing new types of mixed boundary conditions. This would be one of the aims of this study: optimizing the placement of these barriers to prevent stranding areas or propose their arrangement to deflect these devices.
---

**Required Skills**
The candidate should have a Master's degree in Mathematics or equivalent. The candidate must have strong skills in applied mathematics, particularly in Partial Differential Equations analysis. Good proficiency in scientific computing, Python programming (and other compiled languages) is also required.

**Working Context**
The PhD student will benefit from a three-year doctoral fellowship and will be integrated into the "SARG-FORSS" project, a winner of the Pépinière Interdisciplinaire des Antilles Françaises AAP from MITI-CNRS.
The PhD student will be hosted within the Laboratory of Mathematics, Computer Science, and Applications (LAMIA) at the University of the Antilles, on the Fouillole Campus, in Pointe-à-Pitre, Guadeloupe.

Contraintes et risques

There will be close collaboration with Antoine ROUSSEAU, DR INRIA at the Montpellier Institute Alexander Grothendieck (IMAG, UMR CNRS 5149), University of Montpellier, and a member of the SARG-FORSS project. A 5-week mission will be planned in Montpellier in 2026.

**Additional Information**
The PhD student's host laboratory will be UMR CNRS 8053 Powers, History, Slavery, Atlantic Caribbean Environment (PHEEAC), University of the Antilles, BP7209, 97275 Schoelcher Cédex - Martinique.
The PhD student will be enrolled in the Doctoral School Dynamic Environments in the Caribbean-Americas Space (DEECA), ED 636 at the University of the Antilles.
Co-supervision will be provided by Ms. Léa BATTEUX and Professor Priscilla RAMSAMY, both Senior Lecturers at the University of the Antilles. Additionally, co-direction will be ensured by Antoine ROUSSEAU and Jean-Raphaël GROS-DESORMEAUX, respectively, Senior Researchers at UMR 5149 and UMR 8053.