Thèse de doctorat en Océanographie physique et biologique
Sous la direction de Véronique Garçon.
Soutenue en 2005
à Toulouse 3 .
Pas de résumé disponible.
Use of satellite ocean color' data for in coupled physical-biogeochemical modelling of dynamic coastal areas : application to the R'io de la Plata and to the Bay of Biscay
Continental margins play a crucial role in biogeochemical fluxes between the continents and the open ocean. They are also highly sensitive to human perturbations. Primary production is highly dependent on the complex dynamics of these coastal zones, and on the river supply in nutrients, as well as in Suspended Particulate Matter (SPM) which limit light availability. The quantification of the production is a prerequisite in many applications, from the understanding of the role of the coastal ocean in climate change, to eutrophication and fisheries resources issues within the frame of the integrated management of the coastal zone. Coupled physical-biogeochemical models are useful in studying and monitoring primary production. Because of the numerous complex processes and their non-linear interactions, and of some poorly known parameter values, some errors remain in the simulation of the phytoplankton production in coastal areas. Together with these models, satellite `ocean color' data is a major monitoring tool for marine ecosystems. Applying empirical algorithms to SeaWiFS data in coastal areas allows the retrieval of the concentration of chlorophyll a as well as of SPM. This PhD. Work shows the benefits of using `ocean color' data to constrain primary production models. Two study areas are investigated : the Rio de la Plata estuary and its continental margin in the Southwest Atlantic Ocean, and the Bay of Biscay in the Northeast Atlantic. First, the analysis of chlorophyll a concentration data bases for both areas, and of SPM concentration for the Bay of Biscay, reveals major phytoplankton features and processes that should be taken into account in the model formulation. Then a simple five compartments model for the Rio de la Plata, and a more complex one for the Bay of Biscay, have been coupled to the hydrodynamic model MARS-3D of IFREMER. After a first validation step, in which major observed features are reproduced, the models constraint by ocean color data is investigated. For the Bay of Biscay, surface forcing with satellite SPM concentrations improves the accuracy of light availability and then the simulation of late winter phytoplankton blooms. For both areas, we conducted optimisation experiments for some of the biological parameters. . .