Study and parameterization of the vertical distribution of phytoplankton biomass in the global ocean

• Main Author: Sauzède, Raphaëlle
• Other Authors: Laboratoire d'océanographie de Villefranche (LOV), Observatoire océanologique de Villefranche-sur-mer (OOVM), Université Pierre et Marie Curie - Paris 6 (UPMC)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National de la Recherche Scientifique (CNRS)-Université Pierre et Marie Curie - Paris 6 (UPMC)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National de la Recherche Scientifique (CNRS)-Centre National de la Recherche Scientifique (CNRS), Université Pierre et Marie Curie - Paris VI, Hervé Claustre, Julia Uitz
• Format: Doctoral or Postdoctoral Thesis
• Language: French
• Published: HAL CCSD 2015
• Subjects: Phytoplankton; Ocean color; Chlorophyll; Phytoplancton; Flotteurs profileurs Bio; Couleur de l'eau; Coefficient de rétrodiffusion particulaire; Fluorescence; Bio-Optique; Chlorophylle-A; [SDU.OCEAN]Sciences of the Universe [physics]/Ocean; Atmosphere; North Atlantic
• Online Access: https://tel.archives-ouvertes.fr/tel-01342441; https://tel.archives-ouvertes.fr/tel-01342441/document; https://tel.archives-ouvertes.fr/tel-01342441/file/2015PA066625.pdf

This PhD work focuses on the parameterization of the vertical distribution of phytoplankton biomass and community structure in the global open ocean. First we have developed a neural network-based method for the calibration of the fluorescence in chlorophyll a concentration [Chl] associated with the total phytoplankton biomass and with three phytoplankton size classes. This method, (FLAVOR for Fluorescence to Algal communities Vertical distribution in the Oceanic Realm), was trained and validated using a database of ~900 concomitant fluorescence and HPLC-determined pigment profiles. A global database comprising ~49 000 fluorescence profiles was assembled and calibrated with FLAVOR. The resulting database represents a first step towards a global three-dimensional view of phytoplankton biomass and community composition. Second, two neural networks (SOCA for Satellite Ocean Color and Argo data to infer vertical distribution of bio-optical properties) were developed to infer the vertical distribution of two bio-optical proxies of the phytoplankton biomass, [Chl] and the particulate backscattering coefficient, using as input satellite-derived products matched up with a hydrological Argo profile. The SOCA methods were trained and validated using a global database of ~5 000 profiles of bio-optical and hydrological properties collected from Bio-Argo floats with concomitant satellite products. The database used to develop FLAVOR and SOCA originates from various oceanic regions largely representative of the global ocean, making the methods applicable to most oceanic waters. Finally, we proposed a study dedicated to the North Atlantic where the tools developed in this thesis are used in conjunction with a bio-optical primary production model. This allows us to characterize the seasonal cycle of the vertical distribution of the phytoplankton biomass and primary production in various bio-regions of the North Atlantic. Les travaux présentés dans cette thèse concernent la paramétrisation de la distribution verticale de la ...