The overarching goal of this thesis is to develop a diagnostic and predictive model to quantify the estuarine CO2 dynamics across scales – from catchment to the globe – using an approach that explicitly resolves the strong physical and biogeochemical gradients typically observed in these systems.Chapter 1 provides fundamental definitions and descriptions of estuaries, as well as an assessment of their role in the global carbon cycle. It also raises the specific objectives and research questions tackled in the present study. Chapter 2 presents the rationale behind the novel modelling approach (C-GEM, Carbon-Generic Estuary Model) developed in the framework of this thesis. First, the dominant processes that control the estuarine biogeochemistry in estuaries are discussed in detail. Then, the power of reactive-transport models (RTMs) in understanding and quantifying the estuarine biogeochemical functioning is illustrated on the basis of local modelling studies. Finally, trends in estuarine biogeochemical dynamics across different geometries and environmental scenarios are briefly explored with C-GEM and results are discussed in the context of improving the modelling of estuarine carbon dynamics at regional and global scales. In Chapter 3, a detailed description of C-GEM, both in terms of structure and set-up, is provided and model’s performance is successfully evaluated through comprehensive model-data and model-model comparisons in the macro-tidal Scheldt estuary (BE/NL). In Chapter 4, C-GEM is combined with a generic set of forcing conditions and parameter values to quantify the carbon dynamics (net ecosystem metabolism, CO2 exchange at the air-water interface, carbon filtering capacity) in three idealized estuaries subject to temperate climatic conditions. Their hydro-geometrical characteristics span the wide diversity of estuarine morphological characteristics. Model results are used to upscale the estuarine CO2 dynamics under present-day conditions and to quantify the response of the estuarine filter to future atmospheric CO2, land use and climate change scenarios. In Chapter 5, C-GEM is applied to derive estimations of carbon export and CO2 outgassing from all tidal estuaries discharging in the North Sea. Overall, our results suggest that the estuarine carbon filtering capacity and the contribution of these land-ocean transition systems to the atmospheric CO2 budget might not be as high as previously thought. Finally, a conclusive chapter (Chapter 6) provides a synthesis of the key findings and arguments projected by the present research work. Moreover, recommendations are given in the light of further applications of the modelling approach developed during this thesis. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

L'air extrait des glaces polaires du Groenland et d'Antarctique constitue une véritable archive des atmosphères passées de notre planète. L'objectif de cette étude était de comprendre les échanges de carbone en atmosphère naturelle et, à terme appréhender le cycle du carbone futur en atmosphère anthropisée, à partir de l'analyse des teneurs atmosphériques en dioxyde de carbone et méthane au cours de l'Holocène (les derniers 10000 ans). Afin d'interpréter ces profils obtenus dans la glace, nous avons utilisé un modèle atmosphérique du méthane à trois boîtes. Nous avons pu ainsi estimer le gradient interpolaire puis la part relative des sources de CH4 dans les basses latitudes ainsi qu'aux moyennes et hautes latitudes nord au cours de l'Holocène. En comparant l'évolution temporelle de ces sources aux reconstructions paléoclimatiques obtenues sur des enregistrements continentaux (carottes sédimentaires et lacustres) et marins, ainsi qu'à des simulations de modèles climatiques couplant l'atmosphère, l'océan et la biosphère, nous avons pu attribuer les variations de CH4 aux variations de distribution en latitude des sources potentielles de ce gaz. L'étude des variations de CH4 a ainsi permis de contraindre les effets possibles de chacun des réservoirs de carbone sur les variations de CO2 sur la même période. Nous concluons que le rôle de l'océan semble avoir été prépondérant et avoir initié les variations du CO2 sur les 10000 dernières années, tandis que la biosphère continentale pourrait avoir joué un rôle de régulateur .