A 1D numerical reaction-transport model (RTM) that is a coupled system of partial differential equations is created to simulate prominent physical and biogeochemical processes and interactions in limnological environments. The prognostic variables considered are temperature, horizontal velocity, salinity, and turbulent kinetic energy of the water column, and the concentrations of phytoplankton, zooplankton, detritus, phosphate (H3PO4), nitrate (NO3-), ammonium (NH4+), ferrous iron (Fe2+), iron(III) hydroxide (Fe(OH)3(s)), and oxygen (O2) suspended within the water column. Turbulence is modelled using the k-e closure scheme as implemented by Gaspar et al. (1990) for oceanic environments. The RTM is used to demonstrate how it is possible to investigate limnological trophic states by considering the problem of eutrophication as an example. A phenomenological investigation of processes leading to and sustaining eutrophication is carried out. A new indexing system that identifies different trophic states, the so-called Self-Consistent Trophic State Index (SCTSI), is proposed. This index does not rely on empirical measurements that are then compared to existing tables for classifying limnological environments into particular trophic states, for example, the concentrations of certain species at certain depths to indicate the trophic state, as is commonly done in the literature. Rather, the index is calculated using dynamic properties of only the limnological environment being considered and examines how those properties affect the sustainability of the ecosystem. Specifically, the index is calculated from a ratio of light attenuation by the ecosystem’s primary biomass to that of total light attenuation by all particulate species and molecular scattering throughout the entire water column. The index is used to probe various simulated scenarios that are believed to be relevant to eutrophication: nutrient loading, nutrient limitation, overabundance of phytoplankton, solar-induced turbulence, and wind-induced turbulence.

A 1D numerical reaction-transport model (RTM) that is a coupled system of partial differential equations is created to simulate prominent physical and biogeochemical processes and interactions in limnological environments. The prognostic variables considered are temperature, horizontal velocity, salinity, and turbulent kinetic energy of the water column, and the concentrations of phytoplankton, zooplankton, detritus, phosphate (H3PO4), nitrate (NO3-), ammonium (NH4+), ferrous iron (Fe2+), iron(III) hydroxide (Fe(OH)3(s)), and oxygen (O2) suspended within the water column. Turbulence is modelled using the k-e closure scheme as implemented by Gaspar et al. (1990) for oceanic environments. The RTM is used to demonstrate how it is possible to investigate limnological trophic states by considering the problem of eutrophication as an example. A phenomenological investigation of processes leading to and sustaining eutrophication is carried out. A new indexing system that identifies different trophic states, the so-called Self-Consistent Trophic State Index (SCTSI), is proposed. This index does not rely on empirical measurements that are then compared to existing tables for classifying limnological environments into particular trophic states, for example, the concentrations of certain species at certain depths to indicate the trophic state, as is commonly done in the literature. Rather, the index is calculated using dynamic properties of only the limnological environment being considered and examines how those properties affect the sustainability of the ecosystem. Specifically, the index is calculated from a ratio of light attenuation by the ecosystem’s primary biomass to that of total light attenuation by all particulate species and molecular scattering throughout the entire water column. The index is used to probe various simulated scenarios that are believed to be relevant to eutrophication: nutrient loading, nutrient limitation, overabundance of phytoplankton, solar-induced turbulence, and wind-induced turbulence.

A 1D numerical reaction-transport model (RTM) that is a coupled system of partial differential equations is created to simulate prominent physical and biogeochemical processes and interactions in limnological environments. The prognostic variables considered are temperature, horizontal velocity, salinity, and turbulent kinetic energy of the water column, and the concentrations of phytoplankton, zooplankton, detritus, phosphate (H3PO4), nitrate (NO3-), ammonium (NH4+), ferrous iron (Fe2+), iron(III) hydroxide (Fe(OH)3(s)), and oxygen (O2) suspended within the water column. Turbulence is modelled using the k-e closure scheme as implemented by Gaspar et al. (1990) for oceanic environments. The RTM is used to demonstrate how it is possible to investigate limnological trophic states by considering the problem of eutrophication as an example. A phenomenological investigation of processes leading to and sustaining eutrophication is carried out. A new indexing system that identifies different trophic states, the so-called Self-Consistent Trophic State Index (SCTSI), is proposed. This index does not rely on empirical measurements that are then compared to existing tables for classifying limnological environments into particular trophic states, for example, the concentrations of certain species at certain depths to indicate the trophic state, as is commonly done in the literature. Rather, the index is calculated using dynamic properties of only the limnological environment being considered and examines how those properties affect the sustainability of the ecosystem. Specifically, the index is calculated from a ratio of light attenuation by the ecosystem’s primary biomass to that of total light attenuation by all particulate species and molecular scattering throughout the entire water column. The index is used to probe various simulated scenarios that are believed to be relevant to eutrophication: nutrient loading, nutrient limitation, overabundance of phytoplankton, solar-induced turbulence, and wind-induced turbulence.

A 1D numerical reaction-transport model (RTM) that is a coupled system of partial differential equations is created to simulate prominent physical and biogeochemical processes and interactions in limnological environments. The prognostic variables considered are temperature, horizontal velocity, salinity, and turbulent kinetic energy of the water column, and the concentrations of phytoplankton, zooplankton, detritus, phosphate (H3PO4), nitrate (NO3-), ammonium (NH4+), ferrous iron (Fe2+), iron(III) hydroxide (Fe(OH)3(s)), and oxygen (O2) suspended within the water column. Turbulence is modelled using the k-e closure scheme as implemented by Gaspar et al. (1990) for oceanic environments. The RTM is used to demonstrate how it is possible to investigate limnological trophic states by considering the problem of eutrophication as an example. A phenomenological investigation of processes leading to and sustaining eutrophication is carried out. A new indexing system that identifies different trophic states, the so-called Self-Consistent Trophic State Index (SCTSI), is proposed. This index does not rely on empirical measurements that are then compared to existing tables for classifying limnological environments into particular trophic states, for example, the concentrations of certain species at certain depths to indicate the trophic state, as is commonly done in the literature. Rather, the index is calculated using dynamic properties of only the limnological environment being considered and examines how those properties affect the sustainability of the ecosystem. Specifically, the index is calculated from a ratio of light attenuation by the ecosystem’s primary biomass to that of total light attenuation by all particulate species and molecular scattering throughout the entire water column. The index is used to probe various simulated scenarios that are believed to be relevant to eutrophication: nutrient loading, nutrient limitation, overabundance of phytoplankton, solar-induced turbulence, and wind-induced turbulence.

Marine sediments are a key component of the global carbon cycle and climate system. They host one of the largest carbon reservoirs on Earth, provide the only long-term sink for atmospheric CO2, recycle nutrients and represent the most important climate archive. Early diagenetic pro- cesses in marine sediments are thus central to our understanding of past, present and future biogeochemical cycling and climate. Because all early diagenetic processes can be directly or indirectly linked back to the degradation of organic matter (OM), advancing this understand- ing requires disentangling the different factors that control the fate of OM (sedimentation, degradation and burial) on different spatial and temporal scales. In general, the heterotrophic degradation of OM in marine sediments is controlled by the quantity and, in particular, by the ap- parent reactivity of OM that settles onto marine sediments. While the potential ((micro)biological, chemical and physical) controls on OM reactivity are increasingly well understood, their relative significance remains difficult to quantify. Traditionally, integrated data-model approaches are used to quantify apparent OM reactivity (i.e. OM degradation rate constants) at well-studied drill-sites. These approaches rely on Reaction-Transport Models (RTMs) that typically account for transport (advection, molecular diffusion, bioturbation, and bioirrigation) and reaction (pro- duction, consumption, equilibrium) processes, but vary in complexity. Apparent OM reactivity (i.e. the OM degradation rate constant) is generally considered as a free parameter that is used to fit observed depth-profiles, reaction rates or benthic-pelagic exchange fluxes. Currently, no quantitative framework exists to predict apparent OM reactivity in areas where comprehensive benthic data sets are not available.To evaluate the impact of this knowledge gap, the sensitivity of benthic biogeochemical reaction rates, as well as benthic-pelagic exchange fluxes to variations in apparent OM reactivity (i.e. reactive continuum model parameters a and ν) is explored by means of a complex, numerical diagenetic model for shelf, slope and deep sea depositional environments. Model results show that apparent OM reactivity exerts a dominant control on the magnitude of biogeochemical reaction rates and benthic-pelagic exchange fluxes across different environments. The lack of a general framework to quantify OM reactivity thus complicates the parametrization of regional and global scale diagenetic models and, thus, compromises our ability to quantify global benthic-pelagic coupling in general and OM degradation dynamics in particular.To make a first step towards an improved systematic and quantitative knowledge of OM reac- tivity, apparent OM reactivity (i.e. reactive continuum model parameters a and ν) is quantified by inverse modelling of organic carbon, sulfate (and methane) sediment profiles, as well as the location of the sulfate-methane transition zone using a complex, numerical diagenetic model for 14 individual sites across different depositional environments. Model results highlight again the dominant control of OM reactivity on biogeochemical reaction rates and benthic exchange fluxes. In addition, results show that, inversely determined ν-values fall within a narrow range (0.1 < ν < 0.2). In contrast, determined a-values span ten orders of magnitude (1 · 10−3 < a < 1·107) and are, thus, the main driver of the global variability in OM reactivity. Exploring these trends in their environmental context reveals that apparent OM reactivity is determined by a dynamic set of environmental controls rather than traditionally proposed single environmental controls (e.g. water depth, sedimentation rate, OM fluxes). However, the high computational demand associated with such a multi-species inverse model approach, as well as the limited availability of comprehensive pore water data, limits the number of apparent OM reactivity estimates. Therefore, while providing important primers for a quantification of OM reactivity on the global scale, inverse model results fall short of providing a predictive framework.To overcome the computational limitations and expand the inverse modelling of apparent OM reactivity to the global scale, the analytical model OMEN-SED is extended by integrating a nG- approximation of the reactive continuum model that is fully consistent with the general structure of OMEN-SED. The new version OMEN-SED-RCM thus provides the computational efficiency required for the inverse determination of apparent OM reactivity (i.e. reactive continuum model parameters a and ν) on the global scale. The abilities of the new model OMEN-SED-RCM in capturing observed local, as well as global patterns of diagenetic dynamics are rigorously tested by model-data, as well as model-model comparison.OMEN-SED-RCM is then used to inversely determine apparent OM reactivity by inverse modelling of 394 individual dissolved oxygen utilisation (DOU) rate measurements. DOU is commonly used as a proxy for OM reactivity, it is more widely available than comprehensive porewater data sets and global/regional benthic maps of dissolved oxygen utilisation rates (DOU) have been derived based on the growing DOU data set. Sensitivity test show that, while inverse modelling of DOU rates fails to provide a robust estimate of RCM parameter ν, it is a good indicator for RCM parameter a. Based on previous findings, parameter ν was thus assumed to be globally constant. Inversely determined a-values vary over order of magnitudes from a = 0.6 years in the South Polar region to a = 5.6 · 106 in the oligotrophic, central South Pacific. Despite a high intra- as well as interregional heterogeneity in apparent benthic OM reactivity, a number of clear regional patterns that broadly agree with previous observations emerge. High apparent OM reactivities are generally observed in regions dominated by marine OM sources and characterized by efficient sinking of OM and a limited degradation during sinking. In contrast, the lowest apparent OM reactivities are observed for regions characterized by low marine primary production rates, in combination with a great distance to the continental shelf and slope, as well as deep water columns. Yet, results also highlight the importance of lateral transport processes for apparent OM reactivity. In particular, deep sea sediments in the vicinity of dynamic continental margin environments or under the influence of strong ocean currents can receive comparably reactive OM inputs from more productive environments and, thus, reveal OM reactivities that are higher than traditionally expected. Finally, based on the observed strong link between apparent OM reactivity (i.e. RCM parameters a) and DOU rate, a transfer function that predicts the order of magnitude of RCM parameter a as a function of DOU is used to derive, to our knowledge, the first global map of apparent OM reactivity.Finally, we use the new global map of apparent OM reactivity to quantify biogeochemical dynamics and benthic-pelagic coupling across 22 benthic provinces that cover the entire global ocean. To this end, the numerical diagenetic model BRNS model is set-up for each province and forced with regionally averaged boundary conditions derived from global data sets, as well as apparent OM reactivities informed by the global OM reactivity map. The 22 regional model set-ups were then used to quantify biogeochemical process rates, as well as benthic carbon and nutrient fluxes in each province and on the global scale. Model results of regional and global fluxes and rates fall well within the range of observed values and also agree with general globally observed patterns. Results also highlight the role of the deeper ocean for benthic-pelagic cycling and indicate towards a large regional variability in benthic cycling at great depth. This is a first step towards a more refined global estimate of benthic biogeochemical cycling that accounts for the global heterogeneity of the seafloor environment. This aspect is critical to improve our understanding of benthic feedbacks on benthic-pelagic coupling and on the carbon-climate system, which can then be incorporated in benthic processes in Earth System Models. / Les sédiments marins sont un élément clé du cycle mondial du carbone et du système climatique. Ils abritent l’un des plus grands réservoirs de carbone sur Terre, fournissent le seul puits à long terme pour le CO2 atmosphérique, recyclent les nutriments et constituent les archives climatiques les plus importantes. Les processus de la diagénèse précoce dans les sédiments marins sont donc au cœur de notre compréhension des cycles et du climat biogéochimiques passés, présents et futurs. Étant donné que tous les processus diagénétiques précoces peuvent être directement ou indirectement liés à la dégradation de la matière organique (MO), faire progresser cette compréhension nécessite de démêler les différents facteurs qui contrôlent le devenir de la MO (sédimentation, dégradation et enfouissement) à différentes échelles spatiales et temporelles. En général, la dégradation hétérotrophique de la MO dans les sédiments marins est contrôlée par la quantité et, en particulier, la réactivité apparente de la MO qui se dépose sur les sédiments marins. Bien que les contrôles potentiels ((micro) biologiques, chimiques et physiques) de la réactivité de la MO soient de mieux en mieux compris, leur importance relative reste difficile à quantifier. Traditionnellement, des approches de modèle de données intégrées sont utilisées pour quantifier la réactivité apparente de la MO (c’est-à-dire les constantes de vitesse de dégradation de la MO) sur des sites de forage bien étudiés. Ces approches reposent sur des modèles de réaction-transport (RTM) qui tiennent généralement compte des processus de transport (advection, diffusion moléculaire, bioturbation et bio-irrigation) et de réaction (production, consommation, équilibre), mais leur complexité varie. La réactivité apparente de la MO est généralement considérée comme un paramètre libre qui est utilisé pour ajuster les profils de profondeur, les taux de réaction ou les flux d’échange benthique-pélagique observés. À l’heure actuelle, aucun cadre quantitatif n’existe pour prédire la réactivité apparente de la MO dans les zones où aucun ensemble complet de données benthiques n’est disponible.Pour évaluer l’impact de ce manque de connaissance, nous avons exploré la sensibilité des taux de réaction biogéochimiques benthiques, ainsi que des flux d’échange benthique-pélagique aux variations de la réactivité apparente de la MO (c.-à-d. les paramètres du modèle de con- tinuum réactif a et ν) au moyen d’un modèle diagénétique numérique complexe appliqué aux zones de dépôts sur les plateaux, les talus et en haute mer. Les résultats du modèle montrent que la réactivité apparente de la MO exerce un contrôle dominant sur l’ampleur des taux de réaction biogéochimiques et des flux d’échange benthique-pélagique dans différents environ- nements. L’absence d’un cadre général pour quantifier la réactivité de la MO complique donc la paramétrisation des modèles diagénétiques à l’échelle régionale et mondiale et, ainsi, compromet notre capacité à quantifier le couplage benthique-pélagique global en général et la dynamique de dégradation de la MO en particulier.Pour tendre à meilleure connaissance systématique et quantitative de la réactivité de la MO, la réactivité apparente OM (c.-à-d. les paramètres du modèle de continuum réactif a et ν) est quantifiée par modélisation inverse des profils de sédiments organiques de carbone, de sulfate (et de méthane), ainsi que localisation de la zone de transition sulfate-méthane à l’aide d’un modèle diagénétique numérique complexe pour 14 sites individuels à travers différents environnements de dépôt. Les résultats du modèle mettent à nouveau en évidence le contrôle dominant de la réactivité de l’OM sur les taux de réaction biogéochimiques et les flux d’échanges benthiques. De plus, les résultats montrent que les valeurs déterminées inversement déterminées se situent dans une plage étroite (0,1 <ν<0,2). En revanche, les valeurs déterminées s’étendent sur dix ordres de grandeur (1 ·10−3 <ν< 1·107) et sont donc le principal moteur de la variabilité globale de la réactivité OM. L’exploration de ces tendances dans leur contexte environnemental révèle que la réactivité apparente de l’OM est déterminée par un ensemble dynamique de contrôles environnementaux plutôt que par des contrôles environnementaux uniques traditionnellement proposés (par exemple, la profondeur de l’eau, le taux de sédimentation, les flux OM). Cependant, la forte demande de calcul associée à une telle approche de modèle inverse multi-espèces, ainsi que la disponibilité limitée de données complètes sur l’eau interstitielle, limitent le nombre d’estimations apparentes de la réactivité OM. Par conséquent, tout en fournissant des amorces importantes pour une quantification de la réactivité de l’OM à l’échelle mondiale, les résultats du modèle inverse sont loin de fournir un cadre prédictif.Pour surmonter les limites de calcul et étendre la modélisation inverse de la réactivité apparente de l’OM à l’échelle mondiale, le modèle analytique OMEN-SED est étendu en intégrant une approximation nG du modèle de continuum réactif qui est pleinement cohérente avec la structure générale d’OMEN-SED. La nouvelle version OMEN-SED-RCM fournit ainsi l’efficacité de calcul requise pour la détermination inverse de la réactivité apparente de l’OM (c’est-à-dire les paramètres du modèle de continuum réactif a et ν) à l’échelle mondiale. Les capacités du nouveau modèle OMEN-SED-RCM à capturer les modèles locaux et globaux de dynamique diagénétique observés sont rigoureusement testés par les données du modèle, ainsi que la comparaison modèle- modèle.OMEN-SED-RCM est ensuite utilisé pour déterminer inversement la réactivité apparente de l’OM par modélisation inverse de 394 mesures individuelles du taux d’utilisation de l’oxygène dissous (DOU). Le DOU est couramment utilisé comme indicateur de la réactivité de l’OM, il est plus largement disponible que les ensembles de données exhaustifs sur l’eau interstitielle et les cartes benthiques mondiales/régionales des taux d’utilisation de l’oxygène dissous (DOU) ont été dérivées sur la base de l’ensemble de données DOU croissant. Le test de sensibilité montre que, bien que la modélisation inverse des taux de DOU ne fournisse pas une estimation robuste du paramètre RCM ν, c’est un bon indicateur pour le paramètre RCM a. Sur la base des résultats précédents, le paramètre ν a donc été supposé être globalement constant. Les valeurs a déterminées à l’inverse varient selon l’ordre de grandeur, de a = 0,6 an dans la région polaire sud à a = 5, 6 · 106 dans le Pacifique sud oligotrophique central. Malgré une forte hétérogénéité intra et interrégionale dans la réactivité apparente de la MO benthique, un certain nombre de schémas régionaux clairs qui correspondent largement aux observations précédentes émergent. Des réactivités apparentes élevées de l’OM sont généralement observées dans les régions dominées par des sources marines de MO et caractérisées par un naufrage efficace de l’OM et une dégradation limitée pendant le naufrage. En revanche, les réactivités MO apparentes les plus faibles sont observées pour les régions caractérisées par de faibles taux de production primaire marine, en combinaison avec une grande distance du plateau continental et de la pente, ainsi que des colonnes d’eau profonde. Pourtant, les résultats mettent également en évidence l’importance des processus de transport latéral pour la réactivité apparente de l’OM.En particulier, les sédiments des mers profondes au voisinage d’environnements de marge continentale dynamiques ou sous l’influence de forts courants océaniques peuvent recevoir des apports OM de réactivité comparable provenant d’environnements plus productifs et, ainsi, révéler des réactivités OM plus élevées que ce qui était traditionnellement prévu. Enfin, sur la base du lien fort observé entre la réactivité apparente de l’OM (c’est-à-dire le paramètre RCM a) et le taux DOU, une fonction de transfert qui prédit l’ordre de grandeur du paramètre RCM a en fonction de DOU est utilisée pour dériver, pour nos connaissances, la première carte mondiale de la réactivité apparente de l’OM. Les résultats du modèle des flux et des taux régionaux et mondiaux se situent bien dans la gamme des valeurs observées et également d’accord avec les tendances générales observées au niveau mondial. Les résultats mettent également en évidence le rôle de l’océan profond pour le cycle benthique-pélagique et indiquent une grande variabilité régionale du cycle benthique à grande profondeur. Il s’agit d’une première étape vers une estimation mondiale plus précise du cycle biogéochimique benthique qui tient compte de l’hétérogénéité mondiale du milieu marin. Cet aspect est essentiel pour améliorer notre compréhension des rétroactions benthiques sur le couplage benthique-pélagique et sur le système carbone-climat, qui peuvent ensuite être incorporées aux processus benthiques dans les modèles du système terrestre. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

<p dir="ltr">Catalyzing a green hydrogen economy can accelerate progress towards achieving the goal of a sustainable energy map with net-zero carbon emissions by rapid strides. An environmentally benign electrochemical energy conversion system is the Polymer Electrolyte Fuel Cell (PEFC) which uses hydrogen as a fuel to produce electricity and is notably used in a variety of markets such as industries, commercial setups, and across the transportation sector, and is gaining prominence for use in heavy-duty vehicles such as buses and trucks. Despite its potential, the commercialization of PEFCs needs to address several challenges which are manifested in the form of mass transport limitations and deleterious mechanisms at the interfacial scale under severe operating conditions. Achieving a robust electrochemical performance in this context is predicated on desired interactions at the triple-phase boundary of the electrochemical engine of the PEFC – the porous cathode catalyst layer (CCL) where the principal oxygen reduction reaction (ORR) takes place. The liquid water produced as a byproduct of the ORR helps minimize membrane dehydration; however, excess water renders the reaction sites inactive causing reactant starvation. In addition, the oxidation of the carbonaceous support in the electrode and loss of valuable electrochemically active surface area (ECSA) pose major barriers that need to be overcome to ameliorate the life expectancy of the PEFC.</p><p dir="ltr">In this thesis, the multimodal physicochemical interactions occurring inside the catalyst layer are investigated through a synergistic blend of visualization and computational techniques. The spatiotemporal dynamics of capillary force-driven liquid transport that ensues concentration polarization thereby affecting the desired response will be probed in detail. The drop in efficacy of the ORR due to competing catalyst aging mechanisms and the impact of degradation stressors on chemical potential-induced instability will be examined. The reaction-transport-mechanics interplay in core-shell nanoparticles, a robust class of electrocatalysts that promises better mass activity compared to the single metal counterparts is further highlighted. Finally, the influence of electrode microstructural attributes on the electrochemical performance of the reverse mode of fuel cell operation, i.e., Proton Exchange Membrane Water Electrolyzers (PEMWEs) is investigated through a mesoscale lens.</p>