Robust preconditioners on block-triangular and block-factorized form for three types of linear systems of two-by-two block form are studied in this thesis. The first type of linear systems, which are dense, arise from a boundary element type of discretization of crack propagation problems. Numerical experiment show that simple algebraic preconditioning strategies results in iterative schemes that are highly competitive with a direct solution method. The second type of algebraic systems, which are sparse, indefinite and nonsymmetric, arise from a finite element (FE) discretization of the partial differential equations (PDE) that describe (visco)elastic glacial isostatic adjustment (GIA). The Schur complement approximation in the block preconditioners is constructed by assembly of local, exactly computed Schur matrices. The quality of the approximation is verified in numerical experiments. When the block preconditioners for the indefinite problem are combined with an inner iterative scheme preconditioned by a (nearly) optimal multilevel preconditioner, the resulting preconditioner is (nearly) optimal and robust with respect to problem size, material parameters, number of space dimensions, and coefficient jumps. Two approaches to mathematically formulate the PDEs for GIA are compared. In the first approach the equations are formulated in their full complexity, whereas in the second their formulation is confined to the features and restrictions of the employed FE package. Different solution methods for the algebraic problem are used in the two approaches. Analysis and numerical experiments reveal that the first strategy is more accurate and efficient than the latter. The block structure in the third type of algebraic systems is due to a fine-coarse splitting of the unknowns. The inverse of the pivot block is approximated by a sparse matrix which is assembled from local, exactly inverted matrices. Numerical experiments and analysis of the approximation show that it is robust with respect to problem size and coefficient jumps.

Several problems in physics and engineering are modeled by reaction-advection-diffusion (RAD) equations. However, when the diffusive terms are small compared with the other ones, these problems can become difficult to solve numerically. Besides, formulating the unsteady version of these models in a semi-discrete fashion, it can be interpreted that the overall diffusivity gets smaller as the time step decreases. To overcome these drawbacks, this thesis considers the development of Galerkin (or Petrov-Galerkin) finite element methods based on approximation spaces enriched by residual-free bubbles (RFB) or multiscale functions. Beginning with the unsteady reaction-diffusion problem, new methods using multiscale functions are presented which improve the solutions in the reaction-dominated regime and/or when small time steps are adopted. They also give rise to a general concept of stabilizing unsteady problems differently along the time. In the following, it is shown that switching RFB by suitable multiscale functions in the elements connected to the outflow boundaries of the domain increases the accuracy of the solutions in this region for RAD problems with advection. Next, this methodology is further studied for systems of RAD equations. In a final contribution, an extension of the RFB method is introduced for the shallow waters equations. All these methods are tested through benchmark problems and compared with stabilized methods presenting stable and accurate results. / A modelagem de vários problemas físicos e de engenharia envolve a solução de problemas de transporte do tipo reação-advecção-difusão (RAD), porém, estes podem tornar-se singularmente perturbados quando os termos difusivos são pequenos comparados aos demais. Além disso, ao adotar formulações semi-discretas em problemas transientes, observa-se que diminuir o passo de tempo tem um efeito de redução da componente difusiva. Para superar estas dificuldades, esta tese considera o desenvolvimento de métodos de elementos finitos de Galerkin (ou Petrov-Galerkin) baseados em espaços de aproximação enriquecidos por funções bolhas livres do resíduo (RFB) ou funções multiescala. Começando pelo problema de reação-difusão transiente, novos métodos utilizando funções multiescala são apresentados, os quais melhoram as soluções no regime reativo-dominante e/ou quando pequenos passos de tempo são adotados. Com estes métodos, discute-se também o conceito de estabilização variável ao longo do tempo para problemas transientes. Na seqüência, verifica-se que utilizar funções multiescala nos elementos conectados às fronteiras de saída de fluxo do domínio e RFB nos demais elementos aumenta a precisão das soluções nesta região em problemas de RAD com advecção dominante. A seguir, esta metodologia é estudada para sistemas de RAD. Como contribuição final, estende-se o método RFB para o modelo de águas rasas. Todos estes métodos são submetidos a testes de robustez e comparados com métodos estabilizados, apresentando resultados estáveis e precisos.

Devido aos riscos de contaminação dos recursos naturais solo e água, ao alto custo, ao tempo e ao esforço humano nas investigações de campo, os modelos matemáticos, aliados às técnicas numéricas e aos avanços computacionais, constituem uma ferramenta importante na previsão do deslocamento de solutos, contribuindo assim, para o controle de alterações ambientais. No Brasil, a modelação de fluxo e transporte de solutos na zona não-saturada é voltada, quase que exclusivamente, aos problemas relacionados às atividades agrícolas. Entretanto, tão importante quanto a problemática dos produtos químicos nas atividades agrícolas é a questão de poluição e contaminação do solo e da água por chorume, gerado pelos resíduos sólidos domiciliares. Neste trabalho, é desenvolvido e validado um modelo computacional unidimensional para simulação de fluxo e transporte de solutos na zona não-saturada do solo. O modelo matemático é dado pela equação diferencial parcial não-linear de Richards, que rege o movimento de água no solo, e a equação diferencial parcial linear de advecção-dispersão, do transporte de solutos, acompanhadas das condições iniciais e de contorno. A equação de Richards é dada em função do potencial matricial da água e a equação de transporte de solutos estima a evolução temporal da concentração de solutos no perfil do solo. Devido à dificuldade de se obter soluções analíticas destas equações, são resolvidas numericamente pelo método de elementos finitos. As referidas equações são resolvidas utilizando-se malhas uniformes inicialmente. Com a finalidade de obter simulações mais eficientes, a um custo computacional reduzido, é empregada a adaptatividade com refinamento h na malha de elementos finitos. A função interpolação polinomial utilizada é de grau 2 ou maior que garante a conservação de massa. Na equação de Richards, a derivada temporal é aproximada por um quociente de diferença finita e é aplicado o esquema de Euler explícito e na equação de advecção-dispersão, é aproximada por um quociente de diferença finita, aplicando-se o esquema de Euler implícito, devido à linearidade da equação. O sistema operacional é o Linux Ubuntu 32 bits, o ambiente de programação é o PZ, escrito em linguagem de programação C++. Na validação do modelo, utilizam-se dados disponíveis na literatura. Os resultados são comparados, utilizando-se malhas uniformes e malhas adaptativas com refinamento h. Usando-se as malhas uniformes para o problema de Richards e de transporte de potássio, o tempo de execução é de 22 minutos e a memória utilizada de 6164 Kb. Com as malhas adaptadas, o tempo de execução é de 3 minutos e 27 segundos, consumindo 5876 Kb de memória. Houve, portanto, uma redução de 84,32% no tempo de execução, usando-se malhas adaptativas. A utilização da função interpolação polinomial de grau 2 ou maior e o refinamento h, permitem uma boa concordância do modelo na comparação com soluções disponíveis na literatura. / Due to the risks of contamination of soil and water resources, the high cost, time and human effort in the field investigations, the mathematical models, combined with numerical techniques and computational advances, are important tools in forecasting the movement of solutes thereby contributing to the control of environmental alteration. In Brazil, modeling of flow and solute transport in the unsaturated zone is focused, almost exclusively, on problems related to agricultural activities. However, as important as the problematical of chemicals products in agricultural activities is the issue of pollution and contamination of soil and water by leachate, generated by municipal solid wastes. In this work, an one-dimensional computational model for simulation of flow and solute transport in the unsaturated soil has been developed and validated. The mathematical model is given by the Richards\'s non-linear partial differential equation, which determines the movement of water in the soil, and the advection-dispersion linear partial differential equation, of the solute transport, together with initial and boundary conditions. The Richards equation is a function of the water pressure head and the solute transport equation estimate the temporal evolution of the solutes concentration in the soil profile. Due to the difficulty of obtaining analytical solutions of these equations, they are solved numerically using the finite element method. The governing equations are solved using initially a uniform mesh. In order to obtain more efficient simulations with low computational cost, adaptativity with h refinement on the finite element mesh is implemented. The interpolation function is of degree two or higher, assuring mass conservation. In Richards\' equation, the temporal derivative is approximated by Euler explicit finite difference. For the advection-dispersion equation, due to the linearity of the equation, an implicit finite difference scheme is used. The code is written in the programming language C++ based on the programming environment PZ using operating system Linux Ubuntu 32 bit. Model results are validated in comparison with data available in the literature. The results are evaluated using uniform meshes and with h refinement adaptive mesh. Using the uniform meshes for the problem of Richards and transport of potassium, the running time is 22 minutes and 6164 Kb of memory is used. With the adapted meshes, the execution time is 3 minutes and 27 seconds, consuming 5,876 Kb of memory. Therefore there was a reduction of 84.32% in execution time, using adaptive meshes. The interpolation function with degree two or higher and the h refinement, with reduction of the computation time, showed a good agreement in comparison with the literature.

La compréhension du fonctionnement des aquifères karstiques est un enjeu considérable au vu des structures complexes de ces réservoirs. La forte hétérogénéité des écoulements induit une grande vulnérabilité de ces milieux et des comportements variés au cours des crues en lien avec différents processus de recharge. Dans le Massif du Jura, les aquifères karstiques constituent la principale ressource en eau potable et posent la question de leur rôle dans la dégradation de la qualité de l'eau observée depuis plusieurs décennies. Cette thèse propose différentes approches complémentaires pour mieux comprendre les dynamiques de crues dans ces aquifères sous diverses conditions hydrologiques. Plusieurs systèmes karstiques du Massif du Jura, présentant des dimensions variables et dominés par des mécanismes de recharges distincts, sont caractérisés à partir de suivis physico-chimiques et hydrochimiques détaillés.Tout d'abord, les différents systèmes sont comparés à l'échelle du cycle hydrologique et à l'échelle saisonnière afin d'identifier les processus de recharge dominants (infiltrations localisées et/ou diffuses) ainsi que les signatures hydrochimiques caractéristiques (arrivées allochtones, autochtones et/ou anthropiques). Une étude comparative de deux systèmes met en avant la forte variabilité saisonnière de la réponse hydrochimique sur un système marqué par une recharge localisée importante. Les différents systèmes sont ensuite analysés à une échelle de temps plus fine afin de mieux comprendre les dynamiques de crues. Une crue intense d'automne a été ainsi comparée à de plus petites crues précédées par des périodes d'étiages importantes et marquées par des signatures hydrochimiques anthropiques significatives. A partir de ces résultats, la méthode EMMA (End-Member Mixing Analysis) est appliquée afin d'établir les principaux pôles hydrochirniques responsables des contributions caractéristiques des différents systèmes. Ensuite, au vu du transport important de matières en suspension au cours des crues dans ces aquifères, une partie de ce travail vise à mieux comprendre le rôle et l'impact de ces matières sur le transport dissous et colloïdal. Les éléments traces métalliques (ETM) sont utilisés afin de caractériser l'origine et la dynamique des transferts. Ils apparaissent alors comme des outils pertinents pour identifier des phénomènes de dépôts et de remobilisation de particules dans le système. Ces dynamiques s'observent à la fois sur le système de Fourbanne marqué par une infiltration localisée importante et sur le petit système du Dahon, caractérisé par une infiltration diffuse.Finalement, afin de mieux comprendre la variabilité spatio-temporelle des interactions qui ont lieu au cours des crues le long du conduit karstique, une nouvelle approche de modélisation est définit. Elle propose l'utilisation des équations de l'onde diffusante et d'advection-diffusion avec la même résolution mathématique (solution analytique d'Hayarni (1951)) en supposant une distribution uniforme des échanges le long du conduit. A partir d'une modélisation inverse, elle permet alors d'identifier et d'estimer les échanges en termes de flux hydriques et de flux massiques entre deux stations de mesure. Cette méthodologie est appliquée sur le système de Fourbanne le long de deux tronçons caractérisant (1) la zone non-saturée et (2) zone non-saturée et saturée. L'analyse de plusieurs crues permet d'observer des dynamiques d'échanges variées sur les deux tronçons. Elle permet ainsi d'établir un schéma de fonctionnement du système soulignant des interactions importantes dans la zone saturée et également le rôle de la zone non-saturée pour le stockage dans le système karstique.Ce travail de thèse propose donc un ensemble d'outils riches et complémentaires pour mieux comprendre les dynamiques de crues et montre l'importance de coupler l'analyse des processus hydrodynamiques et hydrochimiques afin de mieux déchiffrer le fonctionnement de ces aquifères. / The understanding of karst aquifer functioning is a major issue, given the complex structures of these reservoirs. The high heterogeneity of the flows induces a high vulnerability of these media and implies distinct behaviours during floods because of various infiltration processes. In the Jura Mountains, karst aquifers constitute the main source of water drinking supply and raise the question of their role in the degradation of water quality observed for several decades. This work uses complementary approaches to better understand the dynamics of floods in aquifers under various hydrological conditions. Several karst systems of the Jura Mountains, varying in size and characterized by distinct recharge processes, are investigated by detailed physico-chemical and hydrochemical monitoring.First, the different systems are compared at the hydrological cycle scale and at the seasonal scale to identify the dominant recharge processes (localized and/or diffuse infiltrations) as well as the characteristic hydrochemical signatures (allochtonous, autochthonous and/or anthropogenic). A comparative study of two systems with distinct recharge processes highlights the high seasonal variability of the hydrochemical response. The different systems are then analysed on a finer time scale to shed light on flood dynamics. An intense autumn flood was thus compared to smaller floods preceded by periods of significant low flow and marked by significant anthropogenic hydrochemical signatures. The EMMA (End-Member Mixing Analysis) method is applied to these results in order to establish the main hydrochemical end-members responsible for the characteristic contributions of the different systems.Then, considering the important transport of suspended matter during floods in these aquifers, part of this work aims to better understand the role and impact of these materials on dissolved and colloidal transport. Metal trace elements (ETM) are used to characterize the origin and transfer dynamics. These are relevant tools to identify the processes of storage and remobilization of the particles in the system. These dynamics are observed both on the Fourbanne system with an important localized infiltration, and on the small Dahon system, characterized by diffuse infiltration.Finally, in order to shed light on the spatio-temporal variability of the interactions that occur along the karst network during floods, a new modelling approach is defined. It is based upon the use of the diffusive wave and advection­diffusion equations with the same mathematical resolution (Hayami's analytical solution (1951)) assuming a uniform distribution of the exchanges along the reach. An inverse modelling approach allows to identify and estimate the exchanges in terms of water flows and solute between two measurement stations. This methodology is applied to the Fourbanne system on two sections characterizing (1) the unsaturated zone and (2) unsaturated and saturated zone. The analysis of several floods highlights the different exchange dynamics on the two sections. It thus makes it possible to establish a functioning scheme of the system, bringing to light the important interactions in the saturated zone and also the storage role of the unsaturated zone in the karst system.This work offers a set of rich and complementary tools to better characterize the dynamics of floods and shows the importance of coupling the analysis of the hydrodynamic and hydrochemical processes to better decipher the functioning of these aquifers.

La présente thèse porte sur la prédiction numérique de l’impact des transferts thermique et massique sur la qualité de l’air et sur le confort thermique à l’intérieur des cavités ventilées ou non et remplies de polluant. En effet, les cavités ventilées sont généralement considérées comme étant une approximation pour la modélisation des locaux ventilés.Pour mener à bien cette étude, nous avons choisi un modèle numérique basé sur la résolution des équations régissant les transferts de quantité de mouvement, de chaleur et de masse. La première partie du mémoire est consacrée à quelques généralités sur la ventilation, la qualité de l’air et le confort thermique ainsi qu’à la revue bibliographique des travaux réalisés. La démarche suivie est décrite dans le chapitre 2. Celle-ci est basée sur l’approximation de Boussinesq. Le modèle RNG k-ε est utilisé pour traiter la turbulence. La discrétisation des équations est réalisée à l’aide de la méthode des volumes finis associée à l’algorithme SIMPLEC pour traiter le couplage pression-vitesse. Dans la seconde partie, nous considérons d’abord la convection thermique et la convection thermosolutale dans des cavités fermées. Le principal but visé est: a) de valider le modèle considéré en confrontant nos résultats avec ceux de la littérature et b) d’étudier l’influence du nombre de Rayleigh thermique et du rapport de flottabilité. Les résultats obtenus révèlent que le modèle adopté prédit correctement les transferts thermique et massique.Ensuite, nous appliquons cette approche au cas des cavités bidimensionnelles ventilées soumises à des gradients de température et de concentration. Les indices de qualité de l’air et d’efficacité de la ventilation sont calculés et discutés. Nous achevons ce travail en analysant l’influence de la ventilation sur la qualité de l’air intérieur dans une pièce tridimensionnelle en régime transitoire. Cette analyse concerne différents scénarios de ventilation mécanique simple flux en vue de trouver la meilleure configuration en termes d’efficacité et de qualité de l’air intérieur. / This thesis deals with the numerical prediction of heat and mass transfer impact on the air quality and thermal comfort within either ventilated or not cavities filled with pollutants. Indeed, ventilated areas are first modeled to be as ventilated cavities in a first approximation.To carry out this study, we adopt a numerical model based on solving equations governing momentum, heat and mass transfer. The first part of this thesis is dedicated to some generalities on ventilation, air quality and thermal comfort and the bibliographic review of previous works. The adopted approach is described in Chapter 2. It is based on the Boussinesq approximation. The RNG k-ε model is used to handle turbulence. The finite-volume method (FVM) is used to discretize of the set of equations, and the pressure-velocity coupling is achieved via the SIMPLEC algorithm. In the second part, we consider the thermal convection and thermosolutal convection in closed cavities. The main aim is a) to validate the considered model by comparing our results with those of literature, and b) to investigate influence of the thermal Rayleigh number and the buoyancy ratio. Our findings indicate that the model accurately predicts heat and mass transfer.Then, we apply this approach to the case of two-dimensional ventilated cavities subjected to temperature and concentration gradients. The indices of air quality and ventilation efficiency are calculated and discussed. We end this work by analyzing the influence of ventilation on the quality of indoor air in a three-dimensional room in transient regime. This investigation covers different scenarios from the simple flow mechanical ventilation which aims to find the best configuration in terms of efficiency and quality of indoor air.

This thesis focuses on solute transport upscaling. Upscaling of solute transport is usually required to obtain computationally efficient numerical models in many field applications such as, remediation of aquifers, environmental risk to groundwater resources or the design of underground repositories of nuclear waste. The non-Fickian behavior observed in the field, and manifested by peaked concentration profiles with pronounced tailing, has questioned the use of the classical advection-dispersion equation to simulate solute transport at field scale using numerical models with discretizations that cannot capture the field heterogeneity. In this context, we have investigated the use of the advection-dispersion equation with mass transfer as a tool for upscaling solute transport in a general numerical modeling framework. Solute transport by groundwater is very much affected by the presence of high and low water velocity zones, where the contaminant can be channelized or stagnant. These contrasting water velocity zones disappear in the upscaled model as soon as the scale of discretization is larger that the size of these zones. We propose, for the modeling solute transport at large scales, a phenomenological model based on the concept of memory functions, which are used to represent the unresolved processes taking place within each homogenized block in the numerical models. We propose a new method to estimate equivalent blocks, for which transport and mass transfer parameters have to be provided. The new upscaling technique consists in replacing each heterogeneous block by a homogeneous one in which the parameters associated to a memory functions are used to represent the unresolved mass exchange between highly mobile and less mobile zones occurring within the block. Flow upscaling is based on the Simple Laplacian with skin, whereas transport upscaling is based in the estimation of macrodispersion and mass transfer parameters as a result of the interpretation of the r / Llerar Meza, G. (2009). Upscaling nonreactive solute transport [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/5848 / Palancia

The present work was written as a cumulative dissertation based on peer-reviewed papers and is completed by yet unpublished results. The overall objective was to get a deeper insight into the role of the advective fluxes of sensible heat and latent heat in relation to the energy balance and its imbalance at the earth’s surface (typically the sum of the turbulent fluxes sensible and latent heat does not match the available energy). Data from two advection experiments at four coniferous sites across Europe served as the basis for the analysis. One was the advection experiment MORE II which took place in Tharandt (Germany) and the other advection experiment ADVEX was conducted at three different sites (Ritten/Renon, Italy; Wetzstein, Germany; Norunda, Sweden). An inspection of the available energy (AE) that is redistributed to the atmosphere by the sensible heat flux (H) and latent heat flux (LE) showed that the uncertainty of the available energy itself cannot explain the lack of energy balance closure for these four sites. The mean absolute uncertainty of the available energy was largest during midday and ranged from 41 W m-2 to 52 W m-2 (approx. 12 % of AE). During nighttime, the mean absolute uncertainty was smaller (20 W m-2 – 30 W m-2) but the relative uncertainty was much larger as AE itself is small. Among the investigated storage terms the heat storage change of the biomass was most important. The energy balance closure was improved for all investigated sites when storage terms were included. In principle, storage terms should not be neglected in energy balance studies. An investigation of the budget of sensible heat, not only including the vertical advection and the horizontal advection but also the horizontal turbulent flux divergence, was undertaken for the coniferous site at Tharandt. Inclusion of these fluxes resulted in an enlarged mean daily amplitude and suggests an improvement of the energy balance closure, at least during nighttime. The commonly determined budget (vertical turbulent flux plus storage change) was reduced by about 30 % when advective fluxes were included. Results suggest that the horizontal turbulent flux divergence is of minor importance but further studies are needed for an overall evaluation. First results for the inclusion of the advective fluxes of both sensible heat and latent heat indicate that the lack of energy balance closure is partly reduced but the imbalance still exists. Advective fluxes of sensible heat were also compared to advective fluxes of CO2. It became apparent that the advective fluxes of sensible heat and CO2 are, on average, of opposite sign during nighttime and both share large scatter. Both budgets (sensible heat and CO2) were considerably changed (although differently for different sites) when advective fluxes were included. Results further suggest that advective fluxes of H can be taken as an indicator concerning the presence and sign of advection of CO2. This points towards a coincident non-turbulent transport of heat and CO2. However, all investigated advective fluxes are site-specific. They are characterised by a large uncertainty due to uncertainties in the mean vertical velocity (vertical advection) and in the horizontal differences in scalar magnitude (horizontal advection). Obviously, they are influenced by the limitations of the experimental set-up (spatial resolution) and the local characteristics of the individual measurements. An overall evaluation of advective fluxes with respect to their representativeness and magnitude requires further studies / Die vorliegende Arbeit wurde als kumulative Dissertation verfasst, die auf begutachteten Publikationen beruht. Sie wird um bisher nicht veröffentlichte Daten zur Advektion latenter Wärme ergänzt. Ziel war es, vor allem die Rolle der advektiven Flüsse von sensibler und latenter Wärme in Bezug auf die Energiebilanz und das Problem der Energiebilanzschließung an der Erdoberfläche näher zu untersuchen. Unter der Energiebilanzschließungslücke wird im Allgemeinen das Phänomen verstanden, dass die Summe der gemessenen turbulenten Flüsse von sensibler und latenter Wärme zumeist nicht der gemessenen verfügbaren Energie entspricht. Als Datengrundlage für die Arbeiten dienten hierzu die Datensätze von zwei Advektionsexperimenten, die an vier verschiedenen Nadelwaldstandorten in Europa stattfanden. Das erste dieser Advektionsexperimente MORE II fand an der Ankerstation Tharandt (Deutschland) statt und das zweite (ADVEX) wurde an drei verschiedenen Standorten durchgeführt (Ritten/Renon, Italien; Wetzstein, Deutschland; Norunda, Schweden). Eine Untersuchung der verfügbaren Energie (AE), die über den sensiblen Wärmestrom (H) und den latenten Wärmestrom (LE) wieder an die Atmosphäre abgegeben wird, zeigte, dass die in der Bestimmung der verfügbaren Energie liegende Unsicherheit das Problem der Energiebilanzschließungslücke nicht ausreichend erklärt. Die mittlere absolute Unsicherheit der verfügbaren Energie war dabei mittags am größten (41 W m-2 – 52 W m-2; ca. 12 % der verfügbaren Energie). Nachts war diese kleiner (20 W m-2 – 30 W m-2). Jedoch waren dann die relativen Unsicherheiten deutlich größer, da die verfügbare Energie nachts klein ist. Von den betrachteten Speichertermen der Energiebilanz erwies sich die Speicheränderung von Wärme in der Biomasse als am wichtigsten. Für die vier untersuchten Standorte verbesserte sich die Energiebilanzschließung, wenn die Speicherterme mit einbezogen wurden. Grundsätzlich sollten alle Speicherterme bei der Bestimmung der Energiebilanz mit beachtet werden. Für den Nadelwaldstandort Tharandt wurde die Bilanz der sensiblen Wärme unter Beachtung der advektiven Flüsse und der horizontalen turbulenten Flussdivergenz erstellt. Die Einbeziehung der advektiven Flüsse und der horizontalen turbulenten Flussdivergenz führte zu einer Vergrößerung der Amplitude im mittleren Tagesgang und deutet auf eine Verbesserung der Energiebilanzschließung zumindest nachts hin. Im herkömmlichen Sinne wird die Bilanz für Energie oder Massenflüsse als Summe aus vertikalem turbulenten Fluss und Speicheränderung bestimmt. Die Gesamtsumme dieser Bilanz wurde um 30 % reduziert, wenn die advektiven Flüsse mit einbezogen wurden. Hinsichtlich der horizontalen turbulenten Flussdivergenz kann man noch keine abschließende Einschätzung geben. Die vorliegenden Ergebnisse deuten einen vernachlässigbaren Anteil an der Gesamtbilanz für diesen Term an. Erste Ergebnisse für die Bestimmung der Energiebilanz von Nadelwäldern unter Beachtung der advektiven Flüsse von sensibler und latenter Wärme zeigen eine teilweise Reduzierung der Energiebilanzschließungslücke, jedoch keine vollständige Schließung der Energiebilanz. Weiterhin wurden die advektiven Flüsse sensibler Wärme mit denen von CO2 verglichen. Die Bilanzen für den CO2-Fluss als auch für den Fluss sensibler Wärme änderten sich deutlich unter Einbeziehung der advektiven Flüsse, wenn auch unterschiedlich für verschiedene Standorte. Besonders nachts sind die advektiven Flüsse von sensibler Wärme und CO2 im Mittel durch gegensätzliche Vorzeichen gekennzeichnet. Diese Beziehung eröffnet die Möglichkeit, advektive Flüsse von CO2 auf der Basis von advektiven Flüssen sensibler Wärme hinsichtlich ihres Vorhandenseins und ihrer Richtung abzuschätzen. Dies deutet auf einen gleichzeitigen nicht-turbulenten Transport von Wärme und CO2 hin. Generell ist festzustellen, dass alle untersuchten advektiven Flüsse spezifisch für den jeweiligen Standort und durch eine große Unsicherheit gekennzeichnet sind. Diese ergibt sich zum einen aus der mittleren vertikalen Geschwindigkeit (vertikale Advektion) und zum anderen aus den horizontalen Differenzen (horizontale Advektion) der jeweiligen skalaren Größen. Die betrachteten advektiven Flüsse werden offensichtlich durch Einschränkungen, die sich aus dem experimentellen Aufbau ergeben (z.B. begrenzte räumliche Auflösung), in ähnlicher Weise beeinflusst. Eine abschließende Beurteilung der advektiven Flüsse hinsichtlich ihres Anteils an der Gesamtbilanz und ihrer Repräsentativität erfordert weitere Studien.

Disertační práce se zabývá modelováním znečištění ovzduší, jeho transportních a disperzních procesů ve spodní části atmosféry a zejména numerickými metodami, které slouží k řešení těchto modelů. Modelování znečištění ovzduší je velmi důležité pro předpověď kontaminace a pomáhá porozumět samotnému procesu a eliminaci následků. Hlavním tématem práce jsou metody pro řešení modelů popsaných parciálními diferenciálními rovnicemi, přesněji advekčně-difúzní rovnicí. Polovina práce je zaměřena na známou metodu přímek a je zde ukázáno, že tato metoda je vhodná k řešení určitých konkrétních problémů. Dále bylo navrženo a otestováno řešení paralelizace metody přímek, jež ukazuje, že metoda má velký potenciál pro akceleraci na současných grafických kartách a tím pádem i zvětšení přesnosti výpočtu. Druhá polovina práce se zabývá poměrně mladou metodou ELLAM a její aplikací pro řešení atmosférických advekčně-difúzních rovnic. Byla otestována konkrétní forma metody ELLAM společně s navrženými adaptacemi. Z výsledků je zřejmé, že v mnoha případech ELLAM překonává současné používané metody.

[EN] This thesis concerns the analysis of differential equations with uncertain input parameters, in the form of random variables or stochastic processes with any type of probability distributions. In modeling, the input coefficients are set from experimental data, which often involve uncertainties from measurement errors. Moreover, the behavior of the physical phenomenon under study does not follow strict deterministic laws. It is thus more realistic to consider mathematical models with randomness in their formulation. The solution, considered in the sample-path or the mean square sense, is a smooth stochastic process, whose uncertainty has to be quantified. Uncertainty quantification is usually performed by computing the main statistics (expectation and variance) and, if possible, the probability density function. In this dissertation, we study random linear models, based on ordinary differential equations with and without delay and on partial differential equations. The linear structure of the models makes it possible to seek for certain probabilistic solutions and even approximate their probability density functions, which is a difficult goal in general. A very important part of the dissertation is devoted to random second-order linear differential equations, where the coefficients of the equation are stochastic processes and the initial conditions are random variables. The study of this class of differential equations in the random setting is mainly motivated because of their important role in Mathematical Physics. We start by solving the randomized Legendre differential equation in the mean square sense, which allows the approximation of the expectation and the variance of the stochastic solution. The methodology is extended to general random second-order linear differential equations with analytic (expressible as random power series) coefficients, by means of the so-called Fröbenius method. A comparative case study is performed with spectral methods based on polynomial chaos expansions. On the other hand, the Fröbenius method together with Monte Carlo simulation are used to approximate the probability density function of the solution. Several variance reduction methods based on quadrature rules and multilevel strategies are proposed to speed up the Monte Carlo procedure. The last part on random second-order linear differential equations is devoted to a random diffusion-reaction Poisson-type problem, where the probability density function is approximated using a finite difference numerical scheme. The thesis also studies random ordinary differential equations with discrete constant delay. We study the linear autonomous case, when the coefficient of the non-delay component and the parameter of the delay term are both random variables while the initial condition is a stochastic process. It is proved that the deterministic solution constructed with the method of steps that involves the delayed exponential function is a probabilistic solution in the Lebesgue sense. Finally, the last chapter is devoted to the linear advection partial differential equation, subject to stochastic velocity field and initial condition. We solve the equation in the mean square sense and provide new expressions for the probability density function of the solution, even in the non-Gaussian velocity case. / [ES] Esta tesis trata el análisis de ecuaciones diferenciales con parámetros de entrada aleatorios, en la forma de variables aleatorias o procesos estocásticos con cualquier tipo de distribución de probabilidad. En modelización, los coeficientes de entrada se fijan a partir de datos experimentales, los cuales suelen acarrear incertidumbre por los errores de medición. Además, el comportamiento del fenómeno físico bajo estudio no sigue patrones estrictamente deterministas. Es por tanto más realista trabajar con modelos matemáticos con aleatoriedad en su formulación. La solución, considerada en el sentido de caminos aleatorios o en el sentido de media cuadrática, es un proceso estocástico suave, cuya incertidumbre se tiene que cuantificar. La cuantificación de la incertidumbre es a menudo llevada a cabo calculando los principales estadísticos (esperanza y varianza) y, si es posible, la función de densidad de probabilidad. En este trabajo, estudiamos modelos aleatorios lineales, basados en ecuaciones diferenciales ordinarias con y sin retardo, y en ecuaciones en derivadas parciales. La estructura lineal de los modelos nos permite buscar ciertas soluciones probabilísticas e incluso aproximar su función de densidad de probabilidad, lo cual es un objetivo complicado en general. Una parte muy importante de la disertación se dedica a las ecuaciones diferenciales lineales de segundo orden aleatorias, donde los coeficientes de la ecuación son procesos estocásticos y las condiciones iniciales son variables aleatorias. El estudio de esta clase de ecuaciones diferenciales en el contexto aleatorio está motivado principalmente por su importante papel en la Física Matemática. Empezamos resolviendo la ecuación diferencial de Legendre aleatorizada en el sentido de media cuadrática, lo que permite la aproximación de la esperanza y la varianza de la solución estocástica. La metodología se extiende al caso general de ecuaciones diferenciales lineales de segundo orden aleatorias con coeficientes analíticos (expresables como series de potencias), mediante el conocido método de Fröbenius. Se lleva a cabo un estudio comparativo con métodos espectrales basados en expansiones de caos polinomial. Por otro lado, el método de Fröbenius junto con la simulación de Monte Carlo se utilizan para aproximar la función de densidad de probabilidad de la solución. Para acelerar el procedimiento de Monte Carlo, se proponen varios métodos de reducción de la varianza basados en reglas de cuadratura y estrategias multinivel. La última parte sobre ecuaciones diferenciales lineales de segundo orden aleatorias estudia un problema aleatorio de tipo Poisson de difusión-reacción, en el que la función de densidad de probabilidad es aproximada mediante un esquema numérico de diferencias finitas. En la tesis también se tratan ecuaciones diferenciales ordinarias aleatorias con retardo discreto y constante. Estudiamos el caso lineal y autónomo, cuando el coeficiente de la componente no retardada i el parámetro del término retardado son ambos variables aleatorias mientras que la condición inicial es un proceso estocástico. Se demuestra que la solución determinista construida con el método de los pasos y que involucra la función exponencial retardada es una solución probabilística en el sentido de Lebesgue. Finalmente, el último capítulo lo dedicamos a la ecuación en derivadas parciales lineal de advección, sujeta a velocidad y condición inicial estocásticas. Resolvemos la ecuación en el sentido de media cuadrática y damos nuevas expresiones para la función de densidad de probabilidad de la solución, incluso en el caso de velocidad no Gaussiana. / [CA] Aquesta tesi tracta l'anàlisi d'equacions diferencials amb paràmetres d'entrada aleatoris, en la forma de variables aleatòries o processos estocàstics amb qualsevol mena de distribució de probabilitat. En modelització, els coeficients d'entrada són fixats a partir de dades experimentals, les quals solen comportar incertesa pels errors de mesurament. A més a més, el comportament del fenomen físic sota estudi no segueix patrons estrictament deterministes. És per tant més realista treballar amb models matemàtics amb aleatorietat en la seua formulació. La solució, considerada en el sentit de camins aleatoris o en el sentit de mitjana quadràtica, és un procés estocàstic suau, la incertesa del qual s'ha de quantificar. La quantificació de la incertesa és sovint duta a terme calculant els principals estadístics (esperança i variància) i, si es pot, la funció de densitat de probabilitat. En aquest treball, estudiem models aleatoris lineals, basats en equacions diferencials ordinàries amb retard i sense, i en equacions en derivades parcials. L'estructura lineal dels models ens fa possible cercar certes solucions probabilístiques i inclús aproximar la seua funció de densitat de probabilitat, el qual és un objectiu complicat en general. Una part molt important de la dissertació es dedica a les equacions diferencials lineals de segon ordre aleatòries, on els coeficients de l'equació són processos estocàstics i les condicions inicials són variables aleatòries. L'estudi d'aquesta classe d'equacions diferencials en el context aleatori està motivat principalment pel seu important paper en Física Matemàtica. Comencem resolent l'equació diferencial de Legendre aleatoritzada en el sentit de mitjana quadràtica, el que permet l'aproximació de l'esperança i la variància de la solució estocàstica. La metodologia s'estén al cas general d'equacions diferencials lineals de segon ordre aleatòries amb coeficients analítics (expressables com a sèries de potències), per mitjà del conegut mètode de Fröbenius. Es duu a terme un estudi comparatiu amb mètodes espectrals basats en expansions de caos polinomial. Per altra banda, el mètode de Fröbenius juntament amb la simulació de Monte Carlo són emprats per a aproximar la funció de densitat de probabilitat de la solució. Per a accelerar el procediment de Monte Carlo, es proposen diversos mètodes de reducció de la variància basats en regles de quadratura i estratègies multinivell. L'última part sobre equacions diferencials lineals de segon ordre aleatòries estudia un problema aleatori de tipus Poisson de difusió-reacció, en què la funció de densitat de probabilitat és aproximada mitjançant un esquema numèric de diferències finites. En la tesi també es tracten equacions diferencials ordinàries aleatòries amb retard discret i constant. Estudiem el cas lineal i autònom, quan el coeficient del component no retardat i el paràmetre del terme retardat són ambdós variables aleatòries mentre que la condició inicial és un procés estocàstic. Es prova que la solució determinista construïda amb el mètode dels passos i que involucra la funció exponencial retardada és una solució probabilística en el sentit de Lebesgue. Finalment, el darrer capítol el dediquem a l'equació en derivades parcials lineal d'advecció, subjecta a velocitat i condició inicial estocàstiques. Resolem l'equació en el sentit de mitjana quadràtica i donem noves expressions per a la funció de densitat de probabilitat de la solució, inclús en el cas de velocitat no Gaussiana. / This work has been supported by the Spanish Ministerio de Economía y Competitividad grant MTM2017–89664–P. I acknowledge the doctorate scholarship granted by Programa de Ayudas de Investigación y Desarrollo (PAID), Universitat Politècnica de València. / Jornet Sanz, M. (2020). Mean square solutions of random linear models and computation of their probability density function [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/138394

<p>Doktorska disertacija je posvećena re&scaron;avanju Ko&scaron;ijevog problema odabranih klasa frakcionih diferencijalnih jednačina u okviru Kolomboovih prostora uop&scaron;tenih funkcija. U prvom delu disertacije razmatrane su nehomogene evolucione jednačine sa prostorno frakcionim diferencijalnim operatorima reda 0 &lt; &alpha; &lt; 2 i koeficijentima koji zavise od x i t. Ova klasa jednačina je aproksimativno re&scaron;avana, tako &scaron;to je umesto početne jednačine razmatrana aproksimativna jednačina data preko regularizovanih frakcionih izvoda, odnosno, njihovih regularizovanih množitelja. Za re&scaron;avanje smo koristili dobro poznate uop&scaron;tene uniformno neprekidne polugrupe operatora. U drugom delu disertacije aproksimativno su re&scaron;avane nehomogene frakcione evolucione jednačine sa Kaputovim<br />frakcionim izvodom reda 0 &lt; &alpha; &lt; 2, linearnim, zatvorenim i gusto definisanim<br />operatorom na prostoru Soboljeva celobrojnog reda i koeficijentima koji zavise<br />od x. Odgovarajuća aproksimativna jednačina sadrži uop&scaron;teni operator asociran sa polaznim operatorom, dok su re&scaron;enja dobijena primenom, za tu svrhu&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;<br />u disertaciji konstruisanih, uop&scaron;tenih uniformno neprekidnih operatora re&scaron;enja.<br />U oba slučaja ispitivani su uslovi koji obezbeduju egzistenciju i jedinstvenost<br />re&scaron;enja Ko&scaron;ijevog problema na odgovarajućem Kolomboovom prostoru.</p> / <p>Colombeau spaces of generalized functions. In the firs part, we studied inhomogeneous evolution equations with space fractional differential operators of order 0 &lt; &alpha; &lt; 2 and variable coefficients depending on x and t. This class of equations is solved&nbsp; approximately, in such a way that instead of the originate equation we considered the corresponding approximate equation given by regularized fractional derivatives, i.e. their&nbsp; regularized multipliers. In the solving procedure we used a well-known generalized uniformly continuous semigroups of operators. In the second part, we solved approximately inhomogeneous fractional evolution equations with Caputo fractional derivative of order 0 &lt; &alpha; &lt; 2, linear, closed and densely defined operator in Sobolev space of integer order and variable coefficients depending on x. The corresponding approximate equation&nbsp;&nbsp; is a given by the generalized operator associated to the originate&nbsp; operator, while the solutions are obtained by using generalized uniformly continuous solution operators, introduced and developed for that purpose. In both cases, we provided the conditions that ensure the existence and uniqueness solutions of the Cauchy problem in some Colombeau spaces.</p>