Submitted by Maicon Juliano Schmidt (maicons) on 2015-05-25T14:00:15Z No. of bitstreams: 1 Jorge Augusto Dias Samsonescu.pdf: 443638 bytes, checksum: 0ac887f981377608fa611c1016a91b22 (MD5) / Made available in DSpace on 2015-05-25T14:00:15Z (GMT). No. of bitstreams: 1 Jorge Augusto Dias Samsonescu.pdf: 443638 bytes, checksum: 0ac887f981377608fa611c1016a91b22 (MD5) Previous issue date: 2015-02-20 / Banco do Brasil S.A. / Este trabalho analisa o desempenho fora da amostra de carteiras de mínima variância e baixa volatilidade no mercado de ações brasileiro entre 2003 e 2013, comparativamente ao índice IBOVESPA e a uma carteira igualmente ponderada. As carteiras de mínima variância foram otimizadas com restrição de posições vendidas e limite de peso para os ativos. A matriz de covariância foi estimada pelo método amostral e método shrinkage proposto por Ledoit e Wolf (2003). A carteira de baixa volatilidade foi estruturada de forma similar ao método do índice S&P 500 Low Volatility. O período utilizado para o rebalanceamento das carteiras foi quadrimestral e os ativos elegíveis para as carteiras foram os componentes do IBOVESPA em cada quadrimestre analisado. A comparação das carteiras foi feita através dos indicadores de retorno, desvio padrão e índice de Sharpe anualizados, MVaR e maximum drawdown. Os resultados apontam para a importância na escolha do limite de pesos para os ativos das carteiras de mínima variância. As carteiras de menor risco obtiveram os melhores resultados em todos os indicadores testados. / This study analyzes the out-of-sample performance of minimum-variance and low volatility portfolios in the Brazilian stock market from 2003 to 2013, when compared to IBOVESPA index and an equally weighted portfolio. The minimum variance portfolios have been optimized with short selling restriction and weight limits for the assets. The covariance matrix was estimated by sample method and shrinkage method proposed by Ledoit & Wolf (2003). The low volatility portfolio was structured in a similar way to the S&P 500 Low Volatility index method. The portfolios rebalancing period were quarterly and the eligible assets for the portfolios were IBOVESPA components in each analyzed period. The portfolios performance was evaluated through indicators such return, standard deviation, Sharpe ratio, maximum drawdown and MVAR indicators. The results point to the importance in choosing the weight limits for the assets of minimum-variance portfolios. Lower risk portfolios delivered the best results in all tested indicators.

Resumo: Parâmetros genéticos para a produção de leite no dia do controle (PLDC) de primeiras lactações de vacas da raça Guzerá foram estimados utilizando modelo multicaracterísticas de dimensão finita (TMDO) e modelos de regressão aleatória (MRA). A produção acumulada em 305 dias (P305), duração da lactação (DL) e persistência da lactação (PS) também foram analisadas. Para o TMDO, foram analisadas as PLDC juntamente com a P305 e a DL, considerando como aleatórios, o efeito genético aditivo e o residual e, como fixo, o grupo de contemporâneos e a covariável idade da vaca ao parto. Para os MRA, foram considerados como aleatório, o efeito genético aditivo, de ambiente permanente e residual e como efeito fixo, o grupo de contemporâneos, os efeitos linear e quadrático da covariável idade ao parto e a curva média da população. Para os MRA foram consideradas as funções de ajuste de Wilmink (WL), Ali & Schaeffer (AS), uma combinação entre a função de Wilmink com polinômios ortogonais de Legendre (LM), polinômios ortogonais de Legendre (LEG) e funções B-spline (BS). Os efeitos aleatórios genético aditivo e de ambiente permanente foram modelados por meio destas funções, bem como a curva média da população, com a exceção dos modelos ajustados por funções BS que tiveram a curva média ajustada por polinômio de Legendre ou pela função de Ali & Schaeffer. O resíduo foi ajustado considerando variância homogênea ou em classes heterogêneas de variância residual. O modelo empregando funções BS cúbica com número de coeficientes de regressão aleatória igual cinco tanto para efeito genético aditivo como de ambiente permanente com a curva média modelada pela função de Ali & Schaeffer e resíduo ajustado por seis classes variância residual foi o mais adequado. Entretanto, os melhores MRA para cada função de ajuste, não apresentaram diferenças para ... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Genetic parameters for milk production in the test day model (PLDC) for Guzerat dams' first lactations were estimated by multitrait finite model (TMDO) and random regression models (MRA). The cumulative production at 305 days (P305), lactation length (DL) and lactation persistency (PS) were also analyzed. For TMDO, the PLDC were analyzed together with P305 and DL, considering the additive genetic effect and residual effect as random effects , the contemporary group as a fixed effect, and the age of dam at calving as a covariate. For MRA, additive genetic effect, permanent environmental effect and residual effect were considered as random effects and the contemporary group, the linear and quadratic covariate of age at calving and the average curve of the population as fixed effects. Also for the MRA, the Wilmink (WL) and the Ali & Schaeffer (AS) adjustment functions, a combination of the Wilmink function with Legendre orthogonal polynomials (LM), Legendre orthogonal polynomials (LEG) and B-spline functions (BS) were considered. The random additive genetic and permanent environmental effects were modeled by means of these functions, as well as the population average curve , with the exception of the adjusted models by the BS functions that had the average curve adjusted by the Legendre polynomial or by the Ali & Schaeffer function. The residual error was adjusted considering homogeneous variance or heterogeneous classes of residual variance. The model using cubic BS functions with random regression coefficient numbers equal to five for additive genetic effect as well as for permanent environmental with average curve modeled by the Ali & Schaeffer function and residual error adjusted for six classes of residual variance was the more appropriate. However, the best MRA for each adjustment function presented no differences in the estimates of genetic parameters and for order correlation ... (Complete abstract click electronic access below) / Orientador: Humberto Tonhati / Coorientador: Maria Gabriela Campolina Diniz Peixoto / Banca: Lenir El Faro Zadra / Banca: Fernando Sebastián Baldi Rey / Mestre

The main goal of this study is to propose a methodology that measures directed relations between locally stationary processes. Unlike stationary processes, locally stationary processes may present sudden pattern changes and have local characteristics in specific intervals. This behavior causes instability in measures based on Fourier transforms. The relevance of this study relies on considering these processes and propose robust methodologies that are not affected by outliers, sudden pattern changes or local behavior. We start reviewing the Partial Directed Coherence (PDC) and the Wavelet Coherence. PDC measures the directed relation between components of a multivariate stationary Vector Autoregressive (VAR) model in the frequency domain, while Wavelet Coherence is based on complex wavelets decomposition. We then propose a causal wavelet decomposition of the covariance structure for bivariate locally stationary processes: the Directed Wavelet Covariance (DWC). Compared to Fourier-based quantities, wavelet-based estimators are more appropriate for non-stationary processes and processes with local patterns, outliers and rapid regime changes like in EEG experiments with the introduction of stimuli. We then propose its estimators and calculate its expectation and analyze its variance. Next we propose a decomposition for the variance of multivariate processes with more than two components: the Partial Directed Wavelet Covariance (pDWC). Considering a N-variate locally stationary process, the pDWC calculates the Directed Wavelet Covariance of X_1(t) with X_2(t) eliminating the effect of the other components X_3(t), ... ,X_N(t). We propose two approaches to this situation. First we filter the multivariate process to remove all the exogenous influences and then we calculate the directed relation between the components. In the second case, as in Partial Directed Coherence, we consider the multivariate process as a time-varying Vector Autoregressive Model (tv-VAR) and use its coefficients in the decomposition of the covariance function to isolate the effects of the other components. We also compare results of the PDC, Wavelet Coherence and Directed Wavelet Covariance with simulated data. Finally, we present an application of the proposed Directed Wavelet Covariance and Partial Directed Wavelet Covariance on EEG data. Simulation results show that the proposed measures capture the simulated relations. The pDWC with linear filter has shown more stable estimations than the proposed pDWC considering the tv-VAR. Future studies will discuss the DWC\'s and pDWC\'s asymptotic distributions and significance tests. The proposed Directed Wavelet Covariance decomposition is a different approach to deal with non-stationary processes in the context of causality. The use of wavelets is a gain and adds to the number of studies that can be addressed when Fourier transform does not apply. The pDWC is an alternative for multivariate processes and it removes linear influences from observed external components. / O objetivo deste trabalho é propor uma metodologia para mensurar o impacto direcionado entre processos localmente estacionários. Diferente de processos estacionários, processos localmente estacionários podem apresentar mudanças bruscas e características específicas em determinados intervalos. Tal comportamento pode causar instabilidade em medidas baseadas na transformada de Fourier. A importância deste estudo se dá em englobar processos com tais características, propondo metodologias robustas que não são afetadas pela existência de mudanças bruscas, pontos discrepantes e comportamentos locais. Inicialmente apresentamos conceitos já existentes na literatura, como a Coerência Parcial Direcionada (PDC) e a Coerência de Ondaletas. A PDC mede o impacto direcionado entre componentes de um modelo vetorial autoregressivo (VAR) no domínio da frequência. A coerência de ondaletas é baseada em transformadas complexas de ondaletas. Propomos então uma decomposição no domínio de ondaletas para a estrutura de covariância de processos bivariados localmente estacionários: a Covariância Direcionada de Ondaletas (DWC). Em comparação com as quantidades baseadas na tranformada Fourier, os estimadores baseados em ondaletas são mais apropriados para processos não estacionários com padrões locais, pontos discrepantes ou mudanças rápidas de regime, como em experimentos de eletroencefalograma (EEG) com a introdução de estímulo. Ainda, propomos um estimador para a DWC, calculamos a esperança deste estimador e avaliamos sua variância. Em seguida, propomos uma quantidade análoga à DWC para processos multivariados com mais de duas componentes: a Covariância Parcial Direcionada de Ondaletas (pDWC). Considerando um processo N-variado localmente estacionário, a pDWC calcula a Covariância Direcionada de Ondaletas entre X_1(t) e X_2(t) eliminando o efeito das outras componentes X_3(t), ... , X_N(t). Propomos duas abordagens para a pDWC: na primeira, a pDWC é calculada após a aplicação de um filtro linear que remove o efeito das variáveis exógenas. No segundo caso, a exemplo da Coerência Parcial Direcionada, consideramos o processo multivariado como um Modelo Autoregressivo de Vetorial variante no tempo (tv-VAR) e usamos seus coeficientes na decomposição da função de covariância para isolar os efeitos das demais componentes. Também comparamos os resultados da PDC, Coerência de Ondaletas e Covariância Direcionada de Ondaletas com dados simulados. Por fim, apresentamos uma aplicação da DWC e da pDWC em dados de EEG. Identificamos nas simulações que tanto as medidas já existentes na literatura quanto as quantidades propostas identificaram as relações simuladas. A pDWC proposta com filtros lineares apresentou estimações mais estáveis do que a pDWC considerando os modelos tv-VAR. Estudos futuros discutirão as propriedades assintóticas e testes de significância da DWC e pDWC.

Generalized estimating equations (GEE) are popularly utilized for the marginal analysis of longitudinal data. In order to obtain consistent regression parameter estimates, these estimating equations must be unbiased. However, when certain types of time-dependent covariates are presented, these equations can be biased unless an independence working correlation structure is employed. Moreover, in this case regression parameter estimation can be very inefficient because not all valid moment conditions are incorporated within the corresponding estimating equations. Therefore, approaches using the generalized method of moments or quadratic inference functions have been proposed for utilizing all valid moment conditions. However, we have found that such methods will not always provide valid inference and can also be improved upon in terms of finite-sample regression parameter estimation. Therefore, we propose a modified GEE approach and a selection method that will both ensure the validity of inference and improve regression parameter estimation. In addition, these modified approaches assume the data analyst knows the type of time-dependent covariate, although this likely is not the case in practice. Whereas hypothesis testing has been used to determine covariate type, we propose a novel strategy to select a working covariate type in order to avoid potentially high type II error rates with these hypothesis testing procedures. Parameter estimates resulting from our proposed method are consistent and have overall improved mean squared error relative to hypothesis testing approaches. Finally, for some real-world examples the use of mean regression models may be sensitive to skewness and outliers in the data. Therefore, we extend our approaches from their use with marginal quantile regression to modeling the conditional quantiles of the response variable. Existing and proposed methods are compared in simulation studies and application examples.

O planejamento e o funcionamento do sistema elétrico de potência se baseiam na correta parametrização e caracterização de seus elementos, pois a correta parametrização dos sistemas de proteção permite uma atuação confiável e segura. Ademais, a correta caracterização dos parâmetros de elementos como as linhas de transmissão permite calcular o carregamento ótimo para determinados trechos do sistema interconectado com relação ao fluxo de potência, permitindo um melhor planejamento para expansão e instalação de reforços, dentre outros. Desta forma, foram desenvolvidas metodologias para estimação de parâmetros de sistemas de transmissão, que se baseiam na adoção de um modelo para o sistema e utilização de um método de resolução para se obter uma resposta, ou função de transferência, para as entradas e saídas deste modelo. No entanto, a maioria dos métodos existentes na literatura técnica apresenta certas limitações com relação à presença de ruído nos sinais de medição, ou dificuldade na relação existente entre as correntes longitudinais e as medições terminais, ou imprecisão do método de resolução do sistema de equações para determinada situação. Neste trabalho, foram utilizadas técnicas diferentes para a estimação de parâmetros de uma linha diretamente das medições das correntes e tensões terminais da linha em regime, sendo possível reduzir bastante o erro de estimação, pois não é necessário nenhum método de eliminação dos ruídos das medições. Desta forma, a corrente longitudinal é considerada como equivalente à última componente de corrente terminal, e a metodologia adotada apresenta erros pequenos de estimação independente do modelo adotado. Assim, modelam-se linhas de transmissão em cascata de circuitos ?, obtendo-se as medições ruidosas oriundas desta. Por fim, utiliza-se a metodologia baseada na teoria do Filtro de Kalman Unscented para eliminação do ruído nas grandezas medidas e para estimação dos parâmetros série da linha de transmissão. Através da ferramenta computacional Simulation and model-based design (Simulink), realizam-se a obtenção das medições, inclusão de ruído aleatório a estas e cálculos computacionais para estimação dos estados e parâmetros do sistema em regime permanente. / The planning and operation of a power system are based on the correct parameterization and characterization of its elements. By a correct parameterization of protection systems, a reliable and safe operation is allowed. The correct characterization of the parameters of transmission lines allows calculating the optimum power flow for the interconnected power system, allowing a better planning for expansion and installation of reinforcements, among others. In this sense, methodologies were developed for the estimation of transmission system parameters, which are based on the adoption of a model for the system and usage of a resolution method to obtain a response, or transfer function, for the inputs and outputs of the model. However, the most of existing methods in the technical literature presents certain limitations regarding presence of noise in the measurement signals, or difficulty regarding the relationship between the longitudinal currents and the terminal measurement signals, or imprecision of the solving method for the equations system. In this work, different techniques were applied for the estimation of line parameters directly from the measurements of the terminal currents and terminal voltages of the steady state line, being possible to greatly reduce the estimation error, since no extra method of eliminating measurements noise is necessary. In this way the longitudinal current is considered as equivalent to the last terminal current component, and the adopted methodology presents small estimation errors regardless of the adopted model. Thus, transmission lines are modeled by a pi cascade and noisy measurements are obtained. Finally, the methodology is based on the Unscented Kalman Filter theory for noise elimination in the measurements and for estimation of the transmission line longitudinal parameters. By the computational tool Simulink, measurements are obtained, random noise is inserted and computational calculations for estimation of the states and parameters of the system in steady state are done.

本篇論文的目的，希望以台灣為出發點來看，台灣與國際股市間的互動性，所著重的指標仍以共變數與相關係數為主，但同時也考慮了相關係數具有序列相關的影響，希望藉由這種模型的設計能夠找出台灣與國際股市之間的互動性，進而加以應用。 本文探討台灣與美國 NYSE，台灣與日本 Nikkie 225，台灣與香港，台灣與韓國的股票市場間日相關係數與共變異數，採用兩變數 GARCH 模型，實證結果發現： 1，其實台灣與國際市場間的共變異數矩陣，不是為一常數，是具有時間數列的關係存在的，因而具有可預測性。 2，台灣與韓國股市之間的關連性很小，因為他們之間永久共變數與暫時共變數都不顯著異於零；而台灣與美國 NYSE、台灣與日本 Nikkie225、或台灣與香港恆生指數等，他們之間永久共變數與暫時共變數都顯著異於零，這表示台灣與這些國家門存在有相當的關係。 3，台灣與國際市場的條件日相關係數是具有正向的時間趨勢的。 4，此一兩變數的 GARCH 模型於樣本外的共變數矩陣之預估能力，相較於傳統的 CAPM 模型，較具有準確性，因而可以在國際投資方面，將資金更有效的國際市場上，獲得更好的投資績效。 / we examine the co-movements of equity returns in five major international markets by characterizing the time-varying cross-country covariances and correltions. Using a generalized positive definite multivariate GARCH model, we find that Taiwan and Korea stock markets have zero permanent and transitory covariance.The other pairs of markets examined display significant permanent and transitory covariance. We also find that ,while conditional correlations betweem returns are gernerally small, they change considerably over time. An event analysis suggests that basing diversification strategies on these conditional correlations is potentially beneficial.

Les réacteurs actuels produisent deux types de déchets dont la gestion et le devenir soulèvent des problèmes. Il s'agit d'abord de certains produits de fission et de noyaux lourds (isotopes de l'Américium et du Curium) au-delà de l'uranium appelés actinides mineurs. Deux options sont envisagées : le stockage en site géologique profond et/ou l'incinération de ces déchets dans un flux de neutrons rapides, c'est-à-dire, la transmutation par fission. Ces études font appel à de nombreuses données neutroniques. Malheureusement, les bases de données présentent encore de nombreuses insuffisances pour parvenir à des résultats fiables. L'objectif de ce travail est ici d'actualiser des données nucléaires et de les compléter. Nous avons ainsi mesuré la section efficace de fission de l'243Am (7370 ans) en référence à la diffusion élastique (n,p) afin de fournir des données indépendantes des mesures existantes dans la gamme des neutrons rapides (1 - 8 MeV). La réaction 243Am(n,f) a été analysée en utilisant un modèle statistique décrivant les voies de désexcitation du noyau composé d'244Am. Ainsi les sections efficaces de capture radiative (n,) et de diffusion inélastique (n,n') ont pu être évaluées. La mesure directe des sections efficaces neutroniques d'actinides mineurs constitue très souvent un véritable défi compte tenu de la forte activité des actinides mineurs. Pour cela, une méthode indirecte a été développée utilisant les réactions de transfert dans le but d'étudier certains isotopes du curium. Les réactions 243Am(3He,d)244Cm, 243Am(3He,t)243Cm et 243Am(3He,alpha)242Am nous ont permis de mesurer les probabilités de fission des noyaux de 243,244Cm et de l'242Am. Les sections efficaces de fission des curiums 242,243Cm(162,9 j, 28,5 ans) et de l'américium 241Am sont obtenues en multipliant ces probabilités par les sections efficaces calculées de formation des noyaux composés. Pour chaque mesure, une évaluation précise des erreurs a été réalisée à travers une étude des variances-covariances des résultats présentés. Pour les mesures de la réaction 243Am(n,f), une analyse des corrélations d'erreurs a permis d'interpréter la portée de ces mesures au sein des mesures existantes.

En radiologie interventionnelle, le praticien se sert actuellement de plusieurs acquisitions scanner (coupes 2D) pour pouvoir atteindre sa cible au cours d'une ponction percutanée dans la zone abdominale. Notre objectif est de concevoir un système de guidage minimisant ces acquisitions et donc le temps de l'intervention. Pour cela, nous proposons un système de réalité augmentée superposant des reconstructions 3D pré-opératoires des structures abdominales du patient dans des images vidéo externes de son corps. En plus d'être précis, rapide et fiable, cet outil devra pouvoir être introduit aisément en salle d'opération.<br />Dans notre cas, le patient est intubé et sa ventilation contrôlée, nous pouvons donc négliger les effets de la respiration : un recalage rigide 3D/2D de marqueurs radio-opaques collés sur la peau est suffisant pour atteindre la précision requise. Les hypothèses statistiques des critères classiques n'étant pas adéquates pour notre application, nous avons dérivé un nouveau critère généralisant les approches standard. Une évaluation rigoureuse des performances démontre la supériorité de notre méthode en terme de précision et de robustesse.<br />Pour atteindre le temps réel en salle d'opération, nous avons ensuite développé un ensemble d'algorithmes d'extraction et de mise en correspondance des marqueurs radio-opaques dont nous avons validé la robustesse sur de nombreuses images réelles. La précision du système dépendant de nombreux paramètres (nombre de marqueurs radio-opaques, position des caméras...), elle ne peut pas être établie préalablement de manière définitive. Afin de fournir un système fiable, nous proposons donc une technique de propagation des covariances qui permet d'estimer dynamiquement l'erreur de repositionnement des modèles reconstruits. Une phase de validation méticuleuse, sur des données synthétiques et réelles, démontre que notre prédiction est fiable dans les conditions de notre application.<br />Après cette validation de chacun des modules, nous montrons la faisabilité et l'intérêt de notre système complet en menant une évaluation sur un mannequin : quatre chirurgiens ont réussi à atteindre des cibles en des temps dix fois inférieurs à ceux usuellement nécessaires pour ce type d'intervention et avec une précision supérieure. Finalement, plusieurs expériences cliniques sur des patients démontrent que notre système est utilisable en salle d'opération et suggèrent son utilisation en routine dans un futur proche.

Ce travail de thèse a pour but l'étude expérimentale et théorique de l'interaction de la molécule H2 avec des impulsions laser de durée inférieure à 10fs. L'éjection des deux électrons de la molécule par le champ laser conduit à la fragmentation du système en deux protons. Ce processus est appelé explosion coulombienne. La mesure des spectres d'énergie cinétique des protons permet d'analyser les dynamiques électronique et nucléaire en fonction des différents paramètres laser. Ces dynamiques sont également analysées dans le cadre d'un modèle théorique non perturbatif, à deux électrons actifs, basé sur l'équation de Schrödinger dépendant du temps. Dans ce modèle, la distance internucléaire est traitée de façon quantique.<br /><br /><br />La complémentarité des résultats expérimentaux et théoriques permet de mettre en évidence la translation des spectres d'énergie cinétique vers les énergies plus élevées lorsque la durée de l'impulsion diminue. Cette étude est réalisée pour des impulsions dans la gamme de 40 à 10fs expérimentalement et jusqu'à 1fs théoriquement. Cette étude montre également que, pour des durées d'impulsion laser inférieures à 4fs, la phase absolue devient un paramètre essentiel à prendre en compte. En outre, la dynamique moléculaire de H2 en champ laser intense ultracourt est également sensible à la valeur de l'éclairement crête de l'impulsion. Les résultats théoriques et expérimentaux montrent que les spectres d'énergie sont centrés autour d'une énergie plus élevée quand l'éclairement augmente. Par ailleurs, deux régimes d'ionisation double sont également mis en évidence théoriquement pour des impulsions de 4fs. La sensibilité de H2 à la qualité temporelle de l'impulsion laser permet une détection, par l'intermédiaire des spectres expérimentaux d'énergie cinétique, des pré- ou post-impulsions susceptibles d'apparaître autour de l'impulsion laser principale. Enfin, les différents types d'ionisation double sont étudiés et les résultats mettent en évidence la dynamique électronique attoseconde de la recollision et l'influence de cette dernière sur la dynamique nucléaire femtoseconde.

<p>The overall objective of the current study is to develop a quantitative understanding of the effects of spatial variability in physical and geochemical properties of crystalline rock on the migration of radionuclides along a single fracture in bedrock. A stochastic model was developed to describe the transport of solutes in fractured rock. The model describes the migration of radionuclides along a one-dimensional path and includes the transversal diffusion into the rock matrix and sorption kinetics. By using a Lagrangian method of description we can extend the model to the description of a two-dimensional transport problem in single fractures. </p><p>This study presents the first analysis of the impact of heterogeneous mass transfer on the transport of radionuclides in rock fractures, where most of the relevant rock properties such as aperture, porosity, effective diffusivity, sorption capacity and maximum diffusion depth are defined as being spatially random. The stochastic analysis performed here reflects the uncertainty in our knowledge of the properties associated with a discrete sampling technique in site investigations.</p><p>Geostatistics of the main parameters was determined experimentally on a large number of rock samples taken from the Swedish crystalline basement. The knowledge of the covariance functions of the main rock properties is then used as a basis for a stochastic analysis. By combining the small perturbation approach with the spectral method the problem could be solved in terms of closed form solutions for the central temporal moments of the residence time probability density function. </p><p>In order to be able to distinguish between the effects of various mechanisms from the effects of heterogeneity on the migration of radionuclides, it was necessary to perform independent studies of the effect of the variation of the dispersion coefficient on the aspect ratio of a rectangular flow section and the effect of adsorption kinetics on the migration. </p><p>Finally, the effect of the heterogeneous rock properties on the solute transport observed in a limited number of migration experiments corresponds fairly well to the theoretical effect expected on the basis of the experimentally determined auto-covariance functions. </p>