Rock type is a key concept in modern reservoir characterization that straddles multiple scales and bridges multiple disciplines. Reservoir rock classification (or simply rock typing) has been recognized as one of the most effective description tools to facilitate large-scale reservoir modeling and simulation. This dissertation aims to integrate core data and well logs to enhance reservoir description by classifying reservoir rocks in a geologically and petrophysically consistent manner. The main objective is to develop scientific approaches for utilizing multi-physics rock data at different time and length scales to describe reservoir rock-fluid systems. Emphasis is placed on transferring physical understanding of rock types from limited ground-truthing core data to abundant well logs using fast log simulations in a multi-layered earth model. Bimodal log-normal pore-size distribution functions derived from mercury injection capillary pressure (MICP) data are first introduced to characterize complex pore systems in carbonate and tight-gas sandstone reservoirs. Six pore-system attributes are interpreted and integrated to define petrophysical orthogonality or dissimilarity between two pore systems of bimodal log-normal distributions. A simple three-dimensional (3D) cubic pore network model constrained by nuclear magnetic resonance (NMR) and MICP data is developed to quantify fluid distributions and phase connectivity for predicting saturation-dependent relative permeability during two-phase drainage. There is rich petrophysical information in spatial fluid distributions resulting from vertical fluid flow on a geologic time scale and radial mud-filtrate invasion on a drilling time scale. Log attributes elicited by such fluid distributions are captured to quantify dynamic reservoir petrophysical properties and define reservoir flow capacity. A new rock classification workflow that reconciles reservoir saturation-height behavior and mud-filtrate for more accurate dynamic reservoir modeling is developed and verified in both clastic and carbonate fields. Rock types vary and mix at the sub-foot scale in heterogeneous reservoirs due to depositional control or diagenetic overprints. Conventional well logs are limited in their ability to probe the details of each individual bed or rock type as seen from outcrops or cores. A bottom-up Bayesian rock typing method is developed to efficiently test multiple working hypotheses against well logs to quantify uncertainty of rock types and their associated petrophysical properties in thinly bedded reservoirs. Concomitantly, a top-down reservoir description workflow is implemented to characterize intermixed or hybrid rock classes from flow-unit scale (or seismic scale) down to the pore scale based on a multi-scale orthogonal rock class decomposition approach. Correlations between petrophysical rock types and geological facies in reservoirs originating from deltaic and turbidite depositional systems are investigated in detail. Emphasis is placed on the cause-and-effect relationship between pore geometry and rock geological attributes such as grain size and bed thickness. Well log responses to those geological attributes and associated pore geometries are subjected to numerical log simulations. Sensitivity of various physical logs to petrophysical orthogonality between rock classes is investigated to identify the most diagnostic log attributes for log-based rock typing. Field cases of different reservoir types from various geological settings are used to verify the application of petrophysical rock classification to assist reservoir characterization, including facies interpretation, permeability prediction, saturation-height analysis, dynamic petrophysical modeling, uncertainty quantification, petrophysical upscaling, and production forecasting. / text

In the last couple of years there has been a great need for finding alternative, cleaner burning fuel sources. This search has led to the development of various hydrogen technologies. The reason for this is that when burnt, hydrogen gas only forms water and oxygen as products. One of the methods used in the production of hydrogen gas is that of the electrolysis of sulphur dioxide which is facilitated by a sulphur dioxide depolarized electrolyzer. The electrolysis of sulphur dioxide has the advantage of requiring lower cell voltages in the electrolysis process when compared to the electrolysis of water. This type of electrolyzer unfortunately suffers from low hydrogen gas production volumes. It was thought that by linearly increasing the reactions active area of the electrolyzer, the production volumes can be increased. A linearly upscaled 100cm2 cell was designed by using computer aided design software, such as SolidWorks, Cambridge Engineering Selector, EES and ANSYS. The cell was then constructed and tested to determine the effects of linearly upscaling. The results of the 100cm2 cell were compared to the results of a similar 25cm2 cell and results obtained from the literature. The 100cm2 cell exhibited very poor performance when compared to the other cells. The 100cm2 cell showed lower hydrogen production volumes at higher energy inputs than the 25cm2 cell and an 86cm2 stack assembly. It was concluded that creating stack assemblies with cells with smaller active areas would be much more efficient than linearly upscaling the active area of the cells. / Thesis (M.Ing. (Mechanical Engineering))--North-West University, Potchefstroom Campus, 2012.

In the last couple of years there has been a great need for finding alternative, cleaner burning fuel sources. This search has led to the development of various hydrogen technologies. The reason for this is that when burnt, hydrogen gas only forms water and oxygen as products. One of the methods used in the production of hydrogen gas is that of the electrolysis of sulphur dioxide which is facilitated by a sulphur dioxide depolarized electrolyzer. The electrolysis of sulphur dioxide has the advantage of requiring lower cell voltages in the electrolysis process when compared to the electrolysis of water. This type of electrolyzer unfortunately suffers from low hydrogen gas production volumes. It was thought that by linearly increasing the reactions active area of the electrolyzer, the production volumes can be increased. A linearly upscaled 100cm2 cell was designed by using computer aided design software, such as SolidWorks, Cambridge Engineering Selector, EES and ANSYS. The cell was then constructed and tested to determine the effects of linearly upscaling. The results of the 100cm2 cell were compared to the results of a similar 25cm2 cell and results obtained from the literature. The 100cm2 cell exhibited very poor performance when compared to the other cells. The 100cm2 cell showed lower hydrogen production volumes at higher energy inputs than the 25cm2 cell and an 86cm2 stack assembly. It was concluded that creating stack assemblies with cells with smaller active areas would be much more efficient than linearly upscaling the active area of the cells. / Thesis (M.Ing. (Mechanical Engineering))--North-West University, Potchefstroom Campus, 2012.

Les tourbières ont accumulé des quantités importantes de matière organique (tourbe) depuis le début de l’Holocène. La tourbe séquestre des quantités de carbone (C), d’azote (N) et de phosphore (P) substantielles à l’échelle globale. Cependant, très peu d’études ont estimé les stocks de ces nutriments à l’échelle régionale. De plus, les estimations actuelles sont souvent obtenues sans considérer la variabilité des profondeurs de tourbe et les variables biophysiques et physiques qui dictent la capacité de stockage en nutriments. Ce mémoire vise à estimer les stocks en C, N et P des tourbières d’une région du sud du Québec en modélisant les variations de profondeurs de tourbe et en identifiant quelles variables biophysiques et physiques expliquent la capacité de stockage en nutriments des différents types de tourbières. Le sondage manuel des sites d’études a révélé des différences significatives de profondeurs de tourbe entre les types de tourbières. La profondeur moyenne de tourbe obtenue pour les tourbières ombrotrophes était de 4,4 m alors que celles des tourbières minérotrophes et forestières étaient respectivement de 2,3 et 2,0 m. Les profondeurs ont été modélisées par interpolation spatiale pour chaque bassin d’accumulation de tourbe. La relation entre la superficie des tourbières et les stocks en C était significativement différente entre les tourbières ombrotrophes et les tourbières minérotrophes et forestières. Aucune variable étudiée ne permettait de discriminer les stocks en N et P selon le type de tourbière. Les stocks en C, N et P de la tourbe ont été extrapolés pour la région en utilisant la superficie des tourbières et estimés à 5,65 Tg C, 230,5 Gg N et 21,8 Gg P. Mes travaux ont démontré que l’hétérogénéité du bassin d’accumulation et les caractéristiques de surface influencent les stocks en C, N et P contenus dans la tourbe. / Peatlands have accumulated important amounts of organic matter since the beginning of the Holocene. This organic matter accumulation is described as peat, that globally stores large amounts of carbon (C), nitrogen (N), and phosphorus (P). Regional estimations of C and nutrients stocks have seldom been quantified. Moreover, peat depth variability and biophysical and physical variables are often not included in actual peat C, N, and P pool size estimations. The goal of this thesis is to provide an estimation of peat C, N, and P pool sizes for a regional county of southern Quebec by modeling peat depth variations and by using biophysical and physical variables to discriminated peat C, N, and P storage among peatland types. Manual peat depth probing showed significant differences in peat depth among peatland types. Mean peat depth of bogs was 4.4 m, while fens and forested peatlands mean peat depths were of 2.3 and 2.0 m, respectively. Modeling of peat accumulation basin by spatial interpolation took these peat depth variations into account, among peatland types. The relationship between peatland area and peat C stocks was statistically different between bogs and fens, and bogs and forested peatlands. Peat N and P content among peatland types were not discriminated by the biophysical and physical variables considered in this study. Peat C, N, and P stocks were extrapolated within the study region using peatland area and estimated at 5.65 Tg C, 230.5 Gg N et 21.8 Gg P. This study has shown that peat accumulation basin heterogeneity and peatland geometric shape influenced peat C, N, and P stocks at the regional scale.

Depuis 2006, l’arrêt des pompages d’exhaure dans le Bassin Houiller Lorrain (France) a conduit à l’ennoyage des travaux miniers abandonnés, avec pour conséquence la mise en place d’un nouvel équilibre hydrodynamique régional. De récentes recherches portant sur l’exploitation de la chaleur des réservoirs ennoyés ont suscité de nouvelles interrogations, auxquelles nous nous sommes proposé de répondre. Notre travail avait pour objectif de chercher à comprendre le comportement thermo-hydrodynamique de l’eau de mine au sein d’un système en cours d’ennoyage ou récemment ennoyé. Dans un premier temps, les contextes géographique, géologique et hydrogéologique du Bassin Houiller Lorrain ont fait l’objet d’une synthèse, et une zone d’étude a été choisie. Dans un second temps, des profils de température et de conductivité électrique, complétés par des jaugeages, ont été mesurés dans d’anciens puits de mine du Bassin Houiller Lorrain, offrant une meilleure compréhension du comportement de l’eau à l’échelle d’un ouvrage profond. À partir de l’analyse de ces données, un modèle thermo-hydrodynamique et des simulations numériques ont pu être réalisés à cette échelle. Les résultats permettent d’expliquer les phénomènes observés. Leur stabilité est également étudiée. Dans un troisième temps, un modèle spatialisé maillé a été construit pour aborder la problématique du comportement hydrodynamique d’un réservoir minier entier. La montée de l’eau observée a été correctement reproduite ; le modèle peut par ailleurs être utilisé de façon prédictive pour la période suivant l’ennoyage. Plusieurs outils ont été testés, améliorés ou développés afin de faciliter l’étude des réservoirs ennoyés, notamment concernant l’homogénéisation tridimensionnelle des conductivités hydrauliques et le couplage d’un modèle spatialisé maillé avec un réseau de drains-conduits / Since 2006, cessation of dewatering in Lorraine Coal Basin (France) led to the flooding of abandoned mines, resulting in a new hydrodynamical balance in the area. Recent researches concerning geothermal exploitation of flooded reservoirs raised new questions, which we propose to answer. Our work aimed to understand the thermo-hydrodynamical behaviour of mine water in a flooding or flooded system. Firstly, we synthetized the geographical, geological and hydrogeological contexts of the Lorraine Coal Basin, and we chose a specific area for our studies. Secondly, temperature and electric conductivity log profiles were measured in old pits of the Lorraine Coal Basin, giving a better understanding of the water behaviour at a deep mineshaft scale. We were able to build a thermo-hydrodynamical model and simulate water behaviour at this scale. Flow regime stability is also studied. Thirdly, a hydrodynamical spatialized meshed model was realized to study the hydrodynamical behaviour of a mine reservoir as a whole. Observed water-table rise was correctly reproduced : moreover, the model can be used in a predictive way after the flooding. Several tools were tested, improved or developed to ease the study of flooded reservoirs, as three-dimensional upscaling of hydraulic conductivities and a coupled spatialized meshed model with a pipe network

Les réservoirs de type "tight oil" et "shale gaz" ont une partie importante de leur volume poreux occupée par des micropores (< 2nm) et des mesopores (entre 2 et 50 nm). Ce type d'environnement crée de fortes forces d’interaction dans le fluide confiné avec les parois du pores et entre ses propres molécules, ce qui change fortement la thermodynamique du fluide. Un travail important doit donc être effectué sur le développement de méthodes de mise à l'échelle de la distribution de pore pour effectuer des simulations réservoir à grande échelle. Premièrement, des simulations moléculaires sont effectuées sur des fluides confinés afin d'obtenir des propriétés thermodynamiques de référence à l'équilibre liquide/vapeur pour différentes tailles de pore. Ensuite, une comparaison des données de simulation moléculaire avec les résultats issus des équation d'état utilisées dans la littérature a permis de mettre en valeur la méthode de flash avec pression capillaire et changement du point critique comme la meilleure méthode existante pour décrire la physique du fluide confiné. Des simulations fines d'écoulement matrice/fracture ont donc été effectuées pour différentes tailles de pore. Des modèles de mise à l'échelle en maillage grossier ont été ensuite construits à partir du même cas synthétique et les résultats ont été comparés avec ceux des simulations de référence en maillage fin. Un nouveau modèle de triple porosité considérant fracture, petit pores et grand pores avec une approche MINC a donné des résultats très proches du maillage fin. Finalement un réservoir stimulé hydrauliquement à grande échelle a été simulé pour différentes distributions de pores avec le modèle développé. / Tight oil and shale gas reservoirs have a significant part of their pore volume occupied by micro (below 2nm) and mesopores (between 2 and 50nm). This kind of environment creates strong interaction forces in the confined fluid with pore walls as well as between its own molecules and then changes dramatically the fluid phase behavior. An important work has therefore to be done on developing upscaling methodology of the pore size distribution for large scale reservoir simulations. Firstly, molecular simulations are performed on different confined fluids in order to get reference thermodynamic properties at liquid/vapor equilibrium for different pore sizes. Then, the comparison with commonly used modified equation of state (EOS) in the literature highlighted the model of flash with capillary pressure and critical temperature and pressure shift as the best one to match reference molecular simulation results. Afterwards fine grid matrix/fracture simulations have been built and performed for different pore size distributions. Then, coarse grid upscaling models have then been performed on the same synthetic case and compared to the reference fine grid results. A new triple porosity model considering fracture, small pores and large pores with MINC (Multiple Interacting Continua) approach, has shown very good match with the reference fine grid results. Finally a large scale stimulated reservoir volume with different pore size distribution inside the matrix has been built using the upscaling method developed here.

The complex and non-linear fluvial river dynamics are characterized by repeated periods of fluvial erosion and re-deposition in different parts of the floodplain. Understanding the fluvial architecture (i.e. the three-dimensional arrangement and genetic interconnectedness of different sediment types) is therefore fundamental to obtain well-based information about controlling factors. However, investigating the fluvial architecture in buried floodplain deposits without natural exposures is challenging. We studied the fluvial architecture of the middle Weiße Elster floodplain in Central Germany, an extraordinary long-standing archive of Holocene flooding and landscape changes in sensitive loess-covered Central European landscapes. We applied a novel systematic approach by coupling two-dimensional transects of electrical resistivity tomography (ERT) measurements and closely spaced core drillings with spatially resolved measurements of electromagnetic induction (EMI) of larger floodplain areas at three study sites. This allowed for (i) time and cost-efficient core drillings based on preceding ERT measurements and (ii) spatially scaling up the main elements of the fluvial architecture, such as the distribution of thick silt-clay overbank deposits and paleochannel patterns from the floodplain transects to larger surrounding areas. We found that fine-grained sand and silt-clay overbank deposits overlying basal gravels were deposited during several periods of intensive flooding. Those were separated from each other by periods of reduced flooding, allowing soil formation. However, the overbank deposits were severely laterally eroded before and during each sedimentation period. This was probably linked with pronounced meandering or even braiding of the river. Our preliminary chronological classification suggests that first fine-grained sedimentation must have occurred during the Early to Middle Holocene, and the last phase of lateral erosion and sedimentation during the Little Ice Age. Our study demonstrates the high archive potential of the buried fluvial sediments of the middle Weiße Elster floodplain and provides a promising time and cost-effective approach for future studies of buried floodplain sediments.

<p>Examinator Professor Peter Jansson, Institutionen för naturgeografi</p> / Research school for teachers on climate and environment grant 2017-06037

European forests have experienced vitality loss and dieback due to increasing disturbances, temperatures, droughts, and forest management. Around 160 million cubic meters of woody biomass in Germany were damaged from 2018 to 2020. Protected areas can offer insights into the natural resilience of European forests after disturbances. However, they face conflicts such as mass outbreaks and the spread of insect pests that can threaten biodiversity and natural habitats. The European spruce bark beetle (Ips typographus) is among Europe’s most important and severe forest pest species and can act as an ecological disturbance. Due to ongoing climate change, no weakening is expected in the future. Forest management can directly influence important drivers of disturbance regimes. However, the response times are likely slow, requiring effective and adaptive risk-assessing management. Simulation models can play an important and influential role in such management as decision-making tools or for analysis of important drivers. Such models have been successfully applied in fishery management, disease control, and forest fire management. While there are many models published about Ips typographus, there is still no simulation tool that individually describes trees and beetles with their properties and decisions, tests the effectiveness of management measures, and uses spatially explicit data for natural landscapes. This approach would allow for investigating infestation patterns and development in natural landscapes based on individual beetle and tree traits. It could also be used to study the impact of climate change and forest protection management on all spatial scales - from individual beetles to entire landscapes. The main goal of the thesis is to predict the risk of bark beetle infestations and outbreaks in a national park region in Germany under different climate change and management scenarios as an emergent outcome of individual traits of the European spruce bark beetle and its host tree. To achieve this goal, a combination and improvement of existing models on the life history of Ips typographus and its host tree Picea abies is done to simulate infestations on an individual level. The results are then scaled to the landscape level using a Markov chain metamodel for 20 years into the future. It is shown how (i) management scenarios consisting of different sanitation felling intensities, and (ii) climate change scenarios representing increasing numbers of yearly beetle generations interact and affect the infestation spread. Chapter 2 discusses the creation, testing, and implementation of a new simulation model for the dispersal and infestation of Ips typographus in a natural European landscape. The IPS-SPREADS (Infestation Pattern Simulation Supporting PREdisposition Assessment DetailS) model is an innovative combination of existing models on beetle phenology (PHENIPS), spruce tree predisposition (PAS - predisposition assessment systems), and precise mechanistic dispersal flight in an artificial forest (IPS - Infestation Pattern Simulation). The model is used to study the effectiveness of sanitation felling in five different research areas at the border of a national park in Germany under varying annual beetle generation numbers. Chapter 3 explores the various purposes and types of metamodels used in agentbased modeling and attempts to identify and rank the most suited or efficient model types for each application. This assessment was accomplished through a systematic literature review of 27 scientific publications containing 40 different metamodel applications in an individual-based modeling context. A standardized rating instruction was created and used by the candidate and a group of other Ph.D. students and postdoctoral researchers to objectively rate different metamodel applications, such as upscaling agent-based models or conducting sensitivity analyses. In Chapter 4 of the dissertation, the IPS-SPREADS simulation model is used to predict the risk of bark beetle infestations in a national park region for 20 years into the future. A metamodel approach based on Markov chains is used to scale the computationally demanding IPS-SPREADS model to larger temporal and spatial scales. The metamodel is calibrated and validated using data on infested trees from 2015 to 2017. The model is then used to assess the effectiveness of sanitation felling in preventing the spread of a bark beetle outbreak from the national park to adjacent forests until 2038. This work adds to a series of previous studies investigating several aspects of the Norway spruce (Picea abies) and European spruce bark beetle (Ips typographus) system, such as risk assessment, beetle development, beetle dispersal, and infestation patterns, the effectiveness of forest protection measures against the beetle spread or the impact of bark beetle disturbance on the forest as well as its interaction with other disturbances. The results of the studies presented as thesis chapters 2 to 4 indicate that regardless of the number of yearly beetle generations and the spatial or temporal scale, a sanitation felling intensity of 80 % and above seems to mitigate further mass outbreak propagation. It is also shown that habitat inter-connectivity and individual traits of beetles and host trees substantially affect the infestation patterns. In addition, commonly applied metamodel types and application purposes in an individual-based modeling context are revealed. The most promising model variant varies in regard to the chosen application aim. Further research directions based on the work presented in this thesis incorporate investigating spatial configurations of mixed forest stands and their effect on the dispersal and infestation risk of the European spruce bark beetle. Furthermore, it is discussed how applying and integrating open-access GIS data, such as the European Unions Copernicus program, could improve model validity and applicability. The effects and implications for implementing and analyzing further management measures, such as buffer zones around protected areas, terrestrial detection based on visual clues, and dispensers with anti-attractants in the proposed model framework, are discussed. / Europas Wälder haben aufgrund von zunehmenden Störungen, Temperaturen, Dürren und vergangener Managemententscheidungen an Vitalität verloren und zeigen großflächige Absterbeerscheinungen. Allein in Deutschland betrug das Schadvolumen von 2018 bis 2020 rund 160 Millionen Kubikmeter. Schutzgebiete können Einblicke in die natürliche Widerstandsfähigkeit der Wälder nach Störungen gewähren, bieten aber auch Potenzial für Konflikte wie Massenvermehrungen und Ausbreitung von Schädlingen, die die biologische Vielfalt und die natürlichen Lebensräume bedrohen können. Der Große Achtzähnige Fichtenborkenkäfer (Ips typographus) gehört zu den wichtigsten und schwerwiegendsten Waldschädlingen in Europa. Aufgrund des anhaltenden Klimawandels ist keine Gefährdungsabnahme der Waldbeständen zu erwarten. Während die Forstwirtschaft wichtige Treiber von Störungsregimen direkt oder indirekt beeinflussen kann, sind die Reaktionszeiten sehr langsam und verlangen ein wirksames und adaptives Risikomanagement. Simulationsmodelle können eine wichtige und effektive Rolle für das Management als Entscheidungshilfen oder zur Analyse wichtiger Treiber spielen. Beispiele für erfolgreiche Anwendungen im Entscheidungsprozess stellen Bereiche wie das Fischereimanagement, die Bekämpfung von Epidemien und das Waldbrandmanagement dar. Obwohl es bereits einige Simulationsmodelle für Ips typographus gibt, wurde bisher kein Ansatz untersucht, bei dem sowohl die Bäume als auch die Käfer als Individuen mit Eigenschaften, Wahrnehmung und Interaktionen abgebildet werden. Mit solch einem Modell könnte untersucht werden, wie sich die individuellen Eigenschaften der Käfer und der Bäume auf die Befallsmuster und die Wirksamkeit von Managementmaßnahmen in Abhängigkeit der betrachteten räumlichen oder zeitlichen Skala auswirken. Das Hauptziel der vorliegenden Dissertation besteht daher darin, das Risiko von Borkenkäferbefall und von Ausbrüchen unter verschiedenen Klimawandel- und Managementszenarien als emergentes Ergebnis individueller Eigenschaften des Großen Achtzähnigen Fichtenborkenkäfers und seines Wirtsbaums zu analysieren. Dies geschieht durch die Kombination, Verbesserung und Anwendung bestehender Modelle zur Lebensgeschichte von Ips typographus und dessen Wirtsbaum Picea abies am Beispiel einer Nationalparkregion in Deutschland. Die Ergebnisse des Individuenbasierten Modells werden anschließend auf die Landschaftsebene und für 20 Jahre in die Zukunft skaliert, indem ein auf Markov-Ketten basierendes Metamodel kreiert und validiert wird. Es wird gezeigt, wie (i) Managementszenarien bestehend aus verschiedenen Entnahmeintensitäten befallener Bäume sowie (ii) Klimawandelszenarien bestehend aus einer Zunahme der jährlichen Käfergenerationen interagieren und die Befallsausbreitung beeinflussen. In Kapitel 2 der Dissertation wird die Entwicklung, Überprüfung und Implementierung eines neuen Simulationsmodells für die Ausbreitung und den Befall von Ips typographus in einer realen Landschaft Europas diskutiert. Das IPS-SPREADS (Infestation Pattern Simulation Supporting REdisposition Assessment DetailS) Modell ist eine innovative Kombination bestehender Modelle zur Phänologie des Käfers (PHENIPS), der Prädisposition des Wirtsbaumes gegenüber Borkenkäferbefall (PAS - Prädispositionsabschätzsysteme) und der Individuen-basierten Simulation des Ausbreitungsflugs der Käfer in einem künstlichen Wald (IPS - Infestation Pattern Simulation). Das neue Modell IPS-SPREADS wird dann verwendet, um die Wirksamkeit der Entnahme befallener Bäume in fünf verschiedenen Probeflächen des Nationalparks Sächsische Schweiz während zunehmender jährlicher Käfergenerationen zu untersuchen. Darüber hinaus wird der Einfluss individueller Eigenschaften auf das Befallsgeschehen und auf die Wirksamkeit der Entnahme befallener Bäume analysiert. Kapitel 3 widmet sich der Anwendungsgebiete und Typen von Metamodellen, die für Agenten-basierten Modelle verwendet werden, und versucht, die vielversprechendsten Modelltypen je Anwendung zu identifizieren. Dies geschieht mittels einer systematischen Literaturanalyse, die 27 wissenschaftliche Veröffentlichungen mit insgesamt 40 verschiedenen Metamodellanwendungen für Individuen-basierte Modelle untersucht. Ein standardisiertes Bewertungsverfahen wird erstellt und vom Kandidaten zusammen mit einer Gruppe von Doktoranden und Postdoktoranden angewendet, um die Metamodelle im Bezug auf ihr Anwendungsziel objektiv zu bewerten. In Kapitel 4 der Dissertation wird das entwickelte IPS-SPREADS Modell verwendet, um das Risiko von Borkenkäferbefall in einer Nationalparkregion für 20 Jahre in die Zukunft vorherzusagen. Dafür wird ein Metamodell basierend auf Markov-Ketten entwickelt, welches das rechenintensive IPS-SPREADS Modell auf größere zeitliche und räumliche Skalen hebt. Dieses Metamodell wird mittels Befallsdaten aus der Nationalparkregion für die Jahre 2015 bis 2017 validiert. Im Anschluss wird das Metamodell dann verwendet, um die Wirksamkeit der Entnahme befallener Bäume für die Ausbreitungsbekämpfung eines Borkenkäfermassenbefalls vom Nationalpark in die angrenzenden Wälder während eines 20-jährigen Zeitraums zu untersuchen. Die vorliegende Arbeit baut auf einer Reihe von Studien auf, die verschiedene Aspekte des Systems von Großem Achtzähnigen Fichtenborkenkäfer (Ips typographus) und Gemeiner Fichte (Picea abies) untersuchen: Prädispositionsbewertung, Käferentwicklung, Ausbreitungs- und Befallsmuster sowie die Wirksamkeit von Waldschutzmaßnahmen gegen die Befallsausbreitung. Die Ergebnisse der als Kapitel zwei bis vier vorgestellten Studien zeigen, dass unabhängig von der Anzahl der jährlichen Käfergenerationen und der räumlichen oder zeitlichen Skala eine Entnahmeintensität befallener Bäume von 80 % und mehr die Ausbreitung eines Massenbefalls zu mindern scheint. Es wird auch gezeigt, dass die Habitatvernetzung und die individuellen Eigenschaften der Käfer und Wirtsbäume die Befallsmuster beeinflussen. Darüber hinaus werden häufig angewandte Metamodelltypen und Anwendungszwecke in einem Individuenbasierten Modellierungskontext zusammengetragen und die vielversprechendste Modellvariante je Anwendungsgebiet extrahiert. Zum Schluss werden weitere Forschungsrichtungen basierend auf den vorgestellten Studien diskutiert, wie zum Beispiel die Untersuchung der räumlichen Konfiguration von Mischwäldern und deren Auswirkungen auf das Befallsgeschehen des Großen Achtzähnigen Fichtenborkenkäfers. Des Weiteren wird diskutiert, wie die Anwendung und Integration von frei verfügbaren GIS-Daten, wie beispielsweise des Copernicus Programms der Europäischen Union, die Validität und Anwendbarkeit der entwickelten Modelle verbessern können. Am Ende werden potenzielle Implikationen für die Implementierung und Analyse weiterer Managementmaßnahmen wie Pufferzonen um Schutzgebiete, terrestrische Erkennung von Borkenkäferbefall sowie Dispenser mit Anti-Aggregationspheromonen im Rahmen der vorgestellten Modellstudien diskutiert.

Le béton se présente comme un matériau constitué de granulats, jouant le rôle d'inclusions, enchâssés dans une matrice correspondant à la pâte de ciment hydraté. A l'échelle de la pâte, l'hydratation du ciment génère un milieu multiphasique, constitué d'un squelette solide et de pores remplis ou partiellement remplis d'eau selon leur taille. Le composant principal du squelette solide est le gel de C-S-H, les autres composants étant de nature cristalline. La qualification de gel du composant C-S-H est liée à sa nanostructure dont la schématisation la plus admise consiste en une phase aqueuse adsorbée, en sandwich avec des feuillets solides de nature cristalline. Il est bien admis que la structure du C-S-H est à l'origine de son comportement viscoélastique et donc de celui du béton. Ce comportement viscoélastique peut s'expliquer par un réarrangement de sa nanostructure sous l'effet des contraintes mécaniques appliquées à l'échelle macroscopique. La modélisation macroscopique du fluage du béton ne permet pas d'expliquer la variabilité du fluage d'une formulation de béton à une autre. En effet, les paramètres des modèles macroscopiques ne peuvent être identifiés que par l'analyse de résultats expérimentaux obtenus par des essais réalisés sur des éprouvettes de béton. Ces paramètres ne sont valables que pour une formulation donnée. L'identification de ces paramètres conduit donc à des programmes expérimentaux très coûteux et ne fournit pas suffisamment d'informations sur la sensibilité des paramètres macroscopiques à la variabilité des caractéristiques mécaniques et morphologies des constituants. Dans ce travail, on suppose qu'il existe une échelle microscopique à laquelle les mécanismes moteurs du fluage ne sont pas impactés par la formulation du béton. A cette échelle, celle du C-S-H, les propriétés viscoélastiques peuvent être considérées avoir un caractère intrinsèque. L'influence de la formulation ne concerne alors que les concentrations des différents hydrates. Trois approches, analytiques, semi-analytiques et numériques sont alors proposées pour estimer, par une homogénéisation multi-échelle, les propriétés viscoélastiques macroscopiques du béton à partir des propriétés de ses constituants ainsi qu'à partir de sa microstructure. Ces approches sont basées sur l'extension des schémas d'homogénéisation élastique au cas viscoélastique au moyen du principe de correspondance qui utilise la transformée de Laplace-Carson. Les propriétés effectives sont alors déterminées directement dans l'espace de Carson. Par la suite, celles dans l'espace temporel sont obtenues par la transformée inverse. Les approches proposées apportent des solutions aussi bien dans un cadre général que sous certaines hypothèses restrictives : coefficient de Poisson viscoélastique microscopique ou macroscopique constant, module de compressibilité constant. Sur le plan théorique, deux schémas d'homogénéisation ont été étudiés : le schéma de Mori-Tanaka, basé sur le problème de l'inclusion d'Eshelby, et le schéma auto-cohérente généralisé basé sur la neutralité énergique de l'inclusion. Les résultats obtenus montrent que sous ces hypothèses restrictives, le spectre macroscopique se présente comme une famille de sous ensembles de temps caractéristiques bornés par les temps caractéristiques microscopiques. Par ailleurs, les propriétés thermodynamiques, de croissance monotone et de concavité, des fonctions de retard macroscopiques ne sont préservées par l'homogénéisation que sous certaines conditions de compatibilité des spectres microscopiques. Sur le plan pratique, les méthodes développées ont été appliquées pour construire la complaisance de fluage propre macroscopique du béton en connaissant les données communes de toutes sortes de bétons et celles correspondant à une formulation donnée. Les résultats expérimentaux disponibles sont alors exploités pour analyser le caractère intrinsèque des propriétés viscoélastiques à l'origine du fluage du béton / Concrete can be considered as a multiscale heterogeneous material in which elastic inclusions, i.e., aggregates, are embedded in a viscoelastic matrix which corresponds to the hardened cement paste. At the cement paste scale, the hydration process leads to a partially saturated porous multiphase medium whose solid skeleton is mainly constituted of the C-S-H gel, the other hydrates being crystalline solids. The gel-like properties of the C-S-H are essentially related to the existence of strongly adsorbed structural water in the interlayer nanospace. It is well admitted that this feature is the microscopic origin of the creep mechanism of C-S-H and therefore of concrete at the macroscale. This viscoelastic behavior can be explained by a rearrangement of the nanostructure of the C-S-H component due to a macroscopically applied loading. Therefore, the macroscopic modeling of concrete creep does not allow explaining the variability of creep with respect to the concrete mix. Indeed, the identi?cation of the viscoelastic parameters of macroscopic models requires performing creep tests at the concrete scale. Thus, the validity of the identi?ed parameters is limited to the concrete mix under consideration and the identi?cation process does not permit to carry out the influence of the microstructure and the mechanical constituent properties on the macroscopic creep. In this work, we assume that there exists a microscopic scale at which the driving creep mechanisms are not affected by the concrete mix. At this scale, the C-S-H one, the viscoelastic properties can be considered to be intrinsic. The concrete mix will essentially influence the volume fraction of the hydrates. Analytical, semi-analytical and numerical approaches are developed in order to derive, by multi-scale homogenization, the macroscopic viscoelastic properties of concrete on the basis of the knowledge of the microscopic properties, of the constituents together with the knowledge of the mix-dependent microstructure. These approaches consist in extending results of elastic homogenization schemes to viscoelasticity with the help of the correspondence principle using the Laplace-Carson transform. Therefore, the effective properties are obtained in a straightforward manner in the Carson transform space, which requires the transform inversion in order to determine the time evolution of these properties. The proposed models permit to answer to this requirement, both under some restrictive assumptions (constant Poisson's ratio at either microscopic or macroscopic scale, constant bulk microscopic modulus) and in a general way. At the theoretical level, two homogenization schemes are investigated : the Mori-Tanaka scheme based on the Eshelby solution and the Generalized Self Consistent scheme, based on the energetic neutrality condition of the inclusion. The obtained results show that under the aforementioned restrictive conditions, the macroscopic creep spectrum presents as a family of sets of retardation times which are bounded by two successive microscopic characteristic times. Furthermore, the monotonically increasing and concavity features of the macroscopic creep compliance, which derive from thermodynamic considerations, are not always preserved, except if some compatibility conditions are ful?lled by the microscopic spectrums. From the practical point of view, these methods developed are used to establish the complaisance creep function of concrete from the knowledge of the common data of all types of concrete, as well as the other parameters that characterize a given formulation. Hence, by a back-analysis of creep tests obtained on cement paste or on concrete, the intrinsic basic creep properties of the gel C-S-H are estimated