A study of the Nova Scotia surface air temperature over the last century (1900 to 2015) shows that internal variability on inter-annual, decadal and multi-decadal time scales can be partly explained by ocean-atmospheric climate modes, external and anthropogenic forcings. The Atlantic Multidecadal Oscillation (AMO) and Arctic Oscillation (AO) are shown to be the dominant climate drivers in Nova Scotia. The El Niño Southern Oscillation (ENSO) is also shown to be a dominant climate driver but only during the summer. Multivariate models were generated over the full time period using only natural ocean-atmospheric modes of variability but could not explain the rapid increase in the recent rate of warming (post-1980). The inclusion of anthropogenic greenhouse gas forcing to the models improved their predictive power annually and seasonally. The modelling results show that 11% of the annual variability in Nova Scotia results from natural forcings along with anthropogenic greenhouse gas forcing while seasonally up to 28% of the temperature variability can be explained by natural plus greenhouse gas forcings. The annual and seasonal low explained variance suggests that Nova Scotia is poorly modulated by climate indices, specifically during the winter, the time when relationships between ocean-atmospheric modes and the regional climate should be the strongest. It leads to believe that Nova Scotia is located in a transition zone where large-scale ocean-atmospheric modes of variability are transitioning from being positively correlated in a region to being negatively correlated in another region. The results of this study help to better understand how large-scale ocean-atmospheric modes of variability, external and anthropogenic greenhouse gas forcings affect Nova Scotia’s surface air temperatures and also provide insight into future potential variability under a changing climate.

Dans cette thèse, nous analysons les principales caractéristiques des aérosols atmosphériques sur un site urbain en Afrique de l’Ouest : la ville de Ouagadougou. Cette analyse est suivie de l’évaluation du forçage radiatif produit par cette population d’aérosols au sommet de l’atmosphère, dans la couche atmosphérique ainsi qu’à la surface terrestre. Une étude climatologique des propriétés optiques (épaisseurs optiques, exposant d’Angström, albédo de simple diffusion, facteur d’asymétrie) et microphysiques (distribution granulométrique, indice complexe de réfraction) a été effectuée sur la base des données de mesures et d’inversions photométriques du réseau AERONET. L’analyse de ces données a permis de définir à diverses échelles de temps les différentes variabilités des propriétés étudiées. Ces propriétés ont mis en exergue les effets combinés de l’activité anthropique, du cycle de production des poussières minérales d’origine saharienne, de la succession saisonnière et la dynamique du climat spécifiques à cette région de la terre. La composition du panache d’aérosols a également été déduite de l’analyse des données optiques, et les conclusions tirées se sont avérées être en bon accord avec des mesures chimiques effectuées au LSCE sur des échantillons obtenus par prélèvement manuel sur filtres. Le bilan radiatif a été évalué en utilisant le modèle GAME. Cette simulation a montré une forte corrélation du forçage avec la succession des saisons sèche et humide, avec des valeurs extrêmes au mois d’Août (humide) et de Mars (sec avec poussières maximales). Les résultats traduisent un refroidissement en surface pouvant atteindre -41 W/m 2 en Mars, un réchauffement de la couche atmosphérique qui va de 15 à 35 W/m 2 environ et un refroidissement au sommet de l’atmosphère compris entre -2 et -6 W/m 2 . La représentation du forçage calculé en fonction de l’albédo de simple diffusion a montré une tendance linéaire avec un coefficient de corrélation traduisant une fiabilité de nos résultats de l’ordre de 88%. La cohérence des valeurs simulées a été aussi montrée par une comparaison avec des résultats obtenus dans la région Ouest Africaine par d’autres chercheurs utilisant des techniques différentes. De même, la représentation de l’efficacité radiative simulée en fonction de celle inversée par AERONET a montré un très bon accord. / In this thesis, we analyze the main characteristics of atmospheric aerosols on an urban site in West Africa: Ouagadougou. This analysis is followed by an assessment of the radiative forcing produced by this population of aerosols at the top of the atmosphere, in the atmospheric layer and at the Earth's surface. A climatological study of the optical properties (optical thickness, Angstrom exponent, single scattering albedo, asymmetry factor) and microphysical properties (particle size distribution, complex refractive index) was performed on basis of data obtained from photometric measurement and inversions of AERONET network. The analysis of these data allowed us to define the different variabilities of the properties studied at various time scales. These properties have highlighted the combined effects of human activity, the cycle production of mineral dust from the Sahara region, the seasonal succession and the climate dynamics known in this region of the earth. The composition of the aerosol plume was also deduced from the analysis of optical data, and the conclusions were found to be in good agreement with chemical analysis carried out at LSCE on samples obtained by sampling on filters at Ouagadougou. The radiation balance was assessed using the GAME model. This simulation showed a strong relationship of the radiative forcing with the succesion of wet and dry seasons, with extreme values in August (wet) and Mars (dry with maximum dust emissions). The results showed a cooling at the surface that can reach -41 W / m 2 in March, a warming of the atmosphere, ranging from 15 to 35 W / m 2 about and a cooling at the top of the atmosphere between -2 and -6 W / m 2 . The representation of the radiative forcing calculated as a function of the single scattering albedo showed a linear trend with a correlation coefficient reflecting relatively good reliability of our results (about 88%). The consistency of simulated values was also shown by a comparison with the results obtained in the West African region by other researchers using different techniques. Similarly, the representation of the radiative efficiency simulated as a function of the inverted AERONET one showed a very good agreement.

Ce travail de thèse focalise sur le fonctionnement du réseau microbien de la lagune de Thau à travers une approche « Réseau » ayant permis l’étude simultanée de l’ensemble des groupes microbiens (21 groupes) des virus au microzooplancton. La croissance et la mortalité des micro-organismes ont été étudiées expérimentalement lors de plusieurs saisons puis d’une étude en mésocosme. Nous avons déterminé les proportions de mortalité dont sont responsables le microzooplancton et les virus au travers de la lyse virale. Un suivi in situ puis en mésocosme dans la lagune de Thau nous a permis d'observer la réponse des composantes du réseau microbien à un forçage typiquement Méditerranéen, la crue.Il apparait que les bactéries hétérotrophes ont les plus fortes croissances allant jusqu’à 2,18 jour-1. La croissance pour divers groupes pico- et nano-phytoplanctoniques était inférieure à 1,5 jour-1. La cause majeure de mortalité des différents groupes de micro-organismes étudiés est la prédation du microzooplancton (> 90% en moyenne), tandis que la lyse virale semble être un processus sporadique et nettement moins important. Le fait que les taux de prédation observés soient très proches des taux de croissance de différents micro-organismes suggère que le réseau microbien étudié présente une forte efficacité de transfert de la biomasse vers les niveaux trophiques supérieurs.Dans ce contexte, la crue provoque à court terme une diminution de la croissance microbienne et du broutage, alors qu’elle déclenche la lyse virale chez certains groupes. Cependant, la réactivité de ce réseau permet un rétablissement rapide indiquant une forte capacité de résilience de ce système. / The aims of this thesis work were to shed new lights on the functioning of the microbial food web (MFW) in the Thau coastal lagoon through a “global approach” that permit us to study simultaneously the entire MFW components (21 groups) from viruses to microzooplankton. Growth and mortality of microorganisms were studied experimentally along seasons and during a mesocosm experiment. We particularly focus on mortality due to microzooplankton (<200µm) grazing and viral lysis. Responses of the different components of the MFW under a typical Mediterranean forcing, flash flood, was also studied during an in situ monitoring and a mesocosm experiment in the Thau Lagoon.In conclusion, it appears that heterotrophic bacteria show the highest growth rates (until 2.18 day-1). Growth rates of pico- and nanophytolankton groups were always lower than 1.5 day-1. The major cause of mortality of the different group studied, reaching on average 90%, was due to the predation by the microzooplankton. Mortality due to viral lysis appeared to be a “sporadic processes”, less important than microzooplankton grazing indeed. In addition, the observed grazing rates or bacterivory were close to growth rates of microorganisms, suggesting an efficient trophic transfer of the microbial biomass towards higher trophic levels.River flash flood; an important forcing in the Mediterranean coastal zone; triggered on a short time scale, a decrease in growth and grazing rates of microorganisms whereas viral lysis was enhanced. However, the important dynamics of this studied MFW allowed the rapid reestablishment of growth and grazing rates, indicating a strong resilience capability of this system.

Les besoins énergétiques de la population mondiale ne cessent d’augmenter. Les prévisions indiquent par exemple une forte croissance de la demande du secteur du transport aéronautique. La recherche de systèmes toujours plus performants et moins polluants est nécessaire. Des nouveaux concepts pour la combustion ont été mis au point et appliqués aux turbines à gaz. Parmi eux il existe ceux basés sur la combustion en prémélange pauvre ou en prémélange pauvre pré-vaporisé dans le cas où le carburant utilisé est liquide. Les nouveaux systèmes énergétiques basés sur la combustion en régime pauvre sont prometteurs pour satisfaire les futures normes d’émissions polluantes, mais ils sont plus sensibles aux instabilités de combustion qui limitent leur plage de fonctionnement et peuvent détériorer irréversiblement ces systèmes. Dans ce domaine il reste des questions à aborder. En particulier celle du comportement des flammes tourbillonnaires en combustion diphasique soumises à des perturbations acoustiques. La plupart des moteurs aéronautiques utilisent des flammes de ce type, cependant leur dynamique et leurs interactions mutuelles, quand elles subissent les effets d’une perturbation acoustique, sont loin d’être bien comprises. Ce travail aborde ces questions et apporte des éléments de compréhension sur les mécanismes pilotant la réponse de l’écoulement diphasique et de la flamme, ainsi que des éléments de validation des modèles de prédiction des points de fonctionnement instables. TACC-Spray est le banc expérimental utilisé pour ce travail. Il a été conçu et développé au sein du laboratoire CORIA lors de ce doctorat qui s’inscrit dans le cadre du projet ANR FASMIC. Le système d’injection qui équipe ce banc expérimental reçoit trois injecteurs tourbillonnaires alimentés en combustible liquide (ici n-heptane), développés par le laboratoire EM2C. Ils sont montés en lignes dans le banc, celui-ci représentant ainsi un secteur d’une chambre annulaire. Le montage étant complexe et nouveau, un travail de développement de solutions techniques a été fait pour rendre possible l’équipement du TACC-Spray avec des capteurs de pression, température, photomultiplicateur ainsi que des diagnostiques optiques performants (e.g. LDA, PDA, imagerie à haute cadence). Pour cette étude, le système énergétique, composé par l’écoulement diphasique et la flamme, a été soumis à l’impact d’un mode acoustique transverse excité dans la cavité acoustique. La réponse du système a été étudiée en fonction de son positionnement dans le champ acoustique. Trois bassins d’influence du champ acoustique sur le système énergétique ont été choisis, à savoir: (i) le ventre de pression acoustique caractérisé principalement par des fortes fluctuations de pression, (ii) le ventre d’intensité acoustique présentant de forts gradients de pression et vitesse acoustique, (iii) le ventre de vitesse acoustique avec de fortes fluctuations de vitesse où la fluctuation de pression est résiduelle. L’approche de cette étude a consisté à étudier en premier lieu le système de référence en absence de forçage acoustique, les résultats sont recueillis dans la Partie I de ce manuscrit. En deuxième lieu le système énergétique est placé à chacune des positions d’intérêt dans le champ acoustique et la réponse de l’écoulement d’air sans combustion, la réponse de l’écoulement diphasique avec combustion et finalement celle des flammes, sont étudiées systématiquement. Les résultats de l’étude avec forçage acoustique sont rassemblés dans la Partie II du manuscrit. / The energy needs of population around the word are continuously increasing. For instance, forecasts indicates an important grow of the request of the aeronautic transportation sector. It is necessary to continue the research efforts to get more performants and less contaminating systems. New concepts for combustion have been developed and introduced to the gas turbine industry. Among these concepts it is found technologies based on lean-premixed combustion or lean-premixed prevaporized combustion when liquid fuels are employed. These novel energetic systems, making use of lean combustion, are promising to meet the future norms about pollutant emissions, but this make them more sensitive to combustion instabilities that limit their operating range and can lead to irreversible damage. In this domain, many questions still need to be considered. In particular that of the behavior of two-phase flow swirling flames subjected to acoustic perturbations. Indeed most of aero-engines operate with this type of flames, but the dynamics and mutual interaction of these flames, as they are submitted to acoustic perturbation, are not yet well understood. This work addresses these issues and gives some understanding elements for the mechanisms driving the response of the flow and of the flame to acoustic perturbations and delivers data to validate models predicting unstable operating points.The experimental bench employed for this work is TACC-Spray. It has been designed and developed in the CORIA laboratory during this PhD thesis which is inscribed in the framework of the ANR FASMIC project. The injections system that equips this bench is composed by three swirled injectors fed with a liquid fuel (here n-heptane), developed by the EM2C laboratory. They are linearly arranged in the bench such that this represents an unwrapped sector of an annular chamber. The setup, being new and complex, needed technical solutions developed during this work and applied then in order to equip TACC-Spray with pressure and temperature sensors, a photomultiplier as well as adequate optic diagnostics (LDA, PDA, high speed imaging systems). In this study, the energetic system, composed by the two-phase swirling flow and the spray flame, has been submitted to the impact of a transverse acoustic mode excited within the acoustic cavity. The system response has been studied as a function of its location in the acoustic field. Three basins of influence of the acoustic field on the energetic system have been chosen, namely: (i) the pressure antinode characterized mainly by strong pressure fluctuations, (ii) the intensity antinode where important acoustic pressure and velocity gradients are present, (iii) the velocity antinode with strong velocity fluctuations where the acoustic pressure is residual. The approach of the study presented here is to investigate in first place the energetic system free of acoustic forcing. The results concerning this first study are presented in the Part I of this manuscript. In second place, the energetic system is placed in each of the location of interest within the acoustic field and the response of the air flow without combustion, that of the two-phase flow with combustion and finally that of the spray flames, are systematically investigated. The results of the study under acoustic forcing are shown in Part II of the manuscript.

Persistent condensation trails are clouds, induced by the exhaust of an aircraft engine in a cold and ice-supersaturated environment. These artificial ice clouds can both cool and heat the atmosphere by scattering solar radiation and absorbing terrestrial radiation, respectively. The influence of condensation trails on the Earth-atmosphere energy balance and therewith the answer to the question of the dominating process had been mostly approximated on a global scale by treating the condensation trail as plane parallel layer with constant optical properties. Individual condensation trails and the influence of the solar angle had been analyzed, always using a course spatial grid and never under consideration of the aircraft performance, generating the condensation trail. For a trajectory optimization, highly precise results of the impact of condensation trails on the radiation budget and the influence of the aircraft performance on this impact is needed, so that future air traffic may consider the main factors of flight performance on the environmental impact of condensation trails. That’s why, a model is developed in this thesis to continuously estimate the scattering and absorption properties and their dependence on the aircraft performance.:1 Introduction 3 1.1 Motivation 3 1.2 State of the art 5 1.3 Approach 6 2 Theoretical background 9 2.1 The Earth’s atmosphere 9 2.1.1 The mean vertical structure of the atmosphere 12 2.1.2 Standard atmospheres 14 2.2 Radiation 15 2.2.1 Nature of radiation 15 2.2.2 Important metrics describing radiation 17 2.2.3 Relevant spectra and principles of radiation 19 2.2.4 Solar radiation 20 2.2.5 Terrestrial radiation 21 2.2.6 Radiative transfer and extinction 22 2.2.7 Radiative transfer equation 30 2.2.8 Energy budget of the Earth-atmosphere system 32 2.3 Thermodynamics 33 2.3.1 Atmospheric stability 33 2.3.2 Turbulence 36 2.3.3 Conditions of contrail formation 41 3 Development of a radiative forcing model 45 3.1 Model atmosphere 45 3.2 Flight performance model 46 3.3 Atmospheric radiative transfer model 49 3.3.1 Two Stream Approximation 51 3.3.2 Discrete ordinate radiative transfer solver 52 3.3.3 Methods to calculate broadband radiances and irradiances 53 3.4 Contrail life cycle model 57 3.4.1 Dissipation regime 58 3.4.2 Diffusion regime 63 3.5 Contrail radiative forcing model 74 3.5.1 Consideration of multiple scattering using a Monte Carlo simulation 74 3.5.2 Geometry of the Monte Carlo simulation 75 3.5.3 Interpretation of Beer’s law 76 3.5.4 Procedure of the Monte Carlo simulation 79 3.5.5 The extinguished power per unit length contrail 87 3.5.6 Scattering and absorption efficiencies Qs, Qa and asymmetry parameters gHG 89 3.5.7 Calibration of the Monte Carlo simulation 94 4 Calculations 99 4.1 Contrail properties 99 4.1.1 Conditions of contrail formation 100 4.1.2 Initial dimensions at the end of the dissipation regime 101 4.1.3 Microphysical properties during the diffusion regime 103 4.2 Radiative transport up to the contrail 105 4.2.1 Solar direct and diffuse radiance 106 4.2.2 Terrestrial irradiance 107 4.3 Scattering and absorption properties of radiation within the contrail 109 4.3.1 Monte Carlo simulation for solar radiation 109 4.3.2 Monte Carlo simulation for terrestrial irradiances 112 4.3.3 Relevance of multiple scattering 116 4.4 Radiative extinction 116 4.4.1 Solar zenith and azimuthal angle 118 4.4.2 Flightpath 120 4.4.3 Contrail evolution 122 4.4.4 Turbulence 126 4.4.5 Wavelength specific extinction 129 4.5 Terrestrial energy forcing of a contrail 133 4.6 Verification 135 5 Conclusion and outlook 141 5.1 Conclusion 141 5.2 Outlook 144 List of Figures 147 List of Tables 151 Abbreviations and Symbols 153 Glossary 161 Bibliography 169 Acknowledgements 183 / Langlebige Kondensstreifen sind Eiswolken, welche durch Kondensation von Wasserdampf an Rußpartikeln in einer eisübersättigten Atmosphäre entstehen. Der Wasserdampf entstammt einerseits aus dem Triebwerkabgas und andererseits aus der Atmosphäre. Kondensstreifen können die Atmosphäre durch Rückstreuung solarer Strahlung kühlen und durch Rückstreuung und Absorption terrestrischer Strahlung erwärmen. Der Einfluss von Kondensstreifen auf den Wärmehaushalt der Atmosphäre und damit die Antwort auf die Frage nach dem dominierenden Effekt wurde bisher zumeist auf globaler Ebene ermittelt, wobei der Kondensstreifen als planparallele Schicht mit konstanten optischen Eigenschaften angenähert wurde. Individuelle Kondensstreifen und der Einfluss des Sonnenstandes wurden bisher nur mithilfe eines groben Rasters betrachtet und niemals unter Berücksichtigung der Flugleistung des Luftfahrzeuges, welches den Kondensstreifen generiert hat. Für eine Trajektorienoptimierung sind jedoch präzise Berechnungen des Strahlungseinflusses und eine gewissenhafte Berücksichtigung der Flugleistung notwendig. Nur so kann der zukünftige Luftverkehr die Haupteinflussfaktoren der Flugeigenschaften auf den Strahlungseinfluss der Kondensstreifen berücksichtigen. Aus diesem Grund wurde in dieser Arbeit ein Modell entwickelt, welches die Eigenschaften des Strahlungstransfers durch den Kondensstreifen kontinuierlich bestimmt und die aus der Flugleistung resultierenden Parameter berücksichtigt.:1 Introduction 3 1.1 Motivation 3 1.2 State of the art 5 1.3 Approach 6 2 Theoretical background 9 2.1 The Earth’s atmosphere 9 2.1.1 The mean vertical structure of the atmosphere 12 2.1.2 Standard atmospheres 14 2.2 Radiation 15 2.2.1 Nature of radiation 15 2.2.2 Important metrics describing radiation 17 2.2.3 Relevant spectra and principles of radiation 19 2.2.4 Solar radiation 20 2.2.5 Terrestrial radiation 21 2.2.6 Radiative transfer and extinction 22 2.2.7 Radiative transfer equation 30 2.2.8 Energy budget of the Earth-atmosphere system 32 2.3 Thermodynamics 33 2.3.1 Atmospheric stability 33 2.3.2 Turbulence 36 2.3.3 Conditions of contrail formation 41 3 Development of a radiative forcing model 45 3.1 Model atmosphere 45 3.2 Flight performance model 46 3.3 Atmospheric radiative transfer model 49 3.3.1 Two Stream Approximation 51 3.3.2 Discrete ordinate radiative transfer solver 52 3.3.3 Methods to calculate broadband radiances and irradiances 53 3.4 Contrail life cycle model 57 3.4.1 Dissipation regime 58 3.4.2 Diffusion regime 63 3.5 Contrail radiative forcing model 74 3.5.1 Consideration of multiple scattering using a Monte Carlo simulation 74 3.5.2 Geometry of the Monte Carlo simulation 75 3.5.3 Interpretation of Beer’s law 76 3.5.4 Procedure of the Monte Carlo simulation 79 3.5.5 The extinguished power per unit length contrail 87 3.5.6 Scattering and absorption efficiencies Qs, Qa and asymmetry parameters gHG 89 3.5.7 Calibration of the Monte Carlo simulation 94 4 Calculations 99 4.1 Contrail properties 99 4.1.1 Conditions of contrail formation 100 4.1.2 Initial dimensions at the end of the dissipation regime 101 4.1.3 Microphysical properties during the diffusion regime 103 4.2 Radiative transport up to the contrail 105 4.2.1 Solar direct and diffuse radiance 106 4.2.2 Terrestrial irradiance 107 4.3 Scattering and absorption properties of radiation within the contrail 109 4.3.1 Monte Carlo simulation for solar radiation 109 4.3.2 Monte Carlo simulation for terrestrial irradiances 112 4.3.3 Relevance of multiple scattering 116 4.4 Radiative extinction 116 4.4.1 Solar zenith and azimuthal angle 118 4.4.2 Flightpath 120 4.4.3 Contrail evolution 122 4.4.4 Turbulence 126 4.4.5 Wavelength specific extinction 129 4.5 Terrestrial energy forcing of a contrail 133 4.6 Verification 135 5 Conclusion and outlook 141 5.1 Conclusion 141 5.2 Outlook 144 List of Figures 147 List of Tables 151 Abbreviations and Symbols 153 Glossary 161 Bibliography 169 Acknowledgements 183

No description available.

The dissertation thesis consists of two thematic parts. The first part, i.e. chapters 2, 3 and 4, contains results concerning the topic of a new book of the supervisor and coauthors V. Kanovei and M. Sabok "Canonical Ramsey Theory on Polish Spaces". In Chapter 2, there is proved a canonization of all equivalence relations Borel reducible to equivalences definable by analytic P-ideals for the Silver ideal. Moreover, it investigates and classifies sube- quivalences of the equivalence relation E0. In Chapter 3, there is proved a canonization of all equivalence relations Borel reducible to equivalences de- finable by Fσ P-ideals for the Laver ideal and in Chapter 4, we prove the canonization for all analytic equivalence relations for the ideal derived from the Carlson-Simpson (Dual Ramsey) theorem. The second part, consisting of Chapter 5, deals with the existence of universal and ultrahomogeneous Polish metric structures. For instance, we construct a universal Polish metric space which is moreover equipped with countably many closed relations or with a Lipschitz function to an arbitrarily chosen Polish metric space. This work can be considered as an extension of the result of P. Urysohn who constructed a universal and ultrahomogeneous Polish metric space.

Soot particles are a major absorber of shortwave radiation in the atmosphere. They exert a rather uncertain direct and semi-direct radiative effect, which causes a heating or in some cases a cooling of the atmosphere. The mass absorption coefficient is an essential quantity to describe this light absorption process. This work presents new experimental data on the mass absorption coefficient of soot particles in the troposphere over Central Europe. Mass absorption coefficients were derived as the ratio between the light absorption coefficient determined by multi angle absorption photometry (MAAP), and the soot mass concentration determined by Raman spectroscopy. The Raman method is sensitive to graphitic structures present in the particle samples, and was calibrated in the laboratory using Printex90 model particles. The mass absorption coefficients were determined for a number of seven observation sites, ranging between 3.9 and 7.4 m²/g depending on measurement site and observational period. The highest values were found in an continentally aged air mass in winter, where we presumed soot particles to be present mainly in internal mixture. The regional model WRF-Chem was used in conjunction with a high resolution soot emission inventory to simulate soot mass concentrations and absorption coefficients for the Central European Troposphere. The model was validated using soot mass concentrations from Raman measurements and absorption coefficients. Simulated soot mass concentrations were found to be too low by around 50 %, which could be improved by scaling the emissions by a factor of two. In contrast, the absorption coefficient was positively biased by around 20%. Adjusting the modeled mass absorption coefficient to measurements, the simulation of soot light absorption was improved. Finally, the positive direct radiative forcing at top of the atmosphere was found to be lowered by up to 70% for the model run with adjusted soot absorption behaviour, , indicating a decreased heating effect on the atmosphere.

This thesis presents a comparison of a GPU implementation of the Conjugate Residual method as a sequence of generic library kernels against implementations ofthe method with custom kernels to expose the performance gains of a keyoptimization strategy, kernel fusion, for memory-bound operations which is to makeefficient reuse of the processed data. For massive MIMO the iterative solver is to be employed at the linear detection stageto overcome the computational bottleneck of the matrix inversion required in theequalization process, which is 𝒪(𝑛3) for direct solvers. A detailed analysis of howone more of the Krylov subspace methods that is feasible for massive MIMO can beimplemented on a GPU as a unified kernel is given. Further, to show that kernel fusion can improve the execution performance not onlywhen the input data is large matrices-vectors as in scientific computing but also inthe case of massive MIMO and possibly similar cases where the input data is a largenumber of small matrices-vectors that must be processed in parallel.In more details, focusing on the small number of iterations required for the solver toachieve a close enough approximation of the exact solution in the case of massiveMIMO, and the case where the number of users matches the size of a warp, twodifferent approaches that allow to fully unroll the algorithm and gradually fuse allthe separate kernels into a single, until reaching a top-down hardcodedimplementation are proposed and tested. Targeting to overcome the algorithms computational burden which is the matrixvector product, further optimization techniques such as two ways to utilize the faston-chip memories, preloading the matrix in shared memory and preloading thevector in shared memory, are tested and proposed to achieve high efficiency andhigh parallelism.

Floodplains represent a global hotspot of sensitive socioenvironmental changes and early human forcing mechanisms. In this review, we focus on the environmental conditions of preindustrial floodplains in Central Europe and the fluvial societies that operated there. Due to their high land-use capacity and the simultaneous necessity of land reclamation and risk minimisation, societies have radically restructured the Central European floodplains. According to the current scientific consensus, up to 95% of Central European floodplains have been extensively restructured or destroyed. Therefore, question arises as to whether or when it is justified to understand Central European floodplains as a ‘Fluvial Anthroposphere’. The case studies available to date show that human-induced impacts on floodplain morphologies and environments and the formation of specific fluvial societies reveal fundamental changes in the medieval and preindustrial modern periods. We aim to contribute to disentangling the questions of when and why humans became a significant controlling factor in Central European floodplain formation, and how humans in interaction with natural processes and other chains of effects have modified floodplains. As a conclusion, we superimpose emerging fields of research concerning the onset of the Fluvial Anthroposphere and provide 10 specific thematic objectives for future multidisciplinary work.