京都大学 / 新制・課程博士 / 博士(工学) / 甲第23861号 / 工博第4948号 / 新制||工||1773(附属図書館) / 京都大学大学院工学研究科都市社会工学専攻 / (主査)教授 清野 純史, 教授 三村 衛, 准教授 古川 愛子 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DGAM

Les réservoirs carbonatés sont exploités pour leur réserve d’eau potable, de ressource géothermique ou le stockage géologique du dioxyde de carbone. Ces réservoirs sont difficiles à caractériser à cause d’une histoire diagénétique souvent complexe. Cette thèse offre de nouvelles perspectives sur la caractérisation des propriétés pétrophysiques et élastiques des calcaires urgoniens de Provence. Une approche intégrée et multi-échelle est proposée pour caractériser les propriétés pétrophysiques et élastiques des carbonates. Cette étude est basée sur des mesures de vitesse d’ondes P (Vp) et S (Vs) à l’échelle du laboratoire (centimètre–décimètre) et du terrain (mètre–décamètre). En laboratoire, les Vp et Vs ainsi que l’anisotropie sont mesurées sur des plugs et sur des carottes, en utilisant différentes fréquences centrales ultrasonores. Sur le terrain, l’approche consiste à mesurer les Vp et Vs entre deux puits distant de 2 m sur une profondeur de 14 m. Les mesures sont ensuite interprétées en fonction de la géologie observée aux échelles macro- et microscopiques. Les principaux résultats montrent que les Vp et Vs moyennes sont indépendantes de l’échelle de mesure, car elles sont dictées par la porosité. L’anisotropie causée par les fractures (15%) et les hétérogénéités (5%) se manifestent par une variabilité autour des vitesses moyennes. L’approche adoptée ici a permis de définir les interactions entre les propriétés de la matrice, les hétérogénéités, les fractures et les propriétés élastiques des roches carbonatées. Elle a montré que les propriétés élastiques tout comme les structures géologiques varient en fonction de l’échelle. / Carbonate reservoirs are also exploited for water production, geothermal energy, and carbon geological storage. Their Geophysical characterization remains challenging because of complex diagenetic history. This work offers new insights into the characterization of petrophysical, and elastic properties of the Urgonian limestones in the Provence region. An integrated multi-scale approach is proposed to characterize carbonate rocks petrophysical and elastic properties. This study relies on P- and S-wave velocity (Vp and Vs) measurements carried out at laboratory (centimeter–decimeter) and field (meter–decameter) scales. Laboratory scale Vp, Vs, and anisotropy are measured on plugs and cores, while on the field they are measured between two boreholes (crosshole) over a distance of 2 m and 14 m depth. The measurements are then compared to the geology from the macro- to the microscopic scale. The main results show that the average Vp and Vs are porosity related, and are independent from scale. Anisotropy caused by fractures (15%) and heterogeneities (5%) is responsible for variations around the mean velocities. The approach adopted during this work has enabled to scope out the interplay between matrix properties, heterogeneity, fracturing, and elastic properties in carbonate rocks. It has shown that the elastic properties evolve with scale as well as the geological structures.

Eskers are long winding ridges that originate from gravel that has travelled in meltwater streams in glaciers. At the study site, Virttaankangas plane in southwest Finland, there are esker structures covered by sediments. One reason why it is important to study eskers is because they are used for purifying drinking water. The data used in the study were collected during a seismic survey in July 2014. During the survey a controlled source created seismic waves that travelled down through the earth and then reflected back up again. By detecting the travel time of the waves and estimating the velocity of the geologic layers, the depth to the reflecting structures could be calculated. There are two types of waves that travel through the body of the earth, pressure waves (P-waves) and shear waves (S-waves). In a previous study (Maries et al., 2017) P-wave data from the same survey have been analyzed, so this work focuses on S-wave data but also compares the result from the two. Some structures related to eskers were identifiable, such as fractures in the bedrock from the pressure of the main esker core. By comparing S- and P-wave results it was possible to see hints of the arched esker cores and esker fan lobes. Overall the result confirmed the model that was achieved of the profile in the previous study. The location of the bedrock both matched with the previous study, and added information about its orientation. An additional goal was to demonstrate the insensitivity of S-waves to water content by showing that if there was a water table reflection in the P-wave data, this reflection was missing in the S-wave data. The results showed water table reflections in the P-wave data, but there were no distinguishable water table reflections with appropriate velocity for S-waves in the S-wave data. / Rullstensåsar definieras som långa åsar med storlekssorterat grus som avlagrats av smältvattenströmmar i glaciärer. Vid undersökningsplatsen, Virttaankangasheden i sydvästra Finland, finns rullstensåsstrukturer som är begravda under sediment. En anledning till varför det är viktigt att undersöka rullstensåsar är att de används för filtrering vid framställning av dricksvatten. De data som användes i denna studie inhämtades under en seismisk undersökning i juli 2014. Undersökningen gick till på så vis att en kontrollerad källa skapade seismiska vågor som färdades ner i jorden för att sedan reflekteras tillbaka upp mot ytan. Genom att notera tiden det tog för vågorna att färdas, samt uppskatta hastigheten i de geologiska lagren, kunde djupet till de reflekterande strukturerna beräknas. Det finns två sorters vågor som kan färdas genom jorden, tryckvågor (P-vågor) och skjuvvågor (S-vågor). I en tidigare studie (Maries et al., 2017) analyserades P-vågsdata från samma seismiska undersökning, så detta arbete fokuserar på S-vågsdata men jämför också resultaten av båda två. Vissa strukturer kopplade till rullstensåsar kunde identifieras, så som sprickor i begrgrunden från trycket av den största rullstensåsen. Genom att jämföra resultat från S- och P-vågor kunde man se reflektioner från rullstensåsar och sediment. Sammantaget bekräftade resultatet den modell över profilen som framtagits i den tidigare studien. Berggrundens läge stämde överens med den förra studien och tillförde ny information om dess orientering. Utöver detta försökte man också demonstrera S-vågors okänslighet för vatten genom att visa att om det fanns reflektioner från grundvattenytan i P-vågsdatan så skulle de reflektionerna inte synas i S-vågsdatan. I P-vågsdatan visade det sig att det fanns grundvattenreflektioner, men det gick inte att urskilja några liknande reflektioner i S-vågsdatan.

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第20684号 / 工博第4381号 / 新制||工||1681(附属図書館) / 京都大学大学院工学研究科都市社会工学専攻 / (主査)教授 澤田 純男, 教授 清野 純史, 准教授 後藤 浩之 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DGAM

In this study, the limitations of point-source stochastic simulations are investigated in terms of fundamental geophysical parameters. Within this context, a total of 6000 synthetic ground motions are generated for various magnitude (5.0 &amp / #8804 / Mw &amp / #8804 / 7.5), source-to-site distance (less than 100 km), faulting style (shallow dipping and strike-slip) and site class (soft, stiff and rock) bins. The simulations are performed in two main stages: (1) the acceleration time series at outcropping very hard rock sites are simulated based on the stochastic method proposed by Boore (1983, 2003) and (2) they are modified through 1-D equivalent linear site response analysis to generate the free-field motions at soft, stiff and rock sites. Thus, as a part of this study, a probability-based soil profile model that considers the random variation of S-wave slowness as a function of depth is derived. The synthetic ground motions are assessed with several recent empirical ground-motion models to constitute the limitations of the simulation procedure. It is believed that the outcomes of this study will realistically describe the limitations of stochastic point-source simulation approach that can be employed further for the studies on improvements of this simulation technique.

Spring water inflow is distinct at Pah Tempe Hot Springs (also known as Dixie Hot Springs) situated within the damage zone of the Hurricane Fault in Timpoweap Canyon in Hurricane, Utah. Excising of the footwall by the Virgin River has created Timpoweap Canyon and allowed an unusual opportunity to study the spring inflow in relation to the fault damage zone. While correlation of these springs with the damage zone and visible fracture patterns on the canyon wall has been made, no subsurface faulting has been imaged to verify connection to these visible fractures and spring inflows (Nelson et al., 2009). The stream was logged and contoured to note the varying locations of spring water inflows in contrast with unsaturated Virgin River water. Seismic surveys were conducted and subsurface profiles made to locate offsets and faults. Photogrammetry was conducted and a three-dimensional model of the canyon and cliff wall was created to facilitate remote fracture mapping of this wallSubsurface features correlate to fractures, spring water inflow locations, and surface faults mapped by Biek (2002). This suggests that faulting and fracturing from the Hurricane Fault provides subsurface conduits for these thermal waters to rise. In one area in the stream, thermal inflow correlates with both subsurface offsets and major surface fractures. Numerous correlations between just spring water entry and subsurface offsets or surface fractures are also found. Fracture and fault density is atypical at Pah Tempe as these features do not diminish with distance from the main strand of the fault. This has led to the Sevier Orogeny accounting for creating the observed fracture conduits at Pah Tempe. Fractures in the canyon wall at Pah Tempe open west to east. This is indicative of the maximum horizontal compressive stress of southern Utah being north to south (Zoback and Zoback, 2015). Therefore the spring inflow at Pah Tempe is likely a result of the damage from the Hurricane Fault creating conduits for spring water to rise, rather than the Sevier Orogeny.

Utilizamos neste trabalho duas metodologias distintas, a função do receptor com ondas P e a função do receptor com ondas S, para mapear variações da crosta e interfaces do manto (litosfera-astenosfera, 410 km e 660 km) em diferentes estações sismográficas na placa Sul-Americana. No estudo da interface litosfera-astenosfera, por ser o primeiro realizado nesta região, utilizamos as estações temporárias do IAG/USP em conjunto com as estações permanentes da rede mundial cobrindo toda a placa Sul-Americana. O estudo para as outras interfaces (Crosta-Manto, 410 km e 660 km) foi feito com caráter regional, buscando detalhar características da crosta e manto na região estável da placa. Para ambos os métodos os traços (sismogramas) foram rotacionados para o sistema LQT, deconvolvidos, agrupados por pontos de perfuração e por estações, e finalmente empilhados. Nos traços empilhados as fases convertidas de interesse (Ps, Ppps, Ppss+Psps e Sp) foram identificadas e interpretadas. Para a parte estável da placa obtivemos um valor médio de espessura da crosta de 39.4±0.6 km, variando desde 31.0±0.5 km para a província Borborema, até 41.3±1.0 km para a bacia do Paraná, onde aplicamos uma correção para descontar o efeito do sedimento. A razão de velocidade para a crosta, Vp/Vs, apresentou valores mais altos para a bacia do Paraná (~1.75±0.08) e região litorânea oriental (>1.74), enquanto que as regiões cratônicas (cráton São Francisco e Amazônico) apresentaram valores de Vp/Vs baixos (<1.72), chegando até 1.68. O valor médio de Vp/Vs para todas as estações analisadas foi de 1.73±0.02. As variações dos tempos para as interfaces do manto mostraram boa correlação com resultados de tomografia sísmica de outros trabalhos, indicando alterações de até 5% na velocidade das ondas sísmicas para o manto superior sob os crátons, uma deflexão de até 15 km na interface de 660 km para a região Sul da bacia do Paraná e se mostraram bem correlacionadas com as médias globais para as outras região estudadas. Por fim, a espessura da litosfera apresentou valores desde ~40 km, sob as regiões de ilhas oceânicas, até ~160 km, sob as regiões mais estáveis. Para as regiões oceânicas a espessura da litosfera se mostra correlacionada com a idade da placa. À medida que adentramos a parte continental, o limite litosfera-astenosfera se torna menos proeminente, atingindo profundidades maiores no interior dos continentes e menores para as regiões marginais. Para a zona de subducção, observamos duas possíveis litosferas, uma oceânica, subduzindo junto com a placa de Nazca, e outra pertencente à parte continental. / Two distinct methodologies, the P- and S-wave receiver functions, are used to map variations in the crustal parameters (thickness and Vp/Vs) and mantle interfaces (lithosphere-asthenosphere, 410 km and 660 km) on a number of different seismograph stations located in the South American plate. The results of the S receiver function for the lithosphere-asthenosphere boundary are the first of this kind ever performed in South American continent and showed the large scale variations of this interface. To perform this study we analyze data from various global permanent stations together with all available data from temporary stations operated by the IAG/USP during the last15 years. For both methods the traces (seismograms) were rotated to the LQT system, deconvolved, grouped by piercing points and stations, and finally stacked. In the stacked traces, the converted phases (Ps, Ppps, Ppss+Psps and Sp) were identified and interpreted. Inside the stable part of the plate we found a mean crustal thickness of 39.4±0.6 km, ranging from 31.0±0.5 km in Borborema Province up to 41.3±1.0 km in the Paraná Basin, where we applied a correction to remove the sediment effects on the crustal estimates. The crustal velocity ratios, Vp/Vs, showed higher values for the Paraná Basin (~1.75±0.08) and Ribeira belt (>1.74), while the cratonic regions (São Francisco and Amazon cratons) showed low values of Vp/Vs (<1.72), down to 1.68. The average Vp/Vs obtained for all stations was equal to 1.73±0.02. The observed times of the converted mantle phases presented a good correlation with other tomographic studies, indicating that the upper mantle for the cratonic roots may be characterized by a variation up to 5% in seismic velocities, a 15 km deflection in the South Paraná 660 km discontinuity (probably due to a decreased temperature caused by the subducted slab); for other regions the converted times were close to the global average. As a final result, the lithospheric thickness presented values ranging from ~40 km under oceanic islands, to ~160 km under the stable continental regions. We found that for the oceanic islands the thickness of the lithosphere is correlated with the age of the plate. When we go further inside the continents, the lithosphere-asthenosphere boundary becomes less sharp, reaching larger depths inside the continents and shallower depths near the continental margin. In the Andean subduction area, we observed two possibles lithospheres, one oceanic, subducting together with the Nazca plate, and another belonging to the Continent, parallel to the crust interface.

Neste trabalho avaliou-se a potencialidade do uso da sísmica rasa de reflexão multicomponente para investigação geológica-geotécnica. Foram abordados vários aspectos relacionados à aquisição dos dados sísmicos de reflexão multicomponente, com o objetivo de entender as vantagens e limitações do método para aplicação em investigações de subsuperfície rasa. Os ensaios de campo foram realizados em duas áreas, ambas em terrenos da Bacia Sedimentar de São Paulo, em área urbana da cidade de São Paulo. Para a interpretação dos dados sísmicos multicomponente foram investigados procedimentos para a inversão não-linear dos tempos de trânsito e das amplitudes. O testes realizados orientaram a escolha da aproximação não-hiperbólica mais apropriada para calculo dos tempos de trânsito visando à análise de velocidades do pacote acima do refletor. O estudo numérico desenvolvido para a inversão das amplitudes mostrou a viabilidade da estimativa das velocidades e densidades, acima e abaixo do refletor, empregando-se as equações de Zoeppritz para as ondas refletidas PP, PSv, SvP e SvSv, antes e depois do ângulo crítico. Dada a complexidade da inversão nao-linear das amplitudes, se fez necessário elaborar uma estratégia estocástica de otimização e desenvolver uma nova abordagem para análise da função objetivo multi-dimensional, garantindo confiabilidade ao resultado da inversão não-linear. Os resultados deste trabalho mostraram o potencial da sísmica de reflexão rasa multicomponente para caracterização geológica-geotécnica, possibilitando um melhor entendimento das camadas superficiais. / This thesis aims to evaluate the useful of the multicomponent seismic methods for shallow investigations, mainly its potential for the geotechnical and geological characterization of the nearsurface. Several aspects regarding the acquisition and processing data of multicomponent seismic data are discussed. They were based on data set acquired in the urban area of Sao Paulo city, Brazil. Two different areas were investigated. Both located in sedimentary terrains belonging to the Sao Paulo Sedimentary Basin. We present a non-linear travel time and seismic amplitude inversion scheme to quantitative interpretation of multicomponent seismic data. Several tests were performed to guide the choice of non-hyperbolic equation more suitable for travel time inversion aiming the velocity analysis above the reflector. A numerical experiment developed to solve the nonlinear inversion of seismic amplitudes showed the feasibility to estimate seismic interval velocities and layer densities above and below the reflector using the exact Zoeppritz equations for PP, PSv, SvP e SvSv reflected waves, before and after critical angle. Due to the apparent complexity of the nonlinear seismic amplitude inversion, it was necessary elaborate the strategy for stochastic optimization and develop a new approach to analyze the multi-dimensional objective function, with different implications for the accuracy and efficiency of the non-linear inversion. The study show the benefits of using the multicomponent seismic method for shallow geological-geotechnical characterization, improving the nearsurface understanding, once allows an integrated analyzes of a more complete record of the wave field.

Neste trabalho avaliou-se a potencialidade do uso da sísmica rasa de reflexão multicomponente para investigação geológica-geotécnica. Foram abordados vários aspectos relacionados à aquisição dos dados sísmicos de reflexão multicomponente, com o objetivo de entender as vantagens e limitações do método para aplicação em investigações de subsuperfície rasa. Os ensaios de campo foram realizados em duas áreas, ambas em terrenos da Bacia Sedimentar de São Paulo, em área urbana da cidade de São Paulo. Para a interpretação dos dados sísmicos multicomponente foram investigados procedimentos para a inversão não-linear dos tempos de trânsito e das amplitudes. O testes realizados orientaram a escolha da aproximação não-hiperbólica mais apropriada para calculo dos tempos de trânsito visando à análise de velocidades do pacote acima do refletor. O estudo numérico desenvolvido para a inversão das amplitudes mostrou a viabilidade da estimativa das velocidades e densidades, acima e abaixo do refletor, empregando-se as equações de Zoeppritz para as ondas refletidas PP, PSv, SvP e SvSv, antes e depois do ângulo crítico. Dada a complexidade da inversão nao-linear das amplitudes, se fez necessário elaborar uma estratégia estocástica de otimização e desenvolver uma nova abordagem para análise da função objetivo multi-dimensional, garantindo confiabilidade ao resultado da inversão não-linear. Os resultados deste trabalho mostraram o potencial da sísmica de reflexão rasa multicomponente para caracterização geológica-geotécnica, possibilitando um melhor entendimento das camadas superficiais. / This thesis aims to evaluate the useful of the multicomponent seismic methods for shallow investigations, mainly its potential for the geotechnical and geological characterization of the nearsurface. Several aspects regarding the acquisition and processing data of multicomponent seismic data are discussed. They were based on data set acquired in the urban area of Sao Paulo city, Brazil. Two different areas were investigated. Both located in sedimentary terrains belonging to the Sao Paulo Sedimentary Basin. We present a non-linear travel time and seismic amplitude inversion scheme to quantitative interpretation of multicomponent seismic data. Several tests were performed to guide the choice of non-hyperbolic equation more suitable for travel time inversion aiming the velocity analysis above the reflector. A numerical experiment developed to solve the nonlinear inversion of seismic amplitudes showed the feasibility to estimate seismic interval velocities and layer densities above and below the reflector using the exact Zoeppritz equations for PP, PSv, SvP e SvSv reflected waves, before and after critical angle. Due to the apparent complexity of the nonlinear seismic amplitude inversion, it was necessary elaborate the strategy for stochastic optimization and develop a new approach to analyze the multi-dimensional objective function, with different implications for the accuracy and efficiency of the non-linear inversion. The study show the benefits of using the multicomponent seismic method for shallow geological-geotechnical characterization, improving the nearsurface understanding, once allows an integrated analyzes of a more complete record of the wave field.

The present thesis is devoted to the investigation of novel superconductors by phase-sensitive terahertz transmission and infrared to ultraviolet spectroscopy. In particular, a nominally undoped Pr2CuO4 superconducting thin film, an FeTe0.5Se0.5 thin film, and a LiFeAs single crystal have been investigated. The emphasis is on the low-frequency part of the optical spectrum (i.e., the terahertz and infrared spectrum), as the goal of the study was to shed light on the size and symmetry of the superconducting gaps and also to determine the temperature dependences as well as the absolute values of the penetration depth, which are key input parameters for models applicable for new superconductors. In addition, niobium has been investigated as a reference, so as to see what is expected from conventional superconductors and to clarify the electrodynamics of niobium. A superconducting Nb thin film with Tc of 8.04 K has been investigated by backward wave oscillator-based (BWO-based) and time-domain terahertz (TDT) spectrometers in the frequency range between 4 and 100 cm−1 for temperatures ranging from 2 to 10 K. From these measurements an energy gap of 22.50 cm−1 = 2.79 meV = 4.02kBTc have been determined. The optical conductivity below Tc could nicely be described by calculations according to the Eliashberg theory, with the electron-phonon interaction evaluated from tunneling measurements. Absolute values of the penetration depth have been calculated from phase-sensitive terahertz measurements. The zero-temperature limit of at T = 0 is found to be 115 ± 5 nm. From this value, a London penetration depth of 43 ± 2 nm has been obtained. The overall temperature dependence of the penetration depth follows a behavior typical for conventional s-wave superconductors. A superconducting Pr2CuO4 film with T0 structure and Tc of 27 K has been investigated by use of optical methods in a wide frequency (5 – 55000 cm−1) and temperature (2 – 300 K) range. A Drude-like peak centered at zero frequency is observed in the optical conductivity below 150 K, above which it shifts to finite frequencies. The detailed analysis of the low-frequency conductivity reveals that the Drude peak and a far-infrared (FIR) peak centered at about 300 cm−1 persist at all temperatures. The FIR spectral weight is found to grow at the expense of the Drude spectral weight with increasing temperature. Absolute values of the penetration depth have been obtained from temperature and frequency-dependent measurements. The zero-temperature limit of is estimated to be 1600 ± 100 nm. The overall temperature dependence of follows a behaviour typical for cuprate superconductors. However, a closer look at the penetration depth at T 12 K reveals a flattening in the temperature dependence. A superconducting FeTe0.5Se0.5 thin film with Tc = 19 K has been investigated using a combination of BWO and TDT spectroscopy in the frequency range 4 - 80 cm−1 and between 3 and 150 K. From such measurements, a superconducting energy gap of 30 cm−1, representing a coupling strength = 2.27, is observed. Further, the penetration depth has been derived from the temperature dependence of the imaginary part of complex conductivity with the penetration depth = 530 ± 10 nm at lowest measured temperature. The temperature-dependent normalized superfluid density, just as is the case with most iron-based superconductors, could nicely be described by the so-called two-gap gamma model. Finally, a superconducting LiFeAs single crystal with Tc = 18 K has been investigated by optical spectroscopy in the frequency range 15 - 55000 cm−1 between 5 and 300 K. From these measurements, no clear signature of the superconducting energy-gap opening could be identified in spite of the spectral weight been suppressed in the infrared frequency regime below Tc. This indicates that LiFeAs single crystal is in a clean limit. With the aid of the Ferrell-Glover-Tinkham (FGT) sum rule, an absolute penetration depth of 215 nm has been calculated from the missing area at 5 K.