69 research outputs found
Diversity and Distribution of Methane Functional Microorganisms in Sedimentary Columns of Hongfeng Reservoir in Different Seasons
No full textFreshwater ecosystem is a significant natural source of CH4 emission in the atmosphere. To fully understand the dynamics of methane emissions in reservoirs, it is essential to grasp the temporal and vertical distribution patterns, as well as the factors that influence the methanogenic bacterial communities within the sediments. This study investigates the methane dynamics, carbon isotope fractionation (δ13CH4), and abundance of functional microorganisms along the geochemical gradient in the in situ sedimentary column of Hongfeng Reservoir (China). Notably, the methane concentration in sediment in summer ranged in 15.39–127.22 µmol/L, which is twice as high as wintertime concentrations in the surface layer near the sediment–water interface (0–10 cm depth). Illumina sequencing of the sediments identified 11 genera affiliated with methanogenic archaea, with dominant genus Methanosaeta reaching a relative abundance of 34.95% in summer. The total carbon (TOC) content in sedimentary columns in different seasons is positively correlated with Methanosarcina (P < 0.05). In addition, seasonal discrepancies are observed in the sediment profiles for total nitrogen (TN), sulfate (SO42−), and ferrous iron (Fe2+) concentrations. The concentration of total nitrogen (TN) is higher in summer than in winter. In summer, sulfate accumulates in the middle layer of the sedimentary column, while in winter, the maximum concentration of sulfate in the surface layer reaches 0.65 mmol/L. These geochemical gradients drive the biological transformation of nitrogen, sulfur, and iron, may also be linked to the consumption of methane. Thus, it is established that the temporal and spatial dynamics of methanogenic communities in sediments significantly influence the fluctuations in methane release fluxes within reservoirs, highlighting the necessity to account for seasonal biological variations when assessing greenhouse gas emissions from reservoirs. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024
Li yong hai di di zhen yi shu ju yan jiu Maliya na nan bu fu chong dai gao jing du di zhen fen bu he di zhen su du jie gou
No full textPh.D.In this dissertation, I utilize multiple seismic techniques to investigate the seismicity and seismic structures of the Southern Mariana Subduction Zone through multiple ocean bottom seismic experiments. Data for these studies consist of records from 12 China-made ocean bottom seismographs (OBS) deployed in the Southern Mariana Subduction Zone. It is the first time in the world that successful deployment and recovery of the OBSs have been carried out in the Challenger Deep of the Mariana trench, the deepest region on the Earth. Before this experiment, few OBS experiments can go to depths greater than 6 km. Therefore, the datasets provide valuable opportunities to investigate the seismotectonics and seismic structure of the Southern Mariana Subduction Zone. I correct clock errors of the OBS recordings based on both teleseismic waveforms and ambient noise cross-correlation. I also present detailed approaches to identify potential errors in the polarity of individual OBS instruments, to determine the horizontal orientation of OBS on the seafloor via polarization analysis and waveform modelling.To develop a catalog of small earthquakes in the region, I perform matched filter earthquake detection and find >7000 local earthquakes during the 6-month deployment period. I then locate the detected earthquakes using a newly refined local velocity model that is derived from a companion active source experiment in the same region. In total, 2004 earthquakes are located with relatively high resolution. The earthquakes span the Southwest Mariana rift (SWMR), the megathrust interface, fore-arc and outer-rise regions. While most earthquakes are shallow, maximum depths of the slab earthquakes increase from ∼100 to ∼240 km from west to east towards Guam. The seismicity results delineate the subducting interface with an increasing trend in dip angles from west to east. The observed along-strike variation in slab dip angles and its downdip extents provide new constraints on geodynamic processes of the southernmost Mariana Subduction Zone.To constrain the deformation mechanism, I then use waveform inversion to derive earthquake focal mechanisms. The normal faulting earthquakes near the SWMR are consistent with the historical focal mechanisms, indicating active rifting of the SWMR. The normal faulting earthquakes are also observed within the overriding plate spreading over the fore-arc to the SWMR, indicating extensional deformation over the SWMR and the fore-arc region in the Southern Mariana Subduction Zone. The fore-arc strike-slip faulting earthquakes imply that the strike-slip component due to oblique convergence is not completely relieved aseismically at southern Mariana margin.In order to investigate the hydration (serpentinization) of the Southern Mariana Subduction Zone, I conduct tomographic studies of Rayleigh wave velocity to derive the S-wave structure across the Challenger Deep. The measured group and phase velocities as a function of period are combined to invert the S-wave velocity using a Bayesian Monte Carlo algorithm. The tomographic results show low velocity (3.5-4.0 km/s) within upper ~30-km of the mantle near the trench, indicating hydrated upper mantle. Based on the experimental relationship between seismic velocities and serpentinization degree, shear wave velocities of 3.5-4.0 km/s are likely related to serpentinization percentage of ~20-40 vol%.在本篇論文中，我基於多次海底地震儀實驗獲得的數據，利用多種地震學方法研究馬里亞納南部俯衝帶的高精度地震分佈和地震速度結構。研究數據源自於布设在馬里亞納南部俯衝帶的12台中国生产的海底地震儀。这是世界上首次成功地在地球最深的马里亚纳海沟挑战者深渊区域布设并回收海底地震仪。在此之前，鲜少有水深超过6公里的海底地震仪实验成功。因此，这次实验为研究馬里亞納南部俯衝帶的地震构造和速度结构提供了宝贵的数据。我首先利用遠震波形走時和背景噪聲互相關函數校正時間，然後利用地震波極化分析和波形模擬方法校正水平分量的極性和方位角。為了檢測目錄中缺失的地震，我利用波形匹配濾波方法進行地震檢測，在六個月的數據記錄中檢測到超過7000個地震，最終獲得了2004個地震的精確位置。地震遍佈於西南馬里亞納斷裂（SWMR），弧前和弧後地區，俯衝板片間和俯衝半片內部。雖然大多數地震深度較淺，但板內地震最大深度從西至東逐漸從約100公里增加至約240公里，並且顯示了俯衝角度從西向東逐漸變大。对俯衝板片沿海溝走向上深度和角度變化的約束，有助於了解該區動力學過程。為了約束該區的變形機制，我利用波形反演方法得到地震震源機制解。在SWMR附近觀測到正断层類型的機制解與歷史地震的機制吻合，指示了SWMR目前處於活躍裂解階段。從SWMR至弧前區域，都觀測到正斷層類型的機制解，說明了馬里亞納南部俯衝帶上覆板片從SWMR至弧前的擴張變形。在馬里亞納南部俯衝帶的弧前區域也觀測到走滑地震，說明斜向俯衝分量沒有完全以非地震滑移的方式釋放。為了研究馬里亞納南部俯衝帶的水化（蛇纹石化）特性，我利用遠震地震波和背景噪聲互相關函數提取瑞雷面波，得到瑞雷面波不同週期的群速度和相速度，進而利用貝葉斯蒙特卡洛反演剪切波速度結構。結果顯示，在距離海溝近的外緣地區下方，有約30 km上地幔顯示低速度異常(3.5-4.0 km/s)，可能與上地幔水化有關。依據實驗測得剪切波速度與水化程度的關係，3.5-4.0 km/s剪切波速度可能對應~20-40 vol%的蛇紋石化。Zhu, Gaohua.Thesis Ph.D. Chinese University of Hong Kong 2020.Includes bibliographical references (leaves 87-101).Abstracts also in Chinese.Title from PDF title page (viewed on 12, October 2021).Zhu, Gaohua
Enhancing the anti-colon cancer activity of quercetin by self-assembled micelles
No full textGuangya Xu,1,* Huashan Shi,2,* Laibin Ren,1 Hongfeng Gou,1 Daoyin Gong,1 Xiang Gao,1–3 Ning Huang11Department of Pathophysiology, West China College of Preclinical Medicine and Forensic Medicine, Sichuan University, Chengdu, Sichuan, People’s Republic of China; 2State Key Laboratory of Biotherapy and Cancer Center, 3Department of Neurosurgery, West China Hospital, West China Medical School, Sichuan University, Chengdu, Sichuan, People’s Republic of China*These authors contributed equally to this workAbstract: Colorectal cancer, a type of malignant neoplasm originating from the epithelial cells lining the colon and/or rectum, has been the third most frequent malignancy and one of the leading causes of cancer-related deaths in the US. As a bioflavonoid with high anticancer potential, quercetin (Qu) has been proved to have a prospective applicability in chemotherapy for a series of cancers. However, quercetin is a hydrophobic drug, the poor hydrophilicity of which hinders its clinical usage in cancer therapy. Therefore, a strategy to improve the solubility of quercetin in water and/or enhance the bioavailability is desired. Encapsulating the poorly water-soluble, hydrophobic agents into polymer micelles could facilitate the dissolution of drugs in water. In our study, nanotechnology was employed, and quercetin was encapsulated into the biodegradable nanosized amphiphilic block copolymers of monomethoxy poly(ethylene glycol)–poly(ε-caprolactone) (MPEG–PCL), attempting to present positive evidences that this drug delivery system of polymeric micelles is effective. The quercetin-loaded MPEG–PCL nanomicelles (Qu-M), with a high drug loading of 6.85% and a minor particle size of 34.8 nm, completely dispersed in the water and released quercetin in a prolonged period in vitro and in vivo. At the same time, compared with free quercetin, Qu-M exhibited improved apoptosis induction and cell growth inhibition effects in CT26 cells in vitro. Moreover, the mice subcutaneous CT26 colon cancer model was established to evaluate the therapy efficiency of Qu-M in detail, in which enhanced anti-colon cancer effect was proved in vivo: Qu-M were more efficacious in repressing the growth of colon tumor than free quercetin. In addition, better effects of Qu-M on inducing cell apoptosis, inhibiting tumor angiogenesis, and restraining cell proliferation were observed by immunofluorescence analysis. Our study indicated that Qu-M were a novel nanoagent of quercetin with an enhanced antitumor activity, which could serve as a promising potential candidate for colon cancer chemotherapy.Keywords: quercetin, nanoformulation, colon cancer, cell apoptosis, angiogenesi
The addition of PD-1 inhibitor overcame trastuzumab resistance in patients with HER2 positive, PD-L1 negative metastatic gastric cancer: Case report and review of literature
No full textGastric cancer (GC) is a malignancy with poor prognosis and high heterogeneity. For HER2-positive, PD-L1 negative metastatic GC patients, chemotherapy plus trastuzumab is the first-line therapy. However, such patients soon acquired resistance to treatment, especially to trastuzumab during the treatment. Improving the therapeutic resistance of HER2-positive, PD-L1 negative metastatic GC is still a dilemma. We present the case of a metastatic GC patient with HER2-positive and PD-L1-negative expression who suffered progression after a short remission with trastuzumab plus chemotherapy. The patient exhibited strong heterogeneity in the primary and metastatic lesions. His resistance to trastuzumab was overcome after the addition of a PD-1 inhibitor, after which he received a durable response for more than 8 months. In HER2-positive, PD-L1-negative metastatic GC, the addition of PD-1 inhibitors after first-line chemotherapy and trastuzumab treatment resistance may be an option
The dynamic behavior of gas hydrate dissociation by heating in tight sandy reservoirs: A molecular dynamics simulation study
No full textKnowledge on the kinetics of gas hydrate dissociation in microporous sediments is very important for developing safe and efficient approaches to gas recovery from natural gas hydrate (NGH) deposits. Herein, molecular dynamics (MD) simulations are used to study the dissociation kinetics in microporous sediments. The hydrate phase occupies a confined sandy nanopore formed by two hydroxylated silica surfaces with a buffering water layer between the hydrate and silica phase, meanwhile, this system is in contact with the bulk phase outside the pore. The hydrates in this sediment system dissociate layer-by-layer in a shrinking core manner. The released methane molecules aggregate and eventually evolve into nanobubbles, most of which are spherical cap-shaped on the hydroxylated silica surfaces. At high initial temperatures, a faster decomposition of the hydrate phase is observed, however, fewer methane molecules migrate to the bulk phase from the pore phase. These phenomena may occur because more methane molecules are released from the hydrate phase and facilitate the formation of nanobubbles with large heat injection; these nanobubbles can stably adsorb on the surface of silica and capture the surrounding methane molecules, thereby decreasing the number of methane molecules in the water phase. In addition, the injection speed of heat flow should be significantly increased at high dissociation temperatures when using the thermal stimulation method to extract gas from hydrates in tight sediments. This study provides molecular level insight into the kinetic mechanism of hydrate dissociation and theoretical guidance for gas production by thermal injection from sediments with low permeabilities.Accepted Author ManuscriptEngineering Thermodynamic
Deriving rupture scenarios from the interseismic locking distribution in Nicoya Peninsula, Costa Rica
No full textM.Phil.The largest earthquakes often occur on the shallow plate interface in subduction zones, usually accompanied with devastating ground shakings and tsunamis. For example, during the 2004 Sumatra Mw 9.3 earthquake and 2011 Tohoku-Oki Mw 9.0 earthquake, strong ground shakings and huge tsunamic waves caused large fatalities and property losses. Rupture prediction and hazard assessment in subduction zones are highly needed for society. Geodetic measurements provide an important prerequisite for earthquake study. During the interseismic period, geodetic resultsindicate that most regions on the overriding plate move back together with the oceanic plate landward due to frictional coupling between two plates. Moreover, the high coupling zone obtained from the geodetic observations is sufficient to allow a megathrust earthquake. Such kind of qualitative estimation for potential earthquakes from interseismic locking distribution has been validated by a few recent megathrust earthquakes. For example, the largest slip areas during the 2010 Maule earthquake are consistent with high locking distributions. However, the rupture did not propagate further north to the high locking regions. A lot of questions about the earthquake source are still unclear without quantitative estimation.To explore these questions, we conduct dynamic rupture simulations in Nicoya Peninsula, Costa Rica, where the Cocos plate subducts beneath the Caribbean plate at a rate of ~82 mm/yr. Because the peninsula juts more seaward only ~40 km to the trench, it forms a unique subducting environment to allow near-field geodetic observations directly above the seismogenic zone. On September 5, 2012, an Mw 7.6 earthquake occurred directly under the peninsula, consistent with one of the previously determined locked patches. The well instrumentally recorded earthquake provides us an unprecedented opportunity to quantitatively evaluate coseismic slip from interseismic locking distribution using numerical simulations, as well as do meaningful comparisons with coseismic observations.Based on the structure of Nicoya Peninsula, we adopt a three-dimensional curved fault geometry embedded into an elastic medium to set up the model. The initial stress is fixed by the build-up stress inferred from locking and dynamic friction from fastspeed sliding experiments. Govern by the slip-weakening law, we use a finite element software, PyLith, to obtain the dynamic rupture process and corresponding ground motions. The spontaneous rupture initiates at the epicenter of the 2012 Nicoya earthquake, and bi-laterally propagates along high stress concentration region. The simulated result shows a high coherence with geodetic observations in final slip, moment rate and rupture speed. Our results successfully evaluate earthquake potential based on the interseismic locking distribution. However, due to the heterogeneous initial stress distribution, we do not know where the next large earthquake will start and how much energy is needed for large earthquakes to initiate. Thus, we test all possible nucleation zones on the locking regions. The results show that, in Nicoya region, only 40% nucleation zones could finally become large earthquakes (Mw > 7.2). Rupture initiating in other regions fails or develops into medium-size earthquakes with Mw smaller than 6. Therefore, the epicenter locations affect the rupture scenarios and lead to various sizes of earthquakes giving the heterogeneous stress distribution. Our simulated rupture scenarios have great significance to help the rupture prediction and hazard assessments.大地震經常伴隨著強烈的地表震動和海嘯發生在俯沖帶淺部。例如，2004 年9.3 級的蘇門答臘地震和2011 年9.0 級的日本東北地震，強烈的地面震動和巨大的海嘯讓人們的生命財產受到了巨大的傷害。因此，對於俯沖帶大地震的預報和危險評估是極為重要的。大地測量的方法為地震研究提供了堅實的基礎。在地震發生之前，大地測量的數據顯示由於大洋板塊和大陸板塊之間的藕合作用，他們會受到俯沖的影響粘在壹起向陸地的方向移動。當積累的應力達到壹定強度時，大地震就會在高閉鎖區域發生。這種定性的評估在最近的壹些大地震中被證實： 在2010 年的智利地震中，發生較大滑移的位置與高閉鎖區域吻合的非常好。但是，此次地震並沒有沿著高閉鎖區域壹直傳播下去，並不是高閉鎖區域就會有大滑移，我們不知道地震是產生，傳播和停止的。因此，這就需要定量的去模擬潛在的地震，理解震源的傳播過程。為了去解決以上的問題，我們建立了壹個三維的俯沖帶模型，根據震間閉鎖的程度和區域，模擬了壹次可能的地震。在中美洲，由於Cocos 大洋板塊每年以非常大的速率向加勒比板塊俯沖，哥斯達黎加區域形成了壹個非常活躍的地震帶，歷史記載每五十年就會有壹次7 級以上的大地震，這為我們的地震研究提供了很好的歷史資料。此外，哥斯達黎加西海岸的尼克亞半島距離海溝非常近，科學家們在這個半島上佈置了很多GPS 和地震臺站，對於地震的活動週期做了壹個全面的監測。測量結果顯示，2012 年7.6 級的尼克亞地震恰巧發生在之前確定的震間閉鎖區域。因此，這給我們的地震模型提供了壹個非常好的機會，將模擬出來的地震和真實地震作對比。根據哥斯達黎加的俯沖帶板塊和震間閉鎖區域，我們建立了壹個彈性介質模型，將初始應力設定為震間應力積累和動態應力的和。在滑移弱化準則基礎上，運用有限元的方法得到了壹個自發破裂的地震過程。結果顯示，我們模擬的地震從震級，滑移面積，破裂過程與真實地震相比有很高的壹致性。因此，運用震間閉鎖信息可以定量的估計未來地震的大小和破裂過程。 在這次模擬中我們選擇了2012 年地震的震源作為成核區。然而，在真實的地震中，由於強度和初始應力的不均勻性，我們並不知道地震會從哪裏開始，會以怎樣的方式破裂，所以我們測試了斷層上所有潛在的成核區。我們發現在中等閉鎖（55%）的區域內，僅有40%的成核區可以發展成壹個7.2 級以上的大地震，其他的都小於6 級或者由於初始應力太小不能被觸發。以上的結果表明成核區位置會影響潛在地震的大小。大量的數值模擬實驗可以幫助我們更好評估地震風險和海嘯預警。He, Bing.Thesis M.Phil. Chinese University of Hong Kong 2017.Includes bibliographical references (leaves 66-72).Abstracts also in Chinese.Title from PDF title page (viewed on 03, February, 2020)
Fault zone properties inferred from travel-time data recorded by dense across-fault seismic arrays
No full textM.Phil.Fault zone is important for the understanding of earthquake genesis and rupture process, and is seismically characterised as a low velocity zone (LVZ). The resolution required to study faults, with width in the order of 100s metres, could be achieved with the use of dense, across-fault seismic arrays. Two dense arrays were deployed across Renzhishi Fault in Heyuan, Guangdong; and Chenghai Fault in Binchuan, Yunnan. The arrays consisted of 41 seismometers with 60 – 500 m spacing, and 125 seismometers with 15 – 200 m spacing, respectively. Both were operational for one month. P- and S-wave arrival times of local earthquakes, as well as airgun and teleseismic earthquakes were recorded.By comparing the travel time difference between observed arrival time and calculation from 1-D velocity model, effects of the LVZ could be inferred. The array profiles showed clear delay in travel-time on both faults. This LVZ has a width of 500 – 700 m for the Renzhishi Fault, and 2950 m for the Chenghai Fault.Generalised ray theory was further used to calculate the theoretical arrival time on local earthquakes. This model takes into account the presence of the fault zone by incorporating an LVZ in a homogeneous host rock. Manual fitting of fault zone width and velocity reduction yielded a good fit in the P-wave arrival time. The best-fit model is a 500 m LVZ with P-wave velocity of 3.8 km s-1 for Renzhishi Fault, and a 2950 m LVZ with 3.8 km s-1 for the Chenghai Fault.Both faults showed features of a shallow fault zone. For the Renzhishi Fault, there is a prominent V-shaped travel time delay pattern, caused by a shallow fault that could only affect a small part of the array. From fitting of teleseismic arrival time, the depth is expected to be 1 km. In the Chenghai Fault, there were early arrivals or head wave observed on the opposite side of the fault. These are in agreement with a few tomographical studies, for example, one study showed the Renzhishi Fault have a depth of around 1 km; and a tomographic study of the Chenghai Fault detected the depth to be 2 km. However, another tomography at the Renzhishi Fault showed a much deeper depth of 5 km.The Renzhishi dataset suffered both timing error and channel flip. Teleseismic waveforms were very coherent across the array. By computing the particle displacement waveform, precise relative timing between stations could be obtained. Stations with timing error were identified statistically as outliers, and the magnitude of the error were calculated. Precise relative timing also allowed waveform cross-correlation with the reference permanent station, and flipped channels would yield an anti-correlation. While these errors were successfully detected and characterised, due to the episodic nature of these errors, and the sparcely-spaced teleseismic events, only 11 instances were corrected. All other uncorrected data were discarded.地震的發生和破裂過程，與斷層息息相關。斷層在地震的數據觀測上會令地震波在經過時減慢，故稱低速帶（low velocity zone）。觀測斷層所需的解析度為數百米，需要用到密集的地震台網。我們在兩個斷層上建立了地震台網：人字石和程海斷層。在人字石斷層上的台網有41個地震儀，每台相距60至500米；而程海斷層則有125個台、相距為15至200米的地震儀。兩個台網觀測時間均為一個月。本研究使用區內地震、遠震以及氣鎗的P和S波之到達時間為參考。實地觀測和一維模型計算的到達時間，觀兩者之差別，可見地震波經低速帶時，會令到達時間延遲。從此推測人字石斷層闊度為500至700米；程海斷層為2950米。為去除低速帶的影響，本研究用了Generalised Ray Theory計算到達時間，模型為均稱基岩（bedrock）中加了一條低速帶。我們在計算中用了不同的速度和闊度組合，模擬地震波的到達時間，得出人字石斷層的闊度為500米，秒速3.8公里；而程海斷層為2950米，秒速3.8公里。人字石斷層的到達時間橫截面中出現V字形，此見大部分的地震波無經過低速帶，只有小部分的地震儀觀察到低速帶的影響。從遠震觀測得出深度為1公里；而在程海斷層中可見斷層的背面則有折射波（refracted wave）和首波（head wave）。由此推測，研究的兩個斷層都較預期的淺。這結論和其他斷層成像研究結果一致，在人字石斷層研究中深度大約1公里；而程海斷層在一個未發表的研究中得出深度為2公里。然而，有其他於人字石斷層的研究得出深度為5公里，與本研究結果不符。在人字石斷層的數據中，有不少計時誤差和頻道顛倒（channel flip）。在台網記錄得有高相關性的遠震波形，加上計算出粒子位移（particle displacement）波形，可以得出台站之間的相對時差。有時間誤差的台站在統計方式上為離群值，從而計出其誤差的大小。得出精準的相對時差後，再與台網附近的固定台站，計算出互相關函數，其反相關值可得出頻道顛倒的跡象。這方法雖然有效地計算出時間誤差和頻道顛倒，但基於遠震數目稀少、而這些誤差沒有周期性，僅11個誤差可以修正，其他錯誤數據均不能使用。Lee, Hing Bun Martin."November 2019."Thesis M.Phil. Chinese University of Hong Kong 2020.Includes bibliographical references (leaves 73-77).Abstracts also in Chinese.Title from PDF title page (viewed on 20 May, 2021)
Autonomic data management for extreme scale coupled scientific workflows
No full textAdvanced coupled scientific simulation workflows running at extreme scales are providing new capabilities and new opportunities for high fidelity modeling and insights in a wide range of application areas. These workflows compose multiple physical models along with visualization and analysis services that share and exchange large amounts of data at runtime. Due to the huge I/O overhead, traditional file-based coupling approaches become infeasible. Instead, recent simulation-time data management approaches using in-memory data-staging methods have been explored to address this challenge. However, due to the complexities of emerging coupled applications and the architecture of current and future systems, these data staging based solutions are also presenting several new challenges. First, many of these scientific workflows containing dynamically adaptive formulations, such as Adaptive Mesh Refinement (AMR), which exhibit dynamic runtime behaviors and result in dynamically changing data volumes and imbalanced data distributions. Such dynamic runtime behaviors increase the complexity of managing and processing simulation data. In addition, these behaviors introduce new challenges of managing the staging resources as well as scheduling in-memory data processing while satisfying constraints on (1) the amount of data movement, (2) the overhead on the simulation, and/or (3) the quality of the simulations/analysis. Second, architectural trends indicate that emerging systems will have increasing numbers of cores per node and correspondingly decreasing amounts of DRAM memory per core as well as decreasing memory bandwidth. These trends can significantly impact the effectiveness of the online data management approaches for runtime data processing pipelines, and especially their ability to support data intensive simulation workflows. To address the above dynamic data management challenges, this thesis explores an autonomic approach to enable efficient runtime data management, which can dynamically respond to the varying data management requirements. Specifically, it first formulates an abstraction that can be used to realize autonomic data management runtimes for coupled simulation workflows. To address the dynamic data management challenges in tightly coupled simulation workflows containing dynamically adaptive formulations, this thesis then presents a realization of this autonomic approach that uses runtime cross-layer adaptations. This realization explores autonomic runtime adaptations at application layer, middleware layer, and resource layer. It also exploits a coordinated approach that dynamically combines these adaptations in a cross-layer manner. This thesis also presents an autonomic multi-tiered data management runtime that leverages both DRAM and SSD to support autonomic data management for loosely coupled scientific workflows. It demonstrates how an autonomic data placement mechanism can dynamically manage and optimize data placement across the DRAM and SSD storage levels in this multi-tiered runtime realization. The research concepts and approaches have been prototyped and experimentally evaluated using real application workflows on current high end computing systems, including the Intrepid IBM BlueGene/P system at Argonne National Laboratory and the Titan Cray-XK7 system at Oak Ridge National Laboratory.Ph.D.Includes bibliographical referencesby Tong Ji
- …
