Ye Zhang, Ph.D. - US grants

Evaluation of Uncertainty in CO2 Sequestration Modeling: a Flow Relevance Study using Experimental Stratigraphy and Field Verification (Teapot Dome, Wyoming)

Ye Zhang
EAR-0838250
University of Wyoming

ABSTRACT

Injection of supercritical carbon dioxide (CO2) into deep permeable formations of sedimentary basins has been proposed as a viable approach to greenhouse gas sequestration (geostorage). Both aquifer storage efficiency and potential leakage subsequent to injection are critical factors for consideration. Though numerical modeling provides a key assessment tool, multiple sources of uncertainty exist in the model construct, creating significant uncertainty in predicting CO2 flow in the storage formations. For example, one important conceptual model uncertainty is the multiple levels of homogeneity with which a formation can be represented, which are typically constrained by the quality and accessibility of site-specific data. For a given model, some parameters exert more influence on the prediction outcomes than others, thus in site evaluations, the value and relevance of diverse data types need to be better understood. This proposal aims to address this fundamental assessment issue with a two-pronged strategy. First, CO2 flow simulations will be conducted in a novel, experiment-based synthetic aquifer as well as in three increasingly homogenized models (i.e., facies-scale, facies-assemblage-scale, formation scale representations). To assess parameter uncertainties, the simulations will be conducted within an efficient computation framework based on the Design of Experiment. By comparing model predictions (full range of scenarios) and sensitivity (the most significant parameters impacting CO2 flow), an optimal level of model complexity will be determined. The insights gained will then help guide the development of a site-specific model for a CO2 injection test in the Teapot Dome, Wyoming. In modeling the field test, the analysis workflow will be validated in a dynamic setting by integrating simulation with data collection and field observation. Results will clarify the most relevant data types in CO2 modeling that require better characterization. Since successful implementation of carbon geostorage depends on both the accuracy and cost-effectiveness of the technical assessment studies, our work will be of broad scientific significance as well as high societal relevance.

Proposal Number: EPS-1135482

Lead Institution: Brigham Young University

Project Director: Norman Jones

Linked to: EPS-1135483 (University of Wyoming)

Proposal Title: Collaborative Research: CI-WATER, Cyberinfrastructure to Advance High Performance Water Resource Modeling


The project seeks to establish a consortium of Utah and Wyoming researchers who will acquire and develop hardware and software cyberinfrastructure (CI) to support the development and use of large-scale, high-resolution computational water resources models. These models will enable comprehensive examination of integrated system behavior through physically-based, data-driven simulations. Successful integration requires data, software, hardware, simulation models, and tools to visualize and disseminate results, as well as outreach to engage stakeholders and impart science into policy and management decisions. The project, called CI-WATER, will provide a robust and distributed CI consisting of integrated data services, modeling and visualization tools, and a comprehensive education and outreach program that can revolutionize how computer models are used to support water resources research in the Intermountain West and beyond.


Intellectual Merit: The integrated data-intensive modeling enabled by the proposed CI will lead to better understanding of coupled natural and human water resources systems and their response and sensitivity to changes across space-time scales. Advances in data and modeling systems that enhance HPC usability and access by non-HPC specialists will transform the way hydrologic knowledge is created and provide broader informatics applicability beyond the field of water resources.


Broader Impacts: The project will provide CI that improves access to data and sophisticated models, enable scientists to populate models with readily accessible data, harness HPC resources to perform multi-decadal simulations over large spatial areas with space-time resolution, and transform the way hydrologic knowledge is used in water resource planning and management. The CI enhancements will be integrated into a robust education program focused on improving cyber-literacy throughout the region

? DESCRIPTION (provided by applicant): Astrocytes are a major type of glia that play critical roles in the development and function of the nervous system. Malfunction of astrocytes are involved in neurological disorders including glioma, autism, amyotrophic lateral sclerosis, traumatic brain injury and stroke. How astrocyte proliferation and differentiation are regulated remains poorly understood. Increased astrocyte proliferation in humans contributes to the expansion in brain size in human evolution, and is potentially important for human intelligence. Unchecked proliferation of astrocytes, however, can lead to glioma. The mechanistic differences in the regulation of astrocyte proliferation and differentiation in humans and mice are unknown. Transcription factors specifically expressed by a cell type are key regulators of cell differentiation. Although transcriptional regulations of astrocytes have been studied in the spinal cord and the retina, the transcription factor(s) that regulate astrocyte proliferation and differentiation in the brain remains elusive. In my preliminary studies, I used innovative methods to purify all of the major cell types from mouse brains and obtained sensitive and accurate transcriptome datasets of each of the cell types by RNA-sequencing. I identified three astrocyte-specific transcription factors with this unbiased approach. Mice deficient for one of these factors have substantially reduced expression of astrocyte genes. In addition, I developed the first method to acutely purify astrocytes and their progenitors from human brains and I optimized a culturing condition that prevents these astrocytes from becoming reactive, which is a major limitation of existing methods. Building on these results, I propose to test the hypothesis that th three astrocyte-specific transcription factors are necessary and sufficient for astrocytes differentiation and that the differential regulation of these factors underlies the increase of astrocytes in human brains compared with mouse brains. In the K99 phase, I will test the necessity and sufficiency of these transcription factors in astrocyte proliferation and differentiation using existing knockout mouse lines, and a combination of in vitro and in vivo molecular manipulation techniques including viral infection and in utero electroporation. I will acquire expertise in molecular manipulations from the mentoring labs. I will also examine the regulatory interactions between these three transcription factors and determine whether a transcriptional cascade formed by the three factors sequentially regulate astrocyte specification, proliferation, and maturation. Finally, as an independent investigator, I will utilize K99 phase training in molecular manipulations and examine the role of the three transcription factors in human astrocyte development with the new purification and culturing method I developed. I will also investigate the mechanisms underlying the increase of astrocytes in humans. The proposed research is expected to close a major knowledge gap in brain development, as astrocytes are the last major cell type of the brain without knowledge of the transcriptional regulation of their differentiation. Moreover, knowledge obtained from this project has the potential to advance the treatment of glioma.

Eighty percent of U.S. energy demands are met by subsurface resources while deep geologic formations are also used as waste repository such as in the proposed actions of carbon storage. However, activities in deep zones have the potential to impact potable water in overlying shallow aquifers that is subject to contamination from leaking brine, hydraulic fracturing fluids, or gasses. Hence, to manage extraction and storage operations from deep reservoirs and to minimize the environmental impact on them, both an understanding of the processes that contribute to potential leakage and the methods to monitor such leakage are needed. This information helps to evaluate environmental risks and to assess corrective actions. The methods that are currently available for such understanding require extensive data collection from the subsurface which is very costly for deep formations. This research aims to develop an innovative method to integrate all available data from both shallow aquifers and often limited data from deep geologic zones to improve the monitoring of adverse environmental impacts. This new method, when validated in the laboratory, will use more easily available data from the shallow aquifers, thus reducing the need for costly drilling into deep formations. The new science that will be developed will allow for safer extraction of energy from deep formations and provide more secure subsurface storage of carbon dioxide, a greenhouse gas, in order to mitigate global climate change that has potential human, ecological, and environmental impacts. The training opportunities associated with the execution of this research at two universities will contribute to scientific and technical human capacity building and new workforce development that will address emerging problems at the water-energy nexus.

The primary goal is to develop, and experimentally verify, a novel inverse theory that integrates limited data from deep formations with more abundant or easily obtainable shallow aquifer data for improved characterization of deep geologic zones as well as for the monitoring of connected, overlying aquifers for potential contamination. For data-poor subsurface systems, existing techniques that assume boundary conditions (BC) can result in non-unique and uncertain parameter estimates, leading to inaccurate models. Compared to the earlier techniques, the proposed theory does not use forward simulations to assess model-data misfits. Thus the knowledge of the difficult-to-determine site BC is not required. Instead, it imposes fluid flow and/or solute mass continuities conditioned to limited and noisy measurements. In this research, the theory will be further developed and tested by (1) inverting pressure and flow observations for hydraulic characterization of a deep formation, and (2) jointly inverting flow and water quality data for both deep zone characterization and leakage monitoring. This new theory, which is capable of simultaneous parameter and BC estimation, has been successfully tested with synthetic numerical data. Generation of accurate and comprehensive data for theory validation in the field is however not feasible. Thus, an approach that uses data from intermediate-scale laboratory testbeds is proposed. The experimental method allows for the creation of different aquifer heterogeneities in the laboratory and the accurate control of flow and transport initial and BC that emulate deep zone operations. The theory will be first tested by comparing parameters estimated using measurements made in a laboratory aquifer with accurately known parameters and BC. As a second step, hydraulic and tracer measurements will be made in a two-layered aquifer separated by a leaky aquitard. Data from both the shallow unconfined layer and the deep confined layer (i.e., source of the disturbance) will be jointly inverted to characterize the entire system and to identify leakage pathways and rates from the deep layer. The inversion algorithms will be validated by independent measurements from the same testbeds (i.e., fixed packing), but under different BC and leakage scenarios. After this validation, the theory will be demonstrated using synthetic data taken from a model representing a deep formation with geologically relevant parameters and conditions. The new method will aim to make characterization and monitoring more accurate and efficient for data-poor environments, making this research potentially transformational in both theory development and practical problem solution.

Project Summary/Abstract Astrocytes constitute at least one third of all cells in the human brain and are critical for the development and function of the central nervous system. Reactive astrogliosis is a spectrum of cellular, molecular, and functional changes of astrocytes found in a wide range of injuries and diseases, including epilepsy, brain tumor, Alzheimer?s disease, Parkinson?s disease, stroke, inflammation, and traumatic brain injuries. Based on studies of mouse models, reactive astrocytes play both beneficial and harmful roles in disease progression and neural repair by secreting cytokines that regulate immune cells, producing growth factors, and forming scars that insulate disease tissue from healthy tissue. However, little is known about the molecular and cellular changes of astrocytes in human patients, due in part to the difficulties of purifying and culturing human astrocytes. Previous methods of purifying human astrocytes rely on serum, which induces reactive astrogliosis in the purification procedure, making it difficult to investigate reactive changes of astrocytes in patients. We recently developed a novel purification and culturing method for human astrocytes without serum. Using our new method, we will perform molecular characterization of reactive astrocytes purified from human patients with epilepsy, brain tumor, Alzheimer?s disease, Parkinson?s disease, and arteriovenous malformation. In Aim 1, we will characterize the transcriptome of reactive astrocytes and test the hypothesis that the molecular phenotypes of reactive astrocytes are diverse in humans. In preliminary studies, we found that instead of being a single state, there are diverse reactive states of astrocytes depending on the disease condition. We will examine the function of molecules induced in reactive astrocytes in humans using in vitro cultures of human astrocytes. In Aim 2, we will directly compare the responses of human and mouse astrocytes to a variety of harmful stimuli. Our preliminary data showed that human and mouse astrocytes have different susceptibility to oxidative stress and that harmful stimulus activates different signaling pathways in human vs. mouse astrocytes. These studies has the potential to reveal what reactive astrocytes do or fail to do in human neurological disorders, and provide new therapeutic targets for treating epilepsy, brain tumor, and neurodegenerative disorders.

Project Summary/Abstract Glia (astrocytes, microglia, and oligodendrocytes) and vascular cells are critical for the development and function of the central nervous system. Glial and vascular defects are associated with aging and neurodegeneration. Juvenile brains exhibit remarkable plasticity and resilience that diminish when brains mature. The potential contribution of glia and vascular cells to the plasticity of juvenile brains and the diminution of regenerative potentials in adult and aging brains remain poorly understood. Investigating molecular differences between juvenile and mature glia and vascular cells holds promise for the identification of protective factors in the brain. We recently developed immunopanning methods to purify astrocytes, microglia, oligodendrocytes, neurons, and vascular cells from both human and mouse brains. Cell populations isolated by immunopanning have high purity and produce abundant RNA for transcriptome profiling by RNA-sequencing (RNA-seq). Sequencing of purified populations of cells provides higher sensitivity for the detection of differential gene expression than alternative methods such as single cell RNA-seq. Using immunopanning, we purified astrocytes from a series of developmental stages from both human and mouse brains. In this proposed study, we will first perform bioinformatics analysis of our juvenile and mature astrocyte datasets and identify differentially expressed genes at each stage (Aim 1). Human and mouse evolution separated about 100 million years ago. Translating discoveries made in mouse models into clinics has been challenging. Our human and mouse astrocyte datasets will allow us to identify developmentally regulated molecular pathways preserved through a hundred million years of evolution and therefore likely to be essential. These analyses will generate candidate astrocytic juvenile protective factors that can be tested in future studies. Furthermore, we will expand our RNA-seq comparison of juvenile and mature cells to neurons, microglia, oligodendrocytes, and endothelial cells using immunopanning purified cell populations (Aim 2). These systematic analyses have the potential to reveal candidate juvenile protective factors in each cell type, improve our understanding of the roles played by each cell type in brain maturation and aging, and uncover candidate molecular pathways to target in the treatment of aging and neurodegeneration. The proposed study will build the foundation for a larger scale study that tests the function of candidate juvenile protective factors in neurons, glia, and vascular cells.

Ranking from comparisons is a central problem in a wide range of learning and social contexts. Researchers in various disciplines including psychology, economics, and computer science have made significant contributions to the ranking problem. Despite much progress, many fundamentally important statistical tasks remain unclear. One example arises in quantifying the uncertainty of ranking procedures and in analyzing multi-wise comparison data, which appears naturally in recommendation systems, web search, social choice, and many other areas. This project aims to address these challenges by developing an in-depth understanding of the ranking problem through a systematic statistical and computational investigation. The wide range of important applications of the ranking problem ensures that the progress we make towards our objectives will have a great impact on a broad scientific community. The techniques and methods developed will further advance the interplay between a wide range of areas including statistics, optimization, and machine learning. New courses will be developed incorporating results from the project and graduate students will also be involved in the project through their research work supervised by the PI.

This project focuses on the following two directions. First, the PI will quantify the uncertainty in ranking from pairwise comparisons. To achieve this goal, the PI will carry out a thorough fine-grained and entrywise statistical investigation. Second, the PI will deepen theoretical and methodological understanding of the multi-wise comparisons by investigating various tasks including developing optimal procedures, carrying out entrywise analysis, and others. Advances along these directions will be made towards the goal of laying out a firm theoretic foundation that feeds back into the development of practical methodology and informed applications for ranking problems.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.