COLLABORATIVE RESEARCH: A Multiscale Analysis of Chemotactic Bacteria Transport in Heterogeneous Porous Media

协同研究：趋化性芽孢杆菌在非均质多孔介质中转运的多尺度分析。福特化学工程系，美国哥伦比亚大学。俄勒冈州州立大学伍德化学、生物和环境工程学院趋化性是能动细菌感知其局部环境中化学浓度梯度并游向更高浓度的污染物以降解的能力。在地下，生物修复往往由于注入物质和驻留污染物之间无法实现良好混合而受到阻碍。在这种情况下，可以利用趋化性来加强污染区内细菌种群的混合。本研究的目的是将物理和化学非均匀多孔介质中的趋化性的直接实验测量与适当的机械描述理论相联系。两个主要假设将用于组织研究工作。他们将结合联合收割机（a）多尺度实验观察和（B）多尺度放大分析，以开发一个理论框架，可以描述在微观和宏观尺度上化学和物理异质介质中的趋化性。第一个假设是，在多孔介质与非水相液体（NAPL）污染物的捕获源，趋化性细菌将有可测量的慢比非趋化性细菌的运输。 特别是，趋化性细菌将表现出比非趋化性突变体更大的阻滞和拖尾，因为在流体-NAPL界面处选择性地捕获趋化性细菌。第二个假设是，在大规模异质性的存在下，细菌在趋化性与非趋化性条件下的运输是显着不同的。 虽然趋化性本质上是一个孔隙规模的过程，净影响的趋化性细菌从高到低的非均质介质的电导率区域的运输将是实验测量，并将在该领域的细菌运输相关。该项目将结合联合收割机的数据收集和放大创新的实验设计，开发多尺度模型的趋化性。本研究的目的是发展的宏观尺度（散装）运输行为的多孔介质中的趋化性的理解。 这必然涉及到连接的宏观尺度的运输系统，控制趋化性的基本微观尺度的功能。在小尺度上测量现象的能力的改进导致了实验数据集，其中包含的细节甚至在十年前都几乎是不可想象的。发展如此非凡的数据集也经常公布的问题？如何理解这些数据呢？有没有一种方法可以在其中搜索行为的基本特征？？数据分析的新原型（例如，机器学习算法、数据挖掘）已经被用作评估这种数据集的方法。当应用于从物理系统中产生的大数据集的问题时，升级方法是这些数据分析原型中的一种。这一建议的结果将导致更强大的模型，用于预测地下水系统中的微生物运输，这将导致改进生物修复计划的设计和实施。

COLLABORATIVE RESEARCH: A MULTISCALE ANALYSIS OF THE TRANSPORT OF CHEMOTACTIC BACTERIA IN HETEROGENEOUS POROUS MEDIARoseanne M. Ford Department of Chemical Engineering, University of VirginiaBrian D. Wood School of Chemical, Biological, and Environmental Engineering, Oregon State UniversityChemotaxis is the ability of motile bacteria to sense chemical concentration gradients in their local surroundings and swim toward higher concentrations of pollutants that they degrade. In the subsurface, bioremediation is often hindered by the inability to achieve good mixing between injected substances and the resident contaminants. In such situations, chemotaxis might be exploited to enhance the mixing of bacterial populations within contaminated zones. The goal of this study is to connect direct experimental measurement of chemotaxis in physically and chemically heterogeneous porous media with appropriate mechanistic descriptive theory. Two main hypotheses will be used to organize the research effort. They will combine both (a) multiscale experimental observations, and (b) a multiscale upscaling analysis to develop a theoretical framework that can describe chemotaxis in chemically and physically heterogenous media at microscopic and macroscopic scales. The first hypotheses is that in porous media with trapped sources of nonaqueous phase liquid (NAPL) pollutants, chemotactic bacteria will have measurably slower transport than non-chemotactic bacteria. In particular, chemotactic bacteria will exhibit greater retardation and tailing than will non-chemotactic mutants because of selective trapping of chemotactic bacteria at the fluid-NAPL interfaces. The second hypothesis is that transport of bacteria under chemotactic versus non-chemotactic conditions are dramatically different in the presence of large-scale heterogeneities. Although chemotaxis is inherently a pore-scale process, the net influence of chemotactic bacteria on transport from high to low conductivity regions of a heterogeneous medium will be experimentally measurable, and will have relevance to bacterial transport in the field. This project will combine innovative experimental designs for data collection and upscaling to develop multiscale models for chemotaxis. The purpose of this research is to develop an understanding of the macroscopic scale (bulk) transport behavior of chemotaxis in porous media. This necessarily involves linking the macroscale transport to the essential microscale features of the system that control chemotaxis. Improvements in the ability to measure phenomena at small scales has lead to experimental data sets that contain detail that was nearly unimaginable even a decade ago. The development of such extraordinary data sets has also frequently promulgated the question ?how does one make sense of these data? Is there a way to search for essential features of behavior in them?? New archetypes for data analysis (e.g., machine learning algorithms, data mining) have been employed as methods to assess such data sets. When applied to the problem of large data sets arising from physical systems, upscaling methods are among these data analysis archetypes. Outcomes from this proposal will result in more robust models for predicting microbial transport in groundwater systems, which will lead to improved design and implementation of bioremediation schemes.