EAR-PF: Quantifying heterogeneity in stratigraphy across scales

Andrew J. Moodie博士被授予NSF博士后奖学金，分别与导师保拉帕萨拉夸博士和杰夫·凯斯博士合作，在德克萨斯大学奥斯汀分校和斯坦福大学开展研究和教育计划。本研究旨在研究沉积岩在多个尺度上的性质，以建立大尺度地质变异性（~10 m）和小尺度变异性（~10 cm）之间的联系。地球物理成像无法解决小规模的地质变化，然而，穆迪博士希望使用统计模型和大规模变化的观察来限制这种变化。这项研究对沿海河流三角洲的可持续性至关重要，因为这些环境受到对地下地质和流动路径的有限了解的影响，这些地下地质和流动路径受小规模地质变化的控制。对流体流动路径的更好理解将影响许多领域，特别是，这项研究的结果将为污染物传输建模、地下水资源管理、浅层地热和碳封存操作以及碳氢化合物生产提供信息。穆迪博士的教育计划包括指导学生，组织关于地球科学中机器学习的阅读研讨会，以及为沉积岩地层学主题开发主动学习模块。沿海河流三角洲的可持续性受到许多自然和人为因素的影响，包括对地下水流模式的有限了解。地质异质性强烈影响流动路径，从而影响污染物传输和地下水含水层补给的速率，这限制了我们可持续管理水资源和减轻河流三角洲环境中健康风险的能力。由于通道和床形动力学（小于1米）的较小规模的地下非均质性通常是约束不足，因为它低于现有地球物理技术可以成像的分辨率。理论和一些证据表明，地层序列可能是尺度不变的，这开辟了一条途径，以限制小规模的非均质性，通过观察大规模的非均质性。该项目直接解决了这样一个问题：在一个空间尺度上从地下异质性中收集的模式和信息是否可以用来约束另一个尺度上的不确定性？穆迪博士将能够1）严格研究地层学的尺度不变特性，2）将这些发现整合到一种定量方法中，以限制地下非均质性。他将使用各种方法的组合，包括现场测量，数值建模，统计数据分析和机器学习。信息论的措施将量化地层学中的尺度不变性，现有的生成对抗神经网络方法将被修改，以映射跨空间尺度的异质性。该项目是由地球科学部（Earth Science Division，简称NSF）的沉积地质学和古生物学项目共同资助的。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Dr. Andrew J. Moodie has been awarded an NSF EAR Postdoctoral Fellowship to carry out research and education plans at The University of Texas at Austin and Stanford University in collaboration with mentors Dr. Paola Passalacqua and Dr. Jef Caers respectively. This study aims to investigate the properties of sedimentary rocks across multiple scales, to draw connections between large scale geological variability (~10 m) and smaller-scale variability (~10 cm). Geophysical imaging is unable to resolve small-scale geological variability, however, Dr. Moodie expects to constrain this variability using statistical models and observations of larger-scale variability. This research is critical to the sustainability of coastal river-deltas, because these environments are impacted by a limited understanding of subsurface geology and flow pathways that are controlled by small-scale geological variability. Improved understanding of fluid-flow pathways will influence many fields, in particular, results of this research will inform pollutant transport modeling, groundwater resource management, shallow geothermal and carbon sequestration operations, as well as hydrocarbon production. Dr. Moodie’s education plan includes mentoring students, organizing a reading seminar about machine learning in the geosciences, and developing active learning modules for topics in sedimentary rock stratigraphy. The sustainability of coastal river-deltas is impacted by a multitude of natural and anthropogenic factors, including a limited understanding of subsurface flow patterns. Geological heterogeneity strongly influences flow pathways and thus rates of contaminant transport and groundwater aquifer recharge, which limits our ability to sustainably manage water resources and mitigate health risks in river-delta environments. Smaller-scale subsurface heterogeneity due to channel and bedform dynamics (of less than 1 m) is typically under-constrained, because it is below the resolution that can be imaged by existing geophysical techniques. Theory and some evidence suggest that stratigraphic sequences may be scale invariant, which opens a pathway to constrain smaller-scale heterogeneity via observation of larger-scale heterogeneity. This project directly addresses the question: can patterns and information gleaned from subsurface heterogeneity at one spatial scale be used to constrain uncertainty at another scale? The EAR Postdoctoral Fellowship, will allow Dr. Moodie to 1) rigorously investigate scale invariant properties of stratigraphy, and 2) integrate these findings into a quantitative method to constrain subsurface heterogeneity. He will use a combination of approaches, including field measurement, numerical modeling, statistical data analysis, and machine learning. Measures from information theory will quantify scale invariance in stratigraphy, and an existing generative adversarial neural network method will be modified to map heterogeneity across spatial scales. This project was co-funded by the Sedimentary Geology and Paleobiology program in the Earth Science division (EAR).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.