Combining geoelectrical imaging and X-ray Computed Tomography (CT) for improved hydraulic characterisation of soils

土壤是包气带水文和地球化学过程的宿主。然而，土壤结构和水力特性的变化仍然难以量化，因此，如果我们想真正了解土壤中的流体动力学和营养物质和污染物的命运，在多个尺度上改善物理特性是至关重要的。最近在两个有前途但互不相关的领域，即地电成像和X射线计算机断层扫描（CT），取得了快速进展。现代地球物理技术评价土壤的地球物理性质，以推断水文性质或状态的时空模型。新型仪器与永久传感器阵列允许连续的地球物理监测土壤体积在近实时和实用的分辨率在厘米范围内的土柱。相反，CT映射EM辐射的空间衰减随材料密度的变化，这允许在微观水平上检查土壤多孔结构。最先进的CT系统实现了比电子物理学高得多的空间分辨率（10 mm样本上约10 m体素），然而，土壤图像的准确分割并非微不足道，样本大小和分辨率之间存在权衡，重复测量（例如跟踪水分动态）非常耗时。迄今为止尚未尝试整合这两种方法，然而，它们联合应用于定量土壤表征，为减少优先流成像和非饱和导水率估计中的不确定性提供了巨大的潜力。项目目标：设计并开展实验室实验，利用CT和地球物理测量对土柱或土芯进行同步测量;通过利用互补性评估协同成像的潜力;建立理论和定量建模框架，以解释观测结果。学生将探索从典型的实验设计中产生的机会，这些实验设计用于（i）在BGS进行的土壤水文地球物理实验室研究，和（ii）在生物科学学院（UoN）进行的土壤孔隙结构CT成像研究。考虑到实际限制（规模、分辨率、仪器能力、实验室空间），直径约25 cm、高度约1 m的土柱/岩心可能会成为协同CT和地电成像的起点。预计多传感器电阻率层析成像（ERT）或谱激极化（SIP）[3]将成为主要的地球物理技术。柱将配备电流[2]或电容[8]耦合传感器阵列。BGS将提供地电成像设备，CT成像工作将使用UoNs CT扫描仪进行。将制定实验协议和测量策略来克服实际问题（例如成像伪影）。将评估样品几何形状变化的影响，并对不同地质来源和对比质地（砂为主与粘土为主）和结构（实验室组装与现场结构）的土壤进行研究，以确定实验策略的限制因素。将通过这两种方法对土壤水分（含和不含溶质）分布的变化进行系统成像，以量化其对这些变化的相对敏感性。将评估协同方法对估计非饱和水力特性的影响，并建立定量建模框架。使用图像分析和计算机视觉方法对数据集进行组合分析，将阐明低分辨率地球物理数据与从CT提取的特征之间的关系。

Soils are the host for hydrological and biogeochemical processes in the unsaturated zone. However, variations in soil structure and hydraulic properties remain difficult to quantify, hence improved physical characterisation at multiple scales is vitally important if we want to truly understand fluid dynamics and the fate of nutrients and pollutants in soils.Current soil imaging methodologies operate at different spatial scales, are sensitive to different physical properties, and have distinctive strengths. Rapid advances have recently been made in two promising, but unconnected fields, namely geoelectrical imaging and X-ray Computed Tomography(CT). Modern geophysical techniques evaluate geophysical properties of soils to infer spatiotemporal models of hydrological properties or states. Novel instrumentation with permanent sensor arrays allows continuous geophysical monitoring of soil volumes in near-real time and with practical resolutions in the cm range on soil columns. Conversely, CT maps variations of spatial attenuation of EM radiation with material densities, which allows examination of the soil porous architecture at the microscopic level. State-of-the-art CT systems achieve much higher spatial resolution than geophysics (~10m voxels on 10mm samples), however accurate segmentation of soil images is non-trivial, a trade-off exists between sample size and resolution, and repeat measurements, e.g. to track moisture dynamics, are time-consuming.Integration of both methodologies has not been attempted so far, however their joint application to quantitative soil characterisation offers great potential for reducing uncertainty in the imaging of preferential flow and estimation of unsaturated hydraulic conductivity. This would benefit studies of agricultural and industrial leaching of contaminants in different soil scenarios.Project aims:Design and undertake pioneering laboratory experiments using concurrent CT and geophysical measurements on soil columns or core;Assess the potential of synergetic imaging by exploiting complementarity;Establish theoretical and quantitative modelling frameworks to explain observed results.Programme of research:The student will explore opportunities arising from typical experimental designs employed in (i)hydrogeophysical laboratory studies on soils undertaken at BGS, and (ii)CT imaging studies of soil pore structure undertaken at the School of Biosciences(UoN). Given practical constraints (scale, resolution, instrumental capability, laboratory space), soil columns/core of ~25cm diameter and ~1 m height will likely form a starting point for synergetic CT and geoelectrical imaging. It is expected that multi-sensor electrical resistivity tomography(ERT) or spectral induced polarisation (SIP)[3] will be the dominant geophysical techniques. The columns will be instrumented with galvanically[2] or capacitively[8] coupled sensor arrays. BGS will provide geoelectrical imaging equipment, & the CT imaging work will be undertaken with UoNs CT scanners.An experimental protocol and measurement strategy will be developed to overcome practical issues (e.g. imaging artefacts). The effect of variations in sample geometry will be assessed and soils of different geological provenance and contrasting texture (sand-dominated versus clay-dominated) and structure (lab-assembled versus field-structured) will be investigated to determine limiting factors of the experimental strategy. Changes in soil water (with and without solutes) distribution will be systematically imaged by both methodologies, in order to quantify their relative sensitivities to these changes. The impact of the synergetic approach on estimating unsaturated hydraulic properties will be assessed and a quantitative modelling framework established. Combined analysis of the data sets using image analysis & computer vision approaches will elucidate relationships between lower-resolution geophysical data and features extracted from CT.