LNAPLs污染土壤复介电常数特征与成像研究

Light non-aqueous phase liquids (LNAPLs) contamination in soil has become one of the most prominent problems at present and in the future. According to the electrical changes of the medium caused by contamination, electromagnetic methods are in-situ, non-destructive, and effective tools to detect the contamination. The electrical change of the soil polluted by LNAPLs is very complicated. Generally, when the soil is contaminated by LNAPLs, the conductivity of the medium would decrease with the increasing of the concentration of hydrocarbon, and the dielectric constant would also decrease. However, the organic matter would be decomposed by the degradation produced by microorganisms in the anaerobic environment. This process makes the conductivity increase. Hence, it is very important to describe electrical structure of the contaminated medium accurately. Imaging is deemed to be the most direct and most effective technique for detection and identification of targets in the subsurface. Migration and synthetic aperture imaging are mature imaging techniques while they cannot reconstruct the electrical structure of the subsurface to classify and identify targets. Microwave imaging can achieve classification and identification of targets for it is an electromagnetic inverse scattering problem which can reconstruct the shapes or electrical structure of the targets. In a reference to above problems, this project plans to study electrical properties of LNAPLs contaminated soil by measuring the complex electrical parameters of soil samples; to study electromagnetic wave propagation in contaminated soils through numerical simulation, indoor, and field experiments; to study the microwave imaging by combination of different algorithms to improve the imaging accuracy and reconstruct electrical parameters of targets, and further to support the quantitative analysis and interpretation of the contamination.

利用电磁方法探测轻非水相液体（LNAPLs）造成的土壤污染具有非破坏性、高效、原位等优点。LNAPLs污染土壤电性变化非常复杂。当土壤被LNAPLs物质污染时，会认为随着污染物质浓度升高，电导率会下降，介电常数也会下降。但在厌氧性环境中栖息的微生物会产生降解作用对有机物进行分解导致电导率上升。因此描述受污染介质的电性参数非常重要。成像是最直接和最有效检测和识别地下目标的技术，而较成熟的偏移成像及合成孔径成像等不能重建地下目标体的电性结构，不能实现对目标体的分类识别。微波成像是电磁逆散射问题，能够重构地下目标体的形状或电性参数分布，解决目标分类识别问题。因此，本项目将开展LNAPLs污染土壤复介电常数测量，研究污染土壤的电性规律；通过数值模拟计算和实验研究LNAPLs污染土壤中电磁波传播规律；通过微波成像混合算法研究，提高成像精度，重构目标体电性参数，进而支持污染的定量分析和解释。

轻非水相液体（LNAPLs: Light Non-Aqueous Phase Liquids）引起的土壤和地下水的污染是一个世界范围的环境问题。利用电磁方法探测轻非水相液体（LNAPLs）造成的土壤污染具有非破坏性、高效、原位等优点。为此，本项目开展了研究并在以下四个方面取得成果：（1）通过实验室内污染土壤样品复介电常数测量，给出了LNAPLs污染土壤复介电常数特征，为探地雷达探测频率选择提供依据。建立了新介电关系模型并计算得到水-油-气-介电常数的关系图，能够快捷地确定不同污染程度介质的相对介电常数。（2）项目基于多相流理论建立了LNAPLs在土壤中的渗漏迁移模型，再利用新介电模型将水文参数转换为电性参数；然后基于时域用有限差分法对LNAPLs污染介质进行探地雷达探测的正演模拟，分析了由LNAPLs污染引起的介质电性变化和雷达信号响应特征，为实际探测提供理论依据。（3）实现了电磁波阻抗反演算法，得到了地下介质的介电常数分布信息，有利于对污染区域的判定。实现了全波形反演算法，针对跨孔雷达数据获得了高分辨率的介电常数和电导率分布。实现了基于Pauli极化分解的偏移算法，通过对包含极化特征信息的信号进行重构和成像，获得了高精度的成像结果，提高了探地雷达的探测能力。（4）系统研究了探地雷达属性技术，实现了对电性差异微小的LNAPLs污染的清晰定位。结合参数反演，开发了LNAPLs土壤探测探地雷达属性分析软件，通过提取和分析三瞬属性、极化属性、相似性属性和能量属性等多种属性，提供地下LNAPLs污染的形状、位置、空间展布和物性分布等定量解释信息。本项目的研究成果对石油、化工、工业废弃物处理处置等典型高风险污染场地的探测和监测具有理论指导意义和实际应用价值。