A four-parameter-based thermo-TDR approach to estimate water and NAPL contents of soil liquid
A four-parameter-based thermo-TDR approach to estimate water and NAPL contents of soil liquid
复制标题
DOI：
10.1016/j.geoderma.2022.116263
复制
发表时间：
2023-01
期刊：
影响因子：
6.1
通讯作者：
Yuki Kojima;Kenta Okumura;S. Aoki;K. Noborio;Kohji Kamiya;R. Horton
中科院分区：
文献类型：
--
作者：
Yuki Kojima;Kenta Okumura;S. Aoki;K. Noborio;Kohji Kamiya;R. Horton
土壤非水相液体含量（θNAPL）的现场测定对于早期发现土壤NAPL污染和防止污染扩散是十分必要的。热时域反射仪（thermo-TDR）可以同时测量介电常数、电导率、体积热容和热导率，具有估算θNAPL的潜力。本研究的目的是：（1）建立四种热时域反射测量的土壤性质与土壤含水量（θw）和θ NAPL值之间的关系;（2）评估热时域反射测量的土壤性质对θ和θNAPL的敏感性;（3）开发一种基于四参数的方法来同时确定θ和θNAPL。在含有不同量的水和作为NAPL的芥花油的砂和玻璃珠上进行热-TDR测量。在所有情况下，θ w而不是θ NAPL主导了所有四个热TDR测量特性。敏感性分析还表明，所有四个属性对θ w比θNAPL更敏感。其中，介电常数、电导率和体积热容对θNAPL较为敏感，而热导率对θNAPL不敏感。一种新的方法使用所有四个热TDR测量的属性（四参数的方法）来估计θ NAPL被发现比现有的两个基于属性的方法，即，介电常数和体积热容为主。对于沙子和玻璃珠，基于四参数的方法估计θ NAPL的均方根误差（RMSE）值分别为0.066和0.042 m3 m − 3，而基于两种属性的方法的RMSE值分别为0.180和0.220 m3 m −3。基于四参数的方法能够通过优化过程抑制测量误差的影响，并允许高灵敏度参数覆盖低灵敏度参数中的缺点。利用热时域反射传感器结合四参数法测定θ NAPL，可以作为一种定量分析土壤NAPL含量的方法，为各种NAPL土壤污染研究提供参考。
In-situ determination of soil non-aqueous phase liquid content (θNAPL) is necessary for early detection of soil NAPL contamination and preventing the spread of the contamination. Thermo-time domain reflectometry (thermo-TDR), which can simultaneously measure dielectric constant, electrical conductivity, volumetric heat capacity, and thermal conductivity, has the potential to estimate θNAPL. The objectives of the study are i) to establish a relationship between the four thermo-TDR measured soil properties and soil water content (θw) and θNAPLvalues and ii) to evaluate the sensitivities of the thermo-TDR measured properties to θwand θNAPL, and iii) to develop a four-parameter based approach to simultaneously determine θwand θNAPL. Thermo-TDR measurements were performed on sand and glass beads containing various amounts of water and Canola oil as a NAPL. In all cases θwrather than θNAPLdominated all four thermo-TDR measured properties. A sensitivity analysis also indicated that all four properties were more sensitive to θwthan to θNAPL. Among the four properties, the dielectric constant, electrical conductivity, and volumetric heat capacity were somewhat sensitive to θNAPL, while thermal conductivity was not sensitive. A new approach using all four thermo-TDR measured properties (four-parameter-based approach) to estimate θNAPLwas found to be more accurate than the existing two property-based approaches, i.e., dielectric constant and volumetric heat capacity-based. The root mean square error (RMSE) values for θNAPLestimation with the four-parameter-based approach were 0.066 and 0.042 m3m−3for sand and glass beads, while the two property-based approach had RMSE values of 0.180 and 0.220 m3m−3. The four-parameter-based approach enabled suppression of the effects of measurement errors by the optimization processes and allowed the high sensitivity parameters to cover for shortcomings in the low sensitivity parameters. Use of thermo-TDR sensors with the four-parameter-based approach to determine θNAPLcan contribute to various NAPL soil contamination studies as a NAPL content quantifying approach.