时域反射法是目前唯一能够用于野外实测获取未冻水含量的方法.土壤水分对土壤介电特性的影响很大：水的介电常数为88.2（0℃）,空气的介电常数约为1,土颗粒的介电常数为3～7,冰的介电常数为3.27[17-18].这种巨大的差异表明,可通过测量土壤的介电常数来推测土壤的（体积）含水量.但该技术直接应用到冻土方面来的结果并不理想.
都洋等（2004）将核磁共振法和时域反射法的结果进行了比较,认为不宜直接套用融土的标定曲线来获得冻土的未冻水含量,同时他发现与核磁共振法不同,时域反射法测得的未冻水含量随初始含水量的增大而明显增加[17].Watanabe和Wake（2009）也发现了同样的现象,并做了更详尽的研究.他们认为用四相混合物模型求解未冻水含量适合于砂土,而对于粘土和粉质粘土在改进的模型（考虑了土颗粒吸附作用和冰对土体结构的改变）中才能得到与核磁共振法一致的结果[18].总之,目前还没有一种可靠的获取野外实测未冻水含量的方法.
Up to now, Time Domain Reflectometry is the only method available to obtain unfrozen water content through field measurement. Soil moisture has great impact on soil dielectric property: dielectric constant of water is 88.2 (0 ° C), that of air is about 1, 3 ~ 7 for soil particle, 3.27 for ice [17-18]. The significant variations suggest that it is workable to deduce the (volume) water content by measuring the dielectric constant of the soil. However, the result of applying the technology directly into the frozen soil is not so ideal.
After comparing the result obtained by NMR method and that by Time Domain Reflectometry, Du Yang (2004) supposed that it was not suitable to apply the rating curve of melted soil to obtain the unfrozen water content in frozen soil; meanwhile, he found that unfrozen water content measured by domain reflectometry method was significantly increased with the increase of initial water content [17], which is different from that measured through NMR method. Watanabe and Wake (2009) also found the same phenomenon and had done a more detailed research. They believed that the four-phase mixture model for measuring the unfrozen water content is suitable for sand, whereas for clay and silty clay, it can only gain a result the same with that by NMR method in an improved model (taking the adsorption of soil particles and the change to soil structure caused by ice into consideration) [18]. Therefore, to sum up, there is no reliable way to obtain unfrozen water content through field measurements.