Experimental and statistical studies on the thermal properties of frozen clay in Qinghai-Tibet Plateau

Frozen soil consists of an irregular distribution of solid particles, gas, water and ice lenses; hence, the thermal property of the frozen soil is uncertain and variable. In addition, ice lenses and ice-water phase changes make the soil sensitive to temperature and intensify this variability phenomenon. However, little is known about the extent to which the thermal properties of frozen soil vary and their statistical characteristics. To address this lack of knowledge, we conducted large-scale tests on a clay from the Qinghai-Tibet Plateau (>= 50 samples) and measured the thermal conductivities and heat capacities at five temperatures ( - 10 degrees C, - 5 degrees C, - 2 degrees C, -1 degrees C and 20 degrees C). The measured data show that the two thermal parameters of the fifty clay samples exhibit obvious variability at each soil temperature. In addition, we selected four potential probability distributions to compare with the measured curves and found that the normal distribution, lognormal distribution and gamma distribution agreed well with the measured data and all satisfied the hypothesis testing. Based on goodness-of-fit analyses and a comparison with the normal information diffusion method (NIDM), we proposed that the normal distribution can best describe the thermal parameters of the frozen clay. This study can contribute to a better understanding of the variability of frozen soil properties. Furthermore, the results and conclusions of this study provide a theoretical basis and reference for heat calculations of frozen soil.