中文摘要：

西藏地区自然气候特征、成土母质和成土条件有着显著的地区特性,了解西藏高原地区土壤水动力学特性对于研究高原地区水文和水循环过程具有重要的意义。2015年分别在林芝地区米林县（海拔2 905 m）,八一镇（海拔3 240 m）和林芝县（海拔3 879 m）3个青稞种植区田间取样,在原状土土壤水分特征曲线测试的基础上,研究了西藏林芝地区土壤基质势和土壤含水率之间的关系,并建立了参数的微分求解方程,发展了土壤水分特征曲线非线性函数的参数估计方法;基于实测值构建Jacobi矩阵,采用全局性最优方法确定了van Genuchten模型和Gardner-Russo模型的参数。结果表明,-11.0~-4.0、-42.0~-11.0 k Pa以及-54.0~-42.0 k Pa基质势区段的单位基质势平均含水率变化量分别为5.4×10-3、11.2×10-3和6.7×10-3cm~3/（cm~3·k Pa）,最大差异性发生在-54.0~-42.0 k Pa之间,西藏地区的土壤水分特征曲线形状主要决定于黏粒质量分数,而砂粒质量分数则主要影响土壤水分特征曲线的变异性。van Genuchten模型和模拟6组试验数据的相对误差平均值和最大值分别为0.04和0.09。采用Gardner-Russo模型模拟6组试验数据的相对误差的平均值和最大值分别为0.06和0.14。van Genuchten模型模拟西藏林芝地区的土壤水分特征曲线的精度及有效性显著超过Gardner-Russo模型。

英文摘要：

The genesis of soils in Tibetan plateau differs from those in other areas and knowing their hydraulic properties is critical in understandinghydrological cycle in the plateau. In this paper we measured the hydraulic properties of a number of intact soil samples and evaluated how well the van Genuchten model and the Gardner- Russo model can fit these water retention curves. In particular, twenty-four undistributed soil samples were takenfrom hulless barley fields in Milin, Bayi and Nyingchiat altitude of 2 905 m, 3 240 m and 3 879 m, respectively in Linzhi. The results show that within the matric potential of- 11.0 kPa to -4.0 kPa, -42.0 kPa to -11.0 kPa and -54.0 kPa to -42.0 kPa, the average change of water content with matric potential was 5.4×10.3 cm3/（cm3 · kPa）, 11.2×10.3 cm3/（cm3 ·kPa） and 6.7×10.3 cm3/（cm3 · kPa）, respectively. The maximum variance of the measured water retention curves was within the matric potential of-54.0 kPa to -42.0 kPa. The water retention curves of different soils depend on their clay content, while their variability depends on the content of sand. The mean and maximum relative errors of the results predicted by the van Genuchten formula were 0.04 and 0.09 respectively, whereas for the Gardner- Russo formula, these errors were 0.06 and 0.14, respectively. For the soils investigated in this paper, the van Genuchten formula is superior to the Gardner-Russo model.