黄土丘陵沟壑区坡面尺度土壤水分空间变异及影响因子

Spatial patterns of soil moisture at transect scale in the Loess Plateau of China

土壤水分空间分布特征及其影响因子是土壤前期含水量模拟和小流域产流机制研究的重要内容,也是半干旱地区进行生态建设的重要参考。通过对黄土高原典型坡面雨季前后100 cm深度内土壤含水量进行观测,分析地形、植被和雨季对土壤水分空间分布的影响。基本统计分析显示,土壤水分的空间异质性在上层(<20 cm)较小,在下层(>40 cm)较大。坡面尺度上,土壤含水量的空间差异主要表现在不同植被类型之间,而不是坡位之间。各覆被类型的土壤含水量相对大小为荒草地>8年生刺槐林>20年生刺槐林>沙棘林。即使沙棘林和刺槐林位于更利于获取土壤水分的地形条件下,其土壤含水量仍然明显低于荒草地。地形对土壤水分的影响被植被类型的影响所掩盖。上述规律在雨季前后都有明显表现。因此,完全基于地形指数的土壤水分预测模型在黄土高原应该慎用,植被类型应该作为土壤水分空间预测的一个重要参数。雨季使土壤含水量整体提高,但是土壤水分空间分布格局并没有根本改变,高处仍高,低处仍低,各样点处的土壤含水量在雨季前后达到显著相关水平,说明土壤水分空间格局并不是瞬时状态,而具有明显的时间稳定性。

Understanding the spatial variability and impact factors of soil moisture is important when modeling the antecedent soil moisture status of a catchment. Soil moisture is a limiting factor of ecosystem development in semi-arid and arid areas. Soil moisture spatial patterns and impact factors are particularly important for ecosystem restoration in the semi- arid areas of the Loess Plateau. This study identifies soil moisture spatial patterns and impact factors （terrain, land use types and wet season） in Yangjuangou catchment in north of the Loess Plateau from the measurement of soil moisture in the top 100 cm soil layer along five different transects on a loess hill before and after the wet season. Statistical analysis revealed that the soil moisture spatial variation in deeper soil （〉40cm） was more significant than that in upper soil （〈20cm）, despite the influence of wet season. On a typical slope which had natural grassland at upper position and planted R. pseudoacacia forest land at lower position, the average soil moisture in planted forest land was 3.8% lower than that in grassland in the 20cm tol00cm soil depth, while 5% in the 100cm depth. Comparatively, in the scope of same kind of land use type, the difference of soil moisture between different slope positions was less than 2%, which was obvious smaller than the difference between the land use types. These laws kept consistent before and after the wet season. Moreover, the soil moisture content in different land use types was ranked as natural grassland 〉 planted R. pseudoacacia forest land （8 years） 〉 planted R. pseudoacacia forest land （20 years ） 〉 planted Hippophae reamnoides forest land, no matter before or after the wet season. Although the topography of planted forest land has advantageous conditions （ shading slope, lower position, smaller gradient or terraces） for retaining soil moisture, the soil in planted land was still found to be have a much lower soil moisture content than the soil in natural grasslands, especially in deeper soil. Thus it is considered that the influence of terrain on soil moisture content is masked by the influence of surface vegetation. Therefore, when modeling hydrological processes in the Loess Plateau, it is recommended that the pattern of vegetation types within the simulated area should be paid special attention. Hydrologic models of the Loess Plateau based only on terrain properties should be used cautiously. The soil moisture content at all sample sites increased after the wet season, however, the spatial patterns of soil moisture was not significantly changed. Sites which had high soil moisture content relative to other sites before the wet season retained their higher soil moisture content after the wet season. This was also the case for sites with low soil moisture content. Significant correlation was found between the soil moisture before and after the wet season at each site. Soil moisture spatial patterns displayed temporal stability despite the influence of the wet season. This suggests that soil moisture spatial patterns remain consistent at certain time scales. The knowledge gained from this study is an important addition to soil moisture spatial pattern research.