中文摘要：

以2 m×2 m的网格布点,采用经典统计学和地统计学结合的方法研究了一次灌溉前后膜下滴灌棉田的表层土壤体积含水量的空间变异性特征。结果表明：（1）灌溉前后表层土壤含水量变异系数（CV）分别为28.7%和21.8%,均表现出中等变异特性,颗粒分布能为灌溉前后表层土壤含水量的变异系数变化提供参考。（2）灌溉前后表层土壤含水量的半方差函数均可用高斯模型很好地描述（R^2〉0.9）,对应的空间异质比（C0/（C0＋C））分别为0.201和0.324,表现出强烈和中等的空间依赖性;灌溉后,表层土壤含水量空间自相关性减小,4.373 m可作为该研究区域的参考取样步长。（3）灌溉前后表层土壤含水量的Moran’s I随滞后距离变化的趋势大体一致,且灌前表层土壤含水量的自相关性较灌后显著。（4）Kriging插值结果表明,灌溉前后含水量空间分布较相似,表现出显著的变异性,且灌后含水量分布较灌前复杂。

英文摘要：

In this research, classical statistics and geostatistics were combined to research the characteristics of spatial variability for surface layer soil water content in a area of 20 m - 20 m cotton field by dripping under mulch before and after the irrigation. The primary results showed that the variation coefficients （CV） for surface layer soil water before and after irrigation were 28.7% and 21.8%, respectively, exhibiting moderate variability. The particle distribution could provide a basis for the variations in CV of water content before and after irrigation. Additionally, the semivariogarm function for surface layer soil water content before and after irrigation could be described well by Gaussian model （R^2）. The corresponding ratios of spatial heterogeneity （ C0/（ C0 ＋ C）） were 0. 201 and 0. 324, respectively, indicating strong and moderate spatial dependences, respectively. After irrigation, the spatial autocorrelation of surface layer soil water content became decreased. The reference sample step lag of study region could be adopted by 4. 373 m. Moreover, Moran＇ s I for surface layer soil moisture before and after irrigation followed similar trends with the lag distance. The autocorrelation of surface layer soil water content before irrigation was more significant than that after irrigation. Lastly, the results of Kriging interpolation showed that spatial distributions for water content before and after irrigation were similar and showed obvious variabilities. Water content after irrigation was more complex than that before irrigation.