中文摘要：

本文利用多尺度采样数据,探索了两个深度土壤盐分的空间分异,分析了不同尺度、深度土壤盐分的变异系数和空间相关性（结构方差与基台值之比）的变化,揭示了形成这种空间变异的地貌因素,最后利用普通克立格（ordinary kriging）方法对土壤盐分的分布进行了估测.分析发现,研究区土壤盐分的空间变异具有三个尺度.随着采样间隔的增加和区域的扩大,盐分分布空间相关性增强,且下层比上层具有更高的空间相关性.地貌因素（微地貌类型、坡度和高程）均具有较高的空间相关性,当与地貌因素关系密切时,该尺度及深度的土壤盐分空间相关性就大;反之,则小,这时可能主要受具有较小空间相关性其他因子的影响（如人为活动）.最后对合理的土壤采样提出了建议.

英文摘要：

The Yellow River Delta（YRD） is rich in land resource which has a great potential of development. However, soil salinization is serious in the coastal salinized land which accounts for 63.6% of the region＇s total land area. Soil salinization has become the main factors restricting the regional economic development. In order to utilize the salinized soil resources and realize the sustainable development of agriculture in the YRD, it is important to clarify the distribution of salinized soil and the reasons for the formation and development of salinization. Based on the multi-scale samples, this paper determined the spatial variation of soil salinity at two depths using both statistical and geostatistical methods, comparing the correlation coefficients between soil salinity and three geomorphic factors （e. g. micro-geomorphologic type, slope gradient and elevation）, and finally generating the distribution maps of soil salinity using ordinary kriging at multi-scales and two depths. The analysis revealed that there is higher soil salt content in the embanked former back swamps, salt marshes and tidal flats than in the contemporary flood plains, abandoned river channels and isolated depressions. The variations of soil salinity have three scales. At the small scale （sampling intervals are 0.04-0.2km）, the soil has a higher value of salinity in the north east and the south west. At medium scale （1km）, it is high in F1 in the north west, while in F2 it is higher in the east than in the west. At large scale （6km）, the distribution of soil salinity is characterized by a low value in the central part, where the Yellow River flows through, but it rises gradually toward north and southeast in the upper and lower layers. By analyzing the parameter of spatial dependence （the ratio of spatial variance and sill） of soil salinity in the multi-scales, we found the spatial structure of soil salinity became obvious with the increase of sampling intervals and expanded areas. The paper demonstrates