沿海滩涂土壤盐分空间分布的三维随机模拟与不确定性评价

Three-dimensional stochastic simulation and uncertainty assessment on spatial distribution of soil salinity in coastal region

为揭示滨海滩涂地区土壤盐分三维空间分布特点并提供相关技术方法与思路,以苏北海涂围垦区典型地块为例,综合采用三维克里格和随机模拟方法对土体盐分含量的三维空间分布进行估值、模拟与对比分析,并对土体盐分三维分布的空间不确定性进行评价。结果显示,由克里格法得到的土壤盐分空间分布具有明显平滑效应,减小了数据间的空间差异并改变了数据的空间结构;序贯高斯模拟结果整体分布相对离散,突出了原始数据分布的波动性;研究区土壤盐分随深度增加而升高,存在一定次生盐渍化风险;围垦后研究区土壤盐渍化的发生概率已有所降低,轻度盐化土和中度盐化土的高概率区是改良利用的重点区域。该研究表明随机模拟方法能真实地反映土壤属性的三维空间变异特点,其结果亦为围垦区盐渍障碍耕地的治理利用提供了技术对策。

In order to illustrate the three-dimensional spatial pattern of soil salinity in coastal zones and provide relevant practical method and technical route,ordinary kriging and stochastic simulation method were applied to the estimation,simulation,comparison and uncertainty analysis of the three-dimensional spatial distribution of coastal soil salinity. The study was performed in typical farmlands of coastal reclamation zone in north Jiangsu Province,China. The results indicated that the spatial distribution of soil salinity generated by ordinary kriging was continuous and smooth,and the spatial variability of the kriging salinity data was reduced with the spatial structure changed,while the spatial distribution of soil salinity generated by the SGS（sequential gaussian simulation） was discrete and fluctuant. Soil salinity increased with the depth in soil solum across the study area and the risk of secondary-salinization was observed. The probability of soil salinization risk decreased since reclamation,and high probability region of slightly and medium salinized soil was the main area for amelioration and utilization. The study showed that stochastic simulation method revealed the three-dimensional spatial variability of soil attributes more truly than ordinary kriging method,and the research results can provide countermeasures to the management and utilization of salt-affected land in coastal reclamation zone.