环渤海低平原区土壤盐渍化风险的多元指示克立格评价

Risk assessment of soil salinity by Multiple-variable Indicator Kriging in the low plain around the Bohai Sea

为定量评价制约环渤海低平原水土资源合理高效利用的盐渍化问题，综合运用非参数地质统计学的单元和多元指示克立格法（Indicator Kriging Method）对该区表层土壤（0～20cm）全盐量、地下水位埋深和矿化度进行空间变异性分析，并给出满足一定阈值条件的概率分布图。结果表明，采用指示克立格方法能够获得较为稳健的变异函数并可通过球状、指数模型进行拟合。环渤海低平原地下水矿化度大于2g／L、水位埋深小于3m以及土壤盐分含量大于lg／kg的概率分布呈一致性，自内陆平原廊坊-衡水-邢台-邯郸一线向东部滨海平原天津-沧州-东营-滨州一线依次增大，在东营和滨州的黄河三角洲地区尤为明显。盐渍化与次生盐渍化发生概率从内陆到滨海增大，土壤盐渍化的高概率区集中在黄河三角洲和天津、德州的小部分区域，其中受地下水位埋深影响下的盐渍化与次生盐化发生概率偏大。比较分析发现，在信息描述上多元指示克立格法更加有效。

For quantitatively assessing the salinity problem of the low plain around the Bohai Sea, the Univariate Indicator Kriging and Multiple-variable Indicator Kriging of nonparametric geostatistical methods were combined with GIS technique to analyze the spatial variability and frequency distribution of soil salinity （Salt）, groundwater level （GWL） and total dissolved solid （TDS） in this district, and to delineate and estimate the risk probability. The results show that the Indicator Kriging Method can provide robust variogram and be fitted by spherical and exponential models. The Frequency distribution in the ranges of TDS〉 2g/L, GWL〈3m, and Salt〉1g/kg were almost the same with the tendency of increasing from inland plain （Langfang-Hengshui-Xingtai-Handan） to the east coastal plain （Tianjin-Cangzhou-Dongying-Binzhou）, espe- cially in the Yellow River Delta. Soil salinization and secondary salinization increased from inland to shore- land, and salinization probability of Dongying, Binzhou, Tianjin and Dezhou are higher. The GWL is a more active factor of all the salinization events. By comparing with Univariate Indicator Kriging, the analysis result of Multiple-variable Indicator Kriging is more efficient and reliable.