基于耕地图斑数据的土壤有机质估测方法研究——以大兴区南部耕地为例

摘要

In order to explore an optimal estimation method for soil organic matter content based on map patches data,the surface Kriging and sequential Gaussian conditional simulation methods are used to estimate the content of soil organic matter in southern Daxing District of Beijing,and then evaluate the accuracy based on test pattern data and surface data.The results show that the estimation results of soil organic matters in the study area are consistent by surface Kriging method and sequential Gaussian conditional simulation method.However,the sequential Gaussian conditional simulation method possess weaker smoothing effect and prominent local change characteristics at single stochastic simulation;The correlation coefficient and concordance coefficient of point verification and surface verification show that surface Kriging method and sequential Gaussian conditional simulation method can be used to estimate soil organic matter content of cultivated land.Under the comprehensive action of multiple stochastic simulation and variogram fitting accuracy,the estimation precision sequential of Gaussian conditional simulation is higher.%基于测土配方采样点调查数据和土地利用现状数据,构建常规耕地质量评价中的土壤有机质含量的面状数据;基于此,采用面克里格法和序贯高斯条件模拟法估测北京大兴区南部耕地土壤有机质含量,并采用验证图斑的点数据和面数据形式进行精度评价,进而探寻基于面数据的土壤有机质含量最优估测方法.结果表明,面克里格法和序贯高斯条件模拟法对研究区耕地土壤有机质空间分布的估测效果具有一致性,但序贯高斯条件模拟法的平滑作用较弱,且单次随机模拟的局部特征变化更突出;点验证和面验证的相关系数与一致性系数表明,面克里格法和序贯高斯条件模拟法可以用来估测耕地土壤有机质含量;在多次随机模拟和变异函数拟合准确性的综合作用下,序贯高斯条件模拟估测精度更高.