论文部分内容阅读
通过对江苏省如皋市888个采样点的不重复随机抽样,探讨了采样点数量对土壤肥力指标空间变异解析的影响。从半方差函数估计的可靠性角度考虑,在长江冲积物形成的土壤上,针对县级农业管理和生态环境规划的土壤肥力指标调查采样,采集250个样点较为合适。另外,仅使用交互验证和独立验证评价半方差函数估计的可靠性及确定采样点数量是不完善的,而平均克里格方差理论上随采样点数量的增加而单调递减,可以作为不同采样点数量条件下,综合评价半方差函数估计可靠性及确定采样点数量的补充指标。
Through the non-repeated random sampling of 888 sampling points in Rugao City, Jiangsu Province, the effect of the number of sampling points on the spatial variability analysis of soil fertility index was discussed. From the perspective of the reliability of the semivariogram estimation, it is appropriate to collect 250 soil samples for soil fertility index survey and sampling at county-level agricultural management and eco-environmental planning in the soil formed by alluvial deposits in the Yangtze River. In addition, it is imperfect to evaluate the reliability of semi-variance function estimation and confirm the number of sampling points using only cross-validation and independent verification. However, the average Kriging variance theoretically monotonically decreases with the increase of sampling points, which can be used as different sampling points Under the quantitative conditions, the reliability of the semi-variance function and the additional indexes to determine the number of sampling points are evaluated.