基于地图直接比较的不同插值方法制图效果及影响因素分析——以新乡县为例

Influence Factors and Effects of Different Interpolation Methods for Cartography Based on Map Direct Comparison——A Case Study of Xinxiang County

利用新乡县2008年测土配方施肥项目2 694个土壤农化样点数据,选取有效磷和缓效钾为研究对象,基于地图直接比较,对比研究了常见插值方法(IDW、Kriging、PRP)的土壤养分制图效果及影响因素,以期为土壤调查和制图实践提供参考。结果表明:不同插值方法制图效果受土壤养分属性空间变异性及样点个数影响。对于土壤空间变异结构性较强的缓效钾,当样点数为300个时,IDW、Kriging法制图结果均已达到较高可信度。对于空间变异随机性较强的有效磷,当样点数介于700~900个时,IDW法制图结果比较可信;当样点数为1 100个时,制图结果达到较高可信度。3种方法中,PRP法制图效果均不理想,且尤其不适用于空间随机性变异的土壤养分。总之,3种插值方法的制图优劣为Kriging>IDW>PRP。

Kriging method,inverse distance weight （IDW） method and polygon value represented by point value （PRP） method were used to map for available phosphorus and slow-release potassium based on map direct comparison with sample data from 2 694 soil samples collected from Xinxiang county of Henan province in 2008,and the influence factors and spatial prediction accuracy of interpolation methods were analyzed. The results showed that the spatial prediction accuracy was influenced by spatial variability of soil properties and the number of samples. For slow-release potassium which spatial variability was caused by structural factors, Kriging and IDW methods had high spatial prediction accuracy when the sample number was 300 or higher. For available phosphorus which spatial variability was caused by random factors, the mapping interpolated by IDW method was relatively reliable when the sample number was 700-900, andmore reliable when the sample number was 1 100 or higher. Compared to IDW and Kriging methods,PRP method was not satisfied for the available phosphorus and slow-release potassium, especially for available phosphorus which spatial variability was caused by random factors. In a word,the spatial prediction accuracy order was Kri^in^IDW^PRP.