地学环境变量支持的土壤全钾含量自适应曲面建模——以青海湖流域典型地区为例

Adaptive Surface Modeling of Soil Total Potassium Content Supported by the Environment Variables——A Case Study in Typical Areas of Qinghai Lake Basin

土壤属性空间分布受地学环境要素影响,空间分异特征十分明显,单一的全局插值模型在应用中常受到一定条件的限制。对复杂地貌类型区土壤属性插值所面临的空间不连续、全局插值模型精度有限以及适应性差的缺点,提出了一种融合地学环境信息的土壤属性自适应曲面建模方法(Adaptive surface modeling for soil properties,ASM-SP)。利用2013年采集的110个样点数据,以土壤全钾含量为例,利用ASM-SP、普通克里格法(Ordinary Kriging,OK)、回归克里格法(Regression Kriging,RK)、地理加权回归克里格法(Geographically Weighted Regression Kriging,GWRK)和协同克里格(Ordinary Co-Kriging,OCK)5种插值方法,分别模拟了青海湖流域典型地区土壤全钾含量的空间分布。采用平均误差(Mean Error,ME)、平均相对误差(Mean Relative Error,MRE)、均方根误差(Root Mean Square Error,RMSE)、准确度(Accuracy,AC)、相关系数、回归系数和决定系数7类指标系统评价不同插值方法的预测效果。结果表明:(1)利用常规插值(OK)得到的插值曲面较为平滑,具有弱"牛眼"效应,在刻画土壤全钾含量的空间变异性方面存在明显不足,精度有待提高。(2)在融合地学环境信息的插值方法中,RK,OCK,GWRK和ASM-SP模拟精度较OK有不同程度提高,其中ASM-SP在刻画土壤全钾含量的空间变异和局部细节信息方面表现突出,精度较其他插值方法有较大程度提高,其准确度较OK,RK,GWRK和OCK分别提高9.27%,6.29%,2.66%和7.74%。ASM-SP尤其适合复杂地貌类型区,因其考虑了地学环境变量与土壤属性的非线性关系,并融合了多个模型的适应性优势,其在刻画土壤属性空间分异的复杂性方面也更加符合实际情况,为土壤属性的空间模拟提供了新思路。

The single global interpolation model is limited because of the complexity of spatial distribution of soil properties.The spatial discontinuity often causes a poor accuracy when the single model is used for surface modeling of soil properties faces in complex geomorphic area.This paper presented an adaptive method for surface modeling of soil properties supported by the environment variables(ASM-SP).Four methods,including Ordinary Kriging method(OK),Regression Kriging method(RK),Geographically Weighted Regression Kriging method(GWRK),Ordinary Co-Kriging method(OCK),were used to validate the proposed method.We selected the Qinghai Lake Basin,a typical complex geomorphic area,as the study area.Based on the 110 topsoil samples collected in 2013,the five methods were used to predict the spatial distribution of soil total potassium content,respectively.The mean error(ME),mean relative error(MRE),root mean square error(RMSE),accuracy(AC),correlation coefficient,regression coefficient and adjust coefficient were served as the evaluation indicators.The results show that:(1)the OK interpolation result is spatially smooth and has a weak bull′s-eye effect,it is an obvious deficiency in depicting spatial variability of soil total potassium content,and the accuracy also needs to be improved;(2)ASM-SP is the best method for predicting soil total potassium content with higher accuracy in this study.The result presents more details than other in the abrupt boundary,which can make the result consistent with the true geo-environmental variables.It not only considers the nonlinear relationship between geo-environment variables and soil properties,but also adaptively combines the advantages of multiple models.Compared with OK,RK,GWRK and OCK,the accuracy of ASM-SP increased by 9.27%,6.29%,2.66%and 7.74%,respectively.Therefore,the proposed method could provide significant reference for enriching the surface modelling theory and techniques.