耕地土壤速效钾含量的空间预测方法研究  

Spatial Prediction Method of Available Potassium Contents in Arable Soil

作　　者：王彬武[1] 周卫军[1] 马苏[2] 刘少坤[1] 王金国[1] 于良艺[1] 

机构地区：[1]湖南农业大学资源环境学院,湖南长沙410128 [2]湖南师范大学G1S研究中心,湖南长沙410081

出　　处：《地理与地理信息科学》2011年第5期91-94,F0002,F0003,共6页Geography and Geo-Information Science

基　　金：国家科技支撑项目(2009BADC6B005);湖南省研究生科研创新项目(CX2011B310)

摘　　要：利用插值方法进行土壤属性的空间预测是土壤学科的研究热点之一,以土壤发生学相关信息作为辅助变量的克里格空间插值方法则少有研究。该文以成土母质多样的湖南省石门县为例,基于GS+、ArcGIS和Matlab,结合成土母质信息的克里格(PMK)插值方法,对研究区土壤速效钾空间分布进行了预测。结果表明,PMK法较好地解决了因成土母质间速效钾含量差异给预测带来的误差,预测精度较反距离权重(IDW)、普通克里格(OK)法分别提高了21.68%、16.43%;PMK法制作的空间预测图中速效钾分布突变而非渐变,较IDW、OK预测结果接近研究区实际情况,进而证明在成土母质较为复杂的地区,PMK法适合对土壤速效钾含量的预测。There has been a great concern about using interpolation to predict soil properties in the soil science. At present,little information on using Pedology information as auxiliary variables to predict soil properties is available. In this paper, A study was initiated in Shimen County, Hunan Province, which has variety of parent materials, Based on GS＋, ArcGIS and Matlab, using Kriging method combined with parent material information （PMK） to predict the spatial distribution of soil available K. PMK method could reduce the deviation that caused by the differences of AK content among various soil parent materials. Comparing to Inverse Distance Weighted （IDW）, Ordinary Kriging （OK） method, predicting precision was raised 21.68 % and 16.43%, respectively. The polygons of predicting map generated by PMK method distributes are mutational rather than gradual, which shows more closely to the actual situation of the study area. PMK method is more suitable to predict the AK in the region, where has complicated parent materials.

关 键 词：土壤速效钾 克里格插值 Box-Cox变换 成土母质 

分 类 号：S151.9[农业科学—土壤学]