利用地形和遥感数据预测土壤养分空间分布  

Prediction of spatial distribution of soil nutrients using terrain attributes and remote sensing data

作　　者：张素梅[1,2] 王宗明[1] 张柏[1] 宋开山[1] 刘殿伟[1] 李方[1] 任春颖[1] 黄健[3] 张惠琳[3] 

机构地区：[1]中国科学院东北地理与农业生态研究所,长春130012 [2]中国科学院研究生院,北京100039 [3]吉林省土壤肥料总站,长春130015

出　　处：《农业工程学报》2010年第5期188-194,共7页Transactions of the Chinese Society of Agricultural Engineering

基　　金：国家重点基础研究发展计划(973计划)项目课题(2009CB421103);国家自然科学基金项目(40871187)

摘　　要：在GIS支持下,选择地形因子和遥感植被指数,建立土壤养分空间分布预测模型,应用回归克里格(Kriging)方法,预测吉林省农安县土壤养分(有机质和全氮)的空间分布。结果表明,11个环境因子中,相对高程、坡度、地形起伏度、坡度变率、归一化植被指数(NDVI)与土壤有机质和全氮含量均具有显著的相关性。地面粗糙度和地形湿度指数与有机质具有显著相关性,而与全氮的相关性不显著。相对高程、坡度、地面粗糙度、河流动能指数以及NDVI在土壤养分的多元回归预测模型中贡献较大,是预测土壤养分空间分布的最优因子。有机质和全氮在研究区的空间分布格局呈现由东南向西北逐渐减少的趋势,这种分布格局受地形和植被的综合作用,同时与土壤类型密不可分。精度检验结果表明,回归克里格方法能够提高土壤养分空间分布预测精度,是一种有效的空间分布插值方法。The distribution of the soil organic matter and total nitrogen can provide reliable and useful information for sustainable land management and land use planning.In this study,regression Kriging with environmental predictors was used to predict the spatial distribution of soil nutrients（organic matter and total nitrogen）in Nong＇an County,Jilin Province,Northeast China,considering the disadvantages of conventional geo-statistic methods.Terrain and vegetation indicators were chosen for regression Kriging including ten terrain attributes and one vegetation index.The results indicated that relative elevation（Hr）,gradient（β）,roughness of terrain（QFD）,rate of gradient（SOS）and NDVI had significant correlations with soil organic matter and total nitrogen.M andΨhad higher significant correlation with soil organic matter than those with total nitrogen.Relative elevation（Hr）,gradient（β）,surface roughness（M）,river dynamic index（Ω）and NDVI were the best predictors for describing soil nutrients in the study area for they described the regression equations most.In Nong＇an County,the soil organic matter and total nitrogen distributed regularly from southeast to northwest,and the values were higher in the part of southeast.This distribution pattern was affected by terrain and vegetation factors synthetically,and it had a significant relationship with soil type.Precision assessment results showed that regression Kriging improved the accuracy significantly and it could be an effective method for evaluating the spatial distribution of soil properties.

关 键 词：遥感 土壤 养分 空间变量测量 GIS 地形特征 回归Kriging 地统计学 

分 类 号：S158.2[农业科学—土壤学]