基于电磁感应仪的土壤盐渍化剖面特征解译研究

发布时间：2018-03-11 09:41

本文选题：土壤盐渍化　切入点：空间分布　出处：《土壤通报》2017年03期 　论文类型：期刊论文

【摘要】：为了快速、精准解译区域尺度土壤盐分特征,有必要建立土壤剖面盐分信息精确解译模型。以新疆农灌区不同土壤质地的盐渍化土壤为研究对象,利用电磁感应式大地电导率仪EM38获取土壤表观电导率,构建了基于EM38的两种土壤质地剖面分层盐分解译模型,并对盐分解译回归模型进行了精度检验。结果表明:两种质地的土壤剖面盐分含量变异均较大,中壤土各层电导率变异系数在58%~98%,呈现中等变异强度,而砂土表层(0~40 cm)变异系数达到100%以上,属于强变异强度,深层土壤变异系数介于76%~88%属于中度变异强度。两种质地土壤电导率与磁感表观电导率EMh、EMv间呈显著的相关关系,水平和垂直测量模式都能够对不同深度土盐层分进行预测,且以EMh+EMv为自变量的二元回归解译模型具有较高的精度,相关系数R达到0.94以上。中壤土EM38的盐分预测在不同深度土层的验证结果决定系数达到0.59以上,砂土质地土壤盐分预测验证R2达到0.36以上,预测精度与土壤剖面盐分变异性呈现显著负相关,其相关系数为-0.86,中壤土质地的解译效果优于砂土质地。分析EM38在预测不同土壤质地盐分精度上的差异性,构建了电磁感应式土壤剖面盐分含量的预测模型。研究结果引入土壤质地变量,可为大面积土壤盐渍化的快速精确测定提供理论依据。
[Abstract]:In order to quickly and accurately interpret the characteristics of soil salinity in regional scale, it is necessary to establish an accurate interpretation model of soil profile salinity information. Two layered salt interpretation models of soil texture profile based on EM38 were constructed by using electromagnetic inductive earth conductivity meter (EM38) to obtain soil apparent conductivity. The precision of the regression model of salt interpretation was tested. The results showed that the variation of salt content in soil profile of two textures was greater, and the coefficient of variation of conductivity in each layer of middle loam soil was 58 ~ 98, showing moderate variation intensity. The coefficient of variation of the surface layer of sandy soil is above 100%, which belongs to the strong variation intensity, and the coefficient of variation of the deep soil is between 760.88% belongs to the moderate variation intensity. There is a significant correlation between the conductivity of the two textured soils and the apparent conductivity of the magnetic susceptibility EMH (EMV). Both horizontal and vertical measurement models can predict the soil salt layers at different depths, and the binary regression interpretation model with EMh EMv as independent variable has high accuracy. The correlation coefficient R is more than 0.94. The determining coefficient of salt prediction of EM38 in middle loam soil is more than 0.59 in different depth soil layer, and R2 is more than 0.36 in sandy soil. There was a significant negative correlation between prediction accuracy and salinity variability of soil profile, the correlation coefficient was -0.86, the interpretation effect of middle loam soil texture was better than that of sandy soil texture, and the difference of EM38 in predicting salinity accuracy of different soil texture was analyzed. A prediction model of soil salt content in electromagnetic induction soil profile was established. The results of the study can provide theoretical basis for rapid and accurate determination of soil salinization in large area by introducing soil texture variables.
【作者单位】：新疆石河子大学农学院农业资源与环境系;
【基金】：国际科技合作项目(2015DFA11660) 兵团科技项目(2014AB002) 石河子大学校级项目(RCZX201522)资助
【分类号】：S156.41

【相似文献】