土壤修复过程中盐含量及其光谱特征分析研究

发布时间：2018-02-12 19:38

本文关键词： 盐渍化土壤微生物修复光谱变换偏最小二乘法　出处：《光谱学与光谱分析》2017年05期 　论文类型：期刊论文

【摘要】：基于盐渍土修复过程中盐分含量和同步实测光谱数据,通过对原始光谱数据、平滑光谱数据及平滑后的不同变换光谱数据等八种光谱数据集,分别以相关系数的极值和不同相关系数范围两种方法分析其最佳敏感波段范围,深入分析了不同变换下土壤的光谱响应特征。在此基础上,运用偏最小二乘回归方法,以全波段(400~1 650nm)和分析获得的最佳敏感波段建立了基于修复过程的土壤盐含量和光谱反射率的关系模型。结果表明:针对八种光谱数据集,采用两种方法提取的土壤最佳敏感波段,均集中在947.11~949.31,1 340.27,1 394.11,1 419,1 457.81~1 461.31,1 537.68~1 551.39和1 602.32nm;且最佳波段的土壤盐含量反演模型,以模型评价参数的决定系数(R2)和均方根误差(RMSE),以及赤池信息量准则(akaike’s information criterion,AIC)作为选择最佳模型的标准,均以SGSD(Log R)模型的建模和预测结果比其他光谱变换的模型更为显著。基于全波段的PLSR建模效果总体上稍优于最佳波段的模型,其中以SGSD的预测精度最为突出,其模型的决定系数R2与标准差RMSEP分别为0.673和1.256;基于两种方法获得的最佳波段的PLSR模型与全波段对比在模型精度方面虽有一定差距,但从模型的复杂程度比较,具有模型简单、变量更少及运算量小的特点。该研究可在土壤盐含量及其光谱特征的研究中,为实现土壤盐渍化定量、快速、便捷的监测和检测提供参考。
[Abstract]:Based on the salt content and synchronous measured spectral data in the process of saline soil remediation, eight spectral data sets, such as original spectral data, smoothing spectral data and different transformed spectral data after smoothing, are analyzed. The optimum sensitive band range is analyzed by the extreme value of correlation coefficient and the range of correlation coefficient, and the spectral response characteristics of soil under different transformation are analyzed in depth. On this basis, partial least square regression method is used. The relationship model of soil salt content and spectral reflectance based on remediation process was established by using the whole wave band (1 650 nm) and the best sensitive band. The results show that: for the eight kinds of spectral data sets, the relationship between soil salt content and spectral reflectivity is obtained. The best sensitive bands extracted by the two methods are concentrated at 1 340.27 ~ 1 394.11 ~ 1 4191.457.81 ~ 1 ~ 1 461.68 ~ 1 537.68 ~ 1 551.39 and 1 602.32 nm, respectively, and the inversion model of soil salt content in the optimum band is obtained. The determination coefficient of model evaluation parameters (R2) and root mean square error (RMSE), as well as the red pool information quantity criterion (AICs information criteria) are used as the criteria for selecting the best model. The modeling and prediction results of SGSD(Log R) model are more remarkable than those of other spectral transformation models. The modeling effect of PLSR based on the whole band is slightly better than that of the best band model, and the prediction accuracy of SGSD is the most prominent. The determination coefficient R2 and standard deviation RMSEP of the model are 0.673 and 1.256, respectively. Although there are some differences in the accuracy of the model between the best band PLSR model and the full-band model, the model is simple compared with the complexity of the model. This study can be used as a reference for quantitative, rapid and convenient monitoring and detection of soil salinization in the study of soil salt content and spectral characteristics.
【作者单位】：上海交通大学农业与生物学院低碳农业研究中心;农业部都市农业(南方)重点实验室;上海交通大学船舶海洋与建筑工程学院;
【基金】：高分国土资源遥感应用示范系统(一期)项目(04-Y30B01-9001-12/15) 国家自然科学基金项目(41471120) 社科重大项目(14ZDB139) 上海交大农工交叉项目(Agri-X2015004)资助
【分类号】：S156.4

【相似文献】