基于田间原位土壤含水量估测的可见/近红外光谱建模方法

为实时、准确地获取原位土壤含水量信息,利用可见/近红外光谱技术,分别使用全局偏最小二乘(PLS)建模、局部PLS建模方法,对田间原位土壤含水量进行快速估测。结果表明:全局PLS模型中,其建模集的决定系数(R~2)、交叉验证均方根误差(RMSECV)分别为0.943和1.750%,检验集的决定系数(R~2)、预测均方根误差(RMSEP)分别为0.956和1.260%。局部PLS模型中,分别比较了选取定标子集的2种方法(欧氏距离法和马氏距离法),采用欧氏距离法和马氏距离法选取定标子集进行建模的R~2值分别为0.974和0.979,RMSEP值分别为0.976%和0.943%。因此,将可见/近红外光谱技术应用到田间原位含水量测量是可行的,其中,使用局部建模方法的效果优于全局建模。

Estimation of soil water content in situ by using visible/near infrared spectrum modeling

In order to collect the data of soil water content in situ real-timely and accurately, this study used visible/near infrared (Vis/NIR) spectroscopy technology to estimate soil water content in situ rapidly, by global partial least squares (PLS) modeling and local PLS modeling.The results showed that:For global PLS modeling, the decision coefficient (R²) and root mean square error of cross validation (RMSECV) of the modeling set were 0.943 and 1.750%, respectively.The decision coefficient (R²) and root mean square error of prediction (RMSEP) set were 0.956 and 1.260%, respectively.For local PLS modeling, two methods (i.e., Euclidean distance method and Mahalanobis distance method) for subset selection were used.The R² of the two methods were 0.974 and 0.979, respectively.RMSEP were 0.976% and 0.943% respectively.This study suggests that it is feasible to measure soil water content in situ by using the Vis/NIR spectroscopy technology.And the result of local modeling was better than that of global modeling.