论文部分内容阅读
以新疆艾比湖滨盐渍化土壤为对象,利用磁感应电导仪和光谱仪测得的盐渍土表观电导率和可见光/近红外光谱数据,选取与EM38解译的土壤盐分相关性最好的光谱变换形式和特征波长,分别建立多元逐步回归、偏最小二乘回归和支持向量回归的土壤盐分监测模型。结果表明:(1)表观电导率两种模式相结合建立的盐分含量解译模型的拟合优度达到0.91,即在该区域内电磁感应技术可用于土壤盐分含量的间接监测。(2)一阶微分处理优于二阶微分,经一阶微分变换后的光谱可以较好地预测土壤盐分含量。(3)3种建模方法中,支持向量回归的建模精度最高,偏最小二乘回归和多元逐步回归次之。干旱区湖滨湿地土壤盐分含量的估测模型宜选取基于平滑后的原始一阶微分光谱数据建立的支持向量回归模型。
Taking the salinized soils of Lake Aibi in Xinjiang as an example, the best correlation of soil salinity with EM38 was obtained by using the data of apparent conductivity and visible / near-infrared spectroscopy of saline soils measured by magnetic induction conductivity meter and spectrometer Transform form and characteristic wavelength, respectively, to establish a multiple stepwise regression, partial least-squares regression and support vector regression soil salt monitoring model. The results show that: (1) The goodness of fit of the salt content interpretation model established by the two models of apparent conductivity reaches 0.91, that is, the electromagnetic induction technology can be used for the indirect monitoring of soil salinity content. (2) The first-order differential is better than the second-order differential, and the first-order differential transformed spectrum can predict the soil salt content well. (3) Among the three modeling methods, SVR has the highest modeling accuracy, followed by PLS and multivariate stepwise regression. Arid area lake wetland soil salinity content estimation model should be based on the smoothing of the original first-order differential spectral data to establish the support vector regression model.