在实验室条件下,以地物光谱仪(350~2500 nm)为工具,利用光谱反射率(R)、反射率的一阶微分(R′)、反射率的对数(logR)、反射率的倒数(1/R)、反射率对数的一阶微分(logR)′和去包络线(Rc)6种处理方法,研究了4种不同土种的砂姜黑土含水量与光谱反射率的定量关系.结果表明:砂姜黑土不同土种间光谱特征存在一定差异,且4种土的光谱反射率都随土壤含水量升高而下降.在可见-近红外波段,以R′ 提取的敏感波长与含水量的相关系数最高,敏感波长集中分布在712、807、1142、1570、1850、2221 nm等.基于上述波长反射率所建立的土壤含水量预测模型的决定系数达到了0.89,且其模型检验的均方根误差仅为4.6 g/kg,实现了对砂姜黑土含水量较为准确的预测.本研究为非接触、快速地测定砂姜黑土含水量提供了数据基础.%In order to study spectral features of 4 soil species of Shajiang black soil under different water contents, samples were determined by ASD spectrometer in the lab. And the correlation between spectral reflectance and water content was discussed. To improve the accuracy of the model, these models were established with 6 categories of transformations of reflectance using two kinds of methods, such as multiple linear regression and nonlinear regression. The result showed that the correlation coefficient between first derivative differential of reflectance (R′) and water content was the highest and its relationship was negative. At the bands of 712, 807, 1142, 1570, 1850 and 2221 nm , the model established byR′ was the fittest.
展开▼
