利用傅里叶变换红外光谱仪对绿洲盐渍化土壤进行野外测量, 采用光谱平滑迭代法对温度和发射率进行分离, 得到了盐渍化土壤的热红外发射率数据。 通过对盐渍化土壤发射率光谱的特征分析, 得出8～13 μm土壤发射率随盐分含量的增加而减小, 发射率光谱对盐分因子的响应在8～9.5 μm较敏感。 分析了原始发射率光谱、 一阶导数、 二阶导数和标准化比值与含盐量之间的相关性, 表明土壤发射率与含盐量呈负相关关系, 发射率一阶导数与含盐量的相关性最高, 相关系数最大为0.724 2, 对应波段为8.370 745～8.390 880 μm。 建立了土壤发射率一阶导数与盐分含量的二次函数回归模型, 模型拟合的决定系数为0.741 4, 验证结果的均方根误差为0.235 5, 说明利用热红外发射率光谱反演土壤盐分含量的方法可行。

Studying of soil salinization is of great significance for agricultural production in arid area oasis, thermal infrared remote sensing technology provides a new technology and method in this field. Authors used Fourier transform infrared spectrometer to measure the oasis saline soil in field, employed iterative spectrally smooth temperature/emissivity separation algorithm(ISSTES) to separate temperature and emissivity, and acquired the thermal infrared emissivity data of the saline soil. Through researching the emissivity spectral feature of saline soil, and concluded that soil emissivity will reduce with the increasing of salt content from 8 to 13 μm, so emissivity spectra is more sensitive to salt factor from 8 to 9.5 μm. Then, analyzed the correlation between original emissivity spectra and its first derivative, second derivative and normalized ratio with salt content, the result showed that they have a negative correlation relationship between soil emissivity and salt content, and the correlation between emissivity first derivative and salt content is highest, reach to 0.724 2, the corresponding bands are from 8.370 745～8.390 880 μm. Finally, established the quadratic function regression model, its determination coefficient is 0.741 4, and root mean square error is 0.235 5, the result explained that the approach of using thermal infrared emissivity to retrieve the salt content of saline soil is feasible.