中文摘要：

土壤光谱反射特性的研究是土壤遥感的物理基础，土壤氧化铁含量影响土壤反射率。对采集的174个土样350～2500nm谱段的光谱数据进行分析，着重探讨下述3类光谱变量，寻找对土壤氧化铁含量敏感的光谱特征。1）原始光谱反射率及其各种变换形式；2）基于光谱吸收峰特征的变量，因为已知三价铁在870nm附近有吸收峰存在，提取该吸收峰的面积、宽度、深度等变量；3）基于植被指数的变量，构建土壤氧化铁指数。研究结果表明，各类变量中土壤氧化铁指数对土壤氧化铁含量最敏感，建立相应的回归预测模型，模型拟合度R^2为0．534。

英文摘要：

The study of soil spectral reflectance features is the physical basis for soil remote sensing. Soil Fe2O3 content is one of the main factors which influence the soil spectral reflectance. On the basis of predecessor studies, this article tried to construct different spectral indicators to predict the Fe2O3 content of soil samples. Three kinds of spectral variable were mainly discussed, which are spectral reflectance as well as its various transformations, variables based on spectral absorption fcatures and vegetation index variables. In this paper,one hundred and seventy - four soll samples were collected from the top five centimeters of soil from the study area to represent the range of aspect, slope, and elevation and parent materials within the area. Field spectral measurements were taken with the ASD Pro FR and multivariate statistical analysis was applied. On the basis of data analysis,sane conclusions were drawn. The soil Fe2O3 content has negative oorrelation with soil reflectance. It is the same as SOM which manifest the soil reflectance will decrease with the increasing of soil Fe2O3 content. But the prediction effect of Fe2O3 is not as good as SOM or moisture. The Soil Ferric Oxide Index （SFOI） which consists of the logarithm of average red band and blue band reflectance has good correlation with soil Fe2O3. It is more sensitive than the spectra variable composed by O1DLA（Order 1 Derivative of the Logarithm of Albedo）. SFOI possesses of application FotentM in imaging spectral data interpreting. Correlation analysis is also used between absorption peak around 870 nm and Fe2O3 and the results show that they have no correlation.