中文摘要：

基于漫反射光谱（ DRS）的铁氧化物定量分析已用于北方黄土的研究,而在南方黄土中的适用性尚不清楚.本文选 取江西九江黄土为研究对象,采用柠檬酸钠-重碳酸钠-连二亚硫酸钠（ CBD）方法去除其中的游离铁,以此为基体分别加 人不同比例的赤铁矿和针铁矿,通过多元线性回归方法建立了标样的铁氧化物含量与光谱反射率之间的校准方程.经检 验,校准方程应用于未经CBD处理的剖面天然样品时,获得的赤铁矿含量较准,而针铁矿含量不准.原因可能在于：（1） 混合体系中,赤铁矿与针铁矿互为基体,赤铁矿对针铁矿的干扰远大于针铁矿对赤铁矿的影响,这种差异在南方地区尤为 明显;（2） CBD处理不能完全清除南方黄土中的针铁矿,可能与样品中含有部分结晶较好、粒度较粗的针铁矿有关,这也 增加了针铁矿定量分析的不准确性;（3）铝的同晶替代现象对南方黄土影响更为显著,其中受晶格结构等影响,针铁矿的 铝替代量比赤铁矿更大,可能对定量分析结果的准确性影响也更大.因而在南方黄土中,基于DRS的铁氧化物定量分析方 法对于赤铁矿是适用和有效的,对于针铁矿的应用则需要慎重.

英文摘要：

The diffuse reflectance spectroscopy（ DRS） has been applied successfully in the Chinese Loess Plateau, but it is not clear whether the DRS can be applied in the South China or not. Therefore,we choose the loess located in the South China to study. In order to determine the content of hematite and goethite,we derived regression equations by removing the free iron from the matrix by CBD and adding different proportions of hematite and goethite into it. Thus, the content of the iron oxide was obtained. The results show that the estimate of hematite is consistent with the actual hematite concentration, while the content of goethite fails to be verified. The reasons might be ： （ 1） In the sample, hematite and goethite are mutually matrix to each other. Hametite affects goethite more than goethite affects hematite. The matrix effect,which plays an important role in the regression equations,is noticeable in South China; （2）The CBD cannot remove goethite derived from the rock, which is coarser and more crystalized than the secondary one formed by soil. As a result, the concentration obtained by DRS is different from the actual concentration of goethite mixed; （ 3 ） Because of the hot, humid climate in South China,A1 substitution can take place easily in the soil. The crystal structure of goethite leads to more substituted A1 than hematite,which means that the result might be wrong. To conclude,DRS can be applied to the hematite in South China,but one needs to be cautious when it comes to goethite.