滏阳河两岸农田土壤Fe、Zn、Se元素光谱响应研究

The Study of Spectral Responses of Agricultural Soils for Fe,Zn and Se in the Area of Fuyang Riversides

为了探索遥感技术快速定量化监测土壤元素含量的可行性,本文通过对滏阳河两岸农田51个土壤表层样品的室内光谱反射率及其Fe、Zn、Se含量关系的研究,探索了反射光谱快速预测土壤元素含量的技术途径。结果发现预测Fe的最佳光谱间隔为16nm,Zn和Se的为8nm,这说明在使用经验方法预测没有光谱特征的成分时,光谱分辨率不是一个必要条件;土壤中的Fe、Zn、Se元素与土壤的反射光谱存在较好的相关性,各元素含量与土壤平均反射率负复相关系数(R2)均可达到0.49以上,而与相应TM各波段的平均光谱反射率也都具有较好的负相关关系,与TM7波段的复相关系数最大,Fe、Zn为0.58,Se元素为0.55。本研究结果为今后利用高光谱遥感技术定量监测土壤Fe、Zn、Se元素含量提供了一种新的方法和技术途径,对土地质量变化的快速定量监测具有重要的科学意义和应用前景。

In order to explore the feasibility of remote sensing technology for investigating important elements in soils quickly, the relationships between the element concentrations and reflectance spectra of soils were researched in this paper. Fifty one soil samples were collected for chemical analyses and spectral measurements along both sides of Fuyang river. The results showed that Fe, Zn, and Se could be predicted using reflectance spectroscopy. The best spectral interval for predicting Fe is 168 nm, and 8 nm for both Zn and Se, which indicates that spectral resolution is not necessarily a limitation for predicting soil Fe, Zn and Se. The correlation coefficients (R2) between elements and the mean reflectance are all above 0.49. Moreover, the correlation coefficients between elements and the mean reflectance of the simulated TM bands are all significant. TM 7 has the highest correlation coefficients with all the three elements, i.e, the correlation coefficients for both Fe and Zn are 0.58 and for Se is 0.55. The results of this study provide a novel method for rapid mapping the concentrations of Fe, Zn, and Se in soil using the remote sensing technology. In addition, it also has important scientific significance for investigating the soil quality rapidly and quickly. Future study with real remote sensing data and field measurements is strongly recommended.