作　　者： ; ; ;

机构地区： 桂林理工大学理学院

出　　处： 《分析试验室》 2013年第9期11-15,共5页

摘　　要： 研究了傅里叶近红外（FT—NIR）光谱法和Whittaker平滑在土壤有机质、总氮的定量检测中的应用；Whittaker平滑结合偏最小二乘法，分别对全谱、二倍频、一倍频和合频4个区域的光谱数据建立并优选定标模型；对有机质和总氮的成分含量进行定量分析。结果表明，Whittaker平滑在各个区域的分析中均表现出了提高模型预测精度的能力；经过各个区域的建模效果的比较，找到土壤有机质的优选区域是合频（5500—4000em。），土壤总氮的优选区域是一倍频（7000～5500em。）。经过检验，优选的区域能够为土壤的有机质和总氮建立良好的FTr—NIR定标模型。 Organic matter and total nitrogen contents are important indicators to measurement of the fertility of soil. In this paper, the quantitative determination of soil organic matter and total nitrogen by using Fourier transform near-infrared （FT-NIR） and Whittaker smoother were investigated. Combining Whittaker smoother and partial least squares algorithm, the calibration models were established respectively for the 4 regions of full spectrum, second-overtone, first-overtone, and the combination. And the contents of organic matter and total nitrogen of soil were determined. Results showed that Whittaker smoother had the potential to improve the modeling precise for each of the 4 regions. By comparing the results, it was found that the optimal region for organic matter was the combination, and the region for total nitrogen was the first-overtone. In the validation part, the combination and the first-overtone regions were good for FT-NIR analysis of organic matter and total nitrogen in soil.

关 键 词： 土壤 有机质 总氮 平滑 光谱预处理

领　　域： [理学] [理学]