论文部分内容阅读
利用傅立叶热红外光谱仪探测煤田土壤的光谱数据和土壤有机质含量数据,分析了原始光谱发射率的特征,求算原始发射率的一阶、二阶导数,分别建立原始发射率、一阶导数、二阶导数与有机质之间的模型,检验和评价了每个模型的适用性。结果表明:1土壤的有机质含量随着地表地物类型和周边环境的不同而存在明显差异;2无论有机质含量存在多大的差异,热红外光谱发射率随波长的增加而变化的趋势不变;但土壤热红外光谱的发射率在8~11.5μm范围内最敏感,有机质含量的大小会对其敏感性产生影响;3一阶导数、二阶导数的发射率与有机质的相关性明显优于原始发射率数据,发射率数据的导数处理能够有效的增强与有机质之间的相关性,且在该研究区发射率数据的二阶导数与有机质含量相关性最高;4一阶导数和二阶导数的热红外光谱发射率与有机质含量之间的拟合函数效果均较为良好,尤其二阶导数发射率的指数函数效果最佳;一阶导数的多元逐步回归拟合模型预测能力略优于线性函数模型。
The characteristics of the original spectral emissivity were analyzed by using the Fourier thermal infrared spectroscopy to detect the spectral data of the soil and the data of the soil organic matter content. The first and second derivative of the original emissivity were calculated. The original emissivity, first derivative, The model between the order derivatives and organic matter tests and assesses the applicability of each model. The results showed that: (1) The content of organic matter in soil varied significantly with the types of surface features and the surrounding environment. 2 The tendency of the infrared emissivity changing with the increase of the wavelength did not change, no matter how much the content of organic matter was different. The emissivity of soil thermal infrared spectrum is the most sensitive in the range of 8 ~ 11.5μm, and the content of organic matter will affect its sensitivity. The correlation between the emissivity of the first derivative and the second derivative is much better than the original emission The derivative of rate data and emissivity data can effectively enhance the correlation with organic matter, and the second derivative of emissivity data has the highest correlation with organic matter content in the study area. 4 The first derivative and the second derivative heat The fitting function between infrared emissivity and organic matter content is good, especially the exponential function of the second derivative emissivity is the best. The predictive ability of the multiple derivative regression model of the first derivative is better than the linear function model.