摘要
森林土壤有机碳含量是表征林地土壤营养状况的重要指标,该文建立了土壤有机碳含量的近红外光谱定标模型,并比较了偏最小二乘法（PLS）、支持向量机回归（SVR）、主成分回归（PCR）3种建模方法及Savitzky-Golay平滑＋多元散射校正、Savitzky-Golay平滑＋一阶导数、Savitzky-Golay平滑＋二阶导数、Savitzky-Golay平滑＋多元散射校正＋一阶导数、Savitzky-Golay平滑＋多元散射校正＋二阶导数5种光谱预处理方法对土壤有机碳含量定标模型精度的影响,同时进行了波段优选。结果表明：当光谱区域为1 380～1 450 nm,1 800～1 950 nm,2 050～2 300 nm,光谱数据采用Savitzky-Golay平滑＋多元散射校正＋一阶导数预处理,采用PLS的建模方法,主成分数为8时,建立的校正模型预测效果最佳。校正模型的R、RMSE、SEC分别为0.805 2、0.512 2、0.512 5;预测模型的R、RMSE、SEP分别为0.768 1、0.514 3、0.514 6。因此,利用近红外光谱技术可以实现土壤有机碳含量的快速估测,为林区实时、大面积、快速测定森林土壤有机碳含量提供了技术可行性。
The forestry soil organic carbon（ SOC） content is one of the main chemical components of soil and has a critical effect on soil properties and utilization. The carbon contents of 120 soil samples were determined with national standard of China,and then the near infrared（ NIR） spectroscopy of all samples was collected by LabSpec Pro. The calibration model was built by using partial least squares（ PLS）,support vector regression（ SVR） and principal component regression（ PCR） with different pretreatment methods of un-pretreatment,Savitzky-Golay（ SG） ＋ Multiplicative Scatter Correction（ MSC）,SG ＋ the first derivative,SG ＋ the second derivative,SG ＋ MSC ＋ the first derivative and SG ＋ MSC ＋ the second derivative in different spectral region of 350-2 500 nm and optimal spectral region of 1 380-1 450 nm,1 800-1 950 nm and 2 050-2 300 nm. The results showed that the best model was built by PLS with pretreatment spectral data of SG ＋ MSC ＋ the first derivative and 8 principal components in optimal spectral region. Concerning the prediction accuracy,the correlation coefficient（ R） and root mean square error（ RMSE） and the standard error of calibration and prediction model were 0. 805 2,0. 512 2,0. 512 5 and 0. 768 1,0. 514 3,0. 514 6. Thus,application of NIR spectroscopy technology can achieve rapid prediction of SOC content in real-time and large area in forest.
出处
《安徽农业科学》
CAS
2014年第15期4702-4706,4742,共6页
Journal of Anhui Agricultural Sciences
基金
中央高校基本科研业务费专项(DL12EB07-2)
黑龙江省自然科学基金项目(C201111)
