可见-近红外光谱的土壤养分快速检测

发布时间：2018-11-09 14:50

【摘要】：在测定土壤养分中,可见-近红外光谱技术具有很大的应用空间。该研究探讨了基于可见-近红外光谱(250~950nm)离线、快速测定土壤总氮(TN)、总磷(TP)、总钾(TK)、总碳(TC)等土壤养分的方法及应用。采集青岛三个不同地区土壤样品(异质性较高的山地土壤与河畔土壤)各60份,总计180份,并测定其TN,TP,TK,TC含量及其可见-近红外反射光谱,利用Kennard-Stone法按2∶1比例划分校正集和检验集,采用遗传算法分别提取TN,TP,TK,TC特征波长,以偏最小二乘法建立定量分析模型。TN,TP,TK,TC含量所建光谱模型的相关系数分别为0.970,0.964,0.680和0.967,检验集的相关系数分别为0.980,0.937,0.717和0.972,检验集的RPD值分别为4.570,2.424,1.411和4.135。结果表明,该方法能够对土壤TN,TP,TC含量进行精确预测,对土壤TK含量进行粗略预测。该研究主要依靠可见光波段,较好的预测了异质性较高的土壤的氮磷钾等养分含量,有望降低未来土壤养分速测的成本。此外,该研究还提供了青岛土壤养分的光谱库,为我国土壤大数据库的建立提供技术支撑。
[Abstract]:In the determination of soil nutrients, visible-near-infrared spectroscopy has great application space. The method and application of rapid determination of soil nutrients, such as total nitrogen, (TN), total phosphorus, total potassium, (TK), total carbon (TC), etc. based on visible near infrared spectroscopy (250~950nm) off-line, were discussed in this study. The contents of TN,TP,TK,TC and its visible near infrared reflectance (NIR) were measured in three different regions of Qingdao, including 60 samples from mountainous soils and 60 samples from riverside soils. The Kennard-Stone method is used to divide the correction set and the test set according to 2:1 scale. Genetic algorithm is used to extract the characteristic wavelength of TN,TP,TK,TC, and the partial least square method is used to establish the quantitative analysis model. TN,TP,TK,. The correlation coefficients of the spectral model based on TC content were 0.970, 0.964, 0.680 and 0.967, respectively. The correlation coefficients of the test set were 0.980, 0.937, 0.717 and 0.972, respectively. The RPD values of the test set were 4.570, 2.4241.411 and 4.135, respectively. The results showed that this method could accurately predict soil TN,TP,TC content and roughly predict soil TK content. This study mainly depends on the visible light band, which can predict the nitrogen, phosphorus, potassium and other nutrients content of the soil with high heterogeneity, which is expected to reduce the cost of rapid determination of soil nutrients in the future. In addition, the study also provides a spectral database of soil nutrients in Qingdao, which provides technical support for the establishment of a large soil database in China.
【作者单位】：山东省科学院海洋仪器仪表研究所;
【基金】：山东省重点研发计划项目(2015GNC110016) 国家自然科学基金项目(31300414) 山东省自然科学基金项目(ZR2014DM012)资助
【分类号】：O657.33;S158

【相似文献】