基于ArcGIS和Vis-NIR脐橙园土壤养分含量分布图预测研究

发布时间：2018-05-14 07:41

本文选题：ArcGIS + Vis-NIR　；参考：《光谱学与光谱分析》2016年S1期

【摘要】：以赣南区某脐橙园土壤为研究对象,针对不用层土壤的全氮、全磷及有机质养分信息,采用傅里叶型近红外光谱仪和可见/近红外光谱检测装置不同参数配比下组合,采集土壤样品光谱。用化学方法测定养分真值,结合光谱预处理方法和化学计量学算法建立定量检测数学模型,选择出最优的仪器参数、最合适的预处理方法和算法,确定最终的数学模型。实验采样点的数量是有限的,要想得到整个研究区的养分分布数据就需要采用克里金插值法(Kriging)对这些采样点各养分进行插值,运用ArcGIS软件中的地统计分析模块功能,对采样点的土壤全氮、全磷及有机质的真实化学值和最优模型预测值进行克里金插值,得出采样区的土壤养分空间分布图。
[Abstract]:Based on the information of total nitrogen, total phosphorus and organic matter of a navel orange orchard in south Jiangxi, the Fourier near infrared spectrometer and the visible / near infrared spectrometer were used to combine the soil with different parameters. The spectra of soil samples were collected. Using chemical method to determine the true value of nutrient, combined with spectral pretreatment method and chemometrics algorithm to establish a quantitative detection mathematical model, to select the optimal instrument parameters, the most appropriate pretreatment method and algorithm, and to determine the final mathematical model. The number of experimental sampling points is limited. In order to get the nutrient distribution data of the whole research area, we need to use Kriging interpolation method to interpolate the nutrients of these sampling points, and use the function of geostatistical analysis module in ArcGIS software. The true chemical values of soil total nitrogen, total phosphorus and organic matter and the predicted values of the optimal model were interpolated by Kriging interpolation, and the spatial distribution map of soil nutrients in the sampling area was obtained.
【作者单位】：华东交通大学机电工程学院;
【基金】：国家863计划课题项目(2012AA101906) 国家自然科学基金项目(61178036)资助
【分类号】：S666.4;S158

【相似文献】