土壤粒径对近红外光谱检测土壤养分建模的影响

Influence of Soil Particle Size on Modeling to Detect Soil Nutrient Content Using Near Infrared Spectroscopy

近红外光谱法检测土壤养分的广泛应用对于精细农业合理施肥具有重要意义,然而不同土壤粒径对近红外光谱建模精度带来了较大影响。以土壤全氮含量为检测对象,利用PLS回归分析法研究不同粒径土壤对近红外光谱建模精度的影响和规律性。实验结果表明:相同系列全氮含量土壤样品在不同粒径下所建立的预测模型差异明显,验证模型结果显示粒径越小的土壤含氮量预测值与实际值的相关性越好,模型预测精度越高,对于进一步研究消除粒径影响的校正方法提供了依据。

In this paper taking soil total nitrogen content as test object,PLS regression analysis is used to study modeling accuracy and regularity on the near infrared spectrum for different particle size of soil.The experimental results show that the prediction models,established by the soil samples of same series of total nitrogen content with different particle size,exist obvious differences.