摘要
以不同磷肥施用量的20年树龄库尔勒香梨叶片光谱反射率为研究对象,结合室内叶片全磷含量分析,对库尔勒香梨叶片的全磷含量与叶片原始光谱、一阶导数光谱、高光谱参数之间的相关性进行分析。结果表明:不同磷肥施用量处理的库尔勒香梨叶片光谱反射率大小依次为:P0>P1>P2>P3;库尔勒香梨叶片全磷含量与原始光谱在波长1 290 nm处构建的线性模型,调整决定系数R2值达到0.998;在波长1 360 nm处一阶微分光谱与全磷含量构建的线性模型,调整决定系数R2值为0.938;对于所选的高光谱特征变量、叶片全磷含量与红边面积(SDr)的相关关系极显著,并由此建立的线性模型的调整决定系数R2值达到0.927,说明利用这3个变量预测库尔勒香梨叶片全磷含量具有可行性。对所构建的相关模型进行预测值和实测值之间的符合度检验,确定了基于光谱参数红边面积(SDr)的构建模型为Y=-0.122X+7.282,该模型作为库尔勒香梨叶片全磷含量估测的最优模型。
Set the different treatments of P for 20 years Korla Korla fragrant pear leaves were measured, and combined with fragrant pear in the field, the spectral reflectance of total phosphorus content analysis in the lab, analyze the relationships among total phosphorus content of leaves, the original spectrum, the first derivative spectrum, and the high spectral parameters. The results showed that the spectral reflectance of Korla fragrant pear leaves under different treatments of P was in order of Po 〉 P1 〉 Pz 〉 P3- One linear model was built at 1 290 nm between the total phosphorus content and the original spectrum, the adjusted determination coefficient value R2 was 0. 998. Another linear model was built at 1 360 nm between the total phosphorus content and the first order derivative spectrum, and the adjusted determination coefficient value RE was 0. 938. To select hyperspectral feature variables, the relation- ship between the total phosphorus content of leaves and the sum of derivatives in red edge （SDr） were extremely significant. And built the linear model, the adjusted determination coefficient values R2 was 0. 927. These indica- ted that it was feasible to predict the total phosphorus content of Korla fragran valuation of all the built models, the model based on the spectral parameters t pear leaves by the variables. After e- （SDr） was the best model for predic-ting the total phosphorus content of Korla fragrant pear leaves.
出处
《西部林业科学》
CAS
北大核心
2013年第5期7-12,共6页
Journal of West China Forestry Science
基金
自治区"十二五"科技计划项目(201130102-2)
土壤学自治区重点学科资助
关键词
高光谱
库尔勒香梨
叶片全磷含量
估算模型
hyperspectral
Korla fragrant pear
total phosphorus content of leaves
estimation models
作者简介
第一作者简介：柴仲平（1974-），男，甘肃永昌人，副教授，博士研究生，主要从事土壤质量、植物营养方面的研究。
通讯作者简介：盛建东（1970-），男，甘肃秦安人，教授，博士，主要从事土壤质量空间变异和养分资源高效利用方面的研究。
