首页 | 本学科首页   官方微博 | 高级检索  
     检索      

盐胁迫下匍匐翦股颖高光谱分析与电解质渗透率反演
引用本文:肖国增,吴雪莲,滕珂,晁跃辉,李伟涛,韩烈保.盐胁迫下匍匐翦股颖高光谱分析与电解质渗透率反演[J].光谱学与光谱分析,2016(11):3630-3636.
作者姓名:肖国增  吴雪莲  滕珂  晁跃辉  李伟涛  韩烈保
作者单位:1. 北京林业大学草坪研究所,北京 100083; 长江大学园艺园林学院,湖北 荆州 434025;2. 华中农业大学经济管理学院,湖北 武汉,430070;3. 北京林业大学草坪研究所,北京,100083;4. 滁州学院地理信息与旅游学院,安徽 滁州,239000
基金项目:国家高技术研究发展计划(863计划)项目(2013AA102607),2016年安徽地理信息集成应用协同创新中心开放基金项目(201116Y03)
摘    要:叶片电解质渗透率是反映植物细胞渗透性的一个重要指标,对草坪草遭受盐胁迫的研究有重要意义。针对叶片电解质渗透率传统检测方法,耗时长,损伤叶片,无法大面积监测等弊端,探讨了用高光谱快速无损检测叶片电解质渗透率的方法。以匍匐翦股颖(Agrostis stolonifera)为对象,在温室中水培两周后进行浓度分别为0(对照),100和200mmol·L-1的盐处理,7d后按间隔7d取样3次,共72个样。每次取样时先测量样品的光谱值,然后采用电导率法测定叶片电解质渗透率。分析匍匐翦股颖三种盐处理与光谱反射率之间的关系和差异,对三种盐处理的光谱反射率计算归一化植被指数和差值植被指数,采用差分法计算光谱反射率的一阶微分,同时计算出蓝、绿和红光的三边参数,分析叶片电解质渗透率与光谱反射率、归一化植被指数、差值植被指数和三边参数的相关性。利用叶片电解质渗透率和各光谱数据相关程度高的数据,对校正集采用一元线性回归、多元线性回归和偏最小二乘回归法构建叶片电解质渗透率反演模型,用预测集检验反演模型。结果表明:盐胁迫与叶片高光谱在450~700nm波段呈正相关;叶片电解质渗透率与450~732nm波段反射率在0.01水平上显著相关;三边参数中绿边幅值和绿边面积与叶片电解质渗透率显著相关;采用偏最小二乘回归法建立的反演模型精度最好,建模和反演预测的决定系数分别达到了0.681和0.758,均方根误差分别为7.124和7.079。偏最小二乘法构建的反演模型实现了盐胁迫下匍匐翦股颖叶片电解质渗透率的快速无损检测,也为采用高光谱实时监测盐胁迫对匍匐翦股颖及其同类植物的伤害提供了依据和理论参考。

关 键 词:高光谱  盐胁迫  草坪草  匍匐翦股颖  电解质渗透率  反演模型

Hyperspectral Analysis and Electrolyte Leakage Inversion of Creeping Bentgrass under Salt Stress
Abstract:Leaf electrolyte leakage is an important index of the plant cell permeability which plays an important role in the study of turfgrass salt stress.Traditional methods of measuring leaf electrolyte leakage have many disadvantages such as time-consu-ming,destroying the plants and being unable to monitor salt stress in large area.The aim of this study is to build a hyperspectral inversion model for leaf electrolyte leakage of creeping bentgrass under different salt concentration stresses thus to promote the application of the hyperspectral techniques in turfgrass salt stress monitoring.Creeping bentgrass was used in this study,and it was grown in water for two weeks before salt treatments.Leaves were collected at 7,14 and 21 d under 0(CK),100 and 200 mmol·L-1 NaCl respectively.The spectral values were gathered using Unispec-SC Spectral Analysis System (PP SYSTEMS, USA)before collecting grass leaves.Leaf electrolyte leakage was measured with electrical conductivity method.The relation and differences between salt treatments and spectral reflectance values were analyzed with EXCEL.Normalized difference vegetation index (NDVI)and difference vegetation index (DVI)were calculated using the spectral reflectance values.The first-order differ-ential was calculated with difference method.The trilateral parameters of the blue,green and red rays were calculated at the meantime.The correlation analysis of the Leaf electrolyte leakage,spectral reflectance value,DVI and trilateral parameters was achieved by using EXCEL and Matlab software.Electrolyte leakage inversion model of the calibration set consisted of 48 high correlational samples,was built using unary linear regression,multivariate linear regression and partial least-squares regression methods.The prediction set inspection inversion model was established using the other 24 samples.The results showed that there is a positive correlation between salt stresses and 450~700 nm wave band.The leaf electrolyte leakage was positively asso-ciated with 450~732 nm band region at 0.01.The green edge amplitude and area of green edge were correlated with the foliar e-lectrolyte leakage positively.Models based on partial least squares regression could inversion the foliar electrolyte leakage opti-mally.The calibration R2 reached to 0.681,and the validation R2 reached to 0.758.The calibration RMSE was 7.124,and the validation RMSE reached to 7.079.The inversion model made it possible to detect creeping bentgrass leaf electrolyte leakage un-der salt stress rapidly.This study also provided theoretical reference for monitoring the damage of other creeping bentgrass relat-ed plant species resulted by salt stress.
Keywords:Hyperspectral  Salt stress  Turfgrass  Creeping bentgrass  Electrolyte leakage  Inversion model
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号