| 基于高光谱遥感的花生叶片氮积累量监测模型的研究 |
| |
| 引用本文： | 张晓艳,王丽丽,刘锋,封文杰,刘淑云,朱建华.基于高光谱遥感的花生叶片氮积累量监测模型的研究[J].山东农业科学,2012,44(3):7-12,16. |
| |
| 作者姓名： | 张晓艳 王丽丽 刘锋 封文杰 刘淑云 朱建华 |
| |
| 作者单位： | 山东省农业科学院科技信息工程技术研究中心,山东济南,250100 |
| |
| 基金项目： | 国家科技支撑计划项目(2006BAD21B04-20)资助 |
| |
| 摘 要： | 作物氮素状况是评价长势、提高产量和改善品质的重要指标,因此叶片氮积累量的实时无损估测对作物生产的氮素管理具有重要意义。本研究选用大花生品种丰花1号为试验材料,在大田生产条件下,分析了花生叶片氮积累量与冠层高光谱参数的定量关系。结果表明,叶片氮素含量随生育进程逐渐下降,不同处理之间差异较小;叶片氮素积累量随生育时期推进呈现先升后降的单峰曲线变化趋势,在结荚期达到高峰。花生冠层光谱反射率在740～1 100 nm波段内随叶片氮积累量的增加而增加,叶片氮积累量的敏感波段主要存在于近红外平台和可见光区,其中,"红边"区域表现最为显著。通过微分等技术构造多种植被指数,对高光谱参数和叶片氮积累量进行相关回归分析,红边振幅(Dr)、氮素反射指数(NRI)、归一化植被指数(NDVI)各波段组合平均值及比值植被指数(RVI)与叶片氮积累量关系最密切,方程拟合决定系数分别为0.9194、0.8984、0.8918、0.8899、0.8794、0.8797。经另外一组独立数据的检验表明,对叶片氮积累量的预测以红边位置(REP)和Dr两个参数表现最优,预测的根均方差(RMSE)分别为1.78和1.10,相对误差为5.29%和3.59%。NDVIAverage(1230,1240,1250,1260),640]和土壤调整植被指数(SAVI)两个光谱参数预测的RMSE分别为1.15和1.19,预测相对误差为5.42%和7.41%。比较而言,Dr为自变量建立的模型,可以更好地评估不同条件下叶片氮素积累状况。
|
| 关 键 词： | 花生 高光谱遥感 叶片氮积累量 监测模型 |
Study on Monitoring Model for Nitrogen Accumulation in Peanut (Arachis Hypogasa L.) Leaves Based on Hyper- Spectral Remote Sensing |
| |
| ZHANG Xiao-yan
,
WANG Li-li
,
LIU Feng
,
FENG Wen-jie
,
LIU Shu-yun
,
ZHU Jian-hua.Study on Monitoring Model for Nitrogen Accumulation in Peanut (Arachis Hypogasa L.) Leaves Based on Hyper- Spectral Remote Sensing[J].Shandong Agricultural Sciences,2012,44(3):7-12,16. |
| |
| Authors: | ZHANG Xiao-yan WANG Li-li LIU Feng FENG Wen-jie LIU Shu-yun ZHU Jian-hua |
| |
| Institution: | (S&T Information Engineering Research Center of Shandong Academy of Agricultural Sciences,Jinan 250100,China) |
| |
| Abstract: | Nitrogen condition of crops was an important indicator for evaluating growth vigor,increasing yield and improving quality,so it was significant of real-time and lossless estimation of nitrogen accumulation in leaves for nitrogen management during crop production.In this study,the peanut variety Fenghua 1 with large seed was used as experimental material to study the relationship between nitrogen accumulation in peanut leaves and canopy hyper-spectral remote sensing parameters under field conditions.The results were as follows.The nitrogen content decreased gradually in peanut leaves with the growth process,and there were little differences between different treatments.The variation trend of nitrogen was a single peak showing that the nitrogen content increased firstly and then declined,and reached the peak at the pod-setting stage.The canopy spectrum reflectance of peanut increased with the increase of nitrogen accumulation in the wave band of 740 ~ 1 100 nm,and the sensitive bands to nitrogen accumulation were mainly present in near-infrared and visible ranges,in which,the red edge area showed the most significant.The regression relationship of hyper-spectral parameters with nitrogen accumulation was analyzed through construction of multiple vegetation indices by techniques such as differential.The results showed that the means of amplitude of red edge(Dr),nitrogen reflectance index(NRI) and normalized difference vegetation index(NDVI) and ratio vegetation index(RVI) related most closely with the nitrogen accumulation with the determination coefficients of 0.9194,0.8984,0.8918,0.8899,0.8794 and 0.8797 respectively.The test results conducted by another group of independent data indicated that the two parameters of red edge position(RFP) and Dr could predict nitrogen accumulation the best with the RMSE as 1.78 and 1.10 and the relative error as 5.29% and 3.59% respectively.The RMSE of NDVI and SAVI was 1.15 and 1.19 with the relative error as 5.42% and 7.41% respectively.In a word,the model constructed with Dr as the independent variable could evaluate the status of nitrogen accumulation well. |
| |
| Keywords: | Peanut(Arachis hypogasa L ) Hyper-spectral remote sensing Nitrogen accumulation in leaves Monitoring model |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
