| 本文已被：浏览 6770次 下载 1932次 |
码上扫一扫！ |
| 基于高光谱的水浇地与旱地春小麦拔节期叶绿素含量估测模型对比研究 |
| 靳彦华1,3，熊黑钢2,3，张芳1,3，王莉峰1,3 |
|
|
| ((1.新疆大学资源与环境科学学院, 新疆 乌鲁木齐 830046; 2.北京联合大学应用文理学院, 北京 100083; 3.教育部绿洲生态重点实验室, 新疆 乌鲁木齐 830046）) |
|
| 摘要: |
| 测定拔节期水浇地与旱地春小麦冠层光谱、叶绿素含量、覆盖度、苗高和叶宽，采用回归分析方法建立春小麦叶绿素含量高光谱估测模型，并对模型精度进行检验。结果表明: 阳坡和双面坡地春小麦拔节期叶绿素含量与原始光谱反射率在可见光和近红外波段均呈正相关，水浇地和阴坡地在723 nm以前相关系数为负，723 nm以后为正。各地类春小麦叶绿素含量与各高光谱变量的相关性均较好，均达到了极显著水平(P<0.01)。无论在可见光还是近红外波段，水浇地春小麦叶绿素含量均与倒数之对数lg(1/R)的相关性最好，相关系数最大值可达0.98；阴坡地则与一阶微分的相关性最好，最大为0.94；而与阳坡和双面坡地相关性最好的高光谱指数为归一化植被指数。在
各个波段，倒数之对数模型lg(1/R)、一阶微分模型(p′)和归一化植被指数模型(N)分别是估测水浇地、阴坡地、阳坡和双面坡地春小麦叶绿素含量的最佳模型。虽然各模型R2均超过0.90，精确度均大于0.91，但阴坡地、阳坡和双面坡地的模型精确度和准确度略低于水浇地。以上模型的建立可为今后估测水浇地与旱地春小麦的健康状况提供参考。 |
| 关键词: 高光谱指数 水浇地 旱地 拔节期 叶绿素含量 对比研究 |
| DOI： |
|
| 基金项目:国家自然科学基金(41171165)；北京联合大学人才强校计划人才项目（BPHR2012E01）；北京市属高等学校高层次人才引进与培养计划项目(IDHT20130322) |
|
| Comparative study of estimation model on chlorophyll content at jointing stage of spring wheat under different lands based on hyperspectral indices |
| JIN Yan-hua1,3， XIONG Hei-gang2,3, ZHANG Fang1,3, WANG Li-feng1,3 |
| () |
| Abstract: |
| In order to establish high spectral chlorophyll content model based on correlation analysis between chlorophyll content and five spectral variables of spring wheat under irrigated and dry land, canopy spectral, chlorophyll content, ground cover, seedling height, leaf width at jointing stage of spring wheat under different land were determined, and the precision of these model were tested. The result showed a positive correlation between chlorophyll content and original spectrum at visible light and near infrared band of sunny and double slopes at jointing stage of spring wheat, and a negative correlation below 723 nm and a positive correlation above 723 nm for irrigated land and shady slope. There was markedly significant correlation (P<0.01) between chlorophyll content and hyperspectral variables of spring wheat under different lands. Regardless of visible light or near infrared wave band, lg(1/R) was significantly correlated with the leaf chlorophyll content, with a correlation coefficient of 0.98. Shady slope was correlated with first derivative (r=
0.94), while for sunny and double slopes normalized difference vegetation index showed significant relationship. For each band, lg(1/R), p′, and N were suitable indexes estimating chlorophyll content for all lands of irrigated land, shady slope, and sunny and double-sided slope. Although the regression determination coefficient R2 was more than 0.90, precision and accuracy of the proposed for shady slope, sunny and double slope are slightly lower than that of irrigated land. |
| Key words: hyperspectral variables irrigated land dry land jointing stage chlorophyll content comparative study |
