中国农业科学 ›› 2018, Vol. 51 ›› Issue (18): 3486-3496.doi: 10.3864/j.issn.0578-1752.2018.18.005

• 耕作栽培·生理生化·农业信息技术 • 上一篇    下一篇

基于新型植被指数的冬小麦LAI高光谱反演

束美艳1,2,3,4(), 顾晓鹤2,3,4(), 孙林1, 朱金山1, 杨贵军2,3,4, 王延仓5, 张丽妍2,3,4   

  1. 1山东科技大学测绘科学与工程学院,山东青岛 266590
    2农业部农业遥感机理与定量遥感重点实验室/北京农业信息技术研究中心,北京 100097
    3国家农业信息化工程技术研究中心,北京 100097
    4北京市农业物联网工程技术研究中心,北京 100097
    5北华航天工业学院,河北廊坊 065000
  • 收稿日期:2018-04-18 接受日期:2018-06-13 出版日期:2018-09-16 发布日期:2018-09-16
  • 作者简介:

    联系方式:束美艳,E-mail:2448858578@qq.com

  • 基金资助:
    国家重点研发计划(2016YFD0300609)、国家自然科学基金(41571323)、北京市自然科学基金(6172011)、北京市农林科学院创新能力建设专项(KJCX20170705)、河北省青年基金(D2017409021)

High Spectral Inversion of Winter Wheat LAI Based on New Vegetation Index

MeiYan SHU1,2,3,4(), XiaoHe GU2,3,4(), Lin SUN1, JinShan ZHU1, GuiJun YANG2,3,4, YanCang WANG5, LiYan ZHANG2,3,4   

  1. 1College of Geomatics Shandong University of Science and Technology, Qingdao 266590, Shandong
    2Key Laboratory of Quantitative Remote Sensing in Agriculture of Ministry of Agriculture/Beijing Research Center for Information Technology in Agriculture, Beijing 100097
    3National Engineering Research Center for Information Technology in Agriculture, Beijing 100097
    4Beijing Engineering Research Center for Agriculture Internet of Things, Beijing 100097
    5North China Institute of Aerospace Engineering, Langfang 065000, Hebei
  • Received:2018-04-18 Accepted:2018-06-13 Online:2018-09-16 Published:2018-09-16

摘要:

【目的】本研究旨在分析冠层叶片水分含量对作物冠层光谱的影响,构建新型光谱指数来提高作物叶面积指数高光谱反演的精度。【方法】在冬小麦水肥交叉试验的支持下,分析不同筋性品种、施氮量、灌溉量处理下的冬小麦叶面积指数冠层光谱响应特征,并分析标准化差分红边指数(NDRE)、水分敏感指数(WI)与叶面积指数的相关性,据此构建一个新型的植被指数——红边抗水植被指数(red-edge resistance water vegetable index, RRWVI)。选取常用的植被指数作为参照,分析RRWVI对于冬小麦多个关键生育期叶面积指数的诊断能力,随机选取约2/3的实测样本建立基于各种植被指数的叶面积指数高光谱响应模型,未参与建模的样本用于评价模型精度。【结果】研究结果表明,随着生育期的推进,冬小麦的叶面积指数呈先增加后降低的变化趋势,不同的水肥处理对冬小麦叶面积指数具有较大影响。开花期之后冬小麦LAI显著下降,强筋小麦(藁优2018)在整个生育期叶面积指数均高于中筋小麦(济麦22);不同氮水平下冬小麦冠层光谱反射率在近红外波段(720—1 350 nm)随着施氮量的增加而增大,与氮肥梯度完全一致,其中2倍氮肥处理的近红外反射率达到最高;不同生育期下冬小麦冠层光谱反射率变化波形大体一致;各个关键生育期的NDRE和WI均存在较高的相关性,而 NDRE与LAI的相关性明显优于WI,新构建的植被指数RRWVI与LAI的相关性均优于NDRE、WI;虽然8个常用的植被指数均与LAI存在显著相关,但RRWVI与LAI相关性达到最大,其拟合曲线的决定系数R2为0.86。【结论】通过分析各种指数所构建的冬小麦叶面积指数高光谱反演模型,新构建的RRWVI取得了比NDRE、NDVI等常用植被指数更为可靠的反演效果,说明本研究新构建的红边抗水植被指数可有效提高冬小麦叶面积指数的精度。

关键词: 冬小麦, 高光谱, 红边抗水植被指数, 叶面积指数, 标准化差分红边指数, 归一化植被指数

Abstract:

【Objection】The purpose of this study was to analyze the effect of leaf water content on crop canopy spectra and to construct a new spectral index, so as to improve the accuracy of high spectral inversion of crop leaf area index (LAI). 【Method】 Under the support of winter wheat water-fertilizer cross test, the canopy spectral response characteristics of LAI of winter wheat under different recalcitrant cultivars, nitrogen application rates and irrigation amount were analyzed. Through the correlation analysis among the normalized differential red edge index (NDRE), water sensitivity index (WI) and LAI, the paper developed, a new vegetation index, the red-edge resistance water vegetation index (RRWVI) to inverse winter wheat LAI. Several commonly used vegetation indices were used as a reference to analyze the response ability of RRWVI to diagnose the LAI of many key winter wheat varieties. 2/3 of the measured samples were randomly selected to establish a high spectral response model of LAI based on various vegetation indices and 1/3 of the samples not involved in the modeling were used to evaluate the accuracy of the model. 【Result】 The results showed that with the advancement of growth period, the LAI of winter wheat first increased and then decreased, and different water and fertilizer treatments had a greater effect on it. After the flowering stage, the LAI of winter wheat declined significantly, and the LAI of strong gluten wheat (Gaoyou2018) was higher than that of medium-gluten wheat (Jimmy22) during the whole growth period. The spectral reflectance of winter wheat under different nitrogen levels increased with the increase of nitrogen application rate in the near-infrared band (720-1 350 nm), which was completely consistent with the nitrogen fertilizer gradient. The samples with twice-nitrogen treatment had the highest near-infrared reflectance, and the changed in spectral reflectance of winter wheat canopy under different growth stages were generally consistent. There was a high correlation between NDRE and WI in each key growth period, and the correlation between NDRE and LAI was significantly better than that of WI. The correlation between RRWVI and LAI was better than NDRE and WI. Although 8 commonly used vegetation indices are significantly correlated with LAI, RRWVI has the greatest correlation with LAI, and the coefficient of determination R2 of the fitting curve reached 0.86.【Conclusion】 By analyzing the hyperspectral inversion model of winter wheat LAI constructed by all kinds of indices, the newly constructed RRWVI achieved a more reliable inversion effect than frequently-used vegetation indices, such as NDRE and NDVI, indicating that the newly constructed red edge water-resistant vegetation index could effectively improve the accuracy of monitoring winter wheat LAI.

Key words: winter white, hyperspectral, RRWVI, leaf area index, NDRE, NDVI