| 摘要: |
| 【目的】高光谱特征参数能够突出原始光谱的感兴趣信息，分析冬小麦的高光谱特
征参数随受涝时间的变化特征，提出一种快速识别冬小麦受涝的方法，可为冬小麦涝害监
测提供理论支撑。【方法】文章在冬小麦灌浆期设置持续淹水8 天的处理，采集涝害处理
当天、第3d、第5d、第7d 的反射光谱特征，以高光谱位置参数、振幅参数、面积参数
和反射率参数为研究指标，对比分析了健康和受涝冬小麦18 个高光谱特征参数的变化特
征，并根据高光谱特征参数的差异性指数随受涝时间的变化特征判断冬小麦的受涝程度。
【结果】（1）受涝冬小麦的红边位置发生“蓝移”，红谷、绿峰和黄边位置发生“红移”；4
个振幅参数值均减小；近红外面积和绿峰面积增大，红边、黄边和蓝边面积减小；绿峰反射
率Rg 和红谷反射率Ro 增大，两者的比值Rg/Ro 和归一化值（Rg-Ro）/（Rg+Ro）则减小；
（2）根据不同高光谱特征参数差异性指数的大小及变化特征，提取出红边位置、最小振幅、
近红外面积和红谷反射率为判断冬小麦受涝与否的最佳参数；高光谱反射率参数和面积参
数可在受涝前期快速识别冬小麦受涝与否，高光谱振幅参数能够在受涝后期判断受涝程度。
（3）不同高光谱特征参数识别冬小麦受涝的优劣能力从强到弱依次为：高光谱振幅参数> 高
光谱面积参数> 高光谱反射率参数> 高光谱位置参数。【结论】高光谱特征参数的变化特征
能够用来判断冬小麦受涝与否以及受涝程度，可为冬小麦涝害遥感监测提供理论支撑。 |
| 关键词: 高光谱 光谱特征 光谱参数 冬小麦 涝害 |
| DOI：10.12105/j.issn.1672-0423.20180208 |
| 分类号: |
| 基金项目:2012 年度公益性行业（农业）科研专项项目（201203032）；中国农业科学院创新工程（2016-2020，IARRP2017-727-1）资助 |
|
| Effects of persistent waterlogging on hyperspectral characteristic parameters of winter wheat |
|
Duan Dingding1, Xiong Qinxue2, Liu Li3, He Yingbin※1
|
|
1.Institute of Agriculture Resources and Regional Planning，Chinese Academy of Agricultural Science，Beijing 100081，China;2.Academy of agriculture，Yangtze University，Jingzhou 434025，China;3.Institute of Agriculture Remote Sensing and
Information Application，Zhengjiang University，Hangzhou 301158，China
|
| Abstract: |
| ［Purpose］The hyperspectral characteristic parameters can highlight the interesting
information of the original spectrum，analyzing the variation characteristics of hyperspectral
characteristic parameters with waterlogging time. This paper presents a method to quickly identify
the waterlogging of winter wheat，which can provide theoretical support for waterlogging monitoring
of winter wheat.［ Method］This paper sets 8 days of continuous flooding treatment in the winter
wheat filling period，collects the reflectance spectra of the first day，third day，fifth day and
seventh day，compares and analyzes the variation characteristics of 18 high spectral characteristic
parameters of healthy and waterlogged winter wheat based on four indexes—— the hyperspectral
position parameters，amplitude parameters，area parameters and reflectance parameters，and
judges the waterlogged degree of winter wheat according to the variation of the difference index of
hyperspectral characteristic parameters［. Result］The results showed that：（1）the red edge position
of the waterlogged winter wheat showed "blue shift"，and the red valley，green peak and yellow
edge position showed "red shift". The values of 4 amplitude parameters decreased. The near-infrared
area and green peak increased，while the red edge，yellow edge and blue edge area decreased.
Green peak reflectance（ Rg）and red valley reflectance（ Ro）increased，while the ratio Rg/Ro and
normalized value（ Rg-Ro）/（ Rg+Ro）decreased.（ 2）According to the magnitude and variation
characteristics of the difference index based on different hyperspectral characteristic parameters，the
red edge position，minimum amplitude，near-infrared area and red valley reflectance are selected as
the best parameters to judge whether the winter wheat is waterlogged or not. Hyperspectral reflectance
parameters and area parameters can quickly identify whether winter wheat is affected by waterlogging
during the early stage of waterlogging，while hyperspectral amplitude parameters can determine the
extent of waterlogging during the later stage of waterlogging.（ 3）The ability to distinguish winter
wheat with waterlogging of different hyperspectral characteristic parameters from strong to weak
are as follows：hyperspectral amplitude parameter>hyperspectral area parameter > hyperspectral
reflectance parameter>hyperspectral position parameter.［ Conclusion］The variation characteristics
of hyperspectral characteristic parameters can be used to judge whether the winter wheat is waterlogged
and the degree of waterlogging，which can provide theoretical foundation for winter wheat waterlog
monitoring. |
| Key words: hyperspectral spectral characteristics spectral parameters winter wheat waterlogging |
