欢迎访问《棉花学报》! 今天是

棉花学报 ›› 2020, Vol. 32 ›› Issue (5): 392-403.doi: 10.11963/1002-7807.zzlx.20200716

• 研究与进展 • 上一篇    下一篇

滴灌棉田植株氮营养指数的高光谱诊断研究

张泽1,马露露1,洪帅1,林皎1,张立福1, 2,吕新1*   

  1. 1.石河子大学农学院/新疆生产建设兵团绿洲生态农业重点实验室,新疆 石河子832003;2.中国科学院遥感与数字地球研究所,北京100080
  • 收稿日期:2019-11-03 出版日期:2020-09-15 发布日期:2020-09-30
  • 通讯作者: lxshz@126.com
  • 作者简介:张泽(1983―),男, 博士, zhangze1227@163.com
  • 基金资助:
    石河子大学创新发展专项(CXFZ201903);中国博士后科学基金(2017M623282);国家重点研发计划(2017YFD0700503)

Study on Hyperspectral Diagnosis of Nitrogen Nutrition Index among Different Cotton Varieties  under Drip Irrigation

Zhang Ze1, Ma Lulu1, Hong Shuai1, Lin Jiao1, Zhang Lifu1, 2, Lü Xin1*   

  1. 1. College of Agriculture, Shihezi University/Key Laboratory of Oasis Ecology Agricultural, Xinjiang Production and Construction Group, Shihezi, Xinjiang 832003, China; 2. Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing 100080, China
  • Received:2019-11-03 Online:2020-09-15 Published:2020-09-30

摘要: 【目的】通过研究滴灌棉田地上部植株的氮营养指数,探究建立基于氮营养指数的高光谱指数模型的可行性,为高光谱遥感在农田氮营养快速、准确诊断中的应用提供理论依据。【方法】2年试验,以新疆主栽的5个棉花品种为研究对象,在不同施肥处理条件下,探究氮营养指数和17种光谱指数之间的相关性,建立氮素营养诊断模型并进行验证。【结果】不同品种棉花滴灌条件下氮营养指数之间差异显著,杂交棉能更快地接近氮素营养水平的最佳状态;基于高光谱的氮营养指数多元回归模型中R2最高的为鲁棉研24,达到0.8的高水平;经2年数据验证,鲁棉研24号建立的模型精度最高,R2=0.868,均方根误差为0.059。【结论】建立基于氮营养指数的高光谱监测模型能够很好地监测植株氮素养分状况,大田模型精度R2能达到0.5以上,本试验结果能够为今后农业养分诊断提供理论依据。

关键词: 棉花冠层; 光谱参数; 氮营养指数; 诊断模型

Abstract: [Objective] To explore the feasibility of rapid and accurate diagnosis of nitrogen nutrition index investigeted nitrogen nutrition index and hyperspectral index in different varieties of drip irrigation cotton. [Method] Five cotton varieties with different characters were selected as the research object under different fertilization conditions, the correlation between nitrogen nutrient index and 17 spectral indices was explored. Then the diagnosis model of nitrogen nutrition was established and verified. [Results] The difference of nitrogen nutrition index among different cotton varieties of drip-irrigation was significant, the hybrid cotton could approach the best condition of nitrogen nutrition level more quickly. In the multiple regression model of nitrogen nutrition index based on hyperspectral analysis, Lumianyan 24 had the highest R2, which reached the high level of 0.8. The precision of the model established by Lumianyan 24 was the highest under two years' data verification. [Conclusion] The hyperspectral monitoring model based on nitrogen nutrition index can be used to monitor nitrogen nutrient status of plant. The results of this experiment can provide theoretical basis for precision fertilization in later stage agriculture. 

Key words: cotton canopy; spectral parameter; nitrogen nutrition index; spectral model