欢迎访问《中国农学通报》,

中国农学通报 ›› 2014, Vol. 30 ›› Issue (16): 67-70.doi: 10.11924/j.issn.1000-6850.2013-3231

• 林学 园艺 园林 • 上一篇    下一篇

东洞庭湖湿地指示植被高光谱特征分析

李金钊 杨星仕 孙华   

  • 收稿日期:2013-12-11 修回日期:2014-01-23 出版日期:2014-06-05 发布日期:2014-06-05
  • 基金资助:
    2012年湖南省大学生研究性学习和创新性实验计划项目 “东洞庭湖湿地指示植被高光谱特征分析”(湘教通[2012]402号)。

Analysis on the Hyperspectral Characteristics of Indicative vegitations in the East Dongting Lake Wetland

  • Received:2013-12-11 Revised:2014-01-23 Online:2014-06-05 Published:2014-06-05

摘要: 为了建立东洞庭湖湿地指示植被光谱数据库,笔者通过分析指示植被的光谱特征,得到识别湿地植被高分辨率遥感波段窗口,为湿地健康监测提供技术支撑。选取东洞庭湖植被苔草、泥蒿、芦苇、辣蓼与灯心草5种典型湿地植被为东洞庭湖湿地指示植被,对它们进行研究。根据野外测定5种指示植被光谱数据,分析东洞庭湖湿地指示植被的光谱特性、植被类型识别的特征波段及其基本规律。研究发现:(1)在波段范围610~680 nm、810~880 nm、1030~1100 nm、1170~1240 nm、1635~1705 nm之间具有明显的差异。特别在波段1030~1100 nm之间,5种指示植被光谱差异性最大,光谱曲线区分特别明显;(2)通过对苔草、泥蒿、芦苇、辣蓼与灯心草5种湿地植被光谱反射率(R)采用LOG(1/R)变换,得到1个可以区分5种植被的区间,即660~680 nm之间的波段,5种湿地指示植被倒数的对数值依次降低次序是:苔草、芦苇、辣蓼、灯心草和泥蒿;(3)对同一种湿地指示植被苔草不同长势的光谱数据进行降噪、倒数的对数处理后,得到1个特征区间在500~700 nm内,可以明显的区分苔草的长势情况即:长势良好的苔草光谱反射率倒数的对数值要高于长势不好的苔草。

关键词: 葡萄砧木, 葡萄砧木, 抗寒, 电导率, 主成分分析

Abstract: In order to establish the spectra database of indicative vegetations in the East Dongting Lake Wetland, the author obtained hyperspectral remote sensing window by analyzing the hyperspectral characteristics of indicative vegetations in that area, which could provide technological supports for the wetland health. 5 typical types of wetland vegetations in the East Dongting Lake, those were Carex, Artemisia, Phragmites communis, Polygonum hydropiper, and Juncus effusus., were chosen as the indicator vegetations to be investigated. The spectral characteristics, the distinguish bands, and the fundamental rules of 5 indicated vegetations based on field measurement had been analyzed. The results demonstrated that: (1) there were obvious differences among the waveband of 610-680 nm, 810-880 nm, 1030-1100 nm, 1170-1240 nm, and 1635-1705 nm; especially in the waveband of 1030-1100 nm, the spectrum difference among those 5 indicative vegetations is most distinct, and the distinguish of the spectral curve was shaper; (2) by transmitting spectral reflectance (R) of 5 wetland vegetations, namely, Carex, Artemisia, Phragmites communis, Polygonum hydropiper, Juncus effusus into LOG(1/R), a band of 660-680 nm which could be used to distinguish those 5 plants, was achieved, where the reciprocal’s logarithm of 5 vegetations were listed from lowest to highest, namely, Carex, Artemisia, Phragmites communis, Juncus effusus, Artemisia; (3) after denoising and LOG(1/R) of the same wetland vegetation with different spectral data, the band of 500-700 nm could obviously distinguish the growing state of the sedge, that was, spectral reflectance’s LOG(1/R) of Carex in good condition was higher than that of Carex in bad condition.