川西北高寒草甸常见植物光谱特征分析

Spectral characteristics of Northwest Sichuan's main alpine meadow vegetations

摘要:

【目的】草地光谱反射特征是运用遥感数据研究草地植被理化特性和植被分类的基础，对不同草地植物光谱反射特征开展分析研究，为草地植被的定量遥感与准确识别提供理论参考与技术支持。【方法】采用便携式地物光谱仪，于 2021 年 7-8 月在四川省甘孜州雅江县对川西北 8 种高寒草甸常见植物蕨麻（Argentina anserina）、高山嵩草（Kobresia pygmaea）、花葶驴蹄草（Caltha scaposa）、藏蒲公英（Taraxacum tibetanum）、广叶橐吾（Ligularia euryphylla）、淡黄香青（Anaphalis flavescens）、圆穗蓼 （Polygonum macrophyllum）、黄花棘豆（Oxytropis ochrocephala）进行了野外光谱测量，利用一阶导数、 二阶导数、光谱去包络线以及植被指数法对原始光谱数据进行了处理，揭示了 8 种高寒草甸植被光谱反射特性。【结果】相对于不同草地植物的绿峰位置、绿峰幅值、红谷位置、红谷幅值、红边幅值、红边面积、吸收峰面积、吸收峰对称度等光谱特征参量，高光谱植被指数更容易区分此 8 种草地植物，8 种植物的 NDVI 和 mSR₇₀₅指数具有一致性，指数由大到小依次为：藏蒲公英＞黄花棘豆＞高山嵩草＞蕨麻＞ 圆穗蓼＞花葶驴蹄草＞广叶橐吾＞淡黄香青 ，且各植物 mSR₇₀₅ 指数值差异明显 ，易于相互区分 。 【结论】采用不同方法对草地植物光谱反射特征开展分析，可有效的区分不同的草地植被，为高光谱遥感开展草地资源植被分类、草原基况调查和动态监测等提供参考。

Abstract:

【Objective】The spectral reflectance characteristics of grassland are the basis for using remote sensing data to study the physical and chemical characteristics of grassland vegetation and vegetation classification. Research on the spectral reflectance characteristics of different grassland plants provides theoretical reference and technical support for quantitative remote sensing and accurate identification of grassland vegetation.【Method】 Portable ground object spectrometer was used to measure the field spectra of eight alpine meadow grass species：Potentilla anserina、Kobresia pygmaea、Caltha scaposa、Taraxacum tibetanum、Ligularia euryphylla、Anaphalis flavescens、Polygonum macrophyllum and Oxytropis ochrocephala in Yajiang，Ganzi，Sichuan Province from July to August 2021. The original spectral data were processed using the methods of first derivative，second derivative，continuum removal and vegetation index，and the spectral reflectance characteristics of eight alpine meadow species were revealed.【Result】 The results showed that compared with the spectral characteristic parameters of different grassland plants，such as "green peak position"，"green peak amplitude"，"Red Valley position"，"Red Valley amplitude"，"red edge amplitude"，"red edge area"，"absorption peak area"，"absorption peak symmetry"，the Hyperspectral Vegetation index was easier to distinguish these 8 types of grassland vegetations. NDVI and mSR705 indexes of 8 plants were consistent，and the index values as follow：Taraxacum tibetanum ＞ Oxytropis ochrocephala ＞ Kobresia pygmaea ＞ Potentilla anserina ＞ Polygonum macrophyllum ＞ Caltha scaposa＞Ligularia euryphylla＞Anaphalis flavescens. The mSR705 index values were obviously different among the 8 alpine meadow vegetations and easy to distinguish from each other. 【Conclusion】 Using different methods to analyze the spectral reflectance characteristics of grassland plants can effectively distinguish different types of grassland vegetation. This paper provides theoretical support for applying hyperspectral remote sensing in vegetation classification，grassland condition investigation and dynamic monitoring of grassland resources.