草地学报 ›› 2022, Vol. 30 ›› Issue (9): 2438-2448.DOI: 10.11733/j.issn.1007-0435.2022.09.025

• 研究论文 • 上一篇    下一篇

基于不同阶微分高光谱植被指数的牧区草场地上生物量估算

童新1,2,3, 杨震雷4, 张亦然1, 吴宇辰1, 段利民1,2,3   

  1. 1. 内蒙古农业大学水利与土木建筑工程学院, 内蒙古 呼和浩特 010018;
    2. 内蒙古自治区水资源保护与利用重点实验室, 内蒙古 呼和浩特 010018;
    3. 黄河流域内蒙段水资源与水环境综合治理协同创新中心, 内蒙古 呼和浩特 010018;
    4. 西湖大学工学院, 浙江 杭州 310024
  • 收稿日期:2022-02-22 修回日期:2022-04-22 出版日期:2022-09-15 发布日期:2022-09-30
  • 通讯作者: 童新,E-mail:xintong@imau.edu.cn
  • 作者简介:童新(1989-),男,汉族,江西九江人,博士,讲师,主要从事生态水文与环境遥感研究,E-mail:xintong@imau.edu.cn
  • 基金资助:
    国家自然科学基金项目(51809141、52169002);内蒙古农业大学高层次人才科研启动金项目(NDYB2017-24);内蒙古自治区科技计划项目(2022YFSH0105、2021GG0071);内蒙古自然科学基金项目(2018BS05001、2018ZD05);内蒙古农业大学学生科技创新基金项目(KJCX2019019);教育部创新团队发展计划(IRT_17R60)、科技部重点领域科技创新团队(2015RA4013);内蒙古自治区草原英才产业创新创业人才团队(2012);内蒙古农业大学寒旱区水资源利用创新团队(NDTD2010-6)资助

Estimation of Pasture Aboveground Biomass using Different Orders of Differential Hyperspectral Vegetation Indices

TONG Xin1,2,3, YANG Zhen-lei4, ZHANG Yi-ran1, WU Yu-chen1, DUAN Li-min1,2,3   

  1. 1. College of Water Conservancy and Civil Engineering, Inner Mongolia Agricultural University, Hohhot, Inner Mongolia 010018, China;
    2. Inner Mongolia Key Laboratory of Water Resource Protection and Utilization, Hohhot, Inner Mongolia 010018, China;
    3. Collaborative Innovation Center for Integrated Management of Water Resources and Water Environment in the Inner Mongolia Reaches of the Yellow River, Hohhot, Inner Mongolia 010018, China;
    4. School of Engineering, Westlake University, Hangzhou, Zhejiang Province 310024, China
  • Received:2022-02-22 Revised:2022-04-22 Online:2022-09-15 Published:2022-09-30

摘要: 利用高光谱遥感技术能快速、无损、高效地获取草地地上生物量信息,对牧区牧草高效管理、草畜供求关系平衡以及放牧制度优化等方面具有重要意义。为了寻求估算生长旺盛期草地地上生物量最适宜的微分光谱阶数,本研究在内蒙古天然草场通过原位试验采集了高光谱反射率与地上生物量数据,对原始光谱反射率数据进行一至四阶微分处理,在全波段范围内挑选最佳波段构建简单比值植被指数(Simple ratio vegetation index,SRVI)、归一化植被指数(Normalized difference vegetation index,NDVI)、土壤调节植被指数(Soil adjusted vegetation index,SAVI)和增强型植被指数(Enhanced vegetation index,EVI)4种高光谱植被指数,建立相应地上生物量估算模型并对比评价各模型精度。结果表明:对原始高光谱反射率进行微分处理,有助于提高敏感波段与地上生物量的相关性;红边波段与近红外波段是构建最佳植被指数的重要组成波段,占所有优选波段的82%;基于二阶微分光谱的最佳SRVI和NDVI模型精度最好,R2分别为0.69和0.70,RMSE分别为196.60 g·m-2和196.65 g·m-2,过高的微分阶数反而会降低估算模型的精度。本研究能为利用不同阶微分高光谱估算草地地上生物量提供科学借鉴,为精准快速的牧区天然草场遥感监测提供技术和方法支持。

关键词: 地上生物量, 微分高光谱, 植被指数, 波段优选

Abstract: Hyperspectral remote sensing technology can quickly, non-destructively, and efficiently measure grassland aboveground biomass (AGB), which is of great significance for the efficient management of forage in pastoral areas, the balance between supply and demand of forage and livestock, and the optimization of grazing systems. In this paper, hyperspectral reflectance and AGB data were collected at two sites in natural pastures in Inner Mongolia. To determine the most suitable differential order for estimating the pasture AGB at the vigorous growth stage, original spectral reflectance was transferred to first-, second-, third-, and fourth derivative reflectance. Four forms of vegetation indices, i.e., SRVI, NDVI, SAVI, and EVI, using all possible combinations of narrow-band different orders of derivative reflectance and original reflectance were calculated, and the best derivative vegetation indices, as well as the original reflectance vegetation indices, were chosen by a linear correlation analysis against AGB. The linear regression models using each best derivative and original reflectance vegetation index as input variable were developed and compared for estimating the AGB. It was observed that the relationships between the sensitive individual derivative reflectance and AGB were much better than between the original reflectance and AGB. The red edge wavebands and the near-infrared wavebands were important components for constructing the best vegetation indices, accounting for 82% of all preferred wavebands. The best SRVI and NDVI models based on second-order differential spectral reflectance had the best accuracy, with R2 of 0.69 and 0.70, RMSE of 196.60 g·m-2 and 196.65 g·m-2, respectively. In addition, increasing the differential order too much would reduce the accuracy of the estimated models. This study could provide a scientific reference for the estimation of grassland AGB using derivative hyperspectral techniques, and provide technical and methodological support for accurate and rapid remote sensing monitoring of natural grasslands in pastoral areas.

Key words: Aboveground biomass, Derivative hyperspectral technique, Vegetation index, Band selection

中图分类号: