草甸草原植物beta多样性高光谱遥感估算方法  

Estimation of Plants Beta Diversity in Meadow Prairie Based on Hyperspectral Remote Sensing Technology

作　　者：杨星晨 雷少刚[1] 徐军 苏兆瑞 王维忠 宫传刚[4] 赵义博 YANG Xing-chen;LEI Shao-gang;XU Jun;SU Zhao-rui;WANG Wei-zhong;GONG Chuan-gang;ZHAO Yi-bo(Engineering Research Center of Ministry of Education for Mine Ecological Restoration,China University of Mining and Technology,Xuzhou 221116,China;College of Grassland,Resources and Environment,Inner Mongolia Agricultural University,Huhhot 010011,China;Inner Mongolia Jungar Banner Mining Area Development Center,Ordos 017100,China;College of Spatial Information and Surveying Engineering,Anhui University of Science&Technology,Huainan 232001,China)

机构地区：[1]中国矿业大学矿山生态修复教育部工程研究中心,江苏徐州221116 [2]内蒙古农业大学草原与资源环境学院,内蒙古呼和浩特010011 [3]内蒙古准格尔旗矿区事业发展中心,内蒙古鄂尔多斯017100 [4]安徽理工大学空间信息与测绘工程学院,安徽淮南232001

出　　处：《光谱学与光谱分析》2024年第6期1751-1761,共11页Spectroscopy and Spectral Analysis

基　　金：国家重点研发计划项目(2023YFF1306005),鄂尔多斯科技合作重大专项(2021EEDSCXQDFZ010)资助。

摘　　要：目前,全球生物多样性的状态非常堪忧,因此,利用光谱技术估算生物多样性成为了生态学家和遥感学家共同关注的热点。目前关于植物alpha多样性的研究很多,而关于beta多样性的研究较少,仍存在一些问题值得去探索。为了探究利用遥感技术估算植物beta多样性的最佳光谱指数以及最佳影像空间分辨率,以草甸草原为研究区,基于无人机高光谱遥感影像,从光谱距离、光谱角度、生物多样性概念三个方面计算了6种beta多样性估算指数(4种为我们构建的新指数,2种为已有的指数),并采用Mantel tests和相关系数筛选最佳的光谱指数。然后将筛选出来的指数应用于不同空间分辨率的影像,以期得到最佳观测尺度。另外,为了提高指数的估算能力,对比了一阶导数变换和Savitzky-Golay滤波两种光谱变换方法,以及相关系数法、连续投影法、竞争性自适应重加权法三种特征波段选择方法。结果表明,不论是采用亚尺度观测(像元大小<样方大小)还是等尺度观测(像元大小=样方大小),最佳的光谱指数均为光谱距离指数,且光谱距离指数在不同影像空间分辨率下均表现良好。在草原地区,当影像空间分辨率约为0.25 m时,该指数可以取得最佳的估算结果。经一阶导数变换并用相关系数法提取特征波段后构建的光谱距离指数与beta多样性拥有最强的相关性,今后可利用该指数构建估算模型或者直接表征beta多样性。该研究对于科学的选取光谱指数和影像空间分辨率去估算植物beta多样性具有一定的指导意义。Due to global biodiversity loss,the estimation of biodiversity using spectral technology has become a hot topic for ecologists and remote sensing scientists.There are many studies on alpha diversity but few studies on beta diversity.There are still some problems worth exploring.To explore the best spectral index and image spatial resolution for estimating plant beta diversity using remote sensing technology,this paper took meadow grassland as the research area.It calculated six beta diversity estimation indices from three aspects:spectral distance,spectral angle and biodiversity concept based on UAV hyperspectral remote sensing images.We developed four indices,and two are existing indices.Mantel tests and correlation coefficients were used to select the best spectral index.Then,the selected index was applied to images with different spatial resolutions to obtain the best observation scale.In addition,to improve the estimation ability of the index,this paper compared two spectral transformation methods,the first derivative transform and Savitzky-Golay filter,and three feature band selection methods:correlation coefficient,successive projections algorithm and the competitive adaptive reweighted sampling.The results showed that in both subscale observation(pixel size<quadrat size)and equal scale observation(pixel size=quadrat size),the best spectral index was the spectral distance index,and the spectral distance index performed well under different image spatial resolutions.The best estimation result can be obtained in the grassland area when the image spatial resolution is about 0.25 m.The spectral distance index constructed after the first derivative transformation and extraction of characteristic bands by the correlation coefficient method has the strongest correlation with beta diversity.In the future,this index can be used to build estimation models or directly indicate beta diversity.This paper has guiding significance for scientifically selecting spectral index and image spatial resolution to estimate plant b

关 键 词：beta多样性 高光谱遥感 光谱指数 观测尺度 光谱曲线 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]