详细信息
基于多角度高光谱数据的高寒沼泽湿地植被生物量估算 被引量:7
Estimation the Biomass of Alpine Wetland Vegetation based on Multi-angle and Hyperspectral Data
文献类型:期刊文献
中文题名:基于多角度高光谱数据的高寒沼泽湿地植被生物量估算
英文题名:Estimation the Biomass of Alpine Wetland Vegetation based on Multi-angle and Hyperspectral Data
作者:李伟娜[1] 韦玮[2] 张怀清[1] 刘华[1]
第一作者:李伟娜
机构:[1]中国林业科学研究院资源信息所;[2]中国林业科学研究院湿地研究所
年份:2017
卷号:32
期号:5
起止页码:809-817
中文期刊名:遥感技术与应用
外文期刊名:Remote Sensing Technology and Application
收录:CSTPCD;;北大核心:【北大核心2014】;CSCD:【CSCD2017_2018】;
基金:国家自然科学基金项目(31370712);国家高分重大专项课题(21-Y30B05-9001-13/15-2)
语种:中文
中文关键词:高光谱;多角度;生物量;湿地植被
外文关键词:Hyperion ; Multi-angle ; Biomass ; Wetland vegetation
分类号:TP79
摘要:湿地植被生物量是湿地生态系统健康状况的重要指标,直接反映植被群落的生长状况和生产力高低。估测高寒沼泽湿地植被生物量有助于理解高寒沼泽生态系统与全球气候变化之间的反馈关系。以三江源隆宝滩自然保护区为研究对象,使用欧空局CHRIS/PROBA数据,利用高光谱光谱维信息和多角度立体结构信息,分析不同角度光谱反射率、窄波段植被指数、红边指数以及主成分分量与湿地植被地上干、鲜生物量的相关性,探讨生物量对角度的敏感性。结果表明:高寒沼泽湿地植被地上生物量对角度具有一定的敏感性,前向观测的+36°影像优于0°和-36°影像;-36°红边位点指数REIP对地上干生物量、鲜生物量的拟合效果最好,干生物量的R^2为0.599,F值为37.404;鲜生物量的R^2高达0.685,F值达54.410。研究区内最大地上生物量干重为446.7g/m^2,湿重为2 368.1g/m^2。
Wetland vegetation biomass is a significant indicator of the health condition of wetland ecosystem,which directly mirrors the growth and productivity of vegetation communities.The estimation of vegetation biomass in alpine wetland contribute to understanding the causal feedback relations between alpine swamp ecosystem and global climate change.Longbaotan nature reserve in Sanjiangyuan area was taken as the research region.Taking advantage of hyper spectral dimension and multi-angle stereoscopic structure information of CHRIS/PROBA,we analyzed the correlation between wetland vegetation biomass and some remote sensing factors,including spectral reflectivity,narrow band vegetation indices,red edge indices and principal component.The angular sensitivity of biomass was explored at the same time.The results showed that the biomass of vegetation on the alpine wetland was sensitive to satellite observation angle.The forward observation of the +36°degree image was significantly better than 0 and-36°degree images.Besides,the exponential model was the best regression model in the nonlinear models.Among them,REIP index of-36°degree was the fittest variable to biomass,which R^2 was 0.599 and Fvalue was 37.404 in dry biomass,and R^2 was 0.685 and Fvalue was 54.410 in fresh biomass.The maximum of dry biomass in Longbaotan area is 446.7 g/m^2,and fresh biomass is 2 368.1 g/m^2.
参考文献:
正在载入数据...