基于遥感和无人机数据的草地NDVI影响因子多尺度分析

为探索不同尺度草地植被的影响因子，本研究基于Landsat遥感影像和无人机多光谱影像等，在村落和地块2个尺度分析西藏草地归一化植被指数（Normalized Difference Vegetation Index，NDVI）的影响因子。结果表明：村落尺度的海拔、坡度、坡向、地表起伏度皆对NDVI有显著的非线性影响，解释比例为37.20%；方差分析表明，地形等因子相近的同类土地利用内部NDVI差异仍较大；运用无人机影像细分同种土地利用类型内部异质性，发现遥感影像中无法辨别的点状、线状地物（石堆、水渠、田间路等）对草地NDVI有一定影响；缓冲区分析表明，非硬化水渠和道路对NDVI的负面影响在1~3 m，硬化水渠和路面对草地的负面影响超过4 m。无人机获取高分辨率多光谱影像的便捷性可以推进更小尺度下人类活动强度以及景观破碎化对植被和生态系统功能影响研究的深入。

Multi-scales Analysis of the Impacting Factors of Grassland NDVI Based on Remote Sensing and Unmanned Aerial Vehicle Data

In order to clarify the impacting factors of grassland vegetation on different scales,this paper research on the impacting factors of NDVI of grassland ecosystem at both village and plot scales in a Tibetan village,based on multiple data sources including Landsat remote sensing images and unmanned aerial vehicle multispectral images. The results showed that,altitude,slope,aspect and relief degree of land surface all have impacts on vegetation index,with a nonlinear model with an interpreting level of 37.20%. The variance analysis showed heterogeneity in vegetation indices also existed within one same land use type. The heterogeneity within the same land use type was subdivided by unmanned aerial vehicle multispectral images,we found that the indiscernible the punctiform or linear ground objects (e.g. stone packs,channels and field roads) impact on the vegetation indices largely. The buffer analysis showed that,the negative impacts of non-concreted channels and field road on the around vegetation ranged from 1 to 3 meters,while the negative impacts of concreted channels and field road were more than 4 meters. Our research showed,the convenience of achieving high resolution images by unmanned aerial vehicle would boost the deepen the study of how human activities and landscape fragmentation impacting on the vegetation and ecosystem functions,at the fine scale.