摘要
基于闽南地区(厦门、漳州、泉州)2013~2020年Landsat8(OLI)遥感数据和路网数据,运用Sen+Mann-Kendall法分析闽南地区的植被恢复动态,在此基础上,利用缓冲区分析和剖面线分析探讨植被恢复动态对道路网络的响应规律;并在不同尺度的单元网格(500m×500m、1000m×1000m、1500m×1500m、2000m×2000m)上,采用双变量空间自相关和地理加权回归探讨两者之间的空间耦合关系.结果表明:2013~2020年,闽南地区植被变化动态在空间分布上存在明显差异,呈现东部沿海地区退化、西部内陆改善的趋势,退化和改善面积分别占研究区47.93%、50.07%.缓冲区分析表明,三市的道路密度与植被恢复状态之间的关系大致相似,都是呈现东部沿海地区道路密度高但植被退化,西部内陆地区道路密度低但植被改善的格局;剖析线分析表明,在海岸线附近区域、道路核密度值较高以及核密度动态变化区域,植被指数均存显现降低的趋势.多尺度双变量空间自相关分析表明,在1000m×1000m和1500m×1500m这两种网格单元尺度下的空间集聚性较强,闽南三市的道路网络与植被及其恢复动态的双变量Moran'sI均为负值,道路网络与植被、植被恢复状态呈负相关性关系,且HL、LH聚集类型空间分布逐渐扩张,连片趋势明显.多尺度的地理加权回归分析表明,道路网络对植被恢复动态的影响存在空间分异特征,道路网络对植被的负面影响在漳州、泉州呈现扩张趋势,而道路网络对植被恢复在安溪、华安等局部地区起到正向影响作用,但总体上道路网络对植被的恢复起到抑制作用.研究结果可为研究区的植被恢复以及道路网络建设与发展提供参考.
Based on Landsat8(OLI)remote sensing data and road network data from 2013 to 2020 in southern Fujian(Xiamen,Zhangzhou,Quanzhou),the Sen+Mann-Kendall method was used to analyze the vegetation restoration dynamics in southern Fujian.On this basis,the response of dynamic vegetation restoration to road network was studied by using buffer analysis and profile analysis.In addition,bivariate spatial autocorrelation and geographically weighted regression were used to explore the spatial coupling relationship between cell grids of different scales(500m×500m,1000m×1000m,1500m×1500m,2000m×2000m).The results shows that from 2013 to 2020,there were significant differences in the spatial distribution of vegetation change dynamics in southern Fujian,showing a trend of degradation in the eastern coastal areas and improvement in the western inland areas.The area of degradation and improvement accounted for 47.93%and 50.07%of the study area.The buffer zone analysis shows that the relationship between road density and vegetation restoration state in the three cities is roughly similar,showing a pattern of high road density but vegetation degradation in the eastern coastal area and low road density but vegetation improvement in the western inland area.Along the section line,the vegetation index decreases in the area near the coastline,the area with high density of road kernel and the area with dynamic change of kernel density.The multi-scale bivariate spatial autocorrelation analysis shows that the spatial agglomeration of the two sampling units of 1500m×1500m and 2000m×2000m is higher than that of the other sampling units.The bivariate Moran's I of road network,vegetation and vegetation restoration state in the three cities of southern Fujian are all negative values.The road network is negatively correlated with vegetation and vegetation restoration state as well as negatively correlated with space.The HL(High-Low)-type and LH(Low-High)-type aggregation areas of road network and vegetation gradually expanded,and the trend of linking is obvious.The results of multi-scale geographical weighted regression analysis show that the road network has a spatial differentiation of the impact on the dynamic vegetation restoration.The negative impact of road network on vegetation expansion trend in Zhangzhou and Quanzhou,and the road network has a positive effect on vegetation restoration in local areas such as Anxi and Hua'an,but on the whole,the road network has a inhibitory effect on vegetation restoration.The research results can provide reference for vegetation restoration and road network construction and development in the study area.
作者
吴熙
真诗泳
左雪漫
胡喜生
沈嵘枫
WU Xi;ZHEN Shi-yong;ZUO Xue-man;HU Xi-sheng;SHEN Rong-feng(College of Transportation and Civil Engineering,Fujian Agriculture and Forestry University,Fuzhou 350002,China)
出处
《中国环境科学》
CSCD
北大核心
2023年第S1期288-300,共13页
China Environmental Science
基金
国家自然科学基金资助项目(31971639)
福建省自然科学基金资助项目(2019J01406)
关键词
NDVI
植被恢复
道路网络
空间自相关
地理加权回归
NDVI
vegetation restoration
road network
spatial autocorrelation
geographically weighted regression
