传统景观格局指数在城市热岛效应评价中的适用性  

Applicability of traditional landscape metrics in evaluating urban heat island effect

作　　者：陈爱莲[1,2] 孙然好[1] 陈利顶[1] 

机构地区：[1]中国科学院生态环境研究中心城市与区域生态国家重点实验室,北京100085 [2]中国科学院研究生院,北京100049

出　　处：《应用生态学报》2012年第8期2077-2086,共10页Chinese Journal of Applied Ecology

基　　金：国家自然科学基金项目(40925003);国家水体污染控制与治理科技重大专项(2012ZX07501002);城市与区域生态国家重点实验室自主项目(SKLURE2008-1-02)资助

摘　　要：以北京部分城区为研究对象,以QuickBird影像制作景观类型图,基于同年4个季节的LandsatETM+数据反演地表温度;将120m×120m作为固定窗口,计算其中的景观格局指数,探寻传统景观格局指数解释地表温度的适用性.结果表明:在景观水平计算的24个景观格局指数中,只有景观组成百分比(PLAND)、斑块密度(PD)、最大斑块指数(LPI)、欧氏距离变异系数(ENN_CV)和分离度(DIVISION)与3月、5月、11月的地表温度具有稳定的显著相关.在类型水平计算的24个景观格局指数中,PLAND、LPI、DIVISION、相似邻接百分比、分散与并列指数与4个时相(3月、5月、7月和12月)的温度显著相关,且与7月温度的相关性最强;斑块密度、边界密度、聚簇度、凝聚度、有效MESH大小、分裂度、聚合度、归一化景观形状指数依据不同景观类型而与地表温度呈现相关性.传统景观格局指数可能并不适合评估河流对地表温度的影响.一些景观格局指数可以用来表征城市地表温度,辅助分析城市地表热岛效应,但需要对其进行筛选和甄别.By using 24 landscape metrics,this paper evaluated the urban heat island effect in parts of Beijing downtown area.QuickBird（QB） images were used to extract the landscape type information,and the thermal bands from Landsat Enhanced Thematic Mapper Plus（ETM＋） images were used to extract the land surface temperature（LST） in four seasons of the same year.The 24 landscape pattern metrics were calculated at landscape and class levels in a fixed window with 120 m×120 m in size,with the applicability of these traditional landscape metrics in evaluating the urban heat island effect examined.Among the 24 landscape metrics,only the percentage composition of landscape（PLAND）,patch density（PD）,largest patch index（LPI）,coefficient of Euclidean nearest-neighbor distance variance（ENN_CV）,and landscape division index（DIVISION） at landscape level were significantly correlated with the LST in March,May,and November,and the PLAND,LPI,DIVISION,percentage of like adjacencies,and interspersion and juxtaposition index at class level showed significant correlations with the LST in March,May,July,and December,especially in July.Some metrics such as PD,edge density,clumpiness index,patch cohesion index,effective mesh size,splitting index,aggregation index,and normalized landscape shape index showed varying correlations with the LST at different class levels.The traditional landscape metrics could not be appropriate in evaluating the effects of river on LST,while some of the metrics could be useful in characterizing urban LST and analyzing the urban heat island effect,but screening and examining should be made on the metrics.

关 键 词：景观格局指数 城市热岛 景观格局 地表温度 北京 

分 类 号：P427.31[天文地球—大气科学及气象学] Q149[生物学—生态学]