自然资源学报 ›› 2015, Vol. 30 ›› Issue (3): 480-490.doi: 10.11849/zrzyxb.2015.03.011

• 资源评价 • 上一篇    下一篇

城市景观格局时空特征的热环境效应研究——以广州市花都区为例

黄丽明, 陈健飞   

  1. 广州大学地理科学学院, 广州510006
  • 收稿日期:2014-03-24 修回日期:2014-07-11 发布日期:2015-03-23
  • 通讯作者: 陈健飞,E-mail:cjf@gzhu.edu.cn E-mail:cjf@gzhu.edu.cn
  • 作者简介:黄丽明(1987-),女,广东广州人,硕士研究生,主要从事土地利用规划与管理的研究.Email:huanglm13@163.com
  • 基金资助:

    国家自然科学基金项目“环珠江口区域城市扩张及其环境生态效应分析与模拟”(41171446).

Study on Thermal Environment Effect of Urban Landscape Pattern: A Case Study on Huadu District, Guangzhou

HUANG Li-ming, CHEN Jian-fei   

  1. College of Geographic Sciences, Guangzhou University, Guangzhou 510006, China
  • Received:2014-03-24 Revised:2014-07-11 Published:2015-03-23

摘要:

1999—2009 年,广州市花都区经历了经济腾飞的10 a,不断变迁的城市景观格局时刻影响着热环境的分布.论文基于Landsat TM/ETM+遥感数据,获得地温反演图和土地利用类型图.采用均值-标准差法划分热力等级以分析热力重心变化,并与建设用地重心变化作对比.通过移动窗口得出可视化的景观指数,指数刻画包括景观多样性和建设用地密度、形状和分布状况,进而分析景观指数与地温间的相关性.研究结果表明:区内南部受人类活动影响最为剧烈;土地利用类型并非是影响地温分布的唯一因素;区内城市景观格局多样性与地温虽存在正相关关系,但随着多样性程度加剧且各组合面积比例均匀,降低了两者相关程度;建设用地面积比例、分离度与地温的相关关系较为稳定,且比例较高或分离度较低的所在区域通常为高温区;建设用地平均分维指数与地温的相关性随着其形状的复杂性加剧和分布范围扩大而降低,即影响了高温形成的程度.

关键词: 景观格局, 热环境, 移动窗口, 单窗算法, 花都

Abstract:

Huadu experienced economic soaring during 1999-2009. The ever-changing urban landscape pattern impacted the distribution of the thermal environment. The land surface temperature (LST) and land use types in the study area were obtained from Landsat TM/ETM+ remote sensing images. Using the mean-standard deviation method, land surface temperatures were classified and the thermal centroids of low and high temperatures were calculated. And the moving paths of the thermal centroids during 1999-2009 were compared with that of the construction land centroid. Visual landscape metrics, including SHDI, construction land PLAND, FRAC_MN, and DIVISION, were obtained by moving window to analyze their correlations with LST. The results show that: the southern part of Huadu was strongly impacted by human activities. Land use was not the single factor which influenced the distribution of land surface temperature. SHDI and LST were positively correlated, even though the increasing degree of landscape diversity would cause the decrease of the relevance. The correlation between LST and construction land PLAND/DIVISION was stable. The regions with large construction land PLAND or low construction land DIVISION were usually high temperature regions. With the increase of the shape complexity and the area of the construction lands, the correlationship between LST and construction land FRAC_MN become stronger.

Key words: landscape pattern, thermal environment, moving window, mono-window algorithm, Huadu

中图分类号: 

  • X16