Landscape Classifications for Landscape Metrics-based Assessment of Urban Heat Island: A Comparative Study

35th International Symposium on Remote Sensing of Environment

地球科学;生态环境科学

Zhao, X.F.^1,2 ; Deng, L.^1,2 ; Wang, H.N.^1,2 ; Hua, L.Z.^3 ; Chen, F.^1,2

Key Lab of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China^1

Xiamen Key Lab of Urban Metabolism, Xiamen 361021, China^2

Institute of Spatial Information Technology, Xiamen University of Technology, Xiamen, China^3

关键词: Comparative studies;    Land surface temperature;    Landscape analysis;    Landscape metrics;    Standard deviation method;    Urban heat island;    Urban Heat Island Effects;    Visual interpretation;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/17/1/012155/pdf
DOI  :  10.1088/1755-1315/17/1/012155

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

In recent years, some studies have been carried out on the landscape analysis of urban thermal patterns. With the prevalence of thermal landscape, a key problem has come forth, which is how to classify thermal landscape into thermal patches. Current researches used different methods of thermal landscape classification such as standard deviation method (SD) and R method. To find out the differences, a comparative study was carried out in Xiamen using a 20-year winter time-serial Landsat images. After the retrieval of land surface temperature (LST), the thermal landscape was classified using the two methods separately. Then landscape metrics, 6 at class level and 14 at landscape level, were calculated and analyzed using Fragstats 3.3. We found that: (1) at the class level, all the metrics with SD method were evened and did not show an obvious trend along with the process of urbanization, while the R method could. (2) While at the landscape level, 6 of the 14 metrics remains the similar trends, 5 were different at local turn points of the curve, 3 of them differed completely in the shape of curves. (3) When examined with visual interpretation, SD method tended to exaggerate urban heat island effects than the R method.