Scale-dependent and season-dependent impacts of 2D/3D building morphology on land surface temperature

• Published in: Sustainable cities and society 2023-10, Vol.97, p.104788, Article 104788
• Main Authors: Guo, Fengxiang, Schlink, Uwe, Wu, Wanben, Hu, Die, Sun, Jiayue
• Format: Article
• Language: English
• Subjects: 3D landscape metrics; Land surface temperature; Scales; Seasons; Tree height; Urban built forms
• ISSN: 2210-6707
• DOI: 10.1016/j.scs.2023.104788

•3D landscape metrics for quantifying urban morphology.•Tree height contributed mostly to LST during warm months.•Composition indicators were more sensitive to LST than configuration.•3D and 2D indicators contributed higher to LST at small and large scales, respectively. Urbanization has caused significant changes in urban morphology in three-dimensional (3D) space. Although previous studies found a close association between 2D/3D urban morphology and land surface temperature (LST), the conclusions are various and the reasons lie in three fundamental components of urban morphology: physical form, resolution and time. Beijing's old city includes massive vernacular and modern architecture, which provides an ideal laboratory for the studies. In a new approach, 3D landscape metrics were applied to analyze built forms from the aspect of composition and configuration, and to explain their contributions to LSTs together with tree height data across scales and seasons. Tree height contributed most during summer. Compared to composition, configuration metrics were less efficient for cross-scale LST changes. The relative importance of building features is scale-dependent. At small scales (under 105 m), 3D composition features contributed more to LST, while 2D composition features were dominant at larger scales (over 180 m). At intermediate scales, 2D and 3D composition metrics together affect the LSTs. The scale and seasonal dependencies indicated in this study are derived from the application of our concept to a particular city. More experiments to other cities should be carried out for an in-depth understanding of the associations.