Identifying urban morphological archetypes for microclimate studies using a clustering approach

Abstract

Urban morphology relates to the form, structure, physical characteristics, and arrangement of buildings affecting the urban microclimate. As the morphological characteristics vary across the city, small units such as urban blocks are analysed for microclimate estimation. However, microclimatic analysis of all the blocks in a city is computationally challenging and time-consuming. Therefore, it is vital to identify representative blocks in a city to obtain a general overview of the microclimate. Urban morphological archetypes are the representative units of a homogenous group of blocks based on morphological parameters. Here, we propose a systematic approach for identifying urban morphological archetypes suited for microclimatic analysis. Specifically, we employ a well-defined, PCA-based k-means clustering approach supported by validation using external criterion analysis. We use urban morphological parameters based on form, shape, arrangement, and variations within a block in Liege, Belgium. We use the cubic clustering criterion and pseudo F statistic to identify nine distinct homogenous clusters. Then, we propose a validation approach in the absence of existing typologies using ANOVA analysis on the external criterion of land surface temperature, a proxy for measuring microclimate. The validation suggests that the clusters are significantly different, indicating successful clustering. We also compare our classification to the existing local climate zone (LCZ) classification. We identify relevant sub-classes within the broader LCZ classes essential for capturing microclimatic variation. Finally, the study provides realistic archetypes for performing microclimatic simulations at a city scale. The proposed approach can be effectively applied to other cities for urban microclimate studies.