Exploring the Impact of Architectural Landscape Characteristics of Urban Functional Areas in Xi’an City on the Thermal Environment in Summer Using Explainable Machine Learning

Jiayue Xu, Le Xuan, Cong Li, Mengxue Zhang and Xuhui Wang ()
Additional contact information
Jiayue Xu: College of Landscape Architecture & Arts, Northwest A&F University, Yangling 712100, China
Le Xuan: School of Architecture and Fine Art, Dalian University of Technology, Dalian 116024, China
Cong Li: College of Landscape Architecture & Arts, Northwest A&F University, Yangling 712100, China
Mengxue Zhang: College of Landscape Architecture & Arts, Northwest A&F University, Yangling 712100, China
Xuhui Wang: College of Landscape Architecture & Arts, Northwest A&F University, Yangling 712100, China

Sustainability, 2025, vol. 17, issue 14, 1-27

Abstract: Rapid urbanization has exacerbated the urban heat island effect, posing a significant threat to human health and urban ecosystems. While numerous studies have demonstrated that urban morphology significantly influences land surface temperatures (LSTs), few have systematically explored the impact and contribution of urban morphology on LST across different functional zones. Therefore, this study takes Xi’an as a case and employs an interpretable CatBoost-SHAP machine learning model to evaluate the nonlinear influence of building landscape features on LST in different functional zones during summer. The results indicate the following: (1) The highest LST in the study area reached 52.68 °C, while the lowest was 21.68 °C. High-temperature areas were predominantly concentrated in the urban center and industrial zones with dense buildings, whereas areas around water bodies and green spaces exhibited relatively lower temperatures. (2) SHAP analysis revealed that landscape indicators exerted the most substantial impact across all functional zones, with green space zones contributing up to 62%. Among these, fractional vegetation coverage (FVC), as a core landscape factor, served as the primary cooling factor in all six functional zones and consistently demonstrated a negative effect. (3) Population density (POP) exhibited a generally high SHAP contribution across all functional zones, showing a positive correlation. Its effect was most pronounced in commercial zones, accounting for 16%. When POP ranged between 0 and 250 people, the warming effect was particularly prominent. (4) The mean building height (MBH) constituted a major influencing factor in most functional zones, especially in residential zones, where the SHAP value reached 0.7643. Within the range of 10–20 m, the SHAP value increased sharply, indicating a significant warming effect. (5) This study proposes targeted cooling strategies tailored to six functional zones, providing a scientific basis for formulating targeted mitigation strategies for different functional zones to alleviate the urban heat island effect.

Keywords: central urban area; surface temperature; multi-dimensional urban form; urban functional zone; interpretable machine learning (search for similar items in EconPapers)
JEL-codes: O13 Q Q0 Q2 Q3 Q5 Q56 (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2071-1050/17/14/6489/pdf (application/pdf)
https://www.mdpi.com/2071-1050/17/14/6489/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jsusta:v:17:y:2025:i:14:p:6489-:d:1702400

Access Statistics for this article

Sustainability is currently edited by Ms. Alexandra Wu

More articles in Sustainability from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().