EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Explaining and Reducing Urban Heat Islands Through Machine Learning: Evidence from New York City

Shengyao Liao, Zhewei Liu

Buildings · 01.01.2026 · 16 (1) · 186

DOI: 10.3390/buildings16010186 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Buildings, online erschienen am 01.01.2026. Verfasst von Shengyao Liao und 1 weiteren Autor:innen. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 54 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Urban heat islands (UHIs) have intensified in rapidly urbanizing regions like New York, exacerbating thermal discomfort, public health risks, and energy consumption. While previous research has highlighted various environmental and socioeconomic contributors, most existing studies lack interpretable, fine-scale models capable of quantifying the effects of specific drivers—limiting their utility for targeted planning. To address this challenge, we develop an interpretable machine learning framework using Random Forest and XGBOOST to predict land surface temperature across 1800+ census tracts in the New York metropolitan area, incorporating vegetation indices, water proximity, urban morphology, and socioeconomic factors. Both models performed strongly (mean R2 ≈ 0.90), with vegetation coverage and water proximity emerging as the most influential cooling factors, while built form features played supporting roles. Socioeconomic vulnerability indicators showed weak correlations with temperature, suggesting a relatively equitable thermal landscape. Optimization simulations further revealed that increasing vegetation to a threshold level could lower average surface temperatures by up to 6.38 °C, with additional but smaller gains achievable through adjustments to water access and urban form. These findings provide evidence-based guidance for climate-adaptive urban design and green infrastructure planning. More broadly, the study illustrates the potential of explainable machine learning to support data-driven environmental interventions in complex urban systems.

Zitieren

APA

Liao, S., & Liu, Z. (2026). Explaining and Reducing Urban Heat Islands Through Machine Learning: Evidence from New York City. Buildings, 16(1), 186. https://doi.org/10.3390/buildings16010186

Vancouver

Liao S, Liu Z. Explaining and Reducing Urban Heat Islands Through Machine Learning: Evidence from New York City. Buildings. 2026;16(1):186. doi:10.3390/buildings16010186

BibTeX
@article{Liao2026,
  author = {Liao, Shengyao and Liu, Zhewei},
  title = {Explaining and Reducing Urban Heat Islands Through Machine Learning: Evidence from New York City},
  journal = {Buildings},
  year = {2026},
  volume = {16},
  number = {1},
  pages = {186},
  doi = {10.3390/buildings16010186}
}
Kostenlos starten →