EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Optimizing Courtyard Vegetation for Outdoor Thermal Comfort in Hot-Summer and Cold-Winter Regions: A Machine Learning and SHAP Approach

Bailing Zhou, Ziheng Yu, Niejiaying Xie

Forests · 11.11.2026 · 17 (11) · 1345

DOI: 10.3390/f17111345 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Forests, online erschienen am 11.11.2026. Verfasst von Bailing Zhou und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Humanities and Social Sciences Research Project of the Department of Education of Hubei Province, China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Vegetation can mitigate courtyard heat stress, but nonlinear and interactive effects of canopy amount, layout continuity, and species composition remain poorly quantified for semi-enclosed courtyards in hot-summer and cold-winter regions. Using a single representative courtyard in Wuhan, this study coupled ENVI-met /RayMan Pro simulation with a Random Forest surrogate and SHAP interpretation to examine how tree canopy coverage (TCC), layout pattern (LP), and evergreen-to-deciduous ratio (ETR) affect physiological equivalent temperature (PET). Twenty greening scenarios were simulated over five representative meteorological days (2400 hourly samples). The surrogate achieved R2 = 0.98 and RMSE = 1.28 °C. Meteorological and solar variables contributed >93% of global SHAP attribution, whereas TCC, LP, and ETR together contributed 6.3%. Vegetation–PET relationships were condition-specific, and SHAP interactions were non-monotonic. Matched-pair simulations holding TCC and ETR constant showed that layout effects involve spatial continuity and orientation relative to building shade and airflow, not canopy amount alone. Relative to the baseline ATCP of 11.33%, the best tested configuration reached 19.80% (+8.47 percentage points), and all configurations reduced absolute PET deviation. These findings provide a case-conditional screening framework; generalization requires multi-courtyard validation.

Förderung

  • Humanities and Social Sciences Research Project of the Department of Education of Hubei Province, China

Zitieren

APA

Zhou, B., Yu, Z., & Xie, N. (2026). Optimizing Courtyard Vegetation for Outdoor Thermal Comfort in Hot-Summer and Cold-Winter Regions: A Machine Learning and SHAP Approach. Forests, 17(11), 1345. https://doi.org/10.3390/f17111345

Vancouver

Zhou B, Yu Z, Xie N. Optimizing Courtyard Vegetation for Outdoor Thermal Comfort in Hot-Summer and Cold-Winter Regions: A Machine Learning and SHAP Approach. Forests. 2026;17(11):1345. doi:10.3390/f17111345

BibTeX
@article{Zhou2026,
  author = {Zhou, Bailing and Yu, Ziheng and Xie, Niejiaying},
  title = {Optimizing Courtyard Vegetation for Outdoor Thermal Comfort in Hot-Summer and Cold-Winter Regions: A Machine Learning and SHAP Approach},
  journal = {Forests},
  year = {2026},
  volume = {17},
  number = {11},
  pages = {1345},
  doi = {10.3390/f17111345}
}
Kostenlos starten →