Fei Hou
University of Science Malaysia

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

XGBoost modelling of urban green space morphology and summer daytime land surface temperature Fei Hou; Nooriati Taib; Massoomeh Hedayati Marzbali; Aldrin Abdullah
ARTEKS : Jurnal Teknik Arsitektur Vol 11 No 3 (2026): ARTEKS : Jurnal Teknik Arsitektur | July 2026 ~ September 2026
Publisher : Program Studi Arsitektur Fakultas Teknik Universitas Katolik Widya Mandira

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30822/arteks.v11i3.5285

Abstract

Rapid urbanisation has intensified summer daytime land surface temperature (LST) in many cities, yet the influence of detailed green space morphology on surface heat remains unclear. This study examines how different morphological classes of urban green space affect summer daytime LST in Shanghai. Land use and land cover data from ESRI and ECOSTRESS LST products for 2024 were aggregated into a 1 km × 1 km grid. Morphological Spatial Pattern Analysis was used to classify urban green space into core, edge and branch classes, and an explainable machine learning model with SHAP values was applied to quantify variable importance, nonlinear responses and interactions. The core and edge classes emerged as the main controls and together accounted for more than 60% of total importance, while the branch class provided additional, secondary cooling. Partial dependence results showed that higher values of core and edge classes were associated with an almost monotonic decrease in land surface temperature, and interaction analysis indicated that the lowest temperatures occurred where core and edge were both high. These findings highlight the central role of internal green space morphology in mitigating summer daytime surface heat and offer a structural basis for climate responsive urban planning.