Role of urban green space structure and configuration in regulating land surface temperature in NCT Delhi using explainable artificial intelligence
No Thumbnail Available
Date
2026
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Urbanization-driven heat intensification poses a serious challenge to environmental sustainability
and thermal comfort in megacities such as National Capital Territory (NCT) Delhi. The influence
of green space structure and configuration on Land Surface Temperature (LST) remains under
explored. This study presents a novel framework that integrates Fragstats-derived green space
metrics and explainable artificial intelligence (XAI) to assess the spatio-temporal impact of green
space structure and configuration on LST in NCT Delhi. The LST and nine green space metrics
were derived using the Landsat-8 (OLI/TIRS) imagery. Three machine and deep learning models i.
e., Gradient Boosting Machine (GBM), Distributed Random Forest (DRF), and Deep Learning (DL)
were developed to regress and predict LST using H2O AutoML package in Rstudio environment.
Among these, GBM produced the highest predictive accuracy (R
2
= 0.8859) and was therefore
selected for further interpretation using XAI techniques such as SHapley Additive exPlanations
(SHAP) and Individual Conditional Expectation (ICE) plots. Results show that fragmented green
spaces intensify surface heating, while cohesive and well-connected green space corridors pro
mote cooling through enhanced evapotranspiration and shading. The findings highlight that not
only the amount but also the spatial configuration of vegetation determines its cooling efficiency.
Areas in East Delhi (Shahdara, Seelampur) and North-West Delhi (Narela, Bawana, Rohini
Extension), characterized by high fragmentation, experience higher temperatures, whereas
contiguous green space corridors of Central and South Delhi exhibit stronger cooling benefits.
These insights provide actionable guidance for urban planners and policymakers to prioritize
cohesive green space development in future urban planning for climate adaptation and
mitigation.