Planning for Cooler Cities: A Multimodal AI Framework for Predicting and Mitigating Urban Heat Stress through Urban Landscape Transformation

2025-07-30

AI Summary

Addressing the challenge of escalating heat stress due to climate change and urbanization—and the inability of traditional physical models (e.g., SOLWEIG, ENVI-met) to support citywide, high-resolution thermal environmental planning—this paper proposes GSM-UTCI, a multimodal deep learning framework. GSM-UTCI integrates normalized digital surface model (nDSM), high-resolution land cover, and hourly meteorological data, employing a Feature-wise Linear Modulation (FiLM) architecture for dynamic spatial feature modulation to predict daytime mean Universal Thermal Climate Index (UTCI) at 1-meter resolution. Compared to physics-based models, it achieves two orders-of-magnitude faster inference (≤5 minutes citywide) with R² = 0.9151 and MAE = 0.41°C. Applied in Philadelphia, it quantifies that converting impervious surfaces to tree canopy reduces mean UTCI by 4.18°C, revealing both substantial cooling benefits and pronounced spatial heterogeneity of green infrastructure. This work establishes a novel paradigm for fine-grained, scalable urban climate adaptation planning.

Technology Category

• Machine Learning: Multimodal Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

As extreme heat events intensify due to climate change and urbanization, cities face increasing challenges in mitigating outdoor heat stress. While traditional physical models such as SOLWEIG and ENVI-met provide detailed assessments of human-perceived heat exposure, their computational demands limit scalability for city-wide planning. In this study, we propose GSM-UTCI, a multimodal deep learning framework designed to predict daytime average Universal Thermal Climate Index (UTCI) at 1-meter hyperlocal resolution. The model fuses surface morphology (nDSM), high-resolution land cover data, and hourly meteorological conditions using a feature-wise linear modulation (FiLM) architecture that dynamically conditions spatial features on atmospheric context. Trained on SOLWEIG-derived UTCI maps, GSM-UTCI achieves near-physical accuracy, with an R2 of 0.9151 and a mean absolute error (MAE) of 0.41°C, while reducing inference time from hours to under five minutes for an entire city. To demonstrate its planning relevance, we apply GSM-UTCI to simulate systematic landscape transformation scenarios in Philadelphia, replacing bare earth, grass, and impervious surfaces with tree canopy. Results show spatially heterogeneous but consistently strong cooling effects, with impervious-to-tree conversion producing the highest aggregated benefit (-4.18°C average change in UTCI across 270.7 km2). Tract-level bivariate analysis further reveals strong alignment between thermal reduction potential and land cover proportions. These findings underscore the utility of GSM-UTCI as a scalable, fine-grained decision support tool for urban climate adaptation, enabling scenario-based evaluation of greening strategies across diverse urban environments.

Problem

• Predict urban heat stress at high resolution for city planning
• Overcome computational limits of traditional physical heat models
• Evaluate cooling effects of urban landscape transformation strategies

Innovation

• Multimodal deep learning predicts hyperlocal UTCI
• FiLM architecture fuses spatial and meteorological data
• Enables rapid urban heat mitigation scenario evaluation