EventsThe 1st International Online Conference on Urban Sciences
Published
This submission belongs to the session S5. Urban Resilience and Adaptation of the event The 1st International Online Conference on Urban Sciences
Published date
15 May, 2026
Academic Editor
author-avatarGang Xu
Citation
Sonia Alnajjar, Antonio García-Martínez, Victoria Patricia López-Cabeza, AI-Based Segmentation of Heritage Urban Morphology for Urban Heat Vulnerability (UHV) Assessment in Mediterranean Historic Districts, in Proceedings of The 1st International Online Conference on Urban Sciences, 20 May–22 May 2026, MDPI: Basel, Switzerland
Share
Email
Facebook
Twitter
LinkedIn

AI-Based Segmentation of Heritage Urban Morphology for Urban Heat Vulnerability (UHV) Assessment in Mediterranean Historic Districts

image
image
image
1. Instituto Universitario de Arquitectura y Ciencias de la Construcción IUACC, Escuela Técnica Superior de, Arquitectura, Universidad de Sevilla, 41012 Sevilla, Spain, Spain
2. Departamento de Construcciones Arquitectónicas I, Universidad de Sevilla, Sevilla, Spain, Spain
Abstract

Mediterranean historic districts are commonly perceived as climatically responsive due to compact urban form, shaded street canyons, and courtyard typologies. Yet their actual performance under intensifying heat extremes remains insufficiently evaluated within multidimensional Urban Heat Vulnerability (UHV) frameworks. This study develops an artificial intelligence–assisted spatial modelling approach to examine how heritage urban fabric influences heat vulnerability across two climatically comparable but morphologically distinct contexts: Amman and Seville.

High-resolution spatial data are integrated with three-dimensional urban form metrics derived from digital surface models, alongside thermal indicators and socio-demographic proxies structured under the exposure–sensitivity–adaptive capacity paradigm. Fine-scale segmentation of urban fabric enables the delineation of coherent morphological units, facilitating micro-scale analysis of enclosure ratios, canyon geometry, vegetation distribution, and topographic modulation. These spatial descriptors are subsequently incorporated into predictive vulnerability modelling to quantify the extent to which urban morphology mediates heat exposure and adaptive capacity.

Comparative findings indicate that while compact heritage configurations may attenuate direct solar gain in certain geometries, limited vegetation coverage, material thermal inertia, and socio-economic constraints can amplify the risk of residual heat. Marked differences between the morphology of Amman and the courtyard-dominated fabric of Seville underscore the critical role of three-dimensional structure in shaping intra-urban thermal inequities. By integrating urban fabric analysis within a vulnerability-oriented framework, the study advances a transferable methodology for evaluating climate resilience in historic districts. It supports heritage-sensitive adaptation strategies in rapidly warming Mediterranean cities.

Keywords
Urban Heat Vulnerability (UHV)
Heritage Urban Fabric
Artificial Intelligence
Spatial Modelling
Mediterranean Historic Districts
Climate Resilience
A Participatory Chatbot for Strengthening Urban Flood and Landslide Resilience: Integrating Community Awareness and Planning Decision Support
FairGNN-Climate: Fairness-Constrained Graph Neural Networks for Urban Neural Resilience Under Climate Stress