Date of Defense
27-11-2025 11:00 AM
Location
F1-2126
Document Type
Dissertation Defense
Degree Name
Doctor of Philosophy in Architectural Engineering
College
COE
Department
Architectural Engineering
First Advisor
Dr. Martin Scoppa
Keywords
Urban form classification, Arab Gulf Urbanization, Superblock Densification, Gaussian Mixture Model Clustering, Urban Energy Modelling, Parametric methods
Abstract
Urban form plays a critical role in shaping environmental performance, particularly building energy. In Abu Dhabi and other GCC cities, development is structured around superblocks—large arterial-bounded land parcels units derived from NPU planning principles. Rapid urban expansion, rising energy demand, and the predominance of low-density suburban growth (now exceeding 55% of Abu Dhabi’s footprint) highlight the need for frameworks that can evaluate the energy implications of this morphology and guide both new development and retrofitting strategies. Superblocks therefore offer a consistent and policy-relevant scale for analysing form–energy interactions in the GCC’s hot-desert context, where early-stage, morphology-informed planning tools are essential. Despite extensive research on form–energy relationships, significant conceptual, methodological, and contextual gaps persist—particularly the limited integration of energy analysis within an explicit urban morphological and planning framework. Existing studies rely heavily on single-dimension indicators such as density, which is inconsistently defined and unable to capture the combined influence of geometry and network properties on energy performance. Moreover, no standardized or computationally efficient multivariate index exists for characterizing urban form at the neighborhood scale, and typology studies vary widely in spatial units, indicators, and clustering approaches. These gaps are especially pronounced in Middle Eastern cities, where superblock-based planning models and hot-desert climates produce form–energy dynamics not addressed by existing typological frameworks. Together, these limitations underscore the need for an integrated, morphology-sensitive, and context-specific framework for reliable neighborhood-scale energy evaluation. Applied to 165 superblocks in Abu Dhabi, this study develops a three-step, data-driven simulation framework. First, urban typologies are identified using an unsupervised machine-learning method combining Principal Component Analysis and Gaussian Mixture Model clustering. A multivariate Form Index—comprising nine indicators of density, geometry, and networks—is constructed, enabling representative typologies to be identified using Mahalanobis Distance and profiled through k-means discretisation. Second, a hybrid UBEM workflow translates these typologies into simplified “notional grids” through model-order reduction, enabling computationally efficient neighbourhood-scale energy simulations using established modelling techniques. Third, a parametric modelling framework systematically varies density indicators—floor area ratio, land coverage, and building height—and visualises scenario outcomes using Spacemate and parallel-coordinate plots to support early-stage energy-conscious planning. The analysis yields four clearly distinguishable superblock typologies: dense compact high-rise (T1), dense compact low-rise (T2), less dense open low-rise (T3), and spacious open low-rise (T4). T1, the most vertically and horizontally compact form, exhibits the lowest energy intensity, while T4, the most dispersed suburban form, records the highest. Energy performance improves with higher FSI, GSI, plan depth, obstruction angle, and network density, whereas high OSR, wide streets, and coarse networks increase demand. Balanced densification can reduce energy use (up to 10% total and 24% cooling), while poorly configured scenarios may increase consumption by ~11%. This thesis makes two principal contributions. First, it develops a comprehensive multivariate Form Index integrating density, geometry, and network measures into a coherent, planning-relevant descriptor of urban form—addressing the limitations of single-measure approaches and extending the Spacematrix method. Second, it establishes the first systematic, data-driven typo-morphological framework for classifying superblocks in Middle Eastern cities, complemented by a hybrid UBEM–parametric modelling workflow that offers planners an efficient, interpretable tool for evaluating neighbourhood-scale energy performance and densification strategies.
Included in
Urban Form, Density, and Energy Efficiency: A Typo-Morphology-Based Simulation Framework for Superblock Planning and Densification in Abu Dhabi
F1-2126
Urban form plays a critical role in shaping environmental performance, particularly building energy. In Abu Dhabi and other GCC cities, development is structured around superblocks—large arterial-bounded land parcels units derived from NPU planning principles. Rapid urban expansion, rising energy demand, and the predominance of low-density suburban growth (now exceeding 55% of Abu Dhabi’s footprint) highlight the need for frameworks that can evaluate the energy implications of this morphology and guide both new development and retrofitting strategies. Superblocks therefore offer a consistent and policy-relevant scale for analysing form–energy interactions in the GCC’s hot-desert context, where early-stage, morphology-informed planning tools are essential. Despite extensive research on form–energy relationships, significant conceptual, methodological, and contextual gaps persist—particularly the limited integration of energy analysis within an explicit urban morphological and planning framework. Existing studies rely heavily on single-dimension indicators such as density, which is inconsistently defined and unable to capture the combined influence of geometry and network properties on energy performance. Moreover, no standardized or computationally efficient multivariate index exists for characterizing urban form at the neighborhood scale, and typology studies vary widely in spatial units, indicators, and clustering approaches. These gaps are especially pronounced in Middle Eastern cities, where superblock-based planning models and hot-desert climates produce form–energy dynamics not addressed by existing typological frameworks. Together, these limitations underscore the need for an integrated, morphology-sensitive, and context-specific framework for reliable neighborhood-scale energy evaluation. Applied to 165 superblocks in Abu Dhabi, this study develops a three-step, data-driven simulation framework. First, urban typologies are identified using an unsupervised machine-learning method combining Principal Component Analysis and Gaussian Mixture Model clustering. A multivariate Form Index—comprising nine indicators of density, geometry, and networks—is constructed, enabling representative typologies to be identified using Mahalanobis Distance and profiled through k-means discretisation. Second, a hybrid UBEM workflow translates these typologies into simplified “notional grids” through model-order reduction, enabling computationally efficient neighbourhood-scale energy simulations using established modelling techniques. Third, a parametric modelling framework systematically varies density indicators—floor area ratio, land coverage, and building height—and visualises scenario outcomes using Spacemate and parallel-coordinate plots to support early-stage energy-conscious planning. The analysis yields four clearly distinguishable superblock typologies: dense compact high-rise (T1), dense compact low-rise (T2), less dense open low-rise (T3), and spacious open low-rise (T4). T1, the most vertically and horizontally compact form, exhibits the lowest energy intensity, while T4, the most dispersed suburban form, records the highest. Energy performance improves with higher FSI, GSI, plan depth, obstruction angle, and network density, whereas high OSR, wide streets, and coarse networks increase demand. Balanced densification can reduce energy use (up to 10% total and 24% cooling), while poorly configured scenarios may increase consumption by ~11%. This thesis makes two principal contributions. First, it develops a comprehensive multivariate Form Index integrating density, geometry, and network measures into a coherent, planning-relevant descriptor of urban form—addressing the limitations of single-measure approaches and extending the Spacematrix method. Second, it establishes the first systematic, data-driven typo-morphological framework for classifying superblocks in Middle Eastern cities, complemented by a hybrid UBEM–parametric modelling workflow that offers planners an efficient, interpretable tool for evaluating neighbourhood-scale energy performance and densification strategies.