<![CDATA[Defining urban agglomeration boundaries using conventional administrative approaches often creates a ‘statistical illusion’ that distorts the functional spatial reality, while alternative dynamic models (such as cellular automata and agent-based models) are hindered by an absolute dependency on the availability of historical spatio-temporal data. Addressing this epistemological deadlock, this article proposes a new methodological framework that deconstructs the dynamics of metropolitan agglomeration evolution through the integration of space syntax, fractal dimension, and spatial clustering (DBSCAN) methods, purely by extracting the topological network of existing road intersections. This triadic framework addresses three fundamental analytical dimensions: space syntax diagnoses the ‘seeds’ of historical initiation (time-frame) through centrality metrics; fractal analysis quantifies the level of complexity and objectively establishes the threshold of evolutionary scale (scale-frame); and the DBSCAN algorithm visualizes the transition of agglomeration as an emergent spatial structure (visual-frame). The empirical implementation of the proof of concept was applied to the hierarchy of functional urban areas (FUA) in Indonesia, represented by the megalopolis of Jakarta and the metropolitan areas of Bandung and Yogyakarta. The precise calibration results successfully unveiled the ontological cycle of the city: from the discovery of micro-cluster embryos (postdiction) and mapping the explosion of fragmentation in the present (status quo) to determining the boundaries of macro fusion of urban areas in their entirety (prediction). In conclusion, this integrated framework shifts the paradigm from static delineation to process-oriented agglomeration analysis, offering an analytical instrument with extraordinary data efficiency that liberates spatial planning from the bias of arbitrary administrative jurisdictions.]]>
Copyrights © 2026
