Modeling the Land Use–Transport Interaction in Tehran: Satellite Data Fusion, Machine Learning, and a Hybrid CA–Markov Forecast

Tehran’s rapid and heterogeneous urban expansion has generated complex challenges in the interaction between land use patterns and transportation networks. While previous studies often examined physical urban growth or transport network performance in isolation, simultaneous and data-driven analysis of these two dimensions has remained limited. This research develops an integrated framework to model the spatiotemporal dynamics of Land Use–Transport Interaction (LUTI) by combining satellite data fusion, supervised machine learning (SVM), and a hybrid CA–Markov model. Landsat-۸ and Sentinel-۲ imagery were fused through the Gram–Schmidt algorithm, and land-use classification was performed with an overall accuracy exceeding ۹۷%. Results reveal that between ۲۰۱۳ and ۲۰۲۲, Tehran’s built-up area expanded at a rate of ۳.۶۶ km²/yr (≈۳۳ km² increase), while vegetation cover declined by ۵.۶۱ km²/yr (≈۵۱ km² loss). These transitions were concentrated in peripheral zones, predominantly aligned with major highway corridors. Spatial analyses using Moran’s I and LISA indicate that more than two-thirds of the street network falls within clustered patterns (HH and LL), with HH concentrations in districts ۲۱, ۲۲, and ۴, and LL clusters in central districts (۱۰ to ۱۲), highlighting a significant association between population density gradients and network configuration. Forecasts for ۲۰۳۰ indicate continued peripheral and highway-oriented growth, albeit with a lower horizontal expansion rate compared to previous decades. Collectively, automobile-oriented development and infrastructure concentration in the periphery reinforce a self-amplifying cycle of sprawl and spatial inequality. These findings deliver a data-driven framework for LUTI analysis and underscore the need for Transit-Oriented Development (TOD), multimodal network integration, and spatial justice to achieve sustainable and equitable urban futures in Tehran.

کلیدواژه ها:

Satellite Data Fusion ، Supervised Machine Learning ، Hybrid CA–Markov Model ، Multimodal Transport Integration ، Land Use Change