Spatiotemporal Heterogeneity of AI-Driven Traffic Flow Patterns and Land Use Interaction: A GeoAI-Based Analysis of Multimodal Urban Mobility
arXiv:2603.05581v1 Announce Type: cross Abstract: Urban traffic flow is governed by the complex, nonlinear interaction between land use configuration and spatiotemporally heterogeneous mobility demand. Conventional global regression and time-series models cannot simultaneously capture these multi-scale dynamics across multiple travel modes. This study proposes a GeoAI Hybrid analytical framework that sequentially integrates Multiscale Geographically Weighted Regression (MGWR), Random Forest (RF), and Spatio-Temporal Graph Convolutional Networks (ST-GCN) to model the spatiotemporal heterogeneity of traffic flow patterns and their interaction with land use across three mobility modes: motor vehicle, public transit, and active transport. Applying the framework to an empirically calibrated dataset of 350 traffic analysis zones across six cities spanning two contrasting urban morphologies, four key findings emerge: (i) the GeoAI Hybrid achieves a root mean squared error (RMSE) of 0.119 and
arXiv:2603.05581v1 Announce Type: cross Abstract: Urban traffic flow is governed by the complex, nonlinear interaction between land use configuration and spatiotemporally heterogeneous mobility demand. Conventional global regression and time-series models cannot simultaneously capture these multi-scale dynamics across multiple travel modes. This study proposes a GeoAI Hybrid analytical framework that sequentially integrates Multiscale Geographically Weighted Regression (MGWR), Random Forest (RF), and Spatio-Temporal Graph Convolutional Networks (ST-GCN) to model the spatiotemporal heterogeneity of traffic flow patterns and their interaction with land use across three mobility modes: motor vehicle, public transit, and active transport. Applying the framework to an empirically calibrated dataset of 350 traffic analysis zones across six cities spanning two contrasting urban morphologies, four key findings emerge: (i) the GeoAI Hybrid achieves a root mean squared error (RMSE) of 0.119 and an R^2 of 0.891, outperforming all benchmarks by 23-62%; (ii) SHAP analysis identifies land use mix as the strongest predictor for motor vehicle flows and transit stop density as the strongest predictor for public transit; (iii) DBSCAN clustering identifies five functionally distinct urban traffic typologies with a silhouette score of 0.71, and GeoAI Hybrid residuals exhibit Moran's I=0.218 (p<0.001), a 72% reduction relative to OLS baselines; and (iv) cross-city transfer experiments reveal moderate within-cluster transferability (R^2>=0.78) and limited cross-cluster generalisability, underscoring the primacy of urban morphological context. The framework offers planners and transportation engineers an interpretable, scalable toolkit for evidence-based multimodal mobility management and land use policy design.
Executive Summary
This article introduces a GeoAI Hybrid framework that integrates Multiscale Geographically Weighted Regression, Random Forest, and Spatio-Temporal Graph Convolutional Networks to model the spatiotemporal heterogeneity of traffic flow patterns and land use interactions across three mobility modes. The framework is applied to an empirically calibrated dataset of 350 traffic analysis zones across six cities, achieving a root mean squared error of 0.119 and an R^2 of 0.891. The study identifies land use mix as a key predictor for motor vehicle flows and transit stop density for public transit. The findings have implications for evidence-based multimodal mobility management and land use policy design, offering a scalable toolkit for planners and transportation engineers. While the framework demonstrates significant promise, its generalizability across diverse urban contexts remains a concern.
Key Points
- ▸ The GeoAI Hybrid framework leverages machine learning and geospatial techniques to model complex traffic flow dynamics.
- ▸ The study applies the framework to an extensive dataset across six cities, achieving superior predictive performance.
- ▸ Land use mix and transit stop density are identified as key predictors of motor vehicle and public transit flows, respectively.
Merits
Strength in predictive performance
The GeoAI Hybrid framework outperforms conventional models, achieving a root mean squared error of 0.119 and an R^2 of 0.891.
Interpretability and scalability
The framework offers a scalable and interpretable toolkit for planners and transportation engineers, enabling evidence-based decision-making.
Demerits
Limited generalizability
The study's findings may not transfer well to diverse urban contexts, highlighting the need for further research and adaptation.
Dependence on empirical calibration
The framework's performance relies heavily on the quality and representativeness of the empirically calibrated dataset.
Expert Commentary
The study's GeoAI Hybrid framework represents a significant advancement in the field of urban systems modeling, combining the strengths of machine learning and geospatial analytics. However, its generalizability across diverse urban contexts remains a concern, underscoring the need for further research and adaptation. The framework's dependence on empirical calibration also highlights the importance of high-quality and representative datasets. Nonetheless, the study's findings have significant implications for transportation engineering and urban planning, offering a scalable and interpretable toolkit for evidence-based decision-making.
Recommendations
- ✓ Future research should investigate the framework's generalizability across diverse urban contexts, including smaller cities and rural areas.
- ✓ Developing more robust and transferable datasets will be crucial for the framework's widespread adoption and adaptation.
