Call For Papers [@Computational Urban Science]: Revisiting Spatiotemporal Modeling in the Era of GeoAI

Geographic phenomena are considered complex due to the heterogeneous nature of spatiotemporal dependencies. It is impossible to specify a universal law described in statistical or physical languages that can perfectly characterize a real-world geographic process and explain how it forms certain observed patterns. As an emerging field at the intersection of GIScience, statistics, and AI, GeoAI offers a novel and bold perspective on revisiting and advancing current spatiotemporal methods and models, with the potential of new representation, measurement, analytics, and models to cope with diverse urban challenges.

Despite many efforts that utilize deep learning (e.g. CNN, GCN, LSTM, GAN, Transformer, LLM) for spatiotemporal computational research, not many exist to investigate and discuss the intuitive link between key concepts in spatiotemporal analysis and modeling (e.g. dependence, organization, scale, regularity, distance, heterogeneity, step-changes, etc) and AI principles (e.g. convolution, representation learning, attention, transfer learning, meta-learning). 

As both models and data sets become more complex and advanced, we are facing even more challenges in ensuring the interpretability, explainability, replicability, and fairness of GeoAI outcomes. For example, is the data used for AI training enough to reflect the nature of the geographic data generation process? Similarly, in rural regions with limited data records, how does modeling human activity mechanisms affect existing biases and unfairness? It is critical timing for us to revisit the spatiotemporal modeling in current GeoAI efforts, to elucidate why AI models could facilitate spatiotemporal studies theoretically and empirically, and to help understand the true process behind complex spatiotemporal phenomena. 

This special issue is intended to reflect how spatiotemporal models can be enriched with possibilities when combined with state-of-the-art machine learning/deep learning insights. In light of bridging the gap between artificial intelligence techniques and spatiotemporal analysis methods & models, we welcome all the latest efforts in GeoAI theories, methods, and applications that touch the fundamental analytical principles, with topics not limited to:

- GeoAI for spatiotemporal data generation in urban scenario

- GeoAI for spatiotemporal modeling & prediction

- GeoAI for spatial process modeling  and understanding

- GeoAI and spatial networks 

- GeoAI for spatial interaction modeling

- GeoAI and spatial complexity

- GeoAI for spatiotemporal data fusion

- GeoAI and spatiotemporal statistical modeling

- GeoAI for spatial interpolation & extrapolation

- GeoAI for spatiotemporal data imputation

- Explainability and Interpretability of GeoAI models

- Spatially explicit model with spatiotemporal modeling components

- GeoAI for socioeconomic modeling in urban scenarios

- GeoAI for human mobility & dynamics in cities

- GeoAI for multi-modal spatiotemporal data

- Transferability and generalization of GeoAI models

We welcome submissions from scholars in all relevant fields. Interested authors should first submit a short abstract (400 words max) to Dr. Di Zhu ([email protected]), Dr. Ziqi Li ([email protected]), Dr. Duncan Lee ([email protected]) and Dr. Peng Luo ([email protected]) by Jan. 31, 2025, for a review of its fit to the theme of this special issue. Suitable abstracts will be invited to submit full manuscripts although the invitation does not guarantee acceptance to the SI.

Full manuscripts, including any supporting materials and required data and codes that can reproduce findings reported in the manuscript, should be submitted using the journal's online submission portal by Jul. 31, 2025, and the authors should specify this SI as the target during their submission. All manuscripts will go through a standard review process.

Guest Editors: 

Dr. Di Zhu, Assistant Professor, University of Minnesota-Twin Cities ([email protected])

Dr. Ziqi Li, Assistant Professor, Florida State University ([email protected])

Dr. Duncan Lee, Professor, University of Glasgow ([email protected]) 

Dr. Peng Luo, Postdoctoral Researcher, MIT Senseable City Lab ([email protected])