Spatial Interaction (SI) has been an integral part of geographical research for more than a century, and its models have been used to capture the dynamics of flows for half a century. The most prominent discussion in spatial interaction research is related to its spatial structure; specifically, the model’s inability to capture the spatial dynamics of the SI. This thesis builds on this discussion by exploring the points of the debate from a practical perspective using tools and concepts borrowed from network science to advance the application of Spatial Interaction Models SIM in Geography by exploring the definition, representation, and theoretical underpinning of spatial structure in SI. The contribution is three-fold. First, the thesis questions and redefines the concept of spatial structure in spatial interaction into a more practical one as a multi-layered complex network. Second, it suggests a network structure as a suitable way to capture the spatial structure of SI and demonstrates its ability to capture both local and global changes in the interaction, as well as its temporal dynamics. Third, it questions the causal pathways of SIM in order to establish the relationship between interaction and proximity, which is inherently intertwined. Changing the assumptions about the two main components of the model allows for the exploration of latent space in SIMs that has the potential to be used in the identification and validation of influential variables. Overall, this interdisciplinary research challenges the current understanding of spatial structure in SI from a practical perspective, showing how different assumptions about SI can assist in the search for causally justified factors.
- Spatial Interaction Model
- Networks
- Spatial Structure
Rediscovering the Spatial Structure of Spatial Interaction Through Networks and Advanced Quantitative Methods
Hasova, L. (Author). 1 Oct 2024
Student thesis: Doctoral Thesis › Doctor of Philosophy (PhD)