Developments in Quantitative Human Geography, Urban Modelling, and Geographic Information Science
Pablo Mateos; Michael de Smith; Alexander A. Singleton (2011). Transactions in GIS, 15(3), 249-252. DOI: 10.1111/j.1467-9671.2011.01258.x
Abstract
This special issue of Transactions in GIS presents ten research articles selected from the 18th GIS Research UK (GISRUK) conference that took place at University College London (UK) in April 2010, on the theme of ‘Tackling Global Challenges’. These contributions provide a vivid account of the novel and interdisciplinary approaches being currently developed to tackle a variety of global geospatial challenges. Our selection illustrates ways in which complex and dynamic problems can be tackled through bridging the gaps between long-established technical and epistemological silos within the broad community that we here describe as ‘Quantitative Human Geography, Urban Modelling and Geographic Information Science’.
Extended Summary
This editorial examines how interdisciplinary approaches in geographic information science are being developed to address complex global challenges through innovative spatial analysis methods. The work presents ten research articles from the 18th GIS Research UK conference, covering diverse applications including historical urban growth, business location analysis, crime mapping, health disparities, house prices, agricultural land use, animal tracking, and pedestrian navigation. Despite this apparent diversity, the research converges around three core methodological themes: urban classification and modelling, point pattern detection, and individual trajectory analysis. The articles demonstrate how researchers are breaking down traditional disciplinary boundaries by adapting techniques from biological sciences, image processing, urban planning, physics, earth sciences, crime science, marketing, and economics. Key methodological innovations include cellular automata and agent-based models for urban simulation, Monte Carlo techniques for uncertainty analysis, kernel regression and density estimation for spatial patterns, spatio-temporal scan statistics for hotspot detection, and analytical hierarchical processes for decision support. The research showcases five main spatial data models: surfaces for continuous phenomena like house prices and agricultural suitability, polygons for administrative boundaries and retail catchments, points for discrete events like crime incidents, networks for street-constrained activities, and trajectories for movement analysis. Several studies specifically address persistent methodological challenges, including the effects of spatio-temporal scale on analysis results, the participatory value of involving stakeholders in spatial analysis, and the automatic detection of patterns and associations in geographic data. The work also explores fuzzy membership approaches for handling uncertainty in geographic classification. Urban classification studies demonstrate how spatial relationships between land uses and physical environment remain surprisingly consistent through time, creating what researchers term ‘genetic codes’ that determine urban development probabilities. Point pattern detection research reveals new methods for identifying hotspots when movement is constrained by network topology, such as crime incidents along street networks rather than across open space. Trajectory analysis work highlights the challenges of handling high levels of uncertainty in movement data, particularly issues of measurement error and temporal sampling that can significantly affect interpretation. The research shows how participatory approaches and visual analytics can improve urban modelling validation and provide better understanding of self-organising urban processes. This work demonstrates the maturity and vibrancy of contemporary spatial sciences, showing how quantitative methods can address real-world challenges in urban planning, public health, crime prevention, and environmental management through innovative cross-disciplinary collaboration.
Key Findings
- Research converges around three core themes: urban classification and modelling, point pattern detection, and individual trajectory analysis across diverse application domains.
- Interdisciplinary approaches successfully adapt techniques from biology, physics, economics, and image processing to solve complex geographic problems through innovative cross-pollination.
- Urban spatial relationships remain surprisingly consistent through time, creating ‘genetic codes’ that determine development probabilities in metropolitan areas with path-dependent characteristics.
- Network-constrained spatial analysis reveals more realistic patterns for human activities than traditional two-dimensional approaches, particularly for crime hotspot detection along street networks.
- Participatory simulation and visual analytics provide valuable alternatives to static expert-led approaches for urban system modelling and validation processes.
Citation
@article{mateos2011developments,
author = {Pablo Mateos; Michael de Smith; Alexander A. Singleton},
title = {Developments in Quantitative Human Geography, Urban Modelling, and Geographic Information Science},
journal = {Transactions in GIS},
year = {2011},
volume = {15(3)},
pages = {249-252},
doi = {10.1111/j.1467-9671.2011.01258.x}
}