The map comparison problem: tests for the overlap of geographic boundaries

Area of Science:

• Environmental Science
• Epidemiology
• Geographic Information Systems (GIS)

Background:

• Traditional statistical methods like correlation and regression in environmental studies can be limited by presupposed statistical models.
• Geographically distributed variables often require specialized analytical approaches to understand their relationships.

Purpose of the Study:

• To propose novel statistics for measuring boundary overlap between geographically distributed variables.
• To evaluate the performance of these new statistics using simulations and real-world environmental and health data.

Main Methods:

• Development of four distinct boundary overlap statistics.
• Application of these statistics to analyze ozone concentration data and hospital admissions for respiratory conditions.
• Validation through simulation models and analysis of real-world datasets.

Main Results:

• The proposed boundary overlap statistics are sensitive to different facets of spatial relationships.
• These statistics provide a valuable supplementary diagnostic tool for analyzing environmental and health data.
• The effectiveness of the statistics was demonstrated using real-world ozone and respiratory hospital admission data.

Conclusions:

• Boundary overlap statistics offer a robust alternative to conventional methods for analyzing spatial relationships in environmental studies.
• The increasing availability of Geographic Information Systems (GIS) will likely drive the adoption of these overlap statistics.
• These methods enhance the evaluation of hypotheses linking environmental exposures to health outcomes.