Abstract
Shark attacks have historically been studied from a viewpoint of encounter number per region and so limited to the areas in which the attacks occurred. In this exploratory modeling study, the goal was to examine whether an area-specific cluster analysis algorithm undertaken with a modern cluster analysis tool (SaTScan™ 9.1.0) could enhance our spatial and spatio-temporal understanding of attack patterns. The data used were from Florida's east coast between 1994 and 2009. The program suggests several high- and low-risk areas for shark attacks. The results are discussed from a quantitative rather than qualitative perspective.
Acknowledgments
The authors thank the three anonymous reviewers and the editor for their valuable feedback. They also thank Martin Kulldorff for his assistance with the interpretation of the cluster analysis.