Statistical evaluation of weather patterns on fishing vessel incidents in Atlantic Canada

Commercial fishing is an inherently risky industry and the safety of the people and property involved is always a concern of maritime administrations. Weather conditions, location, activity level, vessel type and numerous other factors are acknowledged to be major contributors to the occurrence of maritime incidents. However, there is a dearth of detailed research to understand the relationship between weather conditions and maritime incidents, including the occurrence, the rate and the severity of the consequences. This thesis attempts to demonstrate that such an association exists, and determine which are the principal weather factors driving this outcome. Relationships between boating activity levels and weather factors are also examined.;The study area, which encompasses a broad extent of Atlantic Canadian waters, includes fishing incidents recorded by the Canadian Coast Guard from1997 to 1999. Weather data from numerous sources were processed to ascertain the conditions during each incident. Data preparation included matching records across disparate databases, and spatial and temporal interpolation/extrapolation. Methodologies used for traffic track generation in a geographical information system and aggregation of all relevant weather data needed for the statistical analyses are presented.;Several risk model hypotheses were tested. Classification trees were built to identify important weather factors associated with the occurrence of incidents (ceteris paribus), and also to examine weather effects on the relative incident rates (given that incidents have occurred). Finally, a logistic regression model was built to calculate probability of incidents being classified as severe or not.;This study has demonstrated that certain weather factors have a larger influence on the occurrence, rate and/or consequences of maritime incidents. These results can be instructive for better preventative measures for boating in hazardous weather conditions. Search and rescue resource planning can also be reviewed in better anticipation of severe incident occurrences as a function of the weather forecasts.