TOOLS USED TO IDENTIFY SPATIAL AND TEMPORAL PATTERNS OF WILDLIFE-VEHICLE COLLISIONS ALONG ROADS AND THEIR APPLICATION FOR MITIGATION PLANNING

摘要

Over the past few decades collisions with wildlife has become an important safety issue fortransportation planners, especially with the increase of ungulate populations in some regions inNorth America. For example, by March 2009 a stretch of Highway 1 in Nova Scotia had alreadyhad over 200 deer-vehicle collisions. In Ontario, collisions with moose and deer resulted in 7fatalities, and 542 injuries for motorists in 2004. Highway maintenance contractors andtransportation planners have been collecting wildlife-vehicle collision (WVC) data (the location,and date) along roads. From this data, geospatial analyses have determined that there arehotspots or clustering patterns with WVCs due to specific road and environmental factors, e.g.traffic volumes and forest cover. This paper reviews a series of exploratory and statistical toolsthat can determine the 1-dimensional space, and time, and 2-dimensional spatiotemporalcharacteristics of collision patterns on roads using the kernel density estimator and Ripley’s Kstatistic. These tools can be applied to assess the spatiotemporal dynamics for any type ofvehicle- collision dataset along roads that contains space and time information, to assisttransportation planners with mitigation planning.