3. Spatial distribution
To display the spatial distribution of fatal traffic accidents in Toronto, we conducted the kernel density analysis in ArcGIS and used the output shapefiles as the base map. By setting different variables as the input field, the result was different. Besides, we use the hot spots analysis and count number tools to amplify the pattern of spatial distribution.
kernel density
count number
hot spot analysis
spatial distribution maps
Figure 1. Flow chart of variables included and presented in the kernel density maps
The overall map did not set any input field, it analyzed the spatial pattern of the fatal accidents themselves only based on their locations (Figure 2). Hotspots were mostly clustered in the downtown area (Fig. A possible explanation is because the downtown had a much more dense road system than any other regions. Also, the narrow Gardiner Expressway passed through downtown, which is easy to cause accidents. Scarborough displayed an enclosed hotspot distribution. This is a suburban industrial area with the famous 401 expressway going through and surrounded by residential area. The roads here were classic a dendritic street system. There were a lot of dead-end streets and often only offered one way from home to surrounding suburban arterials. Fatal traffic accidents may be more likely to occur in the dendritic street system, rather than the interconnected street system. The northern region was the North York borough, where had high-density buildings and intense commercial activities. It was also a vital spot of Yonge Street and Finch Avenue.
Figure 2: Kernel density map of total fatal traffic accidents in Toronto, Canada between 2007 and 2017 showing overall density and point counts of fatal hotspot locations
Figure 3: Kernel density map of total fatal traffic accidents at intersections in Toronto, Canada between 2007 and 2017 showing overall density and point counts of fatal intersection accidents hotspot locations
For the fatal accidents occurred at intersections, the hot spots in the downtown, Scarborough and North York Borough still existed (Figure 3). Scarborough had an absolute increase in the fatal crashes at intersections, while the accidents in the other two areas were not quite significant (Figure 3). Besides, several new clusters formed in East York, Steeles, and Malton respectively. The Steeles and Malton have similar dendritic road structure with the Scarborough, and all three boroughs are featured in heavy industry.
Compared to the Fatal crashes at intersections, the accidents involving speeding issues were more concentrated (Figure 4). The most clustered area is still the Scarborough, and then the downtown area (Figure 4). An interesting finding is that of which, and fatal accidents occurred at the intersection, usually not involving the speeding issue. The deadly speeding often occurred at mid-block of the road.
Figure 4: Kernel density map of total fatal traffic accidents involving speeding in Toronto, Canada between 2007 and 2017 showing overall density and point counts of fatal speeding-related hotspot locations
Figure 5: Kernel density maps of total fatal traffic accidents during morning and evening rush hours in Toronto, Canada between 2007 and 2017 showing overall density and point counts of fatal hotspot locations
The fatal traffic accidents happened in the morning rush hours told a different story than the deadly crashes in the evening rush hours. The disasters in the morning had an overall smaller accidents frequency than the accidents in the evening (Figure 5). The maximum frequency was 7 for the morning, but for the evening, the number increased to 14 (Figure 5). Areas near downtown formed enclosed clusters in the morning, but the clusters disappeared in the evening (Figure 5). Instead, the suburban areas such as Scarborough appeared fatal clusters.