Obtain Data

All data was found on either Vancouver Open Data or ICBC’s website Sources of data can be found here and more information about the data themselves can be found under the data tab. Data downloaded was put in a geodatabase and imported into the contents panel.

Name	Description	Source
Traffic signals	This dataset contains the locations of the City’s traffic signals. Attributes: Pedestrian Actuated Signal, Fixed Time, Semi Actuated, Fully Actuated, RRFB, Special Crosswalk, Bus Actuated Signal, FH, CS	https://opendata.vancouver.ca/explore/dataset/traffic-signals/table/ Opendata Vancouver
ICBC reported crashes	2019 data from ICBC by insured vehicle owners, cars, pedestrians, cyclists, motorcyclists	https://public.tableau.com/profile/icbc#!/vizhome/ICBCReportedCrashes/ICBCReportedCrashes ICBC statistics
Street intersections	Street intersections in Metro Vancouver	https://opendata.vancouver.ca/explore/dataset/street-intersections/information/

Parse Data

ICBC 2019 Crash Report

The incident crash report of 2019 came in an excel file that needed to be properly symbolized on the map to show the location of the crashes recorded. The “Table to point Feature Class” tool was used to export the coordinates. The software was able to recognize the latitude and longitude field that was present in the excel file and used those to place the points. After projecting the “XY” coordinates, points of crash accidents appeared on the map.

Traffic Lights & Intersection File

The Intersection file was added to the contents panel and need to be parsed as I was only interested in the intersections that had the traffic lights of interest located in them. With the Traffic Lights file added to the contents panel as well, the select attributes by location tool was used to select intersections that had Traffic Lights in them. I used the relationship “within a distance” at “30” “meters” as that was the average size of an intersection. A new layer was made from the selection. A new layer was created with the selected intersections called “intersection_of_interest” that contained only traffic lights at intersections where there were traffic lights I was interested in, as shown in the image below. 

Histogram

Histograms were created to show the distribution of the count data. The variable of interest was the “crash count” field.

Heat Map

A heat map was created to show the overall distribution of the intersections of interest weighted with the “crash count” field.  The radius was lowered to 10 to show more raster detail around each intersection point.

Kernel Density Analysis

The main analysis performed was a Kernal Density Analysis to look at how where each traffic light was used the most and which had higher rates of traffic accidents that year. With the Kernal Density Analysis, I was able to select the field that recorded the number of crashes, “crash count”, as the population field. By doing so, I was able to weigh the crash counts more heavily. The search radius was left at a small value, 4.53 because I wanted a raster that showed more detail around each individual traffic signal. If I had selected a larger value, the raster would have been much smoother but I would be able to see the result of individual traffic signals. For the method, the planer option was chosen because this analysis was at a local scale and I wanted to keep the correct distance and area. 

Symbology

The symbology chosen from the heat map was a continuous colour scheme. The transparency of the layer was increased to 40% to show the underlying topology.

For the results of the Kernal density analysis, I choose to use a continuous colour to symbolize areas of high density. The transparency was slightly reduced so that the underlying base map could still be seen in areas of low density. This was important because then specific intersections could be seen. The location of each intersection type was placed on top of the Kernal Density analysis to mark where each specific intersection of interest was in relation to the analysis performed. The Colour scheme was changed to equally distributed.