Methodology

Step C - Converting Polygon to Raster

In order to carry out the distance analysis, it is necessary for the cost surface to be a raster file. Therefore, it was required to convert the vector zone costs into a raster file using Cost as the value field. The cell size was set to 25m.
ArcToolbox: Conversion Tools/To Raster → Polygon to Raster
Click on this link to see an example of the resulting Raster Cost Zone

Calculate the Cost Surface

In order to be able to carry out a distance analysis it is required to calculate a cost surface.

To do this, the Distance/Cost Distance tool within the Spatial Analyst Tools of Arctoolbox was used.The input feature class was Downtown Vancouver, the input cost raster was the Raster Zones Cost. An output distance raster as well as an output backlink raster were created.

Identify the Shortest Pathways

To determine the shortest path from Downtown Vancouver to the different road trip destinations - Cultus Lake, Deep Cove, Harrison Hot Springs, and Whistler, that go through attraction points, a Cost Path had to be created.

To do this the Distance tool within the Spatial Analyst Tools was used. The input feature destination data was Cultus Lake (for example), the input cost distance and the backlink rasters were used from the previous cost surface created. The path type selected was Best_Single. The output was a Costh Path to Cultus Lake.

This was repeated 3 more times for the other 3 remaining road trip destinations.

Convert Raster Cost Paths to Polygons

In order to create a more visually pleasing map, and to make the respective paths wider, the 4 paths were converted from Raster to Polyline, with simplified polylines.

Identify and extract Attractions around the Cost Paths

Part of the goal of the project was also to create a guide for EVO users regarding the different attractions surrounding the shortest path to the day trip destinations. In order to do, it was necessary to create a buffer of 500m around the four different Cost Paths, and Intersect using the Overlay tool to create a new layer with only the attractions surrounding the Cost Paths. The resulting layer’s attribute table containing the list of attractions was exported to an excel table. This was repeated for each of the cost paths.

Create Routes from Network dataset

In order to take the analysis of the project slightly further, the final step was to create Routes from the network datasets to the 4 destination points. By creating such routes it is possible to compare them to the calculated Cost Paths generated by the analysis.

The following sections provide a concise description of the analysis process which was required to find the Least Cost path from Downtown Vancouver to Cultus Lake, Deep Cove, Harrison Hot Springs and Whistler. The map resulting from this analyse can be found under Results.

The Project Flow Chart can be found here.

Filtering data

Cleaning the attribute data

The first steps was to clean up the road dataset in order to delete the attributes that will not be necessary in the analysis.

Part of the roads listed in the Census have no “TYPE” associated to them, meaning that they are not officially categorised as “roads” in the system. It is important to delete those selections as we do not want an EVO car to go to small forests paths, and car insurance works only on the official road network.

Clipping to the area of interest

The Road layer, as well as the Attractions Points, and the Cities were all clipped to the region of interests. It was known that the extent of the road analysis was going to extend to 200km of driving roads from downtown Vancouver. Yet, since no analysis had already been made, it was required to assume a zone area around Vancouver, extending North and East. The data beyond a 400 km radius (estimated using a scale), was selected and removed in order to create a new layer file which only had he 400 km by 400 km zone.

Data mining (GIS Analysis)

Creating a Network

Using the new filtered and clipped road dataset, it was possible, and necessary to create a network dataset (It was required to turn on the Network Analyst extension). The network was added as new layers.

Network Analyst - New Service Areas

The next step was to create 3 respective Service Areas from the central point Downtown Vancouver. The analysis was repeated 3 times for 100,000 m, 150,000m, and 200,000 m. Lines were created, as well a detailed polygons trimmed to 10m.

Creating a Raster Cost Surface

Step A - Combined different zones and polygon into one file

In order to create a cost surface, it was necessary to combined different areas (Roads, Attraction Points, and Non-Road Zones) together, in order to be able to assign costs to each surface.
In order to do this, the following steps were taken:
- A 30 metre dissolved buffer was created around the 200 km road network, added as a new layer,
- A 500 metre dissolved buffer was created around the 200 km road network, added as a new layer
- A 10m dissolved buffer zone was created around the attraction points, added as a new layer.
- The three layers were combined using Union, in the Analysis tool of ArcToolbox.
- The Zone around roads, was isolated using inverse selection and added as a new layer.
The 30m road buffer, the 10m road buffer and the Zone around Roads were all combined together using Union.

Step B - Assign a cost to each polygon zone

Once all the different zones were all within one file and there was no overlap, it was necessary to add a New Field, renamed “COST”, and using the edit tool to assign a different cost to each zone.
The zones around roads was set with a high friction of 500 in order for the Least Cost Path to avoid these non driving zones.
The roads were set a friction of 30, and attraction polygons at 5, since the aim of the analysis is to find a Least Cost Path that goes through attractions points.