Network preparation with QGIS’s Styling methods

Methodology, detailed discussions of the GIS Analysis process and functions used

Creating a Grid (Hexagon) using Singapore’s MP19 Subzone No Sea

For our team’s network analysis, we have developed a grid using a hexagon shape which covers Singapore’s subzone at 500m by 500m. For each hexagon, it will be utilized to depict the shortest distance to reach the nearest Primary School and JC. The shorter the distance, and if it falls under the category of a shorter distance (e.g., 0 to 100m) that is classified using Symbology, the denser the colour (e.g., Green) which will be highlighted on the hexagon. 

Creating a centroid for each hexagon present in Singapore’s Subzones

Using the hexagon grid that is developed based on Singapore’s subzones, a centroid will be generated for each hexagon to depict the centre point. This centroid will eventually be utilized to determine the distance of each individual Primary School and JC in Singapore.

Filtering out specific roads for network analysis

The first impact that came to mind was the potential increase in travel time for students to reach their school. However we realised that not all roads in Singapore would be utilised as students would likely live close by, thus we used the “select features using expression” function and only selected the following roads: motorway, motorway_link, primary, primary_link, secondary, secondary_link, tertiary, tertiary_link, residential, trunk and trunk_link.

Filtering out the existing Primary Schools and JC in Singapore

As our group is performing the network analysis for Primary Schools and JC in Singapore, we will be filtering out Primary Schools and JC in Singapore and excluding the other schools that do not fall under the category that meets the criteria. 

Filtering out the Primary Schools and JC that will be affected by the closure or merger

For this section, as our sub-theme solely focuses on the closure and merger of schools in Singapore, and with our group’s goal to focus on Primary Schools and JC, apart from filtering out all Primary Schools and JC in Singapore, we will also be filtering out schools that will be closing soon and there will be 2 point layers where one will depict all primary schools before the arrangement to close the 22 primary schools, JC, and another layer will depict the Primary schools and JC that are still available even after the closure of the schools. 

Creating matrix table for accessibility map

In order to create the choropleth accessibility map of schools, we first needed to create a hexagon grid covering the whole island of Singapore. Centroids of these hexagons were then generated.

Next, we used the QNEAT3 plugin’s “OD Matrix Layers as Table (m:n)” function with the following inputs: point locations of the schools, centroid layer, and roads layers. This was an extremely taxing calculation which lasted 3 hours each time we ran it, totaling to 6 hours spent just running the function as we ran it twice. The first time it was ran to find the distances from all centroids to all schools before the closure, and the second time was for the distances from all centroids to all schools after the closure

Filtering out shortest distances

After the matrices were obtained, we had to filter out the shortest distances as the matrices contained all the possible ways to get to the destination. We accomplished this by using the execute SQL function.