Sustainability of Railway Networks
The article analyzes the sustainability of a railway network in Korea. The article talks about how failure in rail networks are often incorrectly analyzed when they are assumed to be linear, but treating them as network structures allows for a more thorough analysis. For example, if a popular route with many passengers had a technical failure, the stations closer to the failure would suffer more than stations that are farther away, or have a low degree of centrality. Specifically, the article analyzes how susceptible a train station is to failure using two criteria: local performance and network performance.
The local performance of a station is a function of its node value, or how well connected the station is, and place value, which measures the activity of the location. For example, a remote station with few train connections with many shopping malls and restaurants nearby would have a low node value but a high place value. Having a low node value and high place value is unsustainable, as there are not enough passengers to keep the stores in business. Similarly, a high node value and low place value is unsustainable since the high volume of passengers will overwhelm the small amount of local businesses. Therefore, managing the balance between node and place value is vital to maintain a train station.
The network performance of a station is a function of its degree (number of directly connected stations in the graph) and “betweenness centrality”, which is a measure of how much a train station functions as a path between two different nodes. Stations at the center of clusters tend to have a high network performance, which is an advantage in that passengers will likely contribute to the service sectors of surrounding stations. However, this is also a disadvantage because a delay or technical failure will propagate down the network and affect closer stations. Therefore, measuring “betweenness centrality” is a good risk assessment tool to determine the most vulnerable paths in order to decide where stations should allocate most of their resources in the event of an accident.
It is interesting that converting train maps to graphs can serve as a powerful tool to determine how sustainable a train station is. Analyzing the “value” of a node based on its connectivity with other nodes allows for a more robust risk management than analyzing single stations alone. Additionally, the concept of node and place value offers an insight that I was unaware of; As a child riding trains, I questioned why the fastest express trains would stop at a certain station even though there were no other train connections out of that station, but I later learned that there was a large shopping mall near the station. This can be reinterpreted as a way of increasing passenger traffic to increase the node value to lessen the gap to the high place value, which makes the train station and city more sustainable. I am curious as to how “betweenness centrality” is quantified, and if there are other ways to measure centrality that would be useful in determining sustainability.
Source: https://www.mdpi.com/2071-1050/13/9/4778