From morning to night: Unveiling the variability in public transport services across Great Britain
Our team is currently working on the update of our Public Transport Accessibility Indicators (PTAI) dataset, which we first published in 2022. For this update, we finished calculating the travel times by public transport from each small area in Great Britain to every other similar area for the year 2023.
In this upcoming dataset, we've gone beyond our previous work, providing more detail about where people can go and when they can travel. With the first part of our new dataset, we will be able to have good approximations of the travel times by public transport from each of the 230,000 smallest statistical areas in Great Britain (referred to as output areas) to all the other areas at the morning peak.
The second part of our new dataset looks at travel times depending on the time when people leave. We've focused on two key times: the morning rush hour starting at 7 a.m. and the late evening starting at 9 p.m. We have calculated these for every local statistical area in Great Britain (referred to as LSOA/DZ). In this blog, we explore one of these opportunities, especially looking at how travel times change over the course of the day.
This is important for understanding social fairness issues, as many people who work at night are on low incomes and depend on public transport to get to and from work. In this post, we are giving a quick overview of one way the new data could be used, but we know that you, the readers and potential users, will think of many other applications.
Here is how we did it
The public transport travel times were computed following a “door-to-door” approach. We figured out how long it would take to travel from one place to another, considering combinations of walking and public transport. We used the centres of local statistical areas (LSOAs/DZs) as start and end points, which represent where most people live in these areas. We got our data on public transport schedules from the Rail Delivery Group and the Bus Open Data Service. This information includes times for most national and regional trains, local buses, trams, and ferries.
We used these timetables to approximate how long it would take to travel on a typical working day (the 7th of March 2023). To be more specific, we looked at departures in two main windows: from 7 to 10 a.m. and from 9 p.m. to 12 a.m. Windows are useful because the estimated travel times can change depending on the exact time you depart and how the services run at that time. Then, we used the median (or middle) time out of the several possibilities in a time window as the representative duration. All in all, we figured out travel times for 265 million different routes.
In this blog, we are interested in looking at how public transport services change from the morning to the night. To do this, we figured out the percentage change in the number of places people can get to in a maximum of two and a half hours at night compared to the morning period for each local area. Let's say you can reach 80 places from a particular area at night, but 100 places in the morning - this means there is a 20% drop at night. The map below shows these differences for each local area across Great Britain.
This is what we found
The raw results show a large variation, with some places losing almost all their public transport travel options at night and others barely changing at all. For instance, Uttlesford, a countryside area between London and Cambridge, dropped from having 4,163 reachable places to just 19. Another area in East Hertfordshire saw a decrease from 3,486 reachable places to only 19. That's more than a 99% reduction. On the flip side, an area in Bromley only dipped from 4,992 to 4,740 reachable places, and a location in Sutton from 7,735 to 7,160. Both of these cases show a reduction of less than 10%. We'll now take a step back and look at the broader picture across Great Britain.
Looking at the big picture, we can see a clear split between the city and the countryside. Most rural locations face a severe drop in services, usually between 75% and 100%. Such drastic decreases are not as common in cities.
There are quite a few interesting trends at the city level. For instance, it is noticeable that many zones within English cities witness large service reductions ranging from 25% to 50%, with some zones experiencing even larger decreases, between 50% and 75%. The latter is particularly interesting to note in cities like Sheffield, Nottingham, and Birmingham. In Wales, most city zones grapple with a substantial reduction in services. Meanwhile, numerous zones in Scottish cities are estimated to have experienced large decreases in service availability. Two cases stand out on the map with only moderate reductions, namely: Greater London and Greater Glasgow, which decrease by between 25% and 0%.
The table below summarises the changes for the major cities. They are arranged from the least affected at the top to the most affected at the bottom, based on the percentage change. Slough & Heathrow comes out on top with a modest reduction of 17.6%. At the bottom of the table, we find Leicester, with a reduction of 67% in the number of reachable destinations.
Table 1 Overview of shifts in accessible locations at night via public transport within 150 minutes for major cities.
These simple, yet revealing, figures suggest that the data can provide us with insights into public transport provision across GB. Also, we recognise there are some limitations, such as discrepancies between the actual versus scheduled services, potential gaps in the data sources, or the geographical heterogeneity across different countries. A possible extension of this analysis could take into account the number of residents affected by these changes in public transport services. Digging deeper into the context of the cities in future analyses might shed more light on the findings. For example, does Nottingham have a larger decrease at night but better accessibility during the day, so, the drop is down to similar levels of coverage at night to other places?
We expect to publish the travel time dataset used here during Summer 2023. We are eager to learn about the projects from those who are already utilising our data, as well as those who might consider using it in the future. We also warmly welcome any comments or ideas for potential collaborations. Don't hesitate to reach out to Rafael Verduzco via email at JoseRafael.Verduzco-Torres@glasgow.ac.uk or on Twitter @raffverduzco. We look forward to hearing your thoughts and exchanging ideas!
UBDC blog by J Rafael Verduzco-Torres and David McArthur.
This project is supported by the ESRC grant Urban Big Data (ES/L011921/1).
We are grateful for the extremely helpful comments and feedback of our colleague Dr Fiona Crawford.
The text of this blog was aided by AI tools, specifically ChatGPT. Our hope is that these tools can help bring innovative research topics to a wider audience, making complex data or processes more accessible to a broader population.