The effects of the lockdown on traffic in Glasgow

The current restrictions on mobility (lockdown) have led to clear reductions in traffic volume in all cities across the UK.

Using location data from mobile phones (Google Location History data), Google community mobility report shows substantial reductions in mobility. They provide a breakdown by trip purpose (e.g., retail/recreation, grocery/pharmacy, parks, etc.). In Glasgow, for example, they show an overall reduction of 84% for retail/recreation trips.

Another way of looking at the changes is to explore variations by locations across the city. To do that, we used Scoot data (real-time traffic data giving the estimated traffic flow every 15 minutes) provided by Glasgow City Council. We examined the changes after lockdown as well as how the pattern of change varied across the city. Although the data cannot tell us the different trip purposes, they cover trips by all vehicles at certain intersections, not just people using particular apps or with the location tracking enabled on their mobile phone. It is worth noting that traffic flows from the data are derived so they are not an exact vehicle count. However, they are good enough to examine the changes and spatial variations in traffic volumes.

Changes in the mean traffic flow

Some sensors have missing data at certain times so we used average traffic flows rather than total traffic flows for the analyses. We examined four different periods: the morning peak, the afternoon; the evening peak; and night time. This gives us a rough approximation of what is commuting and non-commuting traffic. Our first figure shows monthly averages, while the second shows daily data in March.

Note: Data include January 1st 2019 to April 1st 2020 until 3pm.

Both figures show that all periods have seen substantial reductions in average traffic volumes and for all times of the day. Looking at the monthly data, there was about a 13% reduction in the average traffic flow during the morning peak in March 2020 compared to that of February 2020 while about a 17 % reduction was shown for the evening peak.

With the daily data, we can also see the variations over the week, with the reduced flows on Saturdays and Sundays, particularly for the morning peak. In the period from March 16th when people were asked to work from home if possible, there was about a 14% reduction in the average traffic flow during the morning and evening peaks. After the lockdown announcement of March 23rd, there was a much more significant reduction of around 37%.

Activity in different parts of the city

The four figures below show the spatial variation of the reductions in the average traffic flows in the four time periods we define (comparison of average traffic volumes between pre-lockdown from March 1st- March 15th and post-lockdown from March 23rd-April 1st). As expected, there were larger reductions in average traffic flows during the morning and evening peaks compared to the afternoon and night-time periods. In addition, there were slightly larger reductions outside the city centre area. This could be because there are still some necessary activities happening in the city centre compared to other areas.

Conclusions

The data from traffic sensors show how we can monitor activity levels in the city to give a very up-to-date picture of what is going on. We see a strong response to the lockdown but one that varies across the city to some extent.

Due to the technical issues, we were unable to use the dataset covering the whole city, although we are working with Glasgow City Council to access more data. Comprehensive and daily updated data covering more intersections would allow us to monitor changes in mobility across different areas in detail. This is important both to assess compliance with current restrictions but also to understand what is happening when restrictions eventually begin to be lifted.

This blog was written by Dr Jinhyun Hong with research contribution from Chau Man Fung.