The media reported a rebellion against lockdown this week after signs of increased motor vehicle use were presented in the Government’s Downing Street briefing on 1st April. The chart presented by Dr Yvonne Doyle, Director for Health Protection and Medical Director, compared the percentage change in all motor vehicle use versus rail, TfL bus and TfL tube use. It is not clear which data are being used to collate these figures but the increase in use of motor vehicles from the 29th to 30th March was called “slightly concerning” and advice to stay at home were reiterated. All modes of public transport use remained stable.
What sources of data can we use to explore the impact of the Covid-19 on traffic?
One approach we can take to establish the general impact of coronavirus on vehicle use is to look at MIDAS (Motorway Incident Detection and Signalling) data collected from 5th March 2020 to 1st April 2020. MIDAS data is collected by Highways England from loops and radar installed on motorways; the number of vehicles which pass the radar each minute is recorded1.
We looked at:
- How daily traffic flow has changed over time
- The percentage reduction in traffic flow
- The change in vehicle proportions
- Comparison of a ‘normal’ (pre-lockdown) day to current day during lockdown
Evident within the figures below is the recurring dip in traffic flow at weekends compared with weekdays. Any ‘uptick’ from day to day needs to be considered in this context; sometimes we expect to see large changes, based on the move from weekday to weekend, and back again. As shown in Figure 2, vehicle use on motorways was declining each weekday after the lockdown was announced on the 23rd March. When the weekend (dip) data points are removed in Figure 2, traffic flow on the following Monday to Wednesday 1st April shows a continued decline in weekday traffic flow, one that appears to be levelling out.
Daily traffic flow
Across all motorways recorded, there has been a general decline in traffic flow as the government announced increasingly stringent restrictions on movement. The decline in flow is larger on motorways which generally have a higher volume of traffic (such as M1, M25 and M6) as opposed to those with lower traffic volumes (such as M56, M602). Recurring weekend dips in traffic flow can be clearly seen.
Change in traffic flow
Below we shows three outputs:
1. Daily traffic flow averaged across all motorways
2. Percentage change in flow compared to a baseline period
3. Same but with weekend data points removed
The baseline period was defined as follows:
Weekday Baseline: average daily traffic flow calculated across all the weekdays between 5th and 15th March 2020
Weekend Baseline: average daily traffic flow calculated across all the weekends between 5th and 15th March 2020
The percentage change was then calculated relative to the particular day of the week. For example, the percentage change in flow for Monday 16th March is calculated relative to the weekday baseline, and the percentage change in flow for Saturday 21st March is calculated relative to the weekend baseline. This accounts for the general differences in weekday and weekend flows.
There is an obvious weekday and weekend trend where traffic flow is lower on weekends compared to weekdays. When daily flow is averaged across all motorways, there was a 45% decrease in flow the day after the lockdown was implemented and this reduces further to 66% by 1st April.
MIDAS data collects traffic flow figures by vehicle length. Here we compare two vehicle classes: Heavy Goods Vehicles (HGVs) and “Car & other vehicles”.
The split by vehicle type shows an expected increase in the proportion of HGVs from about 18% the week before any measures were introduced, to 32% on 1st April. The largest proportion of HGVs to other vehicles is now found on the M6 and M18 where HGVs make up almost 40% of traffic. Clearly demand for deliveries and goods is currently partly offsetting commuters and work and leisure-related travel.
Time of day
The graph below presents comparison of average traffic flow by hour of the day for the M3 only. In this example case, the baseline period was from 9th March to 13th March (shown in green) and the impact of the lockdown was considered from 23rd March to 27th March (shown in red).
These data show how the daily pattern of traffic flow remains, but the impact of lockdown on daily trips is clearly represented.
Exploration of these data give us a clear picture of the decline in traffic as coronavirus mitigation measures were gradually introduced in March 2020. The ‘slightly concerning uptick’ may be natural variation caused by the difference between weekend and weekday flows, although the data we present here do not cover local roads and other parts of the strategic road network. Patterns of travel on these other road types may be different and underpin the “uptick” presented in the Government’s briefing.
The data presented here also suggest that the level of traffic on these motorways is levelling out. If further isolation measures are required then more understanding about these journeys, and how essential they are, will be necessary. It might also be possible to use this as a baseline from which to measure the impact of further policies introduced in the coming weeks.
By Neale Kinnear and Sritika Chowdhury
Acknowledgements: Bob Smith, Jo Hammond, Caroline Wallbank, Shaun Helman and Jill Weekley.
1. Although covering a large proportion of the motorway network in England, MIDAS loops are generally installed on the busiest motorways, therefore, traffic estimates are likely to slightly overestimate traffic on the whole motorway network.