Monitoring air pollution under lockdown
Since the Prime Minister’s announcement of a ‘lockdown’ on 23 March, the UK population has been required to avoid all but essential use of cars and public transport in an effort to slow the spread of the virus. The economic and social impacts of this change are enormous and as yet not fully understood, even in the short term.
Some statistics have been produced on declines in mobility which have resulted. Department for Transport statistics showed motor vehicle volumes about 60% down on the start of March. But as the country entered the second week of the lockdown, concerns were expressed that compliance might be weakening as road traffic rose above the low point achieved on 29 March.
Analysis from Google supported the general view that, across the UK, levels of mobility by all modes were hugely reduced. Using location reporting from mobile phones, they estimated that journeys for retail and recreation were down 85% compared to the baseline while mobility for grocery and pharmacy shopping, for parks and for workplaces were down by less (46% to 55%). Footfall at public transport hubs was down by 75%, suggesting a switch to other modes, notably private cars. On the other hand, mobility around residential areas was up by 15%, perhaps reflecting new daily exercise routines.
Another way we can track changes in mobility is through air pollution levels. According to Public Health England, air pollution is “the biggest environmental threat to health in the UK, with between 28,000 and 36,000 deaths a year attributed to long-term exposure” and transport, particularly road transport, is the major source. In this blog, we focus on Nitrogen Oxides (NOx) because of their close connection with road transport.
We would expect a significant improvement in air quality as a result of reduced mobility but offset to some extent by greater reluctance to use public transport. Many people who had previously used public transport might prefer to reduce risks of exposure by using private cars instead. The costs of car use might also be reduced as congestion charges in London have been waived since 20 March as have parking charges at NHS facilities from 25 March. There may also be rising use of home delivery services for a range of goods, especially foods, also contributing to emissions.
Air pollution data are reported through the UK’s Automatic Urban and Rural Network (AURN), with data made publicly available on a daily basis. Air quality readings are provided from around 150 sites on an hourly basis, along with windspeed and direction. Here, we look at data for 2020 for about 40 sites in 13 cities: the ten largest urban areas in the UK plus the capitals of the devolved nations (Belfast, Cardiff and Edinburgh).
One challenge with using air pollution levels to draw conclusions about road transport is that windspeed has a major impact on pollution. NOx levels shoot up on still days. A simple way to remove the ‘noise’ this creates is to restrict our data to days when the wind is above 4 m/s.
The first figure shows the changes just for Glasgow. The chart shows days of the week (from Monday = 1 through to Sunday = 7), with hourly readings showing the rise and fall of NOx levels across the day. Readings for days after the lockdown are in blue/green and the figure shows clearly how these have tended to be much lower than for the earlier days. There are however some overlaps. The data are also compressed due to the presence of a small number of very high readings. The use of a logarithmic scale would give a more balance picture.
The second figure shows the same information but for all the sites, now using a log scale. There are a lot of sites where the post-lockdown readings are much lower on average but also some (e.g. Edinburgh Nicholson Street) where there does not appear to be a change. Some of these sites may be on major public transport routes where the continued flow of buses means that pollution from traffic remains high even if private car volumes have fallen.
These data are also still distorted to some extent by variations in windspeed. We can control for windspeeds using a simple model, and then show whether the measured NOx levels were higher or lower than expected given the prevailing wind. We also add a smoothed line to show averages before and after lockdown. In the majority of cities, there is a clear reduction, see e.g. Belfast. But in others there is more variation with reductions at some sites but not others, see e.g. London Marylebone Road compared with Eltham.
The models allow us to make a simple statistical test of the reduction, and this shows that, across the 36 sites examined here, there is a clear reduction on average. We can even look at the reductions on a day-by-day basis, relative to all days before lockdown. This shows a gradual adjustment to lockdown over the first week, since when levels have been remarkably consistent.
There are significant caveats with this analysis. First, we would stress that the data should not be used to judge how well the residents in each city are conforming to the lockdown policy. Road traffic is not the only source of NOx and, even with traffic, some of this comes from scheduled public transport services which have continued to run. In some sites, continued high levels of NOx post-lockdown may simply indicate that they sit on busy bus routes. Second, the modelling here is a very simple, first stage effort to assess the changes. The next step is to undertake more formal time series analyses, taking account of annual seasonal variations and longer-term trends, not just the short-run of data from 2020.
Nevertheless, the air pollution data show a strong reduction in NOx levels in cities following lockdown. NOx levels are below the levels we would have expected to see given prevailing windspeed. Although there is some variation between sites, the data clearly indicate a substantial reduction in road traffic across the major UK cities. Furthermore, there is no evidence of any weakening in adherence to lockdown in the period covered here.
There is value in continuing to monitor air pollution as we manage the process of returning to normal freedoms of movement. One particular concern is that the experience of the virus may discourage use of public transport which is so urgently needed if we are to achieve more sustainable – and healthy – mobility patterns.
Nick Bailey is Director of the Centre. He is a Professor in Urban Studies, based in the School of Social and Political Sciences at the University of Glasgow.