Blog update on COVID-19 and changes in air pollution
The first blog on this topic considered some of the early evidence of changes in total nitrogen oxides (NOx) concentrations after the UK lockdown on 23rd March. Since that time, there has been a considerable amount of interest in how air quality has changed around the world. Like the last blog, the results shown here should be considered provisional, not least because the data have still to be fully quality assured. This blog extends our analysis to look at some of the changes we are seeing to nitrogen dioxide (NO2) and ozone (O3).
Winds of change (again)
In my last blog I highlighted how exceptional the weather conditions were in February, this year – windier than any month in over 20 years. Since social distancing was introduced on 16th March, the weather has again been exceptional – this time for the high proportion of easterly winds (45 to 135°). These conditions have seen unusually warm weather for most of the UK, as well as having implications for air quality across the period. Figure 1 shows the distinct contrast in wind direction and speed during 2020 before and after the 16th March. The conditions pre-16th March were characterised by strong south-westerly winds, whereas after March 16th the winds were much lighter and from the east.
Figure 1: Wind rose for Heathrow – January 1st to 3rd May 2020. The data are split between the period before and after the 16th March.
For context, the 6-week period since the 16th March 2020 stands out as an exceptional period of easterlies, when considering wind direction since January 2000 – the period is in the top 0.1% of rolling 6-week periods, with easterly winds prevailing about 51% of the time.
As such the meteorology in 2020 to date has been truly exceptional.
Changes in NO2 and O3
We have updated the cumulative sum analysis from our last blog to include NO2 and O3 concentrations and extended it to include many more AURN sites.
You’ll recall from my earlier blog that the approach taken to explore the lockdown’s impact on air pollution was first to construct a model to estimate what air pollution would have been with ‘business as usual’ without the ‘lockdown intervention’. The work builds upon a range of statistical models, to predict concentrations using meteorological and other data as input (Carslaw and Taylor, 2009, Carslaw et al., 2012; Grange and Carslaw, 2019). The analysis then compares the modelled ‘business as usual’ concentrations that would be anticipated with the actual measured concentrations.
Figure 2 illustrates that most sites showed some deviation from business as usual concentrations around the time of the lockdown. The highlighted sites are those where we have witnessed the greatest absolute deviation from business as usual.
Figure 2: The cumulative sum (or cusum) of measured concentrations minus business as usual NO2 at 29 air pollution monitoring sites across the UK. The lighter shaded rectangle shows the period recommended for social distancing from 16th March, the darker-shaded rectangle shows the period from the start of the lockdown starting 23rd March.
The results for O3 are largely a mirror image of the changes seen for NO2, as seen in Figure 3. For most urban and roadside sites, this is expected behaviour because reduced NOx concentrations will generally lead to high concentrations of O3. It should be noted however, that careful work is needed to understand the changes at the rural sites that might be expected to be more obviously influenced by long-range transport.
The results for O3 also help to emphasise the importance of considering individual pollutants. As can clearly be seen, a blanket statement that air pollution has improved due to COVID-19 is inconsistent with the impact on O3 concentrations. The named sites shown in Figure 3 are those that have shown the greatest absolute increase in O3 concentration.
Figure 3: The cumulative sum (or cusum) of measured concentrations minus business as usual O3 at 29 air pollution monitoring sites across the UK. The lighter shaded rectangle shows the period recommended for social distancing from 16th March, the darker-shaded rectangle shows the period from the start of the lockdown starting 23rd March.
So, what do these changes look like at individual sites? Figure 4 below considers the changes from 16th March to early May 2020 and aims to quantify how concentrations have changed. On average, it is estimated that NO2 has decreased by 34% across all sites, but there is a large variation in responses and some interesting behaviours that will warrant further investigation.
Figure 4: Measured and predicted business as usual concentrations of NO2 at selected AURN sites.
The decreases in NO2 seen in Figure 4 are to a large extent compensated for by corresponding increases in O3 – shown in Figure 5. On average, we estimate that ozone, at a mix of urban background and traffic locations, has increased by 27% since the lockdown.
Figure 5: Measured and predicted business as usual concentrations of O3 at selected AURN sites.
Given the exceptional conditions with respect to easterly winds since the lockdown, it is important to ‘control’ for this in any assessment of the impacts of the COVID-19 lockdown on individual air pollutants.
Carslaw, D.C. and P.J. Taylor (2009). Analysis of air pollution data at a mixed source location using boosted regression trees. Atmospheric Environment. Vol. 43, pp. 3563–3570.
Carslaw, D.C., Williams, M.L. and B. Barratt A short-term intervention study — impact of airport closure on near-field air quality due to the eruption of Eyjafjallajökull. (2012) Atmospheric Environment, Vol. 54, 328–336.
Grange, S. K. and Carslaw, D. C. (2019) ‘Using meteorological normalisation to detect interventions in air quality time series’, Science of The Total Environment. 653, pp. 578–588. doi: 10.1016/j.scitotenv.2018.10.344.