The COVID-19 lockdown in early 2020 had a visible effect on pollution.
21 February 2023
01 May 2020
By Julia Wagemann (University of Marburg and ECMWF), Federico Fierli and Alessandra Cacciari (EUMETSAT)
The vast majority of countries around the world implemented restrictions to limit the spread of the virus, this led to a reduction in the levels of emissions. However, it is still difficult to assess which elements of pollution were most affected.
Satellite observations are an important source of information to help answer this fundamental question, thanks to their global coverage, temporal continuity, and variety of phenomena measured. For example, data from the TROPOMI instrument onboard the Sentinel-5P satellite, enhanced by BIRA-IASB in Belgium (BIRA2020), showed a clear impact in the Wuhan region of China after authorities imposed a lockdown during the Chinese New Year. Nitrogen Dioxide (NO2), a fast reacting chemical emitted by fossil fuel combustion, that is localised close to production areas, showed a decrease on its total column close to 30–40%, compared to the previous year.
One of the main difficulties is to assess how representative the satellite observations are, because of their different vertical sensitivity and resolution, especially in the boundary layer. It is well known among experts that there are issues trying to directly compare column data with the surface measurements that are representative of the pollution conditions in inhabited areas.
In this case, this is also combined with the difficulty of separating the effects of the lockdown from the variability at synoptic scale — due to meteorological conditions that change the vertical distribution of the pollutant concentration. We know, in fact, that NO2 column concentration decreases during spring, as observed by satellites (see for instance, Geourgoulias2019) due to changes in weather and emission regimes. Also, many countries saw a steady negative trend on NO2 concentrations due to the implementation of long-term air quality control measures.
In this case study we took a preliminary analysis on NO2 from the GOME-2 and Sentinel 5P instruments, complemented with Carbon Monoxide (CO) IASI data from LATMOS (LATMOS2020). We wanted to take advantage of the relatively long-term coverage of the GOME-2 and IASI instruments, thanks to the series of Metop satellites, to reduce the role of specific weather conditions, to allow a preliminary view on how the month of March 2020 compared with previous years.
Satellite data analysis
Figure 1 shows the decrease NO2 in March 2020, seen by GOME-2 instruments. The difference was calculated using average March concentrations from 2007–2018.
During the March lockdown, although there were different levels of restrictions in different countries in Europe, there was a visible decrease, more pronounced in the most polluted areas — Po Valley, Ruhr-Rhine, Benelux, London and Paris. The decrease peaked to 6x10^15 mol/cm2 with an average decrease of approximately 30% to 50%, compared to previous years.
A closer look to the geographical distribution, although not directly comparable due to the different resolution and time coverage, is given by Sentinel 5P TROPOMI data. Compared to the previous year the TROPOMI NO2 data (Figure 2) showed a similar geographical distribution over Europe, but with a more pronounced reduction. This can be ascribed to the higher values detected by TROPOMI thanks to its higher resolution. However, TROPOMI also showed prominent decreases in areas far from the most polluted basins (e.g. the Alps) and this may be due to the concurrent effect of specific weather conditions in 2019 and 2020.
The 13 year-long time series provided by GOME-2 and IASI allows the lockdown months to be put into perspective.
Carbon monoxide has a longer lifetime and should be less influenced by synoptic conditions, thus representing a good marker of overall emission reduction over a large area. The CO measured by IASI (Figure 3), showed a marked decrease in the last week of February 2020, compared to the Feb 2007–2019 reference period, both in China and the Po Valley. This confirms there was a quick impact on pollutant levels as a result of the lockdown.
Figure 4 shows the NO2 in the Hubei Chinese region from 2007 to 2017, plus 2019 and 2020. January and March 2020 are both characterised by anomalously smaller concentration of the total column (2x10^15 mol/cm2); this is up to five time smaller than the average concentrations observed in 2010–2013. Although, a sharp decrease in NO2 concentration was visible in the area in the late 2010s (as shown in DeFoy2016). A decrease in the NO2 columns over western China was also observed in a combined dataset of satellite data from 2011 to 2018 (Geourgoulias2019). The 2019 values appear to be lower than the previous year, and the 2020 concentration were half as low again.
The data presented here are elaborated from:
- GOME-2 distributed under the EUMETSAT Satellite Facility on Atmospheric Composition (AC SAF) database
- Sentinel 5P from the Copernicus SciHub
Python Jupyter Notebook Tools for data handling
Coronavirus lockdown leading to drop in pollution across Europe (ESA)
TROPOMI observes the impact of the corona virus on air quality in China (Royal Belgian Institute for Space Aeronomy)
Coronavirus : baisse de la pollution en Chine et en Italie, vue depuis l’espace avec IASI (CNRS) (in French)
Trends, seasonal variability and dominant NOx source derived from a ten year record of NO2 measured from space (Journal of Geophysical Research)
Satellite NO2 retrievals suggest China has exceeded its NOx reduction goals from the twelfth Five-Year Plan (Nature)
Trends and trend reversal detection in 2 decades of tropospheric NO2 satellite observations (Atmospheric Chemistry and Physics)