Please find my comments in the attached pdf file.

The study investigated the impact of 2020 COVID-19 lockdowns on air pollution across Europe by comparing the measurement data of PM, NOx and O3 concentrations with expected values revealed by ensemble forecasts if no COVID-19 epidemic had occurred. PNSD and particle light absorption coefficients in-situ measurement data were also used to assess the impact of lockdowns on air pollution. One of the most important findings revealed by this study is that an enhanced oxidizing capacity of the atmosphere could have boosted the production of secondary aerosol which compensated the decrease in primary PM emissions during lockdowns. Given that research on the impact of COVID-19 lockdowns on air pollution has widely conducted over the world, this finding is not surprised and excited. There was no in-depth discussion on this conclusion, e.g., examining the change of chemical composition of PM during lockdown to highlight the increased production of secondary aerosol. In general, this manuscript is well organized but to some extent short of the scientific significance. A major revision is needed by strengthening the result analysis to highlight the scientific objective and significant of this study.

There are also some specific comments to this manuscript as listed below.

1. The abstract is somewhat verbose. It is suggested to illustrate the most important findings of this study are given directly in the abstract. For example, the statement “Changes in the wavelength dependence of the particle light absorption coefficients and PNSD were also examined at 14 and 13 sites, respectively. Since these variables are not calculated by the CAMS model, expected values were estimated from 2017-2019 measurement data” is suggested to be removed from the abstract.
2. The major objective, particularly the scientific aim, of this study is not well documented in the Introduction Section.
3. Line 110, why “15 nm is reasonably insensitive to new particle formation bursts”? Nanoparticles generated through new particle formation can grow to size larger than 15 nm via condensation and coagulation processes in the atmosphere. The lockdown measures lead to a decrease in primary particle emissions but likely result in an increase in new particle formation owing to the enhanced oxidizing capacity of the atmosphere.
4. Figure 3 shows the mean observed / expected PM concentration ratios. How these concentration ratios related to the PM increase or decrease resulting from lockdown? What were the mean increases in PM2.5 and PM10 mass concentration of +1±42% and +5±33% shown in Section 3.2 calculated from? If the increase in PM mass concentration was calculated from the observed/expected ratios between A and D or between D and P, was it assumed that the mean observed/expected ratios during A, D and P were consistent if there were no lockdown measures? Did the mean observed/expected ratios keep consistent during the same 3 time periods in 2019?
5. The mean Obs2019/CAMS2019 ratios during the 3 time periods A, D and P in 2019 is suggested to be used in Eq.1 to account for the bias between CAMS forecasts and observation data instead of the daily Obs2019/CAMS2019 ratios.
6. Line 144, “originate” should be “originated”.
7. Line 180, a left bracket is missed for “(FR)”.
8. Line 180, Full names of counties are suggested to be given when they first appear in the manuscript.
9. Line 218, what’s CYSTAT?
10. Line 233, the definition of 3 time periods before, during and after the lockdowns is not needed here since it has been described in detail in Section 2. It is the same for the expression in Line 253.
11. Line 325, The mean +17±24% increase in O3 concentration at regional background sites was larger than the increase fraction of +11±23% in cities, thus, the increase was not marginal.