Quantification of SO₂ Emission Variations and the Corresponding Prediction Improvements Made by Assimilating Ground-Based Observations

Round 1

Reviewer 1 Report

1. Please rewrite the abstract to better reflect the content of the work, in particular, I find it important to clearly state that experiments are done in this work and used for the analysis.
2. Make sure all abbreviations are properly explained the first time they are mentioned.
3. More detiled description of your method and state-of-art is neede to better understand your work, methodology and findings.
4. Results need to be re-written. Add the proper discusion
5. In the experimental section, there is lack of real data, FILL IT UP.
6. English Language needs extensive chacking

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript deals with an interesting analysis of emission reduction in the most polluted part of the world. For the air quality modeling community, the COVID situation gave us a one-time-only opportunity to test models and emission inventories. It is always a struggle; the lack of modeling performance is due to meteorology IC-BC or emission inventories. The paper is very interesting, no major flaws are found. After minor revision, the proposed manuscript merits publication in the journal Atmosphere.

 

Minor comment:

 

Ln 92. It would be useful to have information on estimation of emission decrease here or at least to what percentage decrease in man-hours in the industry sector of China happened.

Ln 118. The link directs to some site with no data.

Ln 123. Confusing – the figure caption is much clearer in the description of red dots.

Ln 143. “Backward part” indicate something else, I assume that the authors mean (based on further sentence) that this is module for calculation of “backward trajectories”? I suggest authors be more clear here. How much backward steps are used (on average)? Always the same steps in each day or this number varies? 

Ln 181. Does cost function depends on different species?

Ln 231. As there were many fireworks during this period, which as we see had an influence on SO2 emissions, why did the authors choose that particular period? Why not choose also at least a couple of random periods for simulations? Or at least extend them up to 10 days, as the average lifetime of SO2 in the atmosphere is around 10 days?

Ln 266. These are averaged period differences?

Ln 405. Did the authors mad any sensitivity analysis on usage of other chemical parameterizations in WRF-CAUCE? It is possible that other combinations of parameterizations can lead to different modeling performances. The same comments go for different horizontal resolution, can an increase lead to better modeling performance?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf