General remarks:

The authors ran WRF-Chem simulations over East Asia using three different emission inventories (EDGAR-HTAP v2, EDGAR-HTAP v3, and KORUS v5) and compared the model output to three sets of observations (routine monitoring data, airborne KORUS-AQ data, and ground-based KORUS-AQ data).  They also ran sensitivity tests to doubling CO and VOC emissions and probed how the chemistry changed.  Such comparisons are useful for model development, but I think there is opportunity for the paper to be strengthened in the following ways:

1. Currently, O₃ and NO₂ are treated separately in the comparison. I suggest adding a comparison of odd-oxygen (O_x = O₃ + NO₂) to probe whether the model issue is too much O₃ titration by NO or problems with the O₃ production regime.
2. On a related note, can you use the individual comparisons you’ve done of VOCs, NOx, and O₃ in different regions with each inventory to draw some conclusions about how biases in either or both NO_x or VOC emissions affect O₃ predictions / chemical regimes in the different regions? There is a little bit of this on pages 13-14, but more organized conclusions about this (especially with your sensitivity tests to doubling the CO and VOC emissions) would be very useful for future model interpretation and emissions inventory development. For example, on page 12, can you discuss why the differences in VOCs and NO_x in each of the inventories cause them to simulate O₃ differently?
3. Most of the conclusions in the manuscript are stated as “X is biased low with Y emissions inventory.” These statements would be more useful to the atmospheric chemistry community writ large if those statements were extended to say, “X is biased low with Y emissions inventory, which has Z implications for our understanding of emissions/chemistry.”  For example, on page 20 line 12, can you add something to this sentence about the implications of having larger biases in the Transport case compared to the Local case?  For a second example, on page 20, line 20, can you add something about the implications for NO_x emissions (based on O₃ being wrong but CO and HCHO being largely okay)?  There are many other instances in the manuscript where this would be useful, but hopefully, the two examples I provided here are helpful illustrations.
4. The manuscript includes some contextualization of this work in the context of other emissions inventory comparisons (e.g., for CO on page 8). I think the paper would be strengthened by adding similar contextualization for the other comparisons (NO_x, O₃, VOCs, etc.), especially given how many model-measurement comparisons have been done to date with KORUS-AQ (and related) data.

Specific comments:

Page 5, line 8: are the NMVOCs lumped or speciated?

Page 5, line 9: do you apply any scale factors for using 2010 emissions data in a 2016 simulation?

Page 5, line 17: What does ‘specifically’ mean here?

Page 7, lines 1-2: should read “…toluene and less reactive aromatics…”

Page 7, line 14: what species are you referring to that is larger in South Korea by 263%?

Page 8, line 1: add “respectively” after “(HCHO)”

Page 8, line 13: I think it would be clearer to say “For all emission inventories…” rather than “With all emission…”

Page 8, lines 14-17: It was hard for me to figure out which simulations correspond to which numbers in these sentences.  Reword to clarify?

Page 10, line 19: What kind of interpolation method was used?

Page 11, lines 11-19: Can you say something about how much these meteorological biases affect your comparisons?  For example, how much would temperature-dependent evaporative VOC emissions change over these temperature ranges?

Page 12, line 11: “all emissions inventories” instead of “all emissions”

Page 17, line 3: ISO definition should be moved earlier to where it’s first used.

Page 19, lines 23-24: can you use the biases calculated during the local case to draw some conclusions about the emissions inventory over China?

Title: Unclear what ‘precursor’ refers to here.  Is it O₃ and HCHO precursors?  If so, perhaps rephrasing it as “ozone, formaldehyde, and their precursors” would be clearer.

The authors provide an overview of emission inventories and model simulations in East Asia using WRF-Chem v4.4 with three different emission inventories. To do so, they compare model results with various observations and conduct a sensitivity test by doubling CO and VOC emissions in China and South Korea. This study is significant as it shows the current state of emission inventories and represents points to improve simulated ozone. However, the following comments should be considered to enhance the study.

General comments:

1. In my opinion, the authors need to clarify the main topic and purpose of this study. I think that the research aims to evaluate model simulations using three emission inventories, comparing them with various observations and to conduct an analysis and sensitivity test of underestimated ozone. However, these aspects do not seem to be adequately explained in the introduction and abstract. Additionally, it should be made clear whether the analysis of underestimated ozone will focus on East Asia or be specific to South Korea. If the authors want to focus on East Asia, additional analysis for China has to be included, as discussed for South Korea. The scope of regions needs to be clarified.
2. The paper discusses which emissions should improve and what causes the underestimation of ozone, comparing simulated species with observations. However, the discussion needs to be organized more. It mainly focuses on the underestimation of VOC emissions about underestimated ozone, but the impact of VOC emissions can vary depending on the ozone production regime (relative ratio to NOx emissions). The paper independently compares VOC, NOx, and O3 mixing ratios, but these species are related to each other, and more effort considering them together is needed to understand model performance and emissions. Therefore, additional explanations about these regimes are necessary for the analysis during KORUS-AQ. In addition, NOx emissions could contribute to VOC chemistry, resulting in ozone changes, so it would be great to describe them in the sensitivity analysis.
3. The font sizes in the figures are small, and the figure resolutions in both the manuscript and supplementary document are low. Please enhance their readability.
4. Many paragraphs are overly lengthy, attempting to cover multiple topics within a single paragraph. Please ensure that paragraphs are concise.

Specific comments:

P3 L3-15: In my opinion, the paragraph should be separated at L9. Additionally, before providing an overall description of the paper, it is important to clearly explain what the authors want to convey through the paper and what scientific significance it holds.

P4 L18: The link is not open.

P4 L21-22: Even so, in South Korea, the authors analyzed model performance over China. Therefore, the authors should discuss the effects of fire emissions on China.

P5 L2 What about “Anthropogenic bottom-up emission data” or something similar?

P5 L4: The bottom row of Figure 1 is for only toluene or TOL (toluene +less reactive aromatics)?

Please rearrange the order of figures and tables in the manuscript and supplementary document. Table S3 appears before Table S2. Similarly, Figure 7 needs to be rearranged.

P6 L23: Only toluene or TOL emissions?

P7 L1-4: Is CO a major precursor affecting ozone formation?

P7 L4: It would be great to include boxes of three regions in Figure 1.

P7 L12-18: What about other VOC emissions? Toluene is one of reactive species with OH, but isoprene, ethene, and other species are also reactive with OH, leading to ozone production. It would be helpful to show how different total reactive VOC emissions are between three inventories.

P7 L23-P8 L1: The sentence could potentially mislead about VOC emissions, as formaldehyde is produced by the oxidation of many VOCs. Some VOC species might overestimate or underestimate.

P8 L14-21: Regarding general comment #1, the authors should clarify the scope of this study for ozone underestimations.

P9 L11-P10 L2: The paragraph could be separated at L19, and the sentence at L22-23 might be relocated after the description of NO2 in South Korea. Please revise.

Section3.2:

It would be helpful to explain what causes discrepancies between models and observations. In addition, the authors mentioned that low MDA8 in the models could be related to low VOC emissions. However, determining the regimes to which certain areas belong should involve considering both NOx and VOCs. Underestimated ozone could result from both high NOx concentrations in the model and low VOC emissions. In Section 3.2, the comparison of ozone and NO2 between models and observations was discussed separately. Combining the analysis and discussion would enhance the manuscript and the understanding of emissions.

P13 L16-19: The authors mentioned ground-based NO2 observations over China and South Korea have positive biases in Section 2.3.2. If these positive biases are corrected, FNR might exceed 1, indicating a transition or NOx-limited regime. In addition, it would be useful to provide the range of the FNR to determine ozone production regimes.

Figure S5: Please plot NO2 figures together with FNR and HCHO.

P14 L11-13: In Figure 2, ozone mixing ratios simulated with EDV3, KOV5 are not substantially lower than the observations, and MDA8 ozone also appears as yellow in Figure 3.

P14 L16-17: The lowest bias of -5.5 ppbv was observed for KOV5 in South Korea.

P16 L10: It would be helpful to describe why the authors showed HCHO.

P17 L5-6: Overall ozone production regimes were discussed in Section 3.2. However, to lead to your conclusion that underestimated ozone results from low VOCs during KORUS-AQ, it would be helpful to show FNR during KORUS-Aq. Also, the lifetime of CO is too long to significantly contribute to ozone. How much does the CO oxidation contribute to ozone? Is it related to background ozone concentration, and then how much?

P17 L14-16: During KORUS-AQ, ground-based NO2 was measured from a photolytic converter following the description in P10 L6-8, correct? If data from both molybdenum and photolytic converters were available, why did you choose to use data from molybdenum converter for the comparison? Also, low NOx from the model is a result influenced by various factors, including emissions, chemistry, PBL, and others. It would be helpful to describe which factors are affected rather than just saying the comparison with observations.

P17 L 20-22: Simulated TOL from KOV5 also shows significant differences from observations below 1 km except at the surface.

P18 L13-14: It would be great to explain why surface and vertical columns exhibit different diurnal patterns.

P18 L15: VCD patterns differ between model and Pandora.

P19 L 2-4: How did diurnal profiles contribute to reducing biases?

P19 L18-19: Even though the authors cite Peterson et al. (2019) to separate local and transport cases, please provide a brief description for readers.

P20 L22-24: When considering the contributions of South Korea’s CO and VOCs emissions, calculated as the difference between ch2 and chko2, the increase of CO and VOC emissions does not lead to improvements in CO and VOCs. The difference in CO is only 5 ppbv, while the difference in HCHO is 32 ppb, resulting in overestimation of the model. In addition, in chko2, doubling emissions in China might affect O3, CO, and HCHO in Korea because of transport. Also, local cases may be influenced by the transport of species with relatively long lifetimes. To describe the effects of South Korea’s emissions on ozone, CO, and HCHO, the author could simulate the model with only doubling CO and VOC emissions in Korea.

Figures 2-6: Adding boxes to the plots in the second and third rows would be helpful for easily recognizing the regions.

TableS4: Please provide definitions for the species names in MOZART and SAPRC-99.

Table S7-S8: Please clarify the time period for the data.

Figure 8: Please indicate the color for observations.