The paper assimilated a long-term data which is critical for air quality managers and researchers to understand the long term air quality trends and changes at a fine scale county level. Also made that long-term data available to the stake holders through a dashboard which makes it very easy to visualize the data.

Specific Comments:

Line 107: Is the methodology any different than any of the previous studies? If so the author should provide that information and highlight the differences. 

Line 188: What is the rationale to perform 9 simulations every day rather than the every hour of the day?

Line 189: The first simulation seems to cover more time period that the subsequent simulations. Will that compromise any of the model predictions?

Line 212: Did the author perform any trace gas species performance for the with and without assimilation of the MOPITT to show the indirect effect of CO on the trace gas species?

Line 275: Can the author specify the 10 regions in the text or in the supplementary document? The 10 regions information will be helpful for the reader to understand the regional changes in the data. 

Line 279: Figure 2-5: The time series for all the parameters appears too clustered and it is not clear to see the red and black lines. Is it possible to reduce the temporal resolution to something like daily while plotting?

Line 323: Figure 7: Again the hourly time series looks very overcrowded, can this be changed to daily MDA8?

Line 325-327: The author made a point about the nighttime ozone but there is no figure or data to support the statement. Can the author include figure or a table in the appendix to show this finding?

Line 373: Explain what is trend analysis here, so that the reader can understand the context clearly. 

Line 379: Why did the author pick 2-sigma level for this analysis?

Line 391: Include the number of sites in the parenthesis. 

Line 509: Reword this line "millions of people in counties with no monitors"

This paper describes the methodology and data evaluation for a long-term, high-resolution air quality reanalysis using assimilated AOD, CO, and the CMAQ air quality model. The authors have also developed a dashboard to make the data easily accessible and publicly available to stakeholders. The high-resolution reanalysis offers valuable information, particularly in regions without monitoring stations, and can be useful for local policymakers and public in understanding air quality trends and exceedances under various pollution control policies.

Overall, the paper is clearly written, particularly in the methods section. However, despite the detailed evaluation provided, a more thorough analysis of the reanalysis performance on diurnal cycles, extremes, and challenges in simulating PM2.5 peaks, as well as PM2.5 trends in certain regions, is needed. This additional information would help users better understand the data's quality and potential limitations when interpreting trends or assessing policy effectiveness.

Major Comments:

1. Are Figures 2-5 and Figures 7-8 showing hourly time series? The diurnal cycle is difficult to see from this figure. It would be helpful to also compare the diurnal cycle for four seasons for meteorological variables as well as ozone and PM2.5 concentrations.
2. The black lines representing observations in Figures 7-8 are difficult to see. The authors might consider increasing the transparency of the red lines. Also, additional analysis of the diurnal cycle and annual trends would be valuable as suggested above.
3. The air quality reanalysis seems unable to capture PM2.5 trends in some regions and PM2.5 extremes exceeding the national standard of 35 µg/m³ in most regions (as shown in Figure 8). These discrepancies could significantly impact trend and exceedance evaluations, and thus affecting policy interpretation. What is the main causes of these biases? Is there a reason why assimilating AOD data does not mitigate these biases?
4. Figures 11-12 provide examples of the dashboard displaying annual trends in pollution concentrations and the number of days exceeding national standards at the county level. Model uncertainties in reproducing the observed trends and daily pollution extremes may be an issue. For instance, the mean bias for ozone is 3.7-6.8 ppbv and for PM2.5 is -0.9-5.6 µg/m³, which could significantly impact the calculation of days exceeding 70 ppb or 35 µg/m³. Is there any way to add an uncertainty metric to the datasets and the display?
5. The evaluations are performed and shown at 10 EPA regions, but the dashboard displays data at the county level. It would be helpful to provide uncertainty statistics at the county level. Also, it may be interesting to compare model performance in counties that are currently non-attainment.

Minor Comments:

1. Lines 284-292: Should "RH2" be "RH"?
2. Why did the authors choose the study period of 2005-2018? Meteorological and surface air quality datasets are available up to 2023. Is there any plan to update the data to include more recent years?
3. Can the authors clarify what is meant by "meteorology-dependent anthropogenic emissions"?
4. Line 420: Should "fig. 9" be "fig. 10"?

1. Figure 5 caption, 2 m temperature -> wind direction?
2. In the introduction section on recent trends in surface ozone and PM2.5, these following recent papers also provide detailed analysis on the observed trends and drivers, and include discussion of model problems of simulating these trends and extremes.

• Lin, M., Horowitz, L. W., Payton, R., Fiore, A. M., & Tonnesen, G. (2017). US surface ozone trends and extremes from 1980 to 2014: quantifying the roles of rising Asian emissions, domestic controls, wildfires, and climate. Atmospheric Chemistry and Physics, 17(4), 2943-2970.
• Xie, Y., Lin, M., & Horowitz, L. W. (2020). Summer PM2. 5 pollution extremes caused by wildfires over the western United States during 2017–2018. Geophysical Research Letters, 47(16), e2020GL089429.
• Burke, M., Childs, M. L., de la Cuesta, B., Qiu, M., Li, J., Gould, C. F., ... & Wara, M. (2023). The contribution of wildfire to PM2. 5 trends in the USA. Nature, 622(7984), 761-766.