This paper presents model-measurement comparions for aerosol surface area that affects heterogeneous chemistry. The paper is generally well written. I have a few comments:

1. Figure 13: Since the question is how ratio of modeled to measured aerosol surface area changes with aging, it might be better to plot it with respect to total modeled  SOA particle mass rather than just modeled isoprene oxidation products. Can the authors generate a similar plot comparing to modeled SOA?

2. Some of the model-measurement differences in aerosol surface area might relate to processes not considered in the model. For e.g.:

(a) How does phase state and viscosity of organic aerosols affect heterogeneous loss of organic aerosols in the model? Heterogeneous loss also affects aerosol surface area and vice versa. 

(b) How does model representation of aqueous and cloud chemistry affect model-measurement comparisons of aerosol surface area?

3. The model-measurement comparisons in the paper are focused on an aircraft campaing that samples upto 5 km altitude (e.g. Fig 8). It would be interesting to compare at higher altitudes e.g. in the upper troposphere as well. Are there measurements from other aircraft campaigns that could be used as well?

4. How does relative humidity affect model-measurement comparison of aerosol surface area?

5. A general comment: A more useful approach might be to perturb aerosol formation and loss processes in the model e.g. turning heterogeneous chemistry on/off, aqueous chemistry on/off,  etc. and determine which of these affect model-measurement comparisons of aerosol surface area the most, especially as a function of aging. 

6. The reactive uptake analyses only focused on N2O5. Could the authors look at other reactive uptake coefficients as well e.g. related to IEPOX-SOA?

The authors compared CMAQ simulated particle surface areas (Sa) with those observed during the 2011 DISCOVER-AQ Baltimore-Washington and showed that the model under-predicted the Sa by a factor of ~ 2. Possible reasons for the difference were analyzed and discussed. The authors also studied the impact of the model simulated Sa on heterogeneous chemistry. Overall, the manuscript provided a useful assessment of CMAQ predicted Sa. However, the following comments need to be properly addressed before I can recommend it to be published in ACP.

1. Particle surface area and number concentrations depend strongly on the size distributions of primary particles emitted (especially in the urban region like this study). It is unclear in the manuscript how primary particles of different sources (industry, residential, traffic, etc.) are parameterized in the model. This information is necessary and a table summarizing these will be helpful.
2. Particle size distributions (and thus Sa) are also affected by new particle formation (or nucleation) and growth, especially at higher altitudes and in areas away from the urban center. The authors shall discuss how new particle formation was treated in the model and how the process might have affected the results.
3. The results presented in this paper are based on CMAQ simulations with 12 × 12km latitude and longitude horizontal resolution, and hourly temporal resolution. As the authors themselves acknowledged: “The 12 km ×12 km spatial resolution and 1-hour temporal resolution of CMAQ is significantly larger and longer than the spatial and temporal resolution of the aircraft data, resulting in an inherent contrast in resolution between model and measurement that may play a role in the variance in Sa,mod /Sa,meas”. Any reasons why you did not use a nested domain with higher resolution? Also, why not use a higher temporal resolution? 1-hour temporal resolution is too coarse for this kind of simulation (12 km x 12 km).
4. Figure 1. Is the figure for an average of values at all altitudes? It will be helpful to separate those in the boundary layer from those above the boundary layer. I expect a large difference in particle size distributions at different altitudes. Also maybe you can use the same y-axis scale for observed and model values for easier comparison?
5. The manuscript does not have a summary session. I would suggest that the authors add one.
6. L79-81: Are the median diameters for the three modes fixed in the model? How the size range for each mode is decided? How was the value of geometric standard deviation determined?
7. L100: “>99%”: What is the exact value?
8. L104: How condensation of water is treated?
9. Table 1: Q1, Q3 values from Jaegle et al. 2018 appear to be incorrect.
10. Why did two simulations (2011 &2015) use two different versions of the CMAQ model and chemistry/aerosol schemes (CB5 vs CB6, AER06 vs AER07)?
11. L222: Please explain why the measured values are dry Sa.
12. L145: relative to?
13. L283: Please give some details on how this was measured. Completely dry or at some fixed RH?
14. L305-309: What are the possible reasons? Can the difference be reduced by increasing resolution and/or modifying assumed mode sizes?
15. L310: This is inconsistent with what is stated in the figure caption. Please add more labels to the x-axis to make it clearer.
16. L388: This is speculation here.  Why not use a nested domain with 4 km resolution to look into the effect of resolution?
17. L446: Does the condensation change the sizes of particles represented in the model?
18. L481: Please explain how the reactive uptake coefficient was measured.
19. L485: Why? Winter and summer particle size distribution (and thus Sa) could be quite different.