Implications of Sea Breeze Circulations on Boundary Layer Aerosols in the Southern Coastal Texas Region

Subba, Tamanna; Jensen, Michael P.; Deng, Min; Giangrande, Scott E.; Harvey, Mark C.; Singh, Ashish; Wang, Die; Zawadowicz, Maria; Kuang, Chongai

doi:https://doi.org/10.5194/egusphere-2025-2659

Preprints

https://doi.org/10.5194/egusphere-2025-2659

Preprints

27 Jun 2025

| 27 Jun 2025

Implications of Sea Breeze Circulations on Boundary Layer Aerosols in the Southern Coastal Texas Region

Tamanna Subba, Michael P. Jensen, Min Deng, Scott E. Giangrande, Mark C. Harvey, Ashish Singh, Die Wang, Maria Zawadowicz, and Chongai Kuang

Abstract. The Sea Breeze Circulation (SBC) influences atmospheric processes at multiple scales in coastal regions. Understanding how SBCs impact the aerosol number budget and aerosol impact on perturbation of incoming solar radiation is essential. This study investigates sea breeze-aerosol interactions (SAIs) during 46 summertime SBC events using data from the TRacking Aerosol Convection Interactions Experiment (TRACER) field campaign across rural and urban sites in southern Texas. Weather Research and Forecasting model coupled with Chemistry (WRF-Chem) simulations complement observations to explore spatio-temporal meteorological controls on boundary layer aerosols. During the campaign, Sea Breeze Fronts (SBF) penetrating inland transported cool, moist air over the land, introducing air masses with distinct properties compared to the preexisting continental air. These SAIs cause variability in number concentrations of up to a factor of two, with events typically lasting ~5 hrs before returning to background conditions. SAI’s impact on aerosols varies with site’s proximity to the sea and the preceding sea breeze (SB) history, primarily affecting the marine-influenced accumulation mode. The coastal site reflects stronger marine influence, while the inland site reflects SB air masses that regain continental characteristics. Model outputs show that the regional SAIs extend ~50 km inland and reach up to the boundary layer height. SBC further influences the local aerosol radiative impacts by changing the aerosol number budget. SAIs modify cloud condensation nuclei in ~20 % of events during SBF passage. For the typical TRACER SBF passage, the local aerosol radiative forcing was also found to change by up to 40 %.

Received: 05 Jun 2025 – Discussion started: 27 Jun 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 14018 KB)

Supplement (8545 KB)

Download & links

Preprint (14018 KB)
Metadata XML
Supplement (8545 KB)
BibTeX
EndNote

Tamanna Subba, Michael P. Jensen, Min Deng, Scott E. Giangrande, Mark C. Harvey, Ashish Singh, Die Wang, Maria Zawadowicz, and Chongai Kuang

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-2659', Christopher Nowotarski, 23 Jul 2025
General Comments
            This paper presents an overview of aerosol and (to a lesser extent) meteorological changes associated with the passage of a sea-breeze front at two ARM sites during the DOE TRACER field campaign in the Houston region of SE TX during Summer 2022. WRF-Chem simulations supplement the point-based observational analysis.   In general, the paper is largely descriptive (rather than explanatory ) of broad changes in aerosols, providing analysis of before/after SBF passage changes. A key finding of the paper is the number and relative frequency of SBF passage events where aerosols increase, decrease, or stay the same during the SBF passage. As a general description of the field project and presumably regional climatology this paper has value. Where the paper lacks is in rigorous scientific investigation of the reasons why there are differences between sites or days that lead to increases vs. decreases in aerosols. The paper also seeks to tie changes in aerosol to changes in cloud microphysics and aerosol indirect and direct radiative effects. While these are interesting questions, there is not much depth to the analysis and it feels rather tacked on, and these topics may be better suited by separate more detailed analysis in other papers.    I think a reframing of the paper that provides a detailed analysis of each of the three case study days (July 10, July 17, August 16) where the three regimes are observed that seeks to analyze air trajectories differences in SBF meteorology and timing on these days, and the factors that lead to different antecedent aerosol conditions would be more of more interest.
Finally, two recent papers in the literature focus on very related meteorological (Sharma et al. 2024 https://doi.org/10.1175/MWR-D-23-0243.1) an aerosol changes (Thompson et al. 2025 https://doi.org/10.1029/2025JD043353) in continental and maritime air masses divided by the SBF during TRACER. This may come across as shameless self-promotion since I was involved in both papers, but the analysis here complements those studies nicely since it focuses on analyzing ARM site observations while those papers emphasized TAMU TRACER (Rapp et al. 2024 https://doi.org/10.1175/BAMS-D-23-0218.1) measurements. Where possible, I think this paper should put its findings in the context of the existing literature and provide comparisons/contrasts where appropriate.

Specific Comments
(abstract, line 24) The coastal vs. inland categorization of the AMF1 and ANC sites is somewhat misleading. While AMF1 is closer to Galveston Bay, it is about the same distance from the Gulf of Mexico as the ANC site.

(Page 2, line 50) I suggest adding “in the warm season” behind “along coastal regions.”

(Page 3, line 65) I caution against using the relatively general terms “stable” and “unstable” here to describe the maritime and continental airmasses. While the low-level lapse rates may be more stable on the maritime side, we observed throughout forecasting for the project (and described in Sharma et al. 2024 https://doi.org/10.1175/MWR-D-23-0243.1) that conditional instability (CAPE) was often larger on the maritime side of the SBF given the greater moisture in that airmass (similar to MAHTEs described by others, including Hanft and Houston 2018 https://doi.org/10.1175/MWR-D-17-0389.1).   I suggest either simply removing these descriptors and leaving it at as “cooler” and “warmer” or adding a sentence or two here to add this nuance.

(Page 4, Line 124 and throughout) Here the main site is referred to as coastal, and while it is coastal in the sense that it is very close to Galveston Bay, it is not much closer to the Gulf of Mexico coast than the ANC site. This should be clarified here and elsewhere were appropriate. More generally in the introduction, I think some distinction between airmasses and SBCs associated with Galveston Bay vs. the Gulf of Mexico should be described.   The SBFs associated with each body of water are often distinct at the onset of the SBF, but become more merged later in the afternoon/evening. There is likely considerable meteorological heterogeneity and aerosol variability within the maritime airmass that is heavily modified by Galveston Bay as compared with one only sourced over the Gulf of Mexico.

(Section 2.1) A few points that should be included in the description of the M1 and S3 sites: While the text accurately describes M1 as “urban” I would also suggest discussing the concentration of heavy industry surrounding the M1 site, which one might expect to lead to aerosol populations above and beyond the typical “urban” air mass. Conversely, while S3 is relatively removed from the Greater Houston area pollution, with a typical SSE wind direction, it is still downstream of some heavy industry on the SE TX coastline in the vicinity of Freeport and Lake Jackson (as Figure 2b hints at, there are still some anthropogenic sources upstream of S3). Thus, S3, though rural may not always be expected to be as pristine as a typical rural site.

(Page 7, Line 197) Please discuss the timing of the “simulations” in more detail. Was one simulation begun at the beginning of the 2 month period and run for 2 months, or were daily simulations initialized at some fixed time for each day in that period? If so, what times, and how long was accounted for model spin up for the WRF domain before using the results? How often were the boundary conditions and emission sources updated?

(Page 7, Line 214) The word “feasible” here is vague/unclear. What is meant by feasible? Close enough to trust and use? If so, what criteria were used to judge this? To my eye, it looks like the model significantly under predicts the diurnal temperature cycle (meaning it is likely to also struggle handling the SBC, and it also over predicts the wind speeds during calm periods (presumably night time).   I understand that this paper is more focused on the model performance as related to aerosols (which is described in some detail later in this paragraph), but since the hypothesized mechanism for variability in aerosols (the SBF) is after all a meteorological phenomenon, I think a more detailed discussion of the model’s meteorological skill is needed.

(Page 8, Lines 243-245) The authors state that the two sites have a statistically significant difference, but in what sense? Is every meteorological variable statistically different between the sites, or in the whole? Is the important difference in the magntudes of the variables or in their diurnal cycles? It’s not clear what these differences actually are or why it’s relevant that they are meteorologically distinct.

(Page 8, Lines 247-249) The authors say that “the opposite trend is observed for RH” which is true in the strictest sense, but for a constant amount of absolute moisture, the differences in RH are consistent with the differences in temperature between sites (RH for constant vapor pressure or dewpoint is lower when it is warmer and higher when it is cooler). Because RH depends on both temperature and moisture, I suggest that the authors use an absolute measure for moisture like vapor pressure, mixing ratio, or dewpoint temperature instead of RH. This temperature dependence of RH may obscure moisture changes associate with the diurnal SBC cycle.

(Page 9, Lines 261-269) This paragraph largely describes methodology for identifying SBC from a prior paper that is used later in the analysis of this paper. Some or all of this text is more relevant in the data/methods section. It does bring to mind several relevant questions that this section could/should seek to answer, including: does the SBC have a notable effect on the observed surface meteorological variables at each site? Is this affect more pronounced at one site vs. another? What time does the SBF typically pass each site? There may be some hints of this in Figure 4, and since the paper is motivated by the effects of the SBC on aerosols it seems relevant to me to also analyze/discuss the effects of the SBC on the surface meteorology in this section where possible. Particularly relevant to the later aerosol discussion is addressing the questions of: Is the diurnal meteorological cycle distinctly different on days with a SBC than without (it seems like Figure 4 includes all days regardless of if they feature a pronounced SBC)? This is important, because some of the observed aerosol changes could be diurnally forced rather than SBC forced. For example, as the surface temperature warms in the afternoon vertical mixing may increase which could lower aerosol concentrations independent of the SBC, which may confound later results that are attributed to the SBC. Alternatively, the diurnal inertial oscillation in the flow could affect aerosol variability without a SBC.

(Page 10, lines 286-296) The percentages listed in this paragraph do not match those shown in Figure 5c. Since this figure is not cited here, it is not clear if they should match or if the statistics listed here are computed differently than in the figure. If this is an error, please correct. If it is intentional, please explain why the listed percentages should not match the pie charts.

(Page 11, first paragraph) While this paragraph is a nice summary of the possible changes on a local/regional airmass’ aerosol, one key possibility is missing that should be included. Since much of the analysis relies on point measurements at the S3 and M1 sites, it’s possible that small wind shifts caused by the SBC could lead to very large aerosol changes if the wind shift causes the upstream trajectories to shift away from or towards a very local emissions source. This is a highly localized effect that would lead the point observation sites to be less representative of the overall airmass. The authors somewhat account for this by including WRF simulations to fill in the gaps with approximate aerosol states, but this possibility should be mentioned here, and throughout when one of the site measurements could be unrepresentative of the larger airmass.

(Page 12, lines 353-355) Please provide some explanation for why +/- hour was selected as the before and after times. Is this because, as in the stated example that most enhancement or reduction effects last at least one hour? Or was some other criteria used for this selection?

(Page 12, lines 359-366) The percentages in parentheses (60% and 34% of SB days with a change in aerosol concentration) is somewhat unclear in the first sentence here Does this mean that 40% and 56% of SB days would be considered “neutral influence” days? It’s also not clear if the total SB days are the same at each site. Some of this information is discussed later in this paragraph but it’s a bit hard to decipher. I think that supplementing or replacing some of this text with a table showing the number and percent of SB days at each site showing the total, number with an enhancement, number with a reduction, and the neutral days would be helpful here. Moreover, the average aerosol change is stated for each site during SB days, but it’s not clear if this includes the neutral days with the enhancement/reduction days or only the latter. On line 368 the authors state that neutral days aren’t included in the analysis, yet later in this paragraph they state the number of neutral days at each site, which leaves some ambiguity on which statistics include neutral days and which don’t. This should be clarified. It also seems to mask the characteristic aerosol concentration for each regime (enhancement or reduction) by averaging them together.   It would be more useful to show the average change of the enhancement and reduction days separately for each site. As is currently presented, the enhancement and reduction day concentration changes partially cancel each other out, so the “typical” change in each regime (likely a larger percentage change) is masked.

(Page 12, lines 368-370) I’m glad that the authors differentiate between the Galveston Bay Breeze and the Gulf of Mexico Sea Breeze front here, which will be familiar to TRACER participants, however, as stated in an earlier comment, it would be good to mention some of these particulars earlier in the manuscript when describing the sites and the local meteorology. Moreover, I also suggest discussing how the maritime air masses behind each front may be different (Sharma et al. 2024 describe some meteorological differences), and also atypical for maritime airmasses. For instance, the recently published paper by Thompson et al. 2025 https://doi.org/10.1029/2025JD043353 shows that the “maritime” airmass near Galveston, TX is often heavily polluted compared to the typical maritime airmass, such that a bay breeze passage transitioning to a maritime airmass heavily modified by industry and shipping along Galveston Bay might not be that surprising if it often enhances the aerosol concentration rather than decreasing them.

(Page 10, last paragraph, Page 13, first paragraph). A bit more description of how NPFs are identified from the data is needed here. For instance, it’s not obvious to me from Figure S4 why this event is clearly a NPF event rather than simply representative of aerosol increases through transport/advection. Also, on Page 10, the authors state there are more NPF events during the period than stated on Page 13, where only 11 events are stated to have occurred. Please correct/clarify.   There is also some unclear wording in the NPF paragraph on Page 13. The authors state there were 11 NPF events on SB days, only 5 of which showed changes in the NPF characteristics during the SBF passage. However, later in this paragraph the authors state “the cleaner air mass trailing the SBF passage led to a sharp reduction in the aerosol number concentration” as a general statement without clarifying how often this occurred, which implies to me that the SBF always leads to a reduction in aerosol number after passage on a NPF day… but this is at odds with the beginning of the paragraph. Please clarify the text with more specificity.

(Page 13, lines 388-392) This discussion seems to be at odds with earlier analysis. Here it’s stated that “this transition” which the earlier sentence implies is an overall wind shift towards faster winds and a more SE direction after the SBF is “consistently accompanied by a reduction in aerosol concentrations” however, earlier in the paper the authors state that the M1 site had 16 events with a reduction and 13 events with an enhancement in aerosols, which is anything but a consistent reduction! Please clarify or adjust the text to be more consistent with earlier analysis.

(Page 13, line 394) The authors state that the S3 site responds “similarly” to the M1 site, but I don’t really see the similarity at all. The M1 site more often (though not consistently) has a decrease in aerosol concentration after the SBF, but the S3 site more often has an increase, and the average change is positive at S3 but negative at M1 (as discussed on Page 12). This analysis (showing the aerosol concentration with wind direction and the shift in winds before/after SBF passage) seems designed to explain why the S3 and M1 are different, not similar, yet there is no detalied discussion/analysis of these differences. I think this analysis/discussion could be improved with a more specific analysis of pre/post SBF wind direction and aerosol concentration at each site on days when there is an enhancement vs. reduction. This would seem to be more valuable/important than the current decomposition by month shown in the supplemental figure.

(Lines 400-410) Much of this text seems more relevant to the introduction section since the authors have not conducted any analysis to show that any particular aerosol enhancement days are due specifically to any of these phenomenon.

(Page 14, Lines 422-425) I think this statement is generally true, but this and the following analysis neglect some interesting differences in the size distribution changes at the M1 site that contradict this statement. For instance, at smaller sizes, the number concentration stays the same at the SBF passage time and even increases a few hours after the SBF passage.   This might suggest a shift towards different aerosol species (for instance, the composition time series shows that the reduction at M1 seems mostly in the sulfate and ammonium categories) or sources rather than simply a transition to a “cleaner” maritime airmass. Indeed, the much greater aerosol concentrations after the SBF at M1 compared to S3 suggest that the maritime air mass near Galveston Bay and Houston is still quite dirty compared to the more pristine maritime airmass that comes straight from the Gulf of Mexico. The manuscript would benefit if such subtleties were pointed out and described rather than glossed over, as is currently the case.

(Page 15, line 439) Please provide a more physical explanation with references where relevant for how higher windspeeds would “dilute” the existing airmass in terms of the production of turbulent mixing that could do the “diluting”. Do you mean that mechanical production of turbulence increases mixing with greater wind speed? If so, does such increased mechanical mixing compensate for reduce boundary layer buoyant production of turbulence with a more stable surface layer?

(Page 16, Lines 473-475) I don’t see where this statement is evident in Figure 8. In fact, it looks from the panels in Figure 8f that the normalized change in aerosol concentration is actually stronger at 00Z than earlier times. Please correct or clarify with a figure citation where this effect is demonstrated.

(Page 16, middle paragraph and Figure 9). In this section comparing the July 17 and August 16 cases to July 10, the authors demonstrate that the change in aerosols following the SBF passage at each site is highly sensitive to the initial aerosol distribution, which is intuitive. However, they neglect to show/discuss any differences in the meteorological properties of the SBC in this case. Another issue with this figure/discussion is that the times plotted are much earlier than in Figure 8 for the July 10 case. While the supplemental figures show that the SBF passes the ARM sites earlier on July 17 and August 16, the times plotted for the July 17 and August 16 case don’t even include when the SBF passed the S3 site. For instance supplemental figure S7 shows this was 22Z on July 17 and figure S8 shows this was about 21Z on August 16—both of which are beyond the times shown in Figure 9. I think this comparison would benefit greatly from expanding Figure 9 into two figures that show exactly what is shown in Figure 8 (including meteorological variables and the times before SBF passes both sites, after it passes M1 only, and after passing both sites). This will allow for an apples-to-apples discussion of both the meteorological differences with the SBF on all three days (which may also contribute to the observed shifts in aerosols) as well as the current focus on the antecedent aerosol distributions. In general, the main paper currently focuses on a detailed analysis and discussion of the July 10 case, then casually mentions the other days. I think the paper would benefit from a more direct and comprehensive comparison of July 10 with July 17 to better get at the root of why some SBF passages lead to enhancement vs. reduction of aerosols.

(Page 16, lines 478-481). While this is evident from the figure and somewhat intuitive that it must be true for days to have an aerosol enhancement to have more antecedent aerosols in the maritime airmass, a more interesting question that could be addressed here is why are the antecedent aerosol conditions so different? Since your WRF chem simulations span the entire IOP, can you comment on differences in the synoptic or regional meteorology in the days leading up to these events that would cause the aerosol distributions to be so different prior to SBF passage?

(Page 16, lines 493-500). The reference to the scanning radar wind data is somewhat out of the blue and not clear how it relates to the prior analysis. Do these results refer to a specific day/event or are they more generally true? To this point your analysis does not clearly differentiate between bay breeze vs. gulf-breeze passage, so it’s not clear which results the scanning radar wind data support. Please clarify which findings/figures from the WRF simulations or site analyses are supported by the radar data.

(Page 17, Lines 511-518, Figure b,c) It’s interesting to me that the magnitude of the changes in nu0 and ac0 tend to increase with altitude but the changes in PM2.5 do not. Do the authors have any explanation as to why this is the case? It’s important to address this inconsistency because if the largest changes in nu0 and ac0 are occurring above the layer where the SBC is most prominent, it suggests there is another process causing changes in the aerosol at this time rather than just the SBC, and that not all the changes occurring here are due to the SBC.

(Section 3.7.1 and Figure 11) I find this section and figure 11 to be somewhat redundant and to be less than persuasive as a discussion of changes in aerosol indirect effects (as implied by its inclusion in Section 3.7). First, using N100 as a proxy for CCN is a rough approximation. Secondly, this information is available already in Figure 7. Finally, there are a lot of assumptions required for the change in N100 to translate to a change in the aerosol indirect effect, namely that these aerosol actually form clouds on either side of the SBF and that the change in CCN meaningfully affects the cloud drop size distribution, which are not analyzed here. Thus, I suggest simply removing this section and then when discussing Figure 7 in Section 3.4, you could add a sentence or two that specifically describes the changes in N100 and the implications for CCN.

(Page 18, lines 551-554) The details of these simulations need to be included either here or in the methods section when they differ from those of the other WRF-Chem simulations. How exactly is aerosol chemistry excluded vs. included? Are aerosols simply set to zero everywhere in the simulation NA simulation and the WA simulation is the realistic WRF-chem simulation as described earlier? What physical processes may be affected by this choice? How are CCN prescribed in the microphysics parameterization of the simulation without aerosol chemistry? Were these simulations conducted for the full IOP or only on particular case study days?

(Figure 12 and related discussion) In the second column of Figure 12b, it shows the “after” SBF time for the M1 site as 18Z, however supplemental figure S7 shows the SBF time at M1 on this date is after 18Z. How can this time represent the post SBF time for that day? Please correct or explain the discrepancy in the analysis.

(Section 3.7.2) While this is an interesting discussion of the ARF as it relates to SBC associated aerosol changes, A few additional pieces of analysis could improve this section and add more depth to the analysis. Have you compared the simulations to observations of surface radiative forcing from the M1 site?   While the total ARF cannot be directly compared, a comparison of the simulated surface radiative flux to the observations would be enlightening and add confidence to the model simulations if they agree. Second, an implication of increased ARF is that it can lead to warming of the atmosphere, which would have effects on clouds/convection and other meteorological aspects. Do the simulated changes in ARF correspond to significant changes in the simulated vertical temperature profiles?

(Summary and Conclusions in general) As discussed in earlier comments, some of the summary statistics here for each site that average over both reduction and enhancement days mask some of the nuance/detail. I suggest adjusting this text to reflect any corresponding changes made earlier in the manuscript in response to earlier comments.

(Page 21, line 638) Along the lines of an earlier comment, this paper has only shown the ARF, but it has not demonstrated corresponding heating/cooling in response, which might be implied by this statement in the conclusions. This sentence should be removed if such heating/cooling is not demonstrated in the simulations.

(Page 21, lines 639-641) If, following an earlier comment, the CCN section is removed, this text should be changed accordingly. At the very least, it needs to be stated here that “N100, a proxy for CCN” is used rather than actual CCN measurements.

(Conclusions, general) If my earlier suggestion to discuss existing complementary literature from the TRACER field project is adopted, please include some high-level comparison/contrasts to that literature in the conclusions section.

Technical Comments
(abstract, lines 27-28) I suggest moving the sentence starting with “SAIs modify cloud condensation nuclei…” to before the previous sentence since it is currently sandwiched in between two sentences that discuss the radiative impacts.

(Page 11, line 318) “resulting to” should be “resulting from”

(Figure S2 and S3) Please make it clear in the captions to these figures why some cases are included in the left column and others are in the right column. Presumably these are dates when the aerosol concentration increased (left) or decreased (right) after SBF passage?

(Page 12, line 343) “continental sites” is unclear here since I don’t think this phrase has been used/defined previously. Does this refer to both S3 and M1?

(Page 13, line 377) remove “that” before “occurred”

(Page 14, lines 425-427) This sentence is repeated from text at the beginning of this paragraph and can be removed.

(Page 16, line 485) Insert “an” before “air mass”

(Figure 6a label) The x-axis label here states the time is UTC, but it appears to be the SBF-relative time. Please correct.

(Page 18, line 556) “ARM” should be “ARF”

(Page 19, line 574) I think “preceding” here should be “following” since the air mass behind the SBF (or after it passed) had less aerosol on this date.

(Page 20, line 598) add “is” before “therefore warranted.”

(Page 20, lines 616-617; 618) I suggest replacing with “land-ocean interface” with “Galveston Bay” since M1 isn’t actually closer to the Gulf of Mexico than S3. Likewise replace “downstream” in reference to S3 with a more accurate modifier like “rural” or “farther removed from urban and industrial influences”
Citation: https://doi.org/10.5194/egusphere-2025-2659-RC1
- AC1: 'Reply on RC1', Tamanna Subba, 20 Sep 2025
  
  Response is attached.
  
  Citation: https://doi.org/10.5194/egusphere-2025-2659-AC1
RC2:
'Comment on egusphere-2025-2659', Anonymous Referee #2, 08 Aug 2025
General Comments

This manuscript provides an observational overview of aerosol changes due to sea breeze events, focused at two DOE TRACER sites in Southeastern Texas during the TRACER IOP that took place June 1 - September 30 2022, as well as a modeling component for individual cases to investigate spatial trends. The paper provides a nice overview of observed aerosol evolution during previously identified sea breeze events, with three identified sea breeze changes to the aerosol characteristics. While the authors provide good descriptions of different case study examples, the depth of analyses used throughout are shallow in many areas, leaving the reader with many unanswered questions and/or unclear connections being made between the observations and the explanations for what’s causing observed changes. Houston has a history of air quality/aerosol focused studies that could be beneficial for the authors to tie their results to past works, which could also help clear up areas where things are currently unclear.

While I believe the scientific goals and aims of this paper are sound, I think a major reorganization is necessary and I could see this as being two separate papers. This would allow the authors to expand the observational analyses and go into more depth while a second paper focused on the modeling outcomes could stand as its own paper. Additionally, there are several places where supplemental figures are being referenced and used more like core analyses. In general, supplemental materials should include things that are not integral to the results, but add confidence that the results shared are robust. By breaking into multiple papers, more room to include the currently supplemental figures could be made.

There are many places throughout where the same statements are being made (e.g. SBC influence extends 50km inland) that could be reduced or slightly modified so they’re not so repetitive. Some of these occur in transition areas between paragraphs or sections, suggest revising these transitions.

Specific Comments

Abstract, line 23: Technically the two site’s proximity to the sea are very similar, but their proximity to the bay differ.

Page 2, line 40: ‘These impacts…’. What impacts are you referring to here? The global scale changes to energy balance?

Page 2, lines 47-49: As written this sounds like the only complicating factor is the sea breeze, suggest a rephrase so it isn’t so definitive. Things like urban characteristics can also affect the aerosol processes.

Page 3, line 65: I would change ‘stable’ and ‘unstable’ to ‘more moist’ and ‘drier’ respectively. The stability, as written, is overly generalized here and there’s recent works (Boyer et al., 2025 https://doi.org/10.1175/JAS-D-23-0180.1) showing how the cooler side of an airmass can actually be more ‘unstable’.

Page 3, lines 80-85: First sentence discusses trapped aerosols, but then the next sentence discusses mixing out aerosols. Suggest a rephrase to make things clearer here. Also, ‘The competition between converging winds’ is an awkward phrasing, ‘The often opposing winds associated..’ reads cleaner and avoids personification.

Page 4: lines 99-100: I’m not sure what you mean by ‘aerosol property implications’. I believe this is an example where some rephrasing could help connect this sentence to the following two where you give direct examples.

Page 4, lines 114-118: It’s stated that this work expands on Li et al. 2020 work, however it isn’t very clear how this expands on it, since the Li work is focused on Ozone. A better connection here is needed.

Page 4, lines 123-: The main site is labeled here as ‘urban coastal’ and the second site as ‘rural’. While the M1 site is close to the Bay, I would either add an explanation of this classification or use a different word for it, as both sites are almost equidistant from the Gulf coast (which many people would consider coastal). The evolution of the bay breeze and sea breeze are also different, with the M1 site experiencing a mix of both breezes throughout TRACER. This is a place where further clarification of the general meteorology over the region could be useful for the reader. Also, be consistent with your naming convention throughout. Sometimes the M1 site is ‘coastal’ other times it’s ‘urban coastal’, which can be confusing.

Page 5, line 145: The M1 site is southeast of Houston, not south. This is an important distinction since other TRACER data was collected in Pearland, which is south and later discussions include aerosol sources in proximity to the city.

Page 5, line 151: You’re classifying the S3 site as ‘rural’, but then say it is ‘periphery’ to highly populated and commercial sectors, which is contradictory to a true rural definition.

Page 5-6, lines 155-159: While this sentence is important, it feels out of place here. Would fit better within the introduction when discussing TRACER.

Section 2.2: This section needs some further expansion/added detail. As it stands it is very surface level. How is missing data handled? What is the resolution of the different datasets and/or why is a 5 minute interval used? For composites, how are these being made? Is there data quality control or processing being applied? TCEQ has many sites, why is only one chosen, why only PM2.5, and how close is the actual site to the M1 site? I’d suggest adding the TCEQ site to the figures that you label the TRACER sites on.

Section 2.3, lines 188-201: Additional explanation for choices made are needed here. Why are simulations only conducted for part of TRACER? How are you determining that the grid spacing and configuration is ‘assumed sufficient’? Is there literature you can point to? Is the simulation period one long simulation that began on 1 July? Or is it daily, individual simulations? This isn’t clear and could have significant impacts to the results.

Page 7, lines 202-210: This information feels better suited to the site description section. While it’s using land use/cover information by the model it’s actually describing the site itself, not the model set up. Suggest either rephrasing or moving. Also, lines 204-206 don’t make sense as written, suggest a rephrase to better describe the differences of the land cover at each site.

Section 2.3, lines 211-229: This paragraph is not describing the model setup and feels out of place. There are also a lot of initial results here. Suggest moving to a different section that would focus on model verification/results.

Page 7, line 214: What is ‘feasible’? It isn’t clear what guidelines are being used and feasible is subjective. Looking at S1, the model doesn’t appear to capture the amplitude of the diurnal cycle well and over estimates wind speed, which are both important when considering the meteorology and sea breeze characteristics. Discussing where the model is and isn’t doing well would add context and benefit the paper.

Fig. S1: figure caption needs expanded to include the stats that are within the figures (assuming obs vs simulation/MERRA). No mention of MERRA dataset takes place anywhere, it’s just shown in the figure without an explanation of what it is or why it’s used.

Page 8, lines 221-228: Suggest a rewrite/reconnection between what is shared in the previous sentences and how you connect them with literature. What discrepancies are you referring to, how does the cited literature (i.e. dust storm) compare with what you’re trying to simulate here? The connection feels disjointed.

Page 8, lines 238-240: Cite your figure. Are your wind speed measurements able to be precise to the tenth of a m/s? How are you calculating average wind direction? I’m suspicious of if this is accurate, especially in the overnight periods when offshore flow was common. Ensure you’re breaking the wind direction into u and v components to find the average, because a simple degree average can lead to southeasterly winds a lot of the time, particularly for northerly winds (e.g. (358 + 2)/2 = 180). Personally, I think the first paragraph of 3.1 could be removed in its entirety.

Page 8, line 248: Suggest using water vapor mixing ratio (or another absolute moisture measurement) rather than RH, since RH depends on temperature.

Page 8, line 249: ‘These values are comparable’. What values? RH? Temperature? Try to be clearer when making these statements. Also, what do you mean comparable? Within a certain range? This is vague wording.

Page 9, lines 253-255: How would the differences in proximity to water and land cover account for the meteorological variations? Which met. Variations? Can you tie this to past findings?

Page 9, lines 261-269: This is a big change in focus from the previous paragraphs in this section, I would consider moving this to the data/methods section and providing significantly more information. While the focus of this paper is on aerosols, this paragraph does not give enough information on the sea breeze circulations themselves, which is the other key component to your work. Things like ‘Most of these cases are under the control of anticyclonic systems’ does not provide the necessary background information of how this applies to the two sites. The sea breeze timing needs more information, as diurnal trends in aerosol lifecycles could play a part in your analyses, coupled with the sea breeze timing, which varied throughout the campaign. It also isn’t clear if your sea breeze events for M1 are only sea breeze or bay breeze events. There is no discussion on how many sea breeze events there are (although this is included in the conclusions and elsewhere later). Discussion on how Wang determines the timing would be beneficial, as wind direction and moisture changes don’t always happen simultaneously with the passage of a sea breeze.

Page 9, line 277: Is this supposed to be shown in a figure?

Page 10, lines 286-288: Is the aerosol bulk mass concentration supposed to be shown somewhere? S3 is now being described as within a ‘marine coastal environment’ which is what you’ve previously described the M1 site as. This is confusing to the reader, as S3 has always been called rural prior to this instance.

Page 10, lines 290-294: Percentages don’t match what is shown in Fig. 5 (assuming this is what you’re referring to). Please reference the figure where this would be shown and address the discrepancy.

Page 10, line 295: This statement does not match the math stated above or Fig. 5. Please address this discrepancy. Also suggest moving this sentence before the previous one so that you’re presenting your findings, giving the average change, then connecting to literature rather than back and forth between findings.

Page 10, lines 297-310: Prior to now NPF is not mentioned, which is jarring. Suggest adding some information about NPF (what they are, how they’re determined, etc) in the introduction.

Page 11, lines 334-337: This sentence is valuable, but feels a bit confusing where it’s at, suggest moving up or expanding the point further for clarity.

Page 11, line 340: Why this time (i.e. 5 mins before)? Is this an average over the 5 minutes before SBF passage or an instantaneous value? You refer back to Sect. 3.1, but this information isn’t there either.

Page 12, line 349: What do you mean by ‘clean conceptual model’? I think you just mean a clear trend, but the wording is awkward.

Page 12, line 351: Word choice for ‘disappears’ is not very scientific. Please change. Also, is there any QC taking place to ensure the spikes are legitimate and not erroneous? This goes back to comment 12 above.

Page 12, lines 354-358: Why is 1 hour chosen? Is this purely subjective based off the observed data or is there any objective reasoning for this timeframe choice?

Same area: Some discussion regarding diurnal cycle trends in aerosols would be beneficial. T +/- an hour centered on the sea breeze, without any consideration in the sea breeze timing, is an important piece of the puzzle that should be addressed somewhere within the results.

Page 12, line 357: How many days were considered neutral that are not being accounted for in the analyses?

Page 12, lines 359-360: This sentence is kind of hard to follow as written. What do you mean by ‘more frequent’ changes? Is this increasing or decreasing aerosol concentration? What about the other 40 or 66% of cases at each site? I would suggest being more specific here by including case numbers (e.g. 20 sea breeze cases at the M1 site show increasing aerosols while only 10 cases at S3 show…). By the time I’ve gotten to this point in the paper I don’t remember how many sea breeze cases there are at each site.

Page 12, lines 360-362: Are these averages considering all sea breeze cases, regardless of their classification (enhancement, reduction, neutral)? Just the cases where a change occurs? It isn’t clear. Maybe think of a clearer way to show these results between sites.

Page 12, lines 363-366: Suggest moving this information up, after the first (revised) sentence.

Page 12, line 366: Which discrepancies? The difference in classification or the changes in aerosol number concentration? Be clear here.

Page 12, line 369-374: This is a good start to connecting the results/descriptions of measurements to the ‘why’ things may be happening. I would suggest expanding on this, while trying to tie in some literature to support your reasoning. Discussing the airmass differences (sea vs bay) and possibly pulling in some trajectory analyses could really strengthen this important piece of the science and story.

Page 13, first paragraph: Switched back to NPFs, feels disjointed. Suggest a reworking of this section for flow purposes. Also, are the 11 NPF SB events observed at both sites?

Page 13, line 375: Did the NPF events occur during SB events or on SB days? Later in the paragraph it’s described as occurring before the SBF passes, so it’s a little unclear what is meant here.

Page 13, line 383: Is the reduction purely ‘cleaner’ air masses being advected in or could other factors be at play? How are you sure you can definitively state that it is the SB frontal passage causing the change in this case?

Page 13, lines 387-388: The Houston urban core is northwest of M1, not north and east. There are heavy industry areas directly north and the shipping emissions from the bay to the east, but the Houston urban center isn’t in the directions listed and doesn’t make sense with the observations.

Page 13, line 391: Terminology used is making following the findings hard. Here you say ‘This transition is consistently…’ but earlier you share that things are very inconsistent.

Page 13, lines 393-394: What do you mean when you say ‘However, concentrations are observed to be higher on days associated with a higher aerosol loaded marine air mass.’? It isn’t clear. Next sentence also isn’t very clear. Does S3 respond similarly when higher aerosol loads are present or respond similarly in any scenario?

Page 13, lines 399-401: Can you make the tie in with this literature a little clearer? How does O3 and NOx compare to the total number concentration that you’re looking at?

Page 13, line 403 and beyond: What does it mean for an ‘episodic transboundary aerosol transport’ to take place? Are these scenarios important to your results (i.e. were there dust events taking place during TRACER). This and biomass burning should be something you can verify for the time period and would be important context with your results, rather than just an input of what could be a factor.

Page 14, lines 410-411: This is the first real mention of the Bermuda-Azores High. It feels tacked on and shallow, but a better meteorological discussion earlier on could help to tie this in better.

Page 14, line 433: Be more specific here, east of M1 is the bay, which is at least partially a marine environment, though not the same as what you’d have from the Gulf’s environment.

Page 15, line 338: How would higher wind speed lead to dilution? Further explanation of this theory for the observed results is needed.

Fig. 7: You show the bulk chemical compositions, but don’t really tie this into your discussion of pre/post sea breeze changes. I would suggest either removing these subplots from the figure or adding discussion about the changes. As it stands, it isn’t really adding anything to the findings.

Page 16, lines 469-471: It isn’t clear that there is an increase in particle concentrations ahead of the SBF. Suggest adjusting the colormap and range to make this point clearer to see.

Page 16, line 474: Is the SBF influence diminishing as it moves inland? The normalized change shows more changes as time goes on. Please clarify this.

Page 16, lines 576-489: When did the SBF occur for these cases? Why are no sea breeze characteristics discussed, like they were in the previous paragraph? This is important context for understanding the results. Also, for the July case, how representative is this case (e.g. higher aerosol load over the sea) with other TRACER cases? Is this an outlier or common? How would that influence your overall findings? If these cases are to be discussed and kept in the paper, they need to be expanded and include a similar discussion as to what is shown in Fig. 8. The sea breeze in this region can be quite variable, so simply stating there was one, is not enough to justify what you’re trying to find.

Page 17, line 500: Make the connections to why the concentration would increase in the subsequent 5hr period over M1.

Page 17, first paragraph in 3.6: Prior to this analysis, you’ve focused on either PM2.5 or integrated aerosol number concentration, but now you’re breaking that into Nu0 and ac0. Please clarify why you’re choosing to do so and how these two evolutions (horizontally and vertically) matter in the bigger picture. It isn’t clear to me what the point in this change is for. You’re also missing the ‘why’ here, it’s purely descriptive without connecting why it matters to the science.

Page 17, lines 520-521: I’m not seeing the connection between the literature and the patterns here. What exactly are you trying to tie together by using an example with dust and mountains to compare with SB changes? Please clarify or find a more fitting example to point to.

Page 18, line 535: Include a citation for why N100 is being used (if applicable).

Section 3.7: I feel like this section doesn’t fit with the rest of the paper. A different modeling approach is being used, without being well described. While I think that this work is important, it feels like it’s a big shift from 90% of the rest of the paper. I would think critically on how/if this piece should be within the paper. As I suggested earlier, I think you could break this manuscript into multiple paper, in which this would fit into a second more modeling focused one.

Page 20, line 604: This is the first time the actual sea breeze numbers for each site are given. This information, while still important here, should be given way earlier as well.

Page 21, line 621: Using the average changes defeats the purpose of having three different types of responses to the sea breeze and undersells your work. I would suggest using statistics here that highlight the changes for each of the types of SAI.

Conclusions: I’ve left many comments regarding some changes or things to possibly remove. Obviously, if you are to make those changes than a lot of the conclusions would be reworked. Overall, I think the conclusions are hitting the main points well, but I would try to further highlight the importance of the work here too, not just a recap of the findings.

Technical Comments

Page 2, lines 38-40: Suggest a rephrasing. The sentence starts with Houston and regional health effects, but shifts to global scale energy balance, which reads awkwardly.

Page 2, line 40: Personification is used here with the word choice of ‘felt’. Suggest a change to ‘exhibited’ or something of that nature.

Page 2, line 53: Add commas around ‘such as SBCs’.

Page 2-3, lines 55-73: This paragraph could benefit from a reorganization for readability.

Page 3, line 79: Add citations to studies here.

Page 3, lines 92-93: Change ‘with increase in the concentration of the smaller particles durig passage’ to ‘with increases in the concentration of smaller particles during the passage’

Page 4, line 119: remove ‘recent’

Page 5, lines 126-127: replace ‘we will use targeted… modeling’ with ‘the… model’ and add ‘is used’ after (WRF-Chem).

Page 5, line 131: replace ‘our’ with ‘the’ – there are several other instances of ‘our’ throughout that could be changed to ‘the’ or another word/phrase to be more formal.

Page 5, line 144: Replace ‘and its’ with ‘at the’

Page 5, line 150: Replace ‘this TRACER’ with ‘the TRACER’, remove ‘also’, and change ‘This S3…’ with ‘The S3…’.

Page 6, line 179: Replace ‘those observations…’ with ‘the TRACER observations…’ and remove ‘we used the’.

Page 6, line 181: Add ‘are used’ after ‘(TCEQ) database’.

Page 7, line 188: Remove ‘We use’ and add ‘model is used’ after citations.

Page 7, lines 198-199: Geogrid sentence isn’t necessary, can remove this.

Page 7, line 211: Add ‘spatial’ after ‘physically-reasonable’. The observations still give physically reasonable depictions of the environment, the model adds that spatial component.

Pages 7, lines 216-219: Suggest rewriting this sentence, it’s difficult to read and comprehend as is with so many numbers in the parentheses.

Page 9, line 250: Change 1 to 1-2

Page 9, lines 277-279: Suggest a rewrite of this sentence. You don’t need to reference Fig. 5a at the end, if you do in the beginning and it could be shorter.

Page 10, lines 287: Change ‘These values did not differ from’ to ‘The values are very similar to’, since they are technically different.

Page 10, line 289: Add ‘observations suggest’ after ACSM

Page 10, lines 300-301: Sentence phrasing is awkward. Suggest a reword to something like ‘During summertime, NPF events were identified at both the M1 and S3 sites, consisting of 22 and 18 events respectively.’.

Page 10, line 309: I think you mean SBC, not SAI, since SAI are not purely meteorological.

Page 11, line 319: Remove ‘formation of’

Page 11, lines 321-322: Suggest removing ‘For any inland-penetrating SBF, the authors find it instructive to define’ and just start the sentence with ‘Three scenarios for the influence of an inland…’ and ending it with adding ‘are possible’.

Page 11, line 325: Replace ‘there may be’ with ‘is’

Page 11, line 339: Remove ‘we’ and ‘the’ so it reads ‘normalized aerosol concentration…’

Page 13, line 375: Add ‘events’ after NPF.

Page 13, line 376: Combine sentences. ‘During the passing of the SBF, like the NPF event that occurred…’.

Polar plot figures: The chosen colorbar for these plots can be very misleading, especially in S5. With a white centered colorbar, increasing concentrations close to the white values are hard to discern and we could be missing datapoints. Suggest reproducing with a sequential colormap. Also, please increase text sizes for readability.

Page 13, line 390: Suggest removing ‘and intermediate directions’.

Page 13, line 397: Change ‘These TRACER’ to ‘The TRACER’.

Page 14, line 412: Suggest changing the section name, since this is now looking at examples with modeling results as well.

Page 14, line 425: Remove repetition of first sentence.

Page 14, line 434: Change ‘suggests’ to ‘indicates’

Page 14, line 435: ‘influence’ is a weird word choice for this statement.

Fig. 8: Suggest reversing the water vapor mixing ratio colorbar so that more moist values are blue/green and drier values are brown

Figures in general: Rainbow colorbars can be hard for colorblind folks to interpret. Where possible, I would suggest using different colorbars/maps so that folks who may be colorblind can see the results in the way you’re intending and they can be more accessible. Many of the captions need to be expanded on and revised. The captions should basically include all the information you need to know and point out each thing in a plot.

Page 16, line 467: Remove ‘also’

Fig. S9: Add a small subplot to indicate the location of the cross section.

Page 17, line 520: ‘resembles that observed in SB simulations’ is a confusing statement here, when you’re using both observations and simulations within this paper. Suggest rephrasing to something like ‘resembles that shown in SB simulations…’.

Page 18, line 552: Add parentheses around ARF.

Page 18, line 556: Change ARM to ARF.

Page 21, line 623: Replace ‘enhancements’ with ‘increases’.
Citation: https://doi.org/10.5194/egusphere-2025-2659-RC2
- AC2: 'Reply on RC2', Tamanna Subba, 20 Sep 2025
  
  Response is attached.
  
  Citation: https://doi.org/10.5194/egusphere-2025-2659-AC2

Tamanna Subba, Michael P. Jensen, Min Deng, Scott E. Giangrande, Mark C. Harvey, Ashish Singh, Die Wang, Maria Zawadowicz, and Chongai Kuang

Supplement

https://doi.org/10.5194/egusphere-2025-2659-supplement

Tamanna Subba, Michael P. Jensen, Min Deng, Scott E. Giangrande, Mark C. Harvey, Ashish Singh, Die Wang, Maria Zawadowicz, and Chongai Kuang

Viewed

Total article views: 914 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
815	82	17	914	36	16	24

HTML: 815
PDF: 82
XML: 17
Total: 914
Supplement: 36
BibTeX: 16
EndNote: 24

Views and downloads (calculated since 27 Jun 2025)

Month	HTML	PDF	XML	Total
Jun 2025	52	5	1	58
Jul 2025	154	40	10	204
Aug 2025	151	15	0	166
Sep 2025	423	14	5	442
Oct 2025	35	8	1	44

Cumulative views and downloads (calculated since 27 Jun 2025)

Month	HTML	PDF	XML	Total
Jun 2025	52	5	1	58
Jul 2025	154	40	10	204
Aug 2025	151	15	0	166
Sep 2025	423	14	5	442
Oct 2025	35	8	1	44

Viewed (geographical distribution)

Total article views: 910 (including HTML, PDF, and XML) Thereof 910 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 28 Oct 2025

Short summary

This study highlights how sea breeze circulations influence aerosol concentrations and radiative effects in Southern Texas region. Using TRacking Aerosol Convection Interactions Experiment field campaign observations and model simulations, we show that sea breeze–aerosol interactions significantly impact cloud-relevant aerosols and regional air quality. These findings improve understanding of mesoscale controls on aerosols in complex coastal urban environments.


Total:	0
HTML:	0
PDF:	0
XML:	0