the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Mar 2025
The sources and diurnal variations of submicron aerosols in a coastal-rural environment near Houston, US
Abstract. Aerosol properties were characterized at a rural site southwest of Houston from May to September 2022 during the intensive operation periods (IOP) of the Tracking Aerosol Convection Interactions ExpeRiment (TRACER). Backward trajectory analysis reveals three major air mass types, including marine air mass from the Gulf, urban air mass influenced by urban emissions, and regional air mass. Marine aerosols typically show a bimodal size distribution and have the lowest particle number and mass concentrations of PM1 (particulate matter with an aerodynamic diameter of less than 1 μm), while the aerosols from air masses strongly influenced by urban emissions exhibit the highest concentrations. Organic aerosol (OA) accounts for more than 50 % of PM1 for urban and regional air masses, whereas sulfate is comparable to OA in marine air masses. Positive Matrix Factorization (PMF) analysis of aerosol mass spectra identifies 6 OA factors, including hydrocarbon-like OA (HOA), OA from the oxidation of monoterpenes (91FAC), OA from the reactive uptake of isoprene epoxydiols by acidic sulfate particles (isoprene-SOA), and three oxygenated OA factors with high O:C ratios (OOA1, 2, and 3). We find OOA1, a factor with a high f55 signal and f55/f57 ratio, is related to shipping emissions, instead of cooking emissions suggested in previous studies. OOA3 has the highest O:C ratio and exhibits elevated mass concentration in the afternoon. Similar diurnal variation of highly oxidized OA factors was commonly observed in the Houston area during previous studies and attributed to the SOA formation by photochemistry and mixing from aloft. Here, using air mass backward trajectories and 1-D box model, we show the diurnal trend of OOA3 mass concentration is instead driven by changes in air mass arriving at the rural site. The air mass changes are likely caused by the shift between land breezes and sea/bay breezes. Within the same air mass type (e.g., either urban or marine air mass), OOA3 mass concentration is largely independent of wind direction and shows essentially no diurnal variation, suggesting OOA3 is related to aged OA with minimal influence by local emissions. This study helps identify the major sources of OA in the Houston region and highlights the impacts of both atmospheric chemistry and meteorology on aerosol properties in the coastal-rural environment.
- Preprint
(2793 KB) - Metadata XML
-
Supplement
(1747 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-726', Anonymous Referee #1, 27 Apr 2025
Li et al., present a study of aerosol characteristics conducted at rural site southwest of Houston during the intensive operation periods (IOP) of the Tracking Aerosol Convection Interactions ExpeRiment. They report significant differences in aerosol properties among marine, urban, and regional air masses, and identified six OA factors using PMF analysis. The results emphasize the roles of both atmospheric chemistry and meteorological conditions in this coastal-rural site. Overall, the manuscript is well-written and fits within the scope of the journal. However, the following comments need to be addressed before publication.
Line 185: Although the CWT map of ammonium is consistent with sulfate, the sources of ammonium could not from marine emissions. The authors should provide further discussion or references regarding potential terrestrial or anthropogenic sources of ammonium. Additionally, the organic nitrate mentioned also needs stronger supporting evidence.
Section 3.2: The time series of all six PMF-resolved OA factors display high similarity, particularly during urban air mass periods. This raises questions about the robustness of the factor separation. For example, how to explain the co-variation of primary and secondary factors during the air mass from urban areas. In addition, how is the six-solution result determined? Providing diagnostics for this six-solution is beneficial for reader to understand.
Line 215-225 paragraph: The attribution of 91FAC to monoterpene oxidation is reasonable based on spectral comparisons, but a few concerns remain: (1) the O/C ratio is higher in previous study using HR-ToF-AMS observation; (2) the name of 91FAC may cause confusion and a more descriptive term may improve clarity.
Section 3.2.2: OOA1 is characterized by strong signals from m/z 55 and 57, which are typically associated with primary OA. However, the authors classify it as a secondary factor and link it to ship emissions. This conclusion needs further justification, such as its high O/C ratio and diurnal variation consistent with OOA2 and OOA3.
OOA3: The decomposition of OOA3 by air masses is novel and provides useful insights. However, the explanation for the noon peak being due to longer land residence is somewhat inconsistent. Based on wind direction, southerly winds from the sea dominated during the daytime which contradicts the interpretation of increased urban influence. If the different diurnal variations between different air masses, it is generally not combined into one factor during the decomposition of PMF. In addition, the diurnal pattern of OOA3 within individual air mass is nearly flat, however the box model results show daytime production. Why is this happening?
Citation: https://doi.org/10.5194/egusphere-2025-726-RC1 -
RC2: 'Comment on egusphere-2025-726', Anonymous Referee #2, 26 May 2025
Li et al. investigated aerosol properties in the southwest of Houston, a region that can be categorized as a coastal-rural site. The authors differentiated air masses using back-trajectory analysis and integrated this with OA PMF analysis to thoroughly examine the contributing factors to the observed amount and chemical characteristics of OA in the Houston area. Urban air masses were identified as the dominant driver of elevated aerosol levels, while marine air masses introduced ship emission plumes that led to enhanced OOA1. Notably, the authors highlight that OOA3, which is less likely to be locally influenced and is primarily composed of aged OA, is attributed to changes in air mass. This is demonstrated through back-trajectory and 1-D box model analysis—providing a different perspective from previous studies that pointed to photochemical formation or mixing from aloft. By utilizing multiple analytical approaches, the authors logically support their key findings, and the manuscript is overall well written. Still, I have a few comments that should be addressed before publication.
Scientific Comments
- Line 166: Is the size at which the mode appears consistent with findings from previous studies?
- Did the observed linear correlation align with the results from the CWT analysis? I think a statistical evaluation should come first before moving into trajectory-based analysis.
- Line 214: Even with a strong tracer ion signal (m/z 91), 91FAC appears to have a substantially low f44, which makes it difficult to attribute this factor to SOA. Could you comment on this observation?
- Are there any potential isoprene sources besides the National Forest near the field site?
- Since sulfate is often associated with ship emissions, do you think that maritime air masses with high SO₄ could also contribute to OOA1, not just isoprene OA formation? Given that OOA1 seems to be heavily influenced by coastal ship emissions, I’m curious about your interpretation here.
- You noted that the high f55/f57 signal of OOA in this study is likely linked to ship emissions rather than cooking. Would there be any common chemical species between the two sources? I think additional evidence would help support the classification of this factor as an OOA rather than a primary-like OA.
- Line 237: Are you suggesting that these aerosols originate from different sources but become mixed near the site, subsequently forming isoprene OA?
- How did the size distribution vary during each OOA event, or when one OOA component was dominant over others?
- Figure 7D: Since you conducted a CWT analysis, didn’t that show an OOA1 hotspot over Freeport?
- The distinction between OOA2 and OOA3 would be more convincing if you could provide additional supporting evidence.
- I understand the intent of this analysis is to support the idea that OOA3 originates as a background SOA. If that’s the case, does the analysis yield a different interpretation for OOA1 and OOA2 if a similar approach is applied?
- Line 442: What about OOA2? I think contrasting air mass trajectory characteristics could further support distinguishing OOA2 from OOA3.
Technical Comments
Methods
- Have you deployed a cyclone for the size cut of sampling aerosol? Was ACSM equipped with a standard vaporizer or a capture vaporizer? Please clarify.
- Line 105: Cite Fröhlich et al., 2013.
- Line 111: Could you provide more details on the PMF analysis? Specifically, how was the solution selected, and how did the analysis results vary depending on the number of factors?
Figures & Organization
- Fig. 9 appears before Fig. 8 in the main text. Please reorganize the figures to appear in sequential order.
- Since Fig. 9 is based on box modeling and Fig. 8 uses backtrajectory analysis, I suggest not introducing Fig. 9 before Fig. 8. Instead, mention that Fig. 9 supports the analysis presented in Fig. 8.
- I think this explanatory section should come before the discussion of Figs. 8 and 9 to maintain the current flow. (I’d prefer it in the Method section, but as long as it appears before those figure discussions, I’ll leave that to the authors.)
- Line 247: Please also note that OOA separation typically has greater uncertainties compared to POAs (Zhang et al., 2011; Ng et al., 2011; Hass-Mitchell et al., 2024).
- Line 345: Are you referring to Fig. 9C and 9F? Please clarify which figure panels are being discussed.
Citation: https://doi.org/10.5194/egusphere-2025-726-RC2
Data sets
TRACER observational datasets Atmospheric Radiation Measurement User Facility https://www.arm.gov/data/
HYSPLIT data National Oceanic and Atmospheric Administration https://www.ready.noaa.gov/HYSPLIT.php
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 245 | 49 | 14 | 308 | 26 | 13 | 24 |
- HTML: 245
- PDF: 49
- XML: 14
- Total: 308
- Supplement: 26
- BibTeX: 13
- EndNote: 24
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1