| 11 Sep 2026
Reconciling online and offline source apportionment of brown carbon absorption in wintertime Seoul
Abstract. Brown carbon (BrC) is a leading uncertainty in aerosol radiative forcing, and the source-resolved mass absorption coefficients (MAC) used in climate models come from both real-time (online) and filter-based (offline) measurements. Both are widely trusted, yet have rarely been applied to the same aerosol and compared source by source, leaving the assumption that they are interchangeable untested. Here we ask whether online and offline source apportionment yield the same sources of BrC absorption for wintertime urban aerosol. We combined a seven-wavelength aethalometer with high-resolution aerosol mass spectrometry (online, whole particle) and 24 h PM2.5 filter UV–Vis of water extracts (offline), attributing bulk absorption to positive-matrix-factorization factors by regression on each platform. The two platforms agreed on the bulk attribution (winter BrC carried by secondary and biomass-burning OA) and on consistency with worldwide observations, but inverted the source ranking: online, biomass burning and less oxidized secondary organic aerosol dominated while more oxidized OA (MO-OOA) was negligible (MAC 0.07 m² g⁻¹); offline, MO-OOA dominated (0.88 m² g⁻¹). This reversal reflects how the sample is measured, not an atmospheric change: water-soluble selection and a likely shift of MO-OOA toward reduced-nitrogen chromophores during offline preparation, while bulk composition and factor time series are preserved. The source inferred for BrC absorption is therefore strongly dependent on method and integration time, a critical caution for source-resolved BrC parameterizations that treat online and offline datasets as interchangeable.
Competing interests: At least one of the (co-)authors(Hwajin Kim) is a member of the editorial board of Atmospheric Chemistry and Physics.
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.
Kim and Chen compare online and offline approaches for source-resolved brown carbon (BrC) absorption in wintertime Seoul. By combining aerosol mass spectrometry (AMS), positive matrix factorization (PMF), and optical measurements, the authors attempt to identify the organic aerosol (OA) factors associated with BrC absorption and, in particular, to understand why online and offline approaches yield substantially different source-resolved mass absorption coefficients (MACs). Direct comparison of online and offline BrC source apportionment for the same ambient aerosol is relatively scarce, and such a comparison could provide useful information for interpreting BrC measurements and their application in models.
However, I have fundamental concerns regarding the experimental basis of the central conclusion. In my view, these issues cannot be resolved through additional discussion or statistical analysis alone, because the optical and chemical measurements used for the offline source apportionment do not necessarily represent the same chemical state of the sample. Additional experimental validation would therefore be required before the principal conclusions of this manuscript can be supported.
Major comments
1. Chemical equivalence of the offline BrC and offline-AMS measurements
For the offline analysis, BrC absorption is measured directly from the aqueous filter extract using UV–Vis spectroscopy. The OA composition used for PMF and subsequent multilinear regression, however, is measured only after the water extract has been atomized, dried, and introduced into the AMS. These two measurements therefore correspond to different stages of sample processing.
This distinction is particularly important because the manuscript itself argues that offline preparation alters the nitrogen-containing composition of MO-OOA and invokes enrichment or formation of reduced-N species during sample preparation as an explanation for the high offline MO-OOA MAC. If chemical changes occur during atomization and/or drying, however, those newly formed or selectively enriched compounds cannot automatically be assumed to represent the chromophores responsible for the absorption that was measured previously in the original aqueous extract.
Drying cannot modify the BrC absorption that has already been measured in the liquid extract; it can only modify the aerosol composition subsequently measured by the AMS. Consequently, an association between reduced-N fragments in the dried nebulized particles and absorption measured in the untreated aqueous extract does not by itself demonstrate that those reduced-N compounds caused the measured BrC absorption.
This issue directly affects the MLR source apportionment. The manuscript reports an offline MO-OOA MAC that is approximately one order of magnitude larger than the corresponding online MO-OOA MAC and interprets this large difference as part of the online–offline source-attribution reversal. However, this interpretation relies on the assumption that the offline-AMS PMF factor remains chemically representative of the chromophores present when the aqueous absorption measurement was performed. At present, this assumption has not been demonstrated.
The strong correlation between the online and offline MO-OOA factor time series does not resolve this issue. A systematic preparation-induced transformation could preserve sample-to-sample covariance and therefore retain a high factor correlation while substantially altering the minor molecular components responsible for light absorption. BrC absorption can be controlled by a relatively small chromophoric fraction of OA, so similarity in bulk OA composition or PMF temporal behavior is not sufficient to establish preservation of optical-relevant composition.
The authors therefore need direct experimental validation of the effect of nebulization and drying on BrC. For example, an aliquot of the original water extract could be measured directly by UV–Vis, while another aliquot is subjected to the same atomization and drying procedure used for offline AMS, recollected either into liquid or onto a filter, and subsequently analyzed for absorption. Ideally, corresponding chemical measurements should also be compared before and after processing. Such an experiment would establish whether the offline preparation increases, decreases, or otherwise modifies BrC absorption and whether changes in reduced-N composition are actually associated with those optical changes.
Without such validation, the manuscript can demonstrate that online whole-particle measurements and offline water-extract measurements give different statistical source associations, but it cannot establish the proposed chemical mechanism responsible for that difference. Because this mechanistic interpretation and the resulting online–offline source-attribution reversal constitute the central message of the manuscript, I do not think the current experimental evidence is sufficient for publication in ACP.
2. Online–offline comparison conflates analytical method with solubility selection
There is an additional methodological distinction that needs to be treated much more explicitly. Online BrC represents absorption by the suspended particle population, whereas the offline UV–Vis measurement represents only chromophores recovered by water extraction. Poorly water-soluble or insoluble chromophores are therefore selectively omitted from the offline optical measurement. Such material may include less-oxidized aromatic and nitrogen-containing compounds that can contribute substantially to near-UV absorption.
The manuscript itself indicates that online absorption agrees considerably better with methanol-soluble absorption than with water-soluble absorption. This suggests that solubility selection itself may be a major reason for the online–offline difference.
The present comparison therefore combines at least three effects:
(1) different temporal integration,
(2) water-solubility selection, and
(3) potential chemical modification during nebulization and drying.
These effects are not experimentally separated in the current analysis. Accordingly, I do not think the source-attribution reversal can presently be assigned specifically to atmospheric aging versus preparation-induced browning. The manuscript can report the empirical difference between the two analytical approaches, but the mechanistic explanation requires further validation.
3. Section 3.3: offline MLR attribution is not sufficiently validated
As noted above, I do not think the offline MLR attribution can presently support the mechanistic interpretation developed in Section 3.3 without experimental validation of the nebulization/drying procedure.
The high inferred MAC of offline MO-OOA is central to the manuscript's interpretation of aged or transported BrC. However, this value may reflect some combination of extraction selectivity, altered composition during nebulization/drying, and differences in the material sampled by the online and offline methods.
The authors therefore need to distinguish clearly between an empirical statistical association and a chemically demonstrated source-specific BrC property. Until the preparation-related effects are quantified, the offline PMF factor cannot be assumed to represent the same chromophoric material responsible for the measured aqueous BrC absorption.
4. Section 3.4: methodology and interpretation of the radiative-effect estimates
The radiative-effect calculation is not described in sufficient detail for the reader to evaluate or reproduce the analysis. The equations, input parameters, wavelength treatment, assumptions, and sources of the optical quantities used in the calculation should be provided in the Materials and Methods section rather than introduced only in the Results and Discussion.
In particular, the origin and derivation of the BrC aerosol absorption optical depth (AAOD) used in this analysis need to be clearly explained. AAOD is a column-integrated optical quantity, whereas the principal measurements in this study are surface-based. If AAOD is obtained from AERONET or another external product, the authors should specify the dataset, wavelength, temporal matching, and methodology used to isolate the BrC contribution from total aerosol absorption. If AAOD is instead derived from the surface absorption measurements, the assumptions required to convert surface absorption to a column-integrated quantity need to be described and justified.
I do not necessarily object to the use of a parameterized radiative-efficiency calculation rather than a full radiative-transfer model, provided that the method has been established previously and all assumptions and limitations are clearly stated. However, the authors should make clear what quantity is actually being estimated and avoid implying a more fully constrained atmospheric radiative forcing than is supported by the available observations.
More importantly, the source-resolved radiative estimates inherit the uncertainty associated with the source-specific MAC values. As discussed in Major Comment 1, the unusually high offline MO-OOA MAC may partly reflect water-extraction selectivity and/or chemical changes occurring during nebulization and drying. Until the offline source-resolved MAC values are experimentally validated, propagating them into estimates of source-specific radiative importance is premature. This limitation should be resolved before drawing conclusions about the atmospheric radiative importance or persistence of MO-OOA-associated BrC.
Minor comments
1. Lines 210–230 / Table 2
I found this section difficult to follow conceptually. The authors should first clearly state that Table 2 is intended to compare the inferred contributions of different OA source classes to BrC absorption, rather than simply comparing OA composition among sites.
The manuscript applies MAC values measured in the present Seoul dataset to OA-factor mass fractions reported for other locations and then uses these reconstructed absorption contributions to infer broader geographical and seasonal patterns. This requires much stronger justification. Factor-specific MAC values can vary substantially with source characteristics, atmospheric processing, fuel type, and measurement approach. Applying Seoul-derived MAC values to OA factors measured at unrelated sites is therefore not equivalent to comparing directly measured source-resolved BrC contributions.
The subsequent conclusion that a previously suggested “Asian-secondary versus European-biomass-burning” distinction is largely a seasonal artifact appears considerably stronger than the evidence presented here. The current literature compilation is limited and heterogeneous with respect to season, location, source composition, and measurement methodology. Either substantially more evidence is needed to support this generalization, or the interpretation should be considerably moderated.
It may also be useful to distinguish the absorption attributed to less-oxidized and more-oxidized OOA in Table 2 where such information is available, given that the contrast between these two types of secondary OA is central to this manuscript.
2. Line 237
Please provide the literature source supporting the stated comparison for the HOA MAC.
3. Around Line 287
The argument that chloride provides evidence against a primary combustion influence appears insufficient by itself. Have the authors examined other primary combustion indicators, such as BC/EC or hydrocarbon-like AMS fragments (e.g., CxHy+ ion families), in relation to LO-OOA2 and the nitrogen-containing fragments? Including multiple independent tracers would provide a more convincing distinction between primary emission and secondary formation.
4. Lines 292–297
The authors correctly note that amines and other reduced-N species can have both primary and secondary sources. However, the manuscript subsequently proposes carbonyl condensation involving ammonia/amines and dicarbonyls derived from aromatic and biogenic VOC oxidation as the pathway most consistent with the observed reduced-N fragments. This mechanistic interpretation appears more specific than the measurements allow.
Without precursor-resolved VOC measurements or molecular-level identification of the nitrogen-containing products, the data appear to establish an association between LO-OOA2, reduced-N AMS fragments, and nighttime conditions, but not the specific precursor chemistry responsible for them. Please clearly distinguish the observed relationships from the proposed chemical mechanism and discuss plausible alternative sources.
5. Lines 298–304
The discussion of nitrate-radical formation of nitroaromatics is currently speculative. HR-ToF-AMS fragments alone provide limited molecular specificity, and the manuscript itself states that the measurements cannot constrain the contribution from this pathway. Unless there is additional evidence for nitroaromatic formation, I suggest substantially shortening this discussion and presenting it only as one possible pathway rather than as part of the explanatory framework.
6. Lines 305–308
The manuscript attributes the post-sunrise decreases in LO-OOA2 and BrC primarily to boundary-layer dilution, with oxidative conversion superimposed. However, if LO-OOA2 is indeed strongly associated with secondary BrC chromophores, photochemical bleaching after sunrise should also be considered quantitatively.
Can the observed decrease be separated into dilution and chemical loss? For example, normalization to an appropriate dilution tracer may help determine whether BrC absorption decreases faster than expected from boundary-layer dilution alone. The authors should also discuss whether the expected photochemical lifetime of the proposed reduced-N chromophores is consistent with the observed morning evolution. At present, the statement that the decrease occurs “mainly through dilution” appears insufficiently demonstrated.
7. Section 3.2.3
The manuscript contrasts nitrogen-containing chromophores with non-nitrogenous/conjugated aromatic chromophores. Since photobleaching is central to the interpretation, it would be useful to discuss whether these proposed chromophore classes are expected to exhibit different photochemical lifetimes.
In particular, is there evidence that reduced-N heterocycles, nitroaromatics, and non-nitrogenous conjugated carbonyl/aromatic species should bleach at comparable rates under the atmospheric conditions considered here? This could help determine whether the observed diurnal behavior is chemically consistent with the proposed assignments.
Technical corrections
1. Section title around Line 241
I suggest replacing the current wording with something more chemically specific, for example:
“Biomass burning is the most efficient absorber and a major source of nitrogen-containing organic aerosol.”
The term “nitrogen” alone is ambiguous because the AMS nitrogen signal can include inorganic and organic nitrogen, whereas the discussion here specifically concerns nitrogen-containing OA.
2. Section 3.2.2 / Figure 4
Figure 4 appears to be incorrectly reproduced in the manuscript and is identical to Figure 3 in the version provided for review. As a result, the arguments in Section 3.2.2 that rely on Figure 4 cannot be independently evaluated. The correct figure needs to be provided.
3. Line 347
Please format “babs,BrC(370 nm)” consistently with the remainder of the manuscript, including the appropriate subscripts.
4. Tables and Supplement formatting
Formatting is inconsistent among the tables and figure/table captions, particularly in the Supplement, including the use of boldface, borders/lines, and general table formatting. Please standardize these elements throughout the manuscript and Supplement for readability.