| 07 Oct 2026
Mutual constraints on global aerosol composition from novel PACE data and ModelE
Abstract. Aerosol climate impacts remain uncertain, in part because global aerosol composition observations have been unavailable to constrain Earth System Models (ESMs). We present the first comparison of aerosol composition from the Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) satellite’s Spectropolarimeter for Planetary Exploration one (SPEXone) against an ESM, co-locating Remote sensing of Trace gas and Aerosol Products (RemoTAP) retrievals with NASA GISS ModelE simulations to within ±15 minutes over 2024–2025. In the fine mode, the two correlate well over land, capturing the same seasonal biomass burning and pollution cycles. The coarse mode, however, diverges: mineral dust agrees to within ∼10 % by volume over land, yet SPEXone exceeds ModelE eightfold over ocean, where it covaries with modeled sea salt and increases away from desert sources. Because SPEXone partitions the coarse mode on shape alone, we suggest this excess is nonspherical marine aerosol rather than mineral dust. Finally, ModelE carries a median 3.6 times the water per unit dry sea salt that SPEXone infers over ocean. Substituting reanalysis relative humidity (RH) and an observationally constrained hygroscopicity each move the model toward the retrieval, but neither alone nor together do they reproduce it. This factor amounts to only 0.027 in global median effective marine refractive index, amplified by the nonlinearity of the mixing rule used to infer water. Together, these results provide a first evaluation of modeled aerosol composition against PACE SPEXone retrievals.
Competing interests: At least one of the (co-)authors 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.
General assessment
This manuscript compares PACE SPEXone aerosol retrievals with collocated ModelE-MATRIX simulations. The comparison is valuable, and the marine water analysis and supplementary regional results are useful. However, several substantial discrepancies are reported without adequate explanation. Agreement in global means is also emphasized despite large regional biases and weak temporal agreement.
The authors offer possible explanations for the main organic-aerosol, ocean dust, and marine water differences. The regional discrepancies below require further discussion. I recommend major revision.
Major comments
1 Explain the regional discrepancies
Several important results in Figure 9 and the supplement are not adequately addressed in the discussion.
For each major discrepancy, provide an explanation supported by the analysis, or a clearly identified possible explanation. Where the cause remains unresolved, state this explicitly and limit the conclusion accordingly.
2 Clarify the uncertainty shown in Figure 2e
Figure 2e (p. 11) shows elevated relative AOD uncertainty over the western United States, southwest South America, Australia, and other land regions. The caption identifies these features, but the manuscript does not explain them. Please clarify the reasons for this regional pattern.
The AERONET evaluation in Figure S2 (supplement p. 5) assesses AOD. It does not establish the accuracy of retrieved composition or water. State what uncertainty information is available for these derived quantities and account for it when interpreting their differences from ModelE.
3 Support the physical attribution of the main discrepancies
The authors attribute part of the ocean dust excess to nonspherical sea salt and the fire-season dust signal to nonspherical smoke. The spatial and seasonal associations support these interpretations, but do not establish them. Sections 4.2 and 4.4 acknowledge limitations; the abstract and conclusions should be clear about the same caution. State which findings are demonstrated and which remain hypotheses.
Figure S12 (supplement p. 15) supports an absorption-based reclassification, but uses a different retrieval product, shorter period, and no ModelE collocation. It therefore cannot directly quantify the reduction of the original ocean dust discrepancy (factor 8.4; Sect. 3.2, p. 12, lines 277-278). Match the comparison samples or remove that numerical extrapolation. Justify the absorption threshold and demonstrate that the retrieved index supports the proposed classification.
4 Clarify the retrieval quantities and their comparison with ModelE
Section 2.2 (pp. 4-5, lines 102-127) describes important assumptions, but does not clearly separate fitted parameters from derived products. Add a short table or schematic identifying the parameters fitted to the satellite measurements, the fixed assumptions, and the steps producing the AOD, component volumes, and water used here. State which derivations are part of RemoTAP and which are performed in this study. A full algorithm description is unnecessary, which is included in the references.
Also distinguish internally retrieved parameters from variables supplied in the released product. For example, coarse nonspherical IRI is described as retrieved in Section 2.2 (p. 4, lines 112-114), but not reported in the product in Section 4.2 (p. 31, lines 513-515). Clarify this distinction.
Map the retrieval quantities to the corresponding ModelE species and populations, stating dry or wet definitions and treatment of mixtures. In particular, clarify the correspondence between the model's accumulation, coarse, and mixed salt populations and the retrieved coarse spherical mode. Tables S1-S3 (supplement pp. 1-3) provide optical constants and densities, but do not resolve these distinctions.
Specific comments