the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 May 2026
Lagrangian investigation of GCMs during the 2014–15 Holuhraun eruption reveals large differences in the representation of aerosol size distribution
Abstract. Our ability to understand and predict future climate scenarios remains limited by significant uncertainties in climate modelling, particularly those related to aerosol-cloud interactions (ACI). The Holuhraun eruption (2014–2015) provides an ideal opportunity to investigate ACI, with peak daily sulphur dioxide (SO2) emission rates exceeding that of all anthropogenic sources in Europe. In this study we perform the first Lagrangian evaluation of aerosol processes associated with an effusive volcanic perturbation that combines in-situ aerosol particle number size distribution (PNSD) from rural in-situ sites with air mass back-trajectories to understand differences in general circulation model (GCM) representations. Holuhraun significantly impacted the observed PNSD at the three sites considered, showing a consistent increase in the accumulation modal diameter and evidence of sustained growth during transport from new particle formation (NPF) in the plume. ECHAM6.3-HAM2.3 did not replicate the observed sustained growth from NPF events, instead the volcanic perturbation was associated with growth of pre-existing particles, contributing to the mass of aerosol. Contrastingly, UKESM1.0 demonstrated no increase in the modal diameter during the eruption period. The inclusion of organic-mediated boundary layer nucleation (BLN) into UKESM1.0 (UKESM-BLN), enabled a considerably better representation of PNSD changes. UKESM-BLN replicated the increase in accumulation mode diameter, as well as sustained NPF events, although it considerably overestimated number concentrations of Aitken mode particles. Investigating the perturbation in cloud condensation nuclei during the eruption year demonstrated that UKESM-BLN better replicated the perturbation at the boreal sites, highlighting the importance of BLN processes in accurate representation of ACI in GCMs.
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.- Preprint
(31904 KB) - Metadata XML
-
Supplement
(46290 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-1043', Anonymous Referee #1, 03 Jun 2026
-
RC2: 'Comment on egusphere-2026-1043', Anonymous Referee #2, 01 Jul 2026
This manuscript uses the 2014–2015 Holuhraun effusive eruption as a natural experiment to examine aerosol microphysical processes in two global climate models, ECHAM6.3-HAM2.3 and UKESM1.0. The study compares in situ particle number size distribution (PNSD) measurements from three rural sites with model output sampled using a Lagrangian trajectory framework. A third simulation, UKESM1.0 with organic-mediated boundary-layer nucleation (UKESM-BLN), is also included to test whether this additional nucleation pathway improves the simulated PNSD evolution during plume transport.
The results suggest that the Holuhraun eruption affected the observed PNSD at all three sites. During the eruption period, the observations show a consistent increase in accumulation-mode diameter and particle growth consistent with new particle formation (NPF) during plume transport. ECHAM6.3-HAM2.3 does not reproduce this sustained growth and appears to attribute much of the response to the growth of pre-existing particles. UKESM1.0 also does not capture the observed increase in modal diameter. UKESM-BLN improves the simulated modal-diameter evolution, but this improvement comes with a substantial overestimate of Aitken-mode number concentrations.
The manuscript is clearly written and the analysis is generally comprehensive. My main concern is that the improvement in UKESM-BLN is interpreted mainly through the presence of the BLN parameterization, without fully diagnosing whether the mechanism is physically consistent with the Holuhraun plume environment. I also have concerns about the use of trajectory clusters to interpret microphysical processes without stronger constraints from meteorological history.
Major comments
1. UKESM-BLN includes organic-mediated boundary-layer nucleation following Metzger et al. (2010), but the Holuhraun perturbation is primarily an SO₂ perturbation. For organic-mediated BLN to be efficient, the model also needs suitable organic vapors, favorable boundary-layer structure, and a condensation sink low enough for newly formed particles to survive. The manuscript should therefore show where and when the UKESM-BLN nucleation is actually active relative to the model-defined boundary layer.
This point is important because boundary-layer NPF can depend strongly on mixed-layer evolution, residual-layer influence, and vertical structure, as shown by Zeppelin-based observations of NPF onset and growth in the planetary boundary layer (Lampilahti et al., 2021). It would be useful to know whether the initial BLN source occurs mainly in the mixed layer, and whether the particles then grow or are transported into the residual layer or entrainment zone. Without this diagnostic, the improved performance of UKESM-BLN should be interpreted cautiously and should not, by itself, be taken as evidence for plume-wide organic-mediated BLN.
2. The trajectory clusters appear to be based mainly on horizontal position and altitude. That is useful for identifying transport pathways, but it is not sufficient by itself to support the microphysical interpretation. Processes such as NPF, condensational growth, sulfate production, and coagulation depend on meteorological history (temperature, RH, vertical velocity, liquid-water exposure, and altitude relative to the boundary layer), not only on position.
I suggest the authors should either provide cluster-wise diagnostics of these variables or test whether the clustering changes when these meteorological variables are included.
3. Please clarify how the 27-member trajectory ensemble is used. Is it only used to represent transport uncertainty, or are the individual members treated as independent samples in the cluster statistics?
4. The BLN effect is currently inferred by visually comparing the two panels in Fig. 6. Since the two UKESM simulations differ only by the inclusion of BLN, a direct UKESM-BLN minus UKESM1.0 difference panel for Δ(dN/dlogDp) would be very helpful. This would show more clearly when BLN adds particles.
5. If the observations do not resolve particles below approximately 10 nm, then the modeled 1–10 nm nucleation mode and the initial particle formation step cannot be directly validated. In that case, the observational evidence should be described as particle growth consistent with prior or nearby NPF, rather than as direct observational evidence of nucleation.
6. UKESM-BLN better reproduces the modal-diameter evolution, but it substantially overestimates Aitken-mode number. This raises a key question: is UKESM-BLN producing a more realistic aerosol lifecycle, or is the apparent improvement caused by excessive particle formation being compensated by growth, coagulation, dilution, or deposition? Here, source–sink diagnostics for nucleation, condensational growth, coagulation, and removal would help address this. A tagged volcanic sulfur tracer, if available, would also be valuable for separating plume-derived growth from growth of background aerosol.
Minor comments
-Line 185: “The simulations of the eruption at Holuhraun in 2014 use emission altitude and magnitude described by Malavelle et al. (2017), distributed between the 800 m and 3000 km”- is this a typo? Should this be 3000 m rather than 3000 km?
-Please clarify whether the UKESM-BLN improvements are statistically significant across the three sites, size ranges, and trajectory clusters, or whether they are mainly limited to specific transport regimes.
References
Lampilahti, J. et al. Zeppelin-led study on the onset of new particle formation in the planetary boundary layer. ACP, 21, no. 16 (2021): 12649-12663. https://doi.org/10.5194/acp-21-12649-2021
Metzger, A. et al. Evidence for the role of organics in aerosol particle formation under atmospheric conditions. PNAS, 107, no. 15 (2010): 6646-6651
Citation: https://doi.org/10.5194/egusphere-2026-1043-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 320 | 68 | 13 | 401 | 49 | 19 | 17 |
- HTML: 320
- PDF: 68
- XML: 13
- Total: 401
- Supplement: 49
- BibTeX: 19
- EndNote: 17
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript entitled “Lagrangian investigation of GCMs during the 2014-15 Holuhraun eruption reveals large differences in the representation of aerosol size distribution” is well written, clearly presented, and thoroughly documented. I thank the authors for conducting a comprehensive analysis of the selected GCMs and their ability to simulate aerosol transport and PNSD. The methodology is robust, well explained, and sufficiently detailed to allow reproducibility. The results are clearly presented and discussed. Overall, I recommend publication after minor revisions.
Specific Comments
Section 3.1
The authors may consider using the Sulfur Oxidation Ratio (SOR), defined as SO4/(SO2+ SO4). SOR is a commonly used metric to quantify the conversion of SO2 to sulfate and can provide useful insight into sulfur oxidation processes and aerosol aging. It may also help identify environments where aqueous or heterogeneous oxidation pathways are important. A similar analysis could potentially strengthen the discussion associated with Figure 10. Ref-( Rogers, M. J., Joo, T., Hass-Mitchell, T., Canagaratna, M. R., Campuzano-Jost, P., Sueper, D., Tran, M. N., Machesky, J. E., Roscioli, J. R., Jimenez, J. L., Krechmer, J. E., Lambe, A. T., Nault, B. A., and Gentner, D. R.: Humid summers promote urban aqueous‐phase production of oxygenated organic aerosol in the northeastern United States, Geophys. Res. Lett., 52, e2024GL112005, https://doi.org/10.1029/2024gl112005, 2025.)
In addition, UKESM and UKESM-BLN appear to show similar SO₂ concentrations. How similar are the sulfate concentrations in these two models? Comparing the SOR values between them may help explain differences in sulfate production and the underlying oxidation mechanisms.
Figure 6-UKESM-BLN appears to exhibit some signatures of banana-shaped growth patterns, although these features are relatively short-lived. There also seems to be a recurring burst of nucleation during the early morning hours from days 11–16. This repetitive pattern may suggest a role of photochemistry, with increasing oxidant production after sunrise promoting particle formation. It would be interesting if the authors could discuss the dominant mechanisms controlling nucleation in these simulations.
Line 575- ECHAM has prescribed oxidants from climatology, do they have a diurnal cycle attached? Can this affect the mechanisms that control NPF?
General Figure Presentation
Several figures could potentially be enlarged for improved readability. Reducing label sizes and using common color bars where appropriate may help maximize the plotting area. For example, in Figure 10, the (DD-MM-YYYY) date labels could potentially be shown only on the bottom row of panels, and the rest of the panels may only account for (DD), allowing more space for the plot. If space permits, including a 100 nm tick mark on the y-axis may also help readers identify the approximate boundary of the Aitken mode.
Section 3.3 – Figure 10
The observed volcanic PNSD exhibits a bimodal structure, characterized by a pronounced Aitken mode and a secondary accumulation mode. Several models reproduce this behavior, although with varying degrees of uncertainty. Could the authors discuss the physical processes responsible for this bimodal structure? For example, does it reflect a combination of newly formed particles that have grown into the Aitken mode together with condensational growth on pre-existing particles along the transport pathway?
In addition, ECHAM appears to show an unusually strong nucleation mode at Zeppelin, which is relatively far from the volcanic source. Could the authors comment on whether this feature reflects a systematic model bias or a specific process represented in the model?
Line 505-The authors mention a significant contribution from gas-phase oxidation. Could they clarify whether the precursor gases responsible for this oxidation are primarily volcanic in origin, or whether background sulfur species also contribute?
Line 600-How is the contribution from pre-existing sulfate separated from newly formed sulfate? Is the source of sulfate known in this analysis? Clarifying whether the sulfate was already present in the atmosphere or was produced through new particle formation in the volcanic plume and subsequent growth would be helpful.
Introduction
It would be useful to provide a brief description of the volcanic emissions. For example:
Such information would help place the modeling results into context.
Minor Comments