the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Aug 2026
Evaluating the performance of the thermodynamic equilibrium model ISORROPIA 2.1 for different aerosol compositions
Abstract. The hygroscopic behavior of atmospheric aerosols is controlled by their chemical composition, which determines phase transitions, water uptake and physical state. Aerosol hydration influences particle size, mass, optical and chemical properties, making it a key process in atmospheric science. This study evaluates the performance of ISORROPIA 2.1, a thermodynamic equilibrium model, in predicting aerosol water uptake across a wide range of aerosol types, including indoor and outdoor (urban, extra-urban and marine) aerosols and deposits on electrical insulators. Inorganic composition was determined by ion chromatography, while deliquescence behavior was experimentally characterized using an Aerosol Exposure Chamber, providing a comprehensive dataset for model validation. ISORROPIA 2.1 accurately reproduces deliquescence relative humidity (DRH) and hygroscopic growth in sulfate–nitrate–ammonium dominated systems, with mean deviations of 4.3 ± 2.6 % RH for the onset of deliquescence (DRHs) and 4.4 ± 3.2 % RH for completion (DRHe). In contrast, ammonium-poor mixtures enriched in alkali and alkaline-earth ions (K+, Mg2+, Ca2+, Na+), including marine aerosol and insulator deposits, show non-physical humidification behavior characterized by discontinuous liquid–solid transitions. These anomalies arise from routine-selection instabilities in dust-rich regimes. Targeted code refinements improved model stability and water-uptake predictions, extending the applicability of ISORROPIA 2.1 to sulfate-poor, sodium- and crustal-rich systems. Overall, this work provides an experimental benchmark for aerosol thermodynamics, identifies the compositional limits of ISORROPIA 2.1 and broadens its applicability to complex inorganic aerosol mixtures.
- Preprint
(1435 KB) - Metadata XML
-
Supplement
(957 KB) - BibTeX
- EndNote
Status: open (until 28 Sep 2026)
- RC1: 'Comment on egusphere-2026-3968', Anonymous Referee #1, 25 Aug 2026 reply
-
RC2: 'Comment on egusphere-2026-3968', Anonymous Referee #2, 28 Aug 2026
reply
General assessment
The manuscript presents an interesting and potentially important evaluation of ISORROPIA 2.1 across indoor, urban, extra-urban, and marine aerosols, as well as aerosol deposits and synthetic systems. Combining experimental hygroscopicity measurements with thermodynamic modelling is a particular strength. The identification of discontinuous and apparently non-physical model behaviour in ammonium-poor, sodium- and crustal-rich regimes is scientifically relevant, especially given the widespread use of ISORROPIA in atmospheric chemistry and chemistry-transport models.
Overall, the study could make a useful contribution to aerosol thermodynamics and model evaluation. However, several issues should be addressed before the conclusions can be fully supported. In particular, the experimental methods and model configuration require clearer documentation, some aspects of the phase-transition interpretation need clarification, and claims concerning the predictive performance and broader applicability of the revised model should be more closely aligned with the presented results.
Specific comments
The objectives are generally clear, but the final part of the Introduction remains mostly descriptive. I suggest stating the central research question more explicitly, for example, whether the performance of ISORROPIA varies systematically with aerosol composition, particularly in ammonium-poor, sodium- and crustal-rich systems. This would improve the link between the Introduction, Results, and Conclusions.
The description of the Aerosol Exposure Chamber experiments requires some additional detail. Please report, or provide a clear reference for, the equilibration time at each RH step, the criterion used to determine that equilibrium had been reached, the number of replicate AEC runs, the stability and uncertainty of temperature and RH, and any blank or filter corrections applied. These points are important because the measurements are compared directly with a thermodynamic equilibrium model.
The experiments include both humidification and dehumidification cycles between 20 and 90% RH in 2-percentage-point increments. Was hysteresis observed in the investigated systems? Even though the model evaluation focuses on deliquescence, a brief discussion of the dehumidification branch, if available, would provide useful physical context.
It is not clear what the experimental points in the humidification curves represent: individual measurements, means, or medians across samples or experimental runs. Please state how many measurements contribute to each curve and how they were combined. If averaged curves are shown, some indication of variability, such as SD or interquartile range, would be useful. Individual curves could also be included in the Supplement.
In one case, the manuscript refers to an “average experimental curve” compared with a model run based on average composition, yet uncertainty is not shown over the full RH range. Please also clarify whether the DRHs and DRHe values in Table 1 were calculated separately for each experiment and then averaged, or extracted directly from an averaged humidification curve. For a nonlinear phase transition, these approaches are not necessarily equivalent.
The ISORROPIA configuration should be described more explicitly. Because the results depend on temperature and on the stable- or metastable-assumption, please specify the settings used in each case. A concise table, or equivalent description, could include temperature, reverse/forward mode, stable/metastable option, RH grid, input units, and the treatment of zero or below-detection-limit concentrations. This would improve reproducibility and interpretation.
Please explain the physical or thermodynamic basis for the new RH criteria involving CaCl₂ and MgCl₂. The Supplement states that solid–liquid coexistence is allowed above 20% RH when CaCl₂ is present and above 24% RH when MgCl₂ is present without CaCl₂. Please indicate how these thresholds were derived and over what composition range they are expected to apply.
The manuscript should distinguish more clearly between removing non-physical discontinuities and achieving quantitative agreement with the observed phase transitions. Important DRH differences remain after the revision. For EUREC4A, the experimental DRHs/DRHe values are 62.0/77.8% RH, whereas the revised model predicts 28/50% RH. For DEP-MR2, the corresponding values are 61.3/70% and 34/50%; for DEP-MR3, they are 55.3/66.7% and 28/50%.
Claims concerning improved predictive performance or broader applicability should therefore reflect these remaining discrepancies. The revision appears to improve particular cases, especially DEP-MR1, but the quantitative improvement is less consistent for EUREC4A, DEP-MR2, and DEP-MR3.
The interpretation of the EUREC4A results could be more nuanced. The measured deliquescence interval is approximately 62–78% RH, while the discussion relates it mainly to the DRH of pure NaCl near 75% RH. Although the upper end of the interval is close to the NaCl value, the onset near 62% RH may not be fully explained by comparison with pure NaCl alone.
Because the sample represents a multicomponent marine aerosol, the lower onset may reflect mutual deliquescence or other interactions within the salt mixture. A brief discussion of this possibility would strengthen the physical interpretation.
The reference list should be checked carefully. Kelly et al. (2014) is cited in relation to aerosol hygroscopicity and the role of alkaline and alkaline-earth cations, but the corresponding reference concerns walking, cycling, and mortality and appears unrelated to the cited context. Please verify and correct this entry.
A final check of figure numbering, in-text citations, and cross-references between the main manuscript and the Supplement would also be useful, as some figure labels appear inconsistent.
Citation: https://doi.org/10.5194/egusphere-2026-3968-RC2 -
RC3: 'Comment on egusphere-2026-3968', Anonymous Referee #3, 28 Aug 2026
reply
Comments on “ Evaluating the performance of the Thermodynamic Equilibrium Model ISORROPIA 2.1 for different aerosol compositions”
BY GINI ET AL.
General comments
This paper evaluates the thermodynamic equilibrium model ISORROPIA (version 2.1) with respect to its performance in predicting aerosol water uptake for different aerosols compositions. The model-predicted deliquescence behavior of aerosols is compared with experimental results from an aerosol exposure chamber. The study show that while ISORROPIA accurately reproduces deliquescence relative humidity (DRH) and hygroscopic growth in sulfate–nitrate–ammonium dominated systems, the model simulated a non-physical humidification behavior characterized by discontinuous liquid–solid transitions. While the results provide valuable information and experimental dataset for further investigating aerosol’s deliquescence and hydroscopic growth using ISORROPIA and other thermodynamic equilibrium models, the paper needs to be improved in the presentation and interpretation of the results before it can be accepted for publication in ACP. Specifically, the improvements of ISORROPIA and physiochemical principles based on to make these changes need to be described and discussed in detail.
Specific comments and other issues:
- Line 111-112: While gaseous NH3, HNO3, HCL, and H2SO4 are aerosols precursors, I wonder if ionic Na+, Ca2+, K+, and Mg2+ can be called ‘aerosol precursors’ considering that they should exist either in solid or liquid aerosol phase already.
- Line 120: Indoor aerosol samples were used for chamber experiment and model simulations in this study. Considering that indoor environment is not the main focus of the ACP journal, I would suggest that some comments be added here or somewhere in Sect. 3.1 on whether or not similar aerosol compositions to those for indoor environment (shown in Fig. 1) have been observed in outdoor atmospheric conditions.
- Line 125-126: It would be better to state ‘the typical chemical composition of atmospheric aerosols in urban areas, which is generally dominated by sulfates, nitrates and ammonium’.
- Line 255: This subtitle might be revised to ‘2.6 ISORROPIA and its improvement’ or ‘2.6 ISORROPIA and its update’. Anyhow, I would suggest adding a description of model modification to Sect. 2.6. Specifically, the content of ‘ISORROPIA refinements’ in Supplementary materials can be moved to this section.
- Line 285-286: As mentioned above, it is not appropriate to say Na+, Ca2+, K+, and Mg2+ are aerosol precursors.
- Line 363: Change ‘concentration’ to ‘Aerosol water’ in the Y-axis of Fig. 2? So for other figures that are applicable.
- Line 475-476: What does ‘the previous one’ here mean? The conditions at lower RH?
- Line 488: What do the deliquescence and crystallization phases refer to, specifically in Fig. 6? Can the AEC and ISORROPIA simulate the crystallization process (where RH is decreased)?
- Line 507-508: How about the deliquescence and crystallization points for MgCl2 and Ca(NO3)2 used in ISORROPIA and reported in the literature? Could uncertainties or misuse of their values result in disagreements between ISORROPIA and AEC shown in Fig. 7?
- Line 514: With respect to the phrase ‘detailed in supplementary material’, the content of ‘ISORROPIA refinements’ in Supplementary materials can be moved to Sect. 2.6, as suggested above.
- Line 518: Is it a standard nomenclature here for ISSORROPIA? Perhaps, revised it to ‘named to ISORROPIA 2.1r in this study’?
- Line 552-555: It is suggested to describe the improvements you have made for ISORROPIA in more clear and specific way.
Citation: https://doi.org/10.5194/egusphere-2026-3968-RC3
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 96 | 49 | 15 | 160 | 21 | 22 | 16 |
- HTML: 96
- PDF: 49
- XML: 15
- Total: 160
- Supplement: 21
- BibTeX: 22
- EndNote: 16
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This manuscript presents a rigorous and well-designed study evaluating the performance of the thermodynamic equilibrium model ISORROPIA 2.1 in predicting the water uptake of atmospheric aerosols. The authors couple theoretical modeling with a variegated experimental approach, utilizing an Aerosol Exposure Chamber and ion chromatography to characterize deliquescence behavior across a highly diverse suite of aerosol samples. These samples span indoor environments, various outdoor settings (urban, extra-urban, and marine), and notably, deposits on electrical insulators.
Given the widespread use of ISORROPIA in global chemical transport models and the fundamental role that aerosol hydration plays in determining particle size, mass, and optical properties, this work is of definite and broad interest to the atmospheric science community. By addressing a known gap in modeling complex inorganic aerosol mixtures and providing both an experimental benchmark and tangible code improvements, this manuscript represents a highly valuable and practical contribution to the field.
For all these reasons, I would recommend the publication of this manuscript, but only after the concerns detailed below are adequately addressed.
General comments
-Despite my positive evaluation of the scientific content, I must note that the paper is currently written and structured in a highly technical manner. Because of this, it might be a better fit for a technically focused journal like Atmospheric Measurement Techniques (AMT) or others. However, I ultimately defer to the Editor's judgment regarding the most appropriate venue for this work.
-Language and Structure: The manuscript is quite verbose and somewhat repetitive in several sections, and the quality of the written English requires improvement. I strongly recommend a thorough stylistic and linguistic revision of the text to enhance its clarity. For instance, while the Introduction is highly comprehensive, it is overly long/didactic and occasionally redundant. Streamlining this section would greatly improve the paper's focus and readability.
- Similarly, Section 2 (Materials and Methods) is excessively lengthy and somewhat convoluted. The current text contains several repetitions, and the organization of the subsections is occasionally redundant and confusing. I strongly suggest restructuring these paragraphs to eliminate unnecessary overlap and improve the overall flow. I provide more specific recommendations on how to address this in the detailed comments below.
Detailed Comments:
Title: The current title is highly technical and does not adequately convey the scope or nature of the work performed. While not a strict condition for acceptance, I strongly suggest revising the title to more clearly reflect the methodology used and/or the key scientific findings of the study.
L37-38: the sentence is too long and not clear. I suggest to cut it in two sentences or to rephrase it.
L39: “In addition, aerosol hygroscopicity is highly relevant in cultural heritage context” sounds weird and verbose. Please consider to rephrase.
L159: as an example of the inconsistencies in Section 2, the authors explicitly mention here the type of filters (i.e., Teflon) used to recreate and redeposit the deposit samples. However, earlier in the text, when discussing the ambient aerosol samples, the filter type is completely omitted. Please clarify whether the same type of filter was used for both ambient and deposit samples. Alternatively, consider removing the mention of Teflon at this specific point, as the experimental procedure is detailed much more comprehensively later.
L168-170: as another example of the aforementioned small inconsistencies, please specify the type of filters, collectors, and the sampling duration used for the Milano airport campaign. Since these methodological details are provided for other campaigns later, I highly recommend carefully checking the Section to ensure that a consistent level of detail is reported across all sampling campaigns.
L191: Please specify the degree of purity of the "distilled water" used for sample preparation. Was it ultrapure water (e.g., Milli-Q grade, 18.2 MΩ·cm) or standard distilled water? If standard distilled water was utilized, please clarify whether analytical blanks were evaluated. It is essential to definitively rule out any potential contamination from the water source that could have artificially influenced the measured inorganic composition of the deposit samples. If this information is already detailed in the Technical Specification IEC 60815-1 or in Gini et al. (2023), please explicitly state this and refer the reader to those documents.
L202-214, Subsection 2.2: In line with my general comment regarding the structural redundancy within Section 2, I recommend removing Subsection 2.2 as a standalone section. The information provided here would be better suited if integrated directly into Section 2.1, which would significantly help streamline the methodology description and improve the overall flow.
L294 (ISORROPIA Model Mode): Please clarify the rationale for running ISORROPIA in "reverse" mode. While this may be a perfectly valid methodological choice—likely dictated by the availability of aerosol-only measurements and the lack of concurrent gas-phase data—it is important to explicitly state the reasons in the text. As is well known, in many atmospheric applications, ISORROPIA is typically run in "forward" mode to accurately account for the gas-particle partitioning of semi-volatile species. Providing a brief justification here will help readers better understand the modeling setup and any associated assumptions or limitations.
L 337-338: The authors state that "The modelled humidification curves reproduce the experimental trend identified in AEC quite well...". However, this is not entirely evident from the plots in Figure 2, which actually display clear step-like jumps or different trends in the simulated deliquescence region with respect to the experimental measurements. While I understand the authors' intent here is to demonstrate that the final measured and modeled DRH values are consistent and fall within an acceptable range of uncertainty, they should explicitly acknowledge and briefly discuss these visible discrepancies in the shapes of the curves.
Figure 2: For clarity, please explicitly state the specific type of concentration on the y-axis label (e.g., by changing it to "LWC Concentration", if I understood well). This clarification should be applied to all other figures presenting similar plots. Additionally, in panel (c), the simulated curve is plotted using both lines and markers, which diverges from the formatting used in the rest of the panels. Please harmonize the plot styles to ensure visual consistency across all graphs.