the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Feb 2026
High climatological cloud cover limits its response to aerosols in ICON-HAM
Abstract. Marine low-level clouds substantially cool the climate and are sensitive to aerosols. Observations suggest strong cooling from cloud fraction (CF) adjustments, yet global climate models (GCMs) may underestimate this process. We apply explainable machine-learning with SHapley Additive exPlanations (SHAP) to the ICOsahedral Nonhydrostatic atmosphere model coupled with the Hamburg Aerosol Module (ICON-HAM) to quantify CF sensitivity to cloud droplet number concentration (Nd). We analyze six experiments combining two cloud cover parameterizations, an RH-only scheme (CC–RH) and a cloud-water-dependent scheme (CC–RH–LWC), with three prescribed lower bounds for Nd (Nd,min = 10, 40, 100 cm-3). The lnNd–CF relationships in ICON–HAM are more nonlinear than in satellite observations and exhibit saturation. We quantify sensitivities using piecewise linear regression (PLR) separating low- and high-Nd regimes. Sensitivities are stronger in the low-Nd regime, while inter-scheme and inter-Nd,min differences are negligible at high Nd. Unexpectedly, CC-RH-LWC shows weaker sensitivity than CC-RH at high Nd,min, despite explicit CF–cloud water coupling. We attribute this to a "headroom effect'': higher Nd,min suppresses autoconversion and enhances liquid water path in both schemes, but CC-RH-LWC converts this into higher mean CF, limiting headroom for aerosol-induced CF increases. These findings are robust across three PLR breakpoint strategies. Our results demonstrate that GCM CF responses are state-dependent and strongly influenced by model configuration choices, which can pre-saturate cloud cover responses and mask true CF adjustment magnitudes. We suggest that model–observation comparisons should account for baseline mean CF to avoid misinterpretation.
- Preprint
(9159 KB) - Metadata XML
-
Supplement
(1207 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-569', Anonymous Referee #1, 11 May 2026
-
RC2: 'Comment on egusphere-2026-569', Anonymous Referee #2, 18 Sep 2026
“High climatological cloud cover limits its response to aerosols in ICON-HAM” by Jia et al.
This manuscript investigates how cloud-cover parameterization and prescribed minimum cloud droplet number concentration affect marine low-cloud fraction and its sensitivity to aerosols in ICON-HAM. The topic is relevant to understanding aerosol–cloud interactions and differences between modeled and observation-based cloud adjustments. The comparison of six model configurations is informative, and the headroom effect is clearly demonstrated within the experiments analyzed. However, the implications of model tuning for the magnitude and robustness of this effect require further assessment, and the results and discussion would benefit from a more integrated and concise presentation. Several physical interpretations and broader implications also benefit from clarification or qualification. I recommend revision to address the following major and specific comments.
MAJOR COMMENTS
- Sensitivity of the headroom effect to model tuning
The experiments under different minimum Nd are not retuned, and it is unclear whether all configurations reproduce observed shortwave and longwave radiative fluxes reasonably well. Although retaining the same tuning parameters helps isolate the effects of the experimental changes, the elevated CF in CC-RH-LWC could partly reflect a departure from a realistic radiation budget. Retuning to restore agreement with observations might reduce CF and consequently weaken the headroom effect. Suggest to check the radiative fluxes across the six experiments against observations and discuss how tuning could affect the magnitude and robustness of the proposed mechanism.
- Structure and presentation of Section 3
The overall message of Sect. 3 is clear, and the headroom effect is clearly demonstrated. However, Section 3 is overly long, and a more integrated structure and concise presentation would greatly improve readability. Section 3.1 provides a useful foundation, but Sects. 3.2 and 3.3.1–3.3.3 could be better integrated to connect the model–observation comparison, the illustrative regional example, and the global results. The current organization repeatedly returns to the same mean-state differences and headroom explanation, interrupting the progression of the argument and making the distinct contribution of each subsection less apparent. For example, the precipitation–LWP–CF explanation in Sect. 3.1 is repeated at the beginning of Sect. 3.3.2, while the SHAP baseline discussion largely revisits the mean-state contrasts already shown in Fig. 1c. Likewise, the explanation that elevated baseline CF limits further increases recurs in Sects. 3.3.1, 3.3.2, and 3.3.3. Explaining the mechanism once, then emphasizing the additional evidence provided by each subsequent analysis, would reduce repetition and strengthen the argument. Section 3.3.4 could be shortened to emphasize the robustness findings without reiterating methodological details or conclusions presented elsewhere. Section 3.4 is informative, but some passages—particularly the speculative extrapolation to modified Lohmann and Feichter (1997) experiments—obscure the main implications. A leaner discussion focused on what the experiments establish, what remains uncertain, and how these uncertainties could be addressed would strengthen the presentation.
SPECIFIC COMMENTS
- Lines 35–36: The phrase “independent of surface air temperature changes” is introduced without explaining its purpose. If retained, please clarify that it distinguishes aerosol-induced cloud adjustments from cloud feedbacks mediated by global-mean surface temperature change; it does not imply that the underlying cloud processes are insensitive to temperature.
- Lines 59–60: The conclusion that LWP adjustments are overestimated and CF adjustments underestimated in GCMs appears too broad. Their magnitudes and relative importance remain debated and depend on the methods, models, and cloud regimes considered. Please indicate that some studies suggest this tendency, rather than presenting it as a general feature of GCMs.
- Lines 151–153: An imposed minimum droplet concentration of 100 cm⁻³ appears less realistic than the lower values considered.
- 4. Line 195: A closing parenthesis appears to be missing after “Sect. 2.1.4”.
- 5. Lines 272–276 and 568–569: The decline in CF in CC-RH at higher Nd,min is substantial and may warrant further investigation. Enhanced cloud-top entrainment would tend to reduce cloud water, whereas Fig. 1b shows increasing LWP. Suppressed precipitation could outweigh this drying effect, but the balance is not demonstrated. Could changes in radiative fluxes also contribute? For example, reduced surface shortwave radiation might alter evaporation and the hydrological cycle, potentially affecting CF.
- Fig. 2d: Please include the corresponding MODIS-based SHAP dependence plot, preferably alongside the ICON-HAM result with consistent axes, to allow direct comparison of their nonlinearity without requiring readers to consult J24.
- Lines 332–337 and 573–575: The smoother observational relationship is attributed to spatial aggregation, but spatial aggregation is also performed in the ICON-HAM analysis. Please explain why this effect would operate differently in the two datasets and provide evidence supporting this interpretation.
- Lines 392–397 and Fig. 4a: Figure 4a conveys essentially the same mean-state contrast as Fig. 1c. Please consider combining or shortening the associated discussion to avoid repetition.
- Line 399: Please specify “at Nd,min = 10 cm⁻³” to make clear that this refers to the imposed minimum droplet concentration, rather than the actual droplet concentration.
- Sections 3.3.2–3.3.3 and throughout: Terms such as “global distribution” and “global patterns” can suggest geographical maps, whereas the figures discussed often show statistical distributions across geographical windows. Please distinguish spatial patterns from statistical distributions and use consistent terminology throughout.
- Lines 510–516: It is unclear how the speculative discussion of modifying the Lohmann and Feichter (1997) experiments advances the interpretation of the present results. Suggest removing this passage and keeping the discussion focused on implications directly supported by the current experiments.
- Lines 517–518: How does the high CF produced by CC-RH-LWC affect agreement with observed shortwave and longwave radiative fluxes? Please clarify whether the configurations were independently tuned or retained identical tuning parameters. If restoring agreement with observations would require retuning, could this reduce CF and alter the proposed headroom effect? This bears on the broader applicability of the conclusion, as discussed in major comment 1.
- Lines 518–525 and Sect. 3.4.2: Although the proposed weakening of CF-adjustment forcing is a reasonable expectation, paired PI–PD experiments would be highly valuable for testing whether it actually occurs. Such experiments would also strengthen the discussion of the implications for observational sensitivity-based forcing estimates in Sect. 3.4.2.
Citation: https://doi.org/10.5194/egusphere-2026-569-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,003 | 632 | 106 | 1,741 | 165 | 139 | 187 |
- HTML: 1,003
- PDF: 632
- XML: 106
- Total: 1,741
- Supplement: 165
- BibTeX: 139
- EndNote: 187
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Please see the comment attached.