Using the ICON-D2-ART model, this study examined the "sub-grid parameterization" for dusty cirrus that occasionally formed and spread over Europe. They utilized CERES SSF Level 2 Edition-4A to validate the "ecRad in ICON" predicted radiative fluxes for different conditions (e..g., all-sky, dusty/cloudy). The authors also show that the dominant longwave cooling at the cloud top could influence turbulent mixing and convective overturning, which could, in turn, help in the maintenance and thickening of the dusty cirrus aloft. The study is interesting, and the method is sound. This study can be published after addressing all the comments below.

General comments.

1. Figure 1 does not include the shortwave radiative effect. How does the shortwave heating rate at the thin ice cloud base/top altitude influence the destabilization effect?
2. The author assumed the same empirically determined thresholds for different dusty-cirrus events over Europe. How will these constant threshold values considered for different cases affect the testing of the sub-grid parameterization? Could you apply the same threshold condition for a similar dusty cirrus event in Asia?
3. Besides the radiative part, what is the role of evaporative cooling in maintaining the dusty cirrus deck aloft?

Specific comments:

1. Line 399f: Please rephrase this statement, "Since today's numerical weather prediction models neither predict mineral dust.....". There are regional models (e.g., the dust-coupled TAQM; Chen et al., 2004) which predict short-term 5-day forecasting of a dust event, including studying the dust effects on cloud microphysics using TAQM-KOSA. A similar statement in the abstract can also be rephrased.
2. Line 406f: Instead of "Once a cirrus cloud has formed at the interface between moist and dusty air, the longwave cooling at cloud top generates turbulence and mixing which thickens the cloud layer.." I would suggest writing as "Once a cirrus cloud has formed at the interface between moist and dusty air, the dominant longwave cooling at cloud top generates turbulence and mixing which thickens the cloud layer."
3. All the supplementary Figures can be referenced in the main text. The excess supplementary Figures can be taken out. If authors prefer to include all the supplementary Figures, they can include them in few .gif files or movies.
4. For clarity, the regional geographical images (Figures 1, 4-5, 11-12, and others) should include latitude and longitude labeling.
5. Figure 6 caption should be elaborated. Please add a table or extend the caption to specify the "clear/clean" & "cloudy/dusty" terms. 
6. Please add legends in Fig. 7c,d and Fig. 8a,b.
7. Quotation marks can be properly closed throughout.