This paper presents an inversion algorithm for retrieving the shapes of ice cloud particles using far infrared spectral radiance measurements. The algorithm's performance is evaluated using simulated measurements from the new far infrared sensor, Far-infrared Outgoing Radiation Understanding and Monitoring (FORUM), in various scenarios: tropics, mid-latitudes, and polar regions. This study is a continuation of a previous paper by Di Natale and Palchetti (2022) that focuses on the algorithm's development and validation.

Overall, the results indicate a successful convergence of the inversion algorithm in tropical and mid-latitude scenarios. However, some challenges were encountered in the polar scenario, particularly when considering retrieval affected by FORUM measurement noise. Notably, there were significant differences in the simulated outgoing longwave radiation when using pre-defined fixed shapes compared to retrieved shapes. These findings have implications for improving ice cloud parameterization and enhancing our understanding of ice particle habits in different locations.

While the paper is logically structured and well-organized, I have a few suggestions to enhance its clarity:

1. Figure 1: Why does the difference between the radiance computed with the rigorous approach and the approximation increase with OD before OD < 2.0 but decrease when OD > 2.0? The author should provide an explanation for this trend.

2. Section 2: I am wondering which database of the optical properties of ice crystals was used in this study. Yang et al. (2013) was continuously updated based on the improvement of computational techniques. For example, Bi and Yang (2017) updated the database based on the invariant imbedding T-matrix method  (Bi and Yang, 2014) and improved ray tracing technique for absorbing particles.

3. Section 5, Lines 399-410: The author conducted test retrievals using the pre-defined habits of King et al. 2004 and the retrieved habits in this study, assuming that the atmospheric, surface, and cloud parameters were known. However, if the cloud parameters were also influenced by the habits in the retrieval when considering simultaneous retrieval of cloud parameters, I wonder if the value of 2.7 W/ would be lower. Additionally, it is possible that the difference caused by the habits would be partially compensated by adjusting the cloud parameters.
4. In Figure 1, the y-axis tick labels of "0,001" should be "0.001". This typo also exists in Figures 4, 6, 7, and 8.
5. Line 278: "the a priori..." should be corrected.
6. Line 324: "...a part 4 out..." should be corrected.
7. Line 319: "Fig.s 3 and 4..." should be corrected.
8. Line 410: “(0.6+-0.4)W/m^2 Wild et al. (2013)” should be corrected.

References:

Di Natale, G., & Palchetti, L. (2022). Sensitivity studies toward the retrieval of ice crystal habit distributions inside cirrus clouds from upwelling far infrared spectral radiance observations. Journal of Quantitative Spectroscopy and Radiative Transfer, 282, 108120. https://doi.org/10.1016/j.jqsrt.2022.108120

Bi, L., P. Yang, (2017). Improved ice particle optical property simulations in  the ultraviolet to far-infrared regime. Journal of Quantitative Spectroscopy and Radiative  Transfer, 189, 228-237.

Bi, L., P. Yang ( 2014). Accurate simulation of the optical properties of atmospheric ice crystals with invariant imbedding T-matrix method. Journal of Quantitative Spectroscopy and Radiative Transfer, 138,17-35.

King MD, Platnick S, Yang P, Arnold GT, Gray MA, Riedi JC, et al. Remote sens- ing of liquid water and ice cloud optical thickness and effective radius in the arctic: application of airborne multispectral MAS data. J Atmos Oceanic Technol 2004;21:857–75. doi: 10.1175/1520-0426(2004)021 < 0857:RSOLWA > 2.0. CO;2 .