Testing remote sensing on artificial observations : impact of drizzle and 3-D cloud structure on effective radius retrievals

Remote sensing of cloud effective particle size with passive sensors like the Moderate Resolution Imaging Spectroradiometer (MODIS) is an important tool for cloud microphysical studies. As a measure of the radiatively relevant droplet size, effective radius can be retrieved with different combinations of visible through shortwave and midwave infrared channels. In practice, retrieved effective radii from these combinations can be quite different. This difference is perhaps indicative of different penetration depths and path lengths for the spectral reflectances used. In addition, operational liquid water cloud retrievals are based on the assumption of a relatively narrow distribution of droplet sizes; the role of larger precipitation particles in these distributions is neglected. Therefore, possible explanations for the discrepancy in some MODIS spectral size retrievals could include 3-D radiative transport effects, including sub-pixel cloud inhomogeneity, and/or the impact of drizzle formation. For three cloud cases the possible factors of influence are isolated and investigated in detail by the use of simulated cloud scenes and synthetic satellite data: marine boundary layer cloud scenes from large eddy simulations (LES) with detailed microphysics are combined with Monte Carlo radiative transfer calculations that explicitly account for the detailed droplet size distributions as well as 3-D radiative transfer to simulate MODIS observations. The operational MODIS optical thickness and effective radius retrieval algorithm is applied to these and the results are compared to the given LES microphysics. We investigate two types of marine cloud situations each with and without drizzle from LES simulations: (1) a typical Correspondence to: T. Zinner (tobias.zinner@lmu.de) daytime stratocumulus deck at two times in the diurnal cycle and (2) one scene with scattered cumulus. Only small impact of drizzle formation on the retrieved domain average and on the differences between the three effective radius retrievals is noticed for both cloud scene types for different reasons. For our, presumably typical, overcast stratocumulus scenes with an optical thickness of 8 to 9 and rain rates at cloud bottom up to 0.05 mm/h clear drizzle impact on the retrievals can be excluded. The cumulus scene does not show much drizzle sensitivity either despite extended drizzle areas being directly visible from above (locally>1 mm/h), which is mainly due to technical characteristics of the standard retrieval approach. 3-D effects, on the other hand, produce large discrepancies between the 1.6 and 2.1 μm channel observations compared to 3.7 μm retrievals in the latter case. A general sensitivity of MODIS particle size data to drizzle formation is not corroborated by our case studies.

The presented work is of great importance to those using cloud properties retrievals, as it hints on the applicability of these retrievals in case of precipitating water clouds. The authors test the effect of drizzle for stratiform and convective clouds cases, for which they conclude that partcile size retrievals are almost insenstive to drizzle. The importance of this conclusion depends on the occurence of these clouds occur in nature. To my opinion the number of presented cases is insufficient to draw conclusions on the impact of drizzle and 3-D cloud effects on particle size retrievals. The authors obviously have all the appropriate tools to evaluate the impacts of both drizzle and 3-D cloud structures. Therefore, the paper can be made stronger by using a more cases, and quantify the impact for different depths of the drizzle layer, different drizzle intensities and different cloud optical thicknesses.
Although the English writing of the paper is well, not all steps conducted in this research are presented clearly. To improve the readability the paper some clarifications and reorganization of the paper is needed. The manuscript needs some major revisions before it can be published. Below the major points of criticisms are indicated, then followed by a chronological list of minor points of criticisms.

MAJOR CRITICISMS Point A
The description of the mehod for evaluating the results of drizzling clouds with respect to their effective radius needs more explaination. The effective radii of the drizzling clouds as presented in table 1 and 4 are calculated for a bimodal distribution. First, the "true" droplet distribution of the cumulus scenes seems different from the droplet distribution that is used in the retrievals (described in section 3). In order to match the "true"size distribion the reff value of the bimodal distribution is conserved by playing arround with reff1 (cloud droplets) and reff2 (drizzle droplets). Therefore the reff1 value seems to be lowered as compared to the reff1 value of the "true" size distribution. Since the majority of the droplets comprise cloud droplets, small modification to reff1 will largly effect the retrieval results. Therefore the authors should aim for a parameterization that does not change the value of reff1 as compared to the "true" value of reff1. Second, effective radius is calculated as the ratio of the third over the second moment of the size distribution. This calculation is most meaningfull for clouds with monomodal droplet size distributions, and less suited to describe the particle size of the bimodal distributions of precipitating clouds. Third, assuming a vertical homogeneous droplet distribution is already unrealistic for no-drizzling water clouds, but even more unrealistic for drizzling water clouds. This should be explained clearly in the paper.

Point B
As written by anonymous Referee #1, there are several studies that find a relationship between drizzle and particle size retrievals. There seems to be disagreement between the findings of these papers. Some see an increase in effective radius, while others hardly find any influence. The impact of drizzle on the particle size retrievals often depends on the set up of the theoretical experiment. In order to verify the sensitivity to drizzle the authors present results of a theoretical study in Figure 7. This is an important Figure, which provides information on the assumptions made in this study. The authors present a droplet size parameterization for drizzling clouds assuming vertical homogeneity. This assumption is verified for 50 cases. However, to verify its applicability for different depth of the precipitating layer within the cloud profile more information is needed. First, the statistics of the 50 samples used (tau, reff_drizzle, reff_droplets) are missing. Second, error bars for the classical and fitted bimodal distributions. Third, an analysis relative to cloud optical thickness, or even better depth of the precipitation layer relative to the cloud top, would be meaningful. Please clarify these points in this section.

Organization of the manuscript
The research method conducted in this paper can be presented more clearly. One needs to read the paper several times before the study set-up becomes clear. Consider to present the research method in the following order: • LES simulations (section 2) • Satellite reflectance simulation (section 3) • Satellite cloud properties retrieval (section 4.3) • Evaluation method (scattered over several sections) A schematic representation of the study set-up would be very helpful.

Introduction
-Page 1224: " This is why some studies suspect …." Can the authors shortly quantify the effects of drizzle on particle size retrievals as found in previous studies? -Page 1225 (line 26): This important paragraph of the introduction misses references to the investigations mentioned in this paragraph. Moreover, in this paragraph the authors should present the objective of their paper, so as to emphasis the unique aspects of this paper relative to work done in the past.

Cloud model -Page 1228 (line 25):
Give for the trade cumulus simulations also the range of optical thicknesses that is considered.

Results
-This section needs some re-organization. Section 4.1: Results of theoretical study Section 4.2 and 4.3 (introductory part): present no results but rather descriptions of the operational retrieval method. As with the LES and parameterized droplet size distributions, these sub-sections may be introduced earlier. clear what settings the authors used for their RTM. Do they assume PP clouds? Do they assume vertical homogeneous clouds? Do they use the droplet spectra as parameterized after the cloud model analysis? - Table 2 and 4: The authors find a small decrease in effective radius retrieval for precipitating clouds. This is opposite of the expected effect. This might be related to the parameterization of the droplet size distribution. Where the size of the water droplets is reduced so as to conserve the effective radius value. Is this realistic with what will happen in nature? In order to match the effective radius size, a large number of cloud droplets is reduced in size (e.g. from 12 to 8.5 micron) so as to conserve the effective radius of the drizzling cloud. How would a precipitating layer affect an effective radius retrieval for a "traditional" PP retrieval as done by the MODIS team?
Discussion and outlook -Discuss how the finings of this study are related to findings of earlier studies and explain reasons for the observed differences. -In order to translate the findings of this study to users of cloud properties retrievals the authors are encouraged to spend some works on the frequency of occurrence of these clouds in nature.