Cloud Characteristics and Their Effects on Solar Irradiance According to the ICON Model, CLOUDNET and BSRN Observations
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsGeneral comments to the Authors
To explore the liquid water path (LWP) and solar irradiance at ground (Q), the ICON model, CLOUDNET and BSRN observations over several sites have been conducted. The results show that LWP is underestimated, while global irradiance has been overestimated. This research also concluded that LWP is not the key factor of Q estimation and the main factor is an inaccurate description of the cloud spatial structure. In general, it can be published in Atmosphere by addressing several comments below.
Specific comments
1、Abstract: “solar irradiance at ground (Q)” should be “solar irradiance (Q) at ground”
2、Model and Methods: Section 2.1, ICON model is not the common weather and climate forecasting model, how about its performance compared with other common model, such as WRF?
3、Model and Methods: Section 2.2, line 126, “higher than”?
4、For the equation (1) (2) (3) (4), the equation should not to be italic, which should be consistent with it in the text.
5、Line 200, it should be “less than”
6、For all tables, it’s better to utilize three-line table.
7、For all figures, the font size is too small, and the picture DPI is low (600 may be better).
Comments on the Quality of English Language
The English should be checked through the whole manuscript carefully before it can be published.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsBRIEF SUMMARY
The authors evaluate the performance of the version of the ICON numerical weather prediction model used at the Hydrometeorological Research Center of Russia in simulating stratiform clouds at three CLOUDNET sites in Germany. It is found that the model underestimates the cloud water path and cloud droplet number concentration, and overestimates the downward solar
irradiance at the surface compared to BSRN data at Lindeberg. The role of
prescribed cloud condensation nuclei concentration is discussed based on a sensitivity experiment. Also, it is argued based on overestimated direct solar radiation that a key factor for the overestimation of total downward solar irradiance is likely to an inaccurate description of cloud spatial structure.
GENERAL COMMENTS
This paper provides valuable information for the ICON modelling community and the research appears to be conducted reasonably. However, there is much room for improving the clarity and the quality of the English language. (I made a list of suggested corrections, but it is probably incomplete).
SPECIFIC COMMENTS
1. It should be noted already in the abstract that this study focused on cases with stratiform non-precipitating clouds.
2. lines 15-16: This is not very clear. Do you really mean to say that lower frequency of cloud occurrence (rather than underestimation of liquid water content) is related to the reduced sensitivity to CCN? Also, it is not quite clear what "reduced sensitivity to CCN" refers to. That the cloud droplet concentration (CDNC) is too small, or that too small a fraction of CCN are activated to cloud droplets? It is totally expected that not all CCN activate to cloud droplets, so a CCN concentration of 250 cm-3 won't result in CDNC = 250 cm-3.
3. lines 53-55: There are various bulk schemes and various bin schemes, so
I'm not sure how generally valid this statement is. Please add some details
here, like which model was used and which kind of clouds were simulated.
4. Overall, the introduction contains relatively few references to previous
work. There is an extensive literature on evaluating clouds in NWP (and even more so) climate models. Most importantly, it should be indicated whether there are previous studies that evaluate the performance of ICON in simulating clouds, and what is new in the present work.
5. line 80: "A certain configuration of the model". This is vague. It would be better to give version number, or if not available, to describe what are the differences to some default version of the model.
6. lines 90-98. The nature of ICON's cloud scheme should be described a bit better, because it differs from most other models. In particular, start with what is meant with grid-scale and subgrid-scale clouds.
7. lines 108-109: "a correction is applied for the upper troposhere ice cloud
fraction and the typeof atmospheric startification". This is not clear. Do you mean that the correction depends on atmospheric stratification?
8. lines 184-187: "We took into account only non-zero simulated and observed values ...". This sounds important. Does this mean that you screened out entirely cloud-free cases, or was the screening done layer by layer?
9. line 193: "as well as the aerosol contribution to cloud condensation nuclei (CCN)". This is not clear. Do you indeed mean to say that CCN influences the uncertainty of the Nd retrievals, or just that Nd depends on CCN?
10. lines 212--213: "we can assume that ... (Nd) equals to minimum CCN
(CCNmin) at the cloud base". This is not entirely clear either. Do you mean
that Nd provides a lower limit for the CCN at the cloud base?
11. lines 228-231, caption of Fig. 2: How exactly are the LWC values defined here? Do the LWC values represent grid-mean values or in-cloud values (i.e., hourly average grid-mean LWC divided by hourly-average cloud fraction)?
These can differ, because while the instantaneous grid-scale cloud fraction in ICON is either 0 and 1, the hourly average grid-scale cloud fraction may be between 0 and 1.
12. lines 238, 240: Please also mention the relative (per cent) error in the average LWP.
13. line 267: Should "observed" be "simulated"?
14. line 270: This reasoning seems a bit questionable. If there is X gm-2 of
rainwater in the column, one cannot conclude that without the formation of
rain, the cloud liquid water path would be X gm-2 higher. The difference
would probably be larger (if the time scale for precipitation falling to the
ground or evaporating is shorter than the time scale for cloud formation).
What one could possibly do is to compare the simulated precipitation rate
with the observed one, although even this would only give a qualitative
indication.
15. line 309: It would be more informative to say that "only a small fraction
of the CCN were activated".
16. lines 327-328: Not only is Nd smaller than indicated by the satellite
observations, it is also mostly below the value of Nd=140 cm-3, which is
the number of CCN activated at the minimum allowed updraft speed 0.1 m/s.
Can you comment on what causes this difference?
17. lines 385-387: This begs the question, how to improve the quality
of the forecast R? I think there are basically two possibilities. Either
to improve the simulation of cloud properties, or modify the assumptions
about subgrid cloud structure (cloud overlap, horizontal variability of
cloud condensate) in the radiation scheme. Also, one should try to avoid
a situation where Q is predicted accurately but LWP is underestimated,
since this indicates compensation of errors.
18. lines 395-396: What was the correlation between LWP and Q?
19. line 415: It is not clear what "cloud moisture" means here. In ICON, the subgrid-scale cloud fraction (eq. 2) is diagnosed based on the ratio of total water content to saturation specific humidity. Since the large majority of total water content is water vapour, the key variable for cloud fraction is relative humidity.
20. lines 457-458: It is quite plausible that ICON underestimates the updraft speeds in the calculation of CCN asctivation. However, it is also of note that for stratiform boundary layer clouds, the updraft speeds are most often below 1 m/s also in reality. See Tonttila et al. (2011 reference 84)
or, for example,
Ghate, V. P., Albrecht, B. A., and Kollias, P. (2010), Vertical velocity structure of nonprecipitating continental boundary layer stratocumulus clouds, J. Geophys. Res., 115, D13204, doi:10.1029/2009JD013091.
21. line 470: I have some difficulty understanding why uncertainty in liquid
water content would be a major factor in the prediction of cloud fraction.
According to Eq. (2), the subgrid-scale cloud fraction is a function of
qtot/qsat; hence it is dominated by the relative humidity (see comment 19).
22. line 475: iIt is unclear what is meant with "errors of a general nature".
That the errors are common to both prognostic and diagnostic schemes,
or that prognostic and diagnostic schemes each have their typical (but different) errors?
23. lines 480-481: "We showed that the global solar irradiance is more sensitive to the prediction of R than to the liquid water or cloud fraction
forecast." While this is an interesting result, it would be better to
express it differently: "the global solar irradiance correlates more strongly
with the forecast R than with the forecast of liquid water path or cloud
fraction (do you mean: total cloud cover?)". The reason for this is, as pointed out in comment 17, that the predicted R is dependent on the cloud properties, along with the cloud overlap assumption.
24. line 486: In fact, I'm not aware of attempts to actually forecast the cloud overlap in any NWP model. Generally if not always, the overlap assumption is fixed. Although, there have been several studies addressing how cloud overlap depends on environmental conditions, e.g.:
Naud, C.M., A. Del Genio, G.G. Mace, S. Benson, E.E. Clothiaux, and P. Kollias, 2008: Impact of dynamics and atmospheric state on cloud vertical overlap. J. Climate, 21, 1758-1770, doi:10.1175/2007JCLI1828.1.
25. line 746: "Tontilla" should be "Tonttila".
1. line 15: replace "compare" with "compared".
2. line 22: replace "plays the most important role compared to ..." with "plays a more important role than the ...".
3. line 27: add "the" before "underlying surface".
4. line 36: add "rather" before "than".
5. line 46-47: Rewrite this sentence. Suggestion: "Today, cloud-aerosol interaction processes are considered in the development of mesoscale
numerical weather predition (NWP) models".
6. line 56: add "the" before "simulated".
7. line 58: delete "The" before "diagnostic".
8. line 65: replace "providing" e.g. with "which results in".
9. Table 1: replace "Droplets" with "Droplet"
10. line 167: add "the" before "global".
11. line 170: add "the" before "warm".
12. lines 238-239: replace "leads to the liquid water path increase" with "leads to an increase in the liquid water path".
13. line 242: replace "successfully with successful".
14. line 252: add "in" after "decrease".
15. line 258: add "The" before "ICON".
16. line 262: replace "can be not only near cloud base" e.g. with "may occur higher than at the cloud base".
17. line 263: simpler might better: "However, even in this case, the Nd concentration exceeds 77cm-3 in only 25% of the ICON sample".
18. line 273: replace with "proving the level of this process" e.g. with "evaluating the strength of this process".
19. line 281: remove "a" before "lower" and "higher".
20. line 287: "more close" should be "closer".
21. line 303: replace "to compare results of specified CCNs" with "when comparing results for the specified CCNs".
22. line 305: replace "with this CCN growth" with "with this increase in CCN" or "when increasing CCN from 250 to 1700 cm-3". Note that "growth of cloud condensation nuclei" usually refers to the process in which the physical size of CCN particles increases due to condensation.
23. lines 328-331: I find this sentence confusing. Consider reorganizing it, or splitting it into parts, to improve the clarity.
24. line 335: Replace "With the growth of cloud condensation nuclei" with "when the number of cloud condensation nuclei is increased".
25. line 344: replace "provides the solar irradiance decrease .." with "results in a 9 Wm-2 decrease in the solar irradiance".
26. line 348: use the definite article here: "the liquid water content and
the effective radius"
27. line 353: replace "LWP successful" with "successful LWP".
28. line 362: Something missing? "a more difficult parameter ..."?
29. line 368: replace "hourly average procedure" with "an/the hourly averaging procedure".
30. line 369: replace "characterizes" with "characterize".
31. line 381: "at different averaging intervals (horizontally)". This gives the false impression that you do horizontal averaging. What about: "With averaging intervals corresponding to larger R ...".
32. line 390: replace the part starting "while..." with "while a more pronounced decrease of CLCL with increasing R is seen in the measurements than in the simulations".
33. line 392: replace "almost everywhere" with "for almost all R intervals".
34. line 405: replace "be observed" with "occur". Also, add "an" before
"inaccurate".
35. lines 415 and 488: "Cloud moisture" is not a conventionally used term
in the atmospheric science literature. Use "cloud liquid water content",
"cloud condensate amount", "cloud liquid water path" etc. instead.
36. line 424: "assumption, which is a difficult task" is not grammatically correct. Replace e.g. with "assumption, which is an uncertainty factor".
37. line 427: replace "is developing" with "is being developed".
38. lines 428-429: This sentence is not clear. Please try to clarify it.
39. line 443: replace "model'" with "model's".
40. line 450: replace "provided the decreasing" with "reduced the"
41. lines 452-453: replace "cloud condesation nuclei growth" with
"increasing the cloud condesation nuclei concentration".
42. line 453: replace "promote to reducing" with e.g. "help to alleviate"
43. line 465: delete "for" after "account".
44. lines 468 and 503: replace "with successful liquid water path prediction" with "in those cases in which the liquid water path was predicted successfully"
45. line 484: "scheme's"? Do you mean "the cloud scheme's"?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsI thank the authors for their efforts in revising the paper. Most of my earlier comments have been addressed adequately. However, there are still some rather small issues requiring clarification.
SPECIFIC COMMENTS
1. lines 94-106: As a follow-up to my previous comment 6, you are right that a detailed description of ICON's cloud scheme would become too long. Howewer, as it stands now, I still find this text rather confusing. To reduce the confusion, I recommend three changes. First and most importantly, adding one sentence on line 94: "ICON considers separately grid-scale clouds, which are prognostic, and subgrid-scale clouds, which are treated diagnostically". Second, I suggest combining the two sentences on lines 103-105 as follows: "The diagnostic cloud scheme in ICON is used to describe liquid and ice water content and cloud fraction for subgrid-scale clouds, that is, clouds that cannot be simulated explicitly due to the model's spatial resolution." And third, immediately after that, consider adding: "The diagnostic clouds are used by the radiation scheme, but not in the microphysical calculations."
2. lines 337-341: Mention also here that the CCN value is lower above the 2 km level (it is mentioned on lines 98-99, but don't count on the reader remembering that!).
3. lines 432-434: This difference (in the correlations between CLCL and Q,
and between LWP and Q) seems very minor, perhaps even encouragingly small. If you prefer to keep this sentence, add "slightly" before "more pronounced".
4. line 447: Underestimation of relative humidity probably influences the
cloud fraction errors more than the underestimation of cloud water content
does (even if cloud water also contributes to qtot in Eq. 2).
5. line 478: So far I understood the paper correctly, you did not actually analyze CCN data. Rather, it was demonstrated the default CCN content of
250 cm-3 is close to the typical value of Nd (cloud droplet concentration, i.e. not CCN) over Europe.
6. lines 530-535: Mention that this paragraph refers to a case study for
a single day.
1. line 17: replace "a default" with "the default".
2. line 65: "with deployed consideration of" appears a bit strange. I would suggest: "through consideration of".
3. line 91: replace "which operationally works" with "which is used operationally".
4. line 99: replace "lower" with "lowest".
5. line 211: move "satellite" between "MODIS" and "data".
6. line 340: "the saturation adjustment scheme may provide the errors of cloud evaluation". Do you mean to say that the scheme contributes to the
underestimation of Nd, or something else?
7. line 341: "fast realization of the water vapor excess". This is also not clear. Do you refer to the fact that the saturation scheme eliminates excess water vapor (i.e., supersaturation) very rapidly?
8. line 412: replace "overestimating" with "overestimation".
9. line 459: It is still not clear to me what is meant with "The scheme
generally corresponds to the global trend of complicating microphysical
processes". That the scheme is similar to many other recent schemes in
the sense that several microphysical processes are considered? Or that
it attempts to capture the net effect of many complex microphysical processes? Please clarify this.
10. line 491: it should be "60% of CCN".
11. line 537: add "the" before "SOCRATES".
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
