Impact of the One-Stream Cloud Detection Method on the Assimilation of AMSU-A Data in GRAPES
Round 1
Reviewer 1 Report
Manuscript remotesensing-987000
Impact of the One-stream Cloud Detection Method on the Assimilation of AMSU-A Data in GRAPES
Zhengkun Qin, Zhiwen Wu, and Juan Li
The paper describes and evaluates a method intended for application in the detection of cloud-affected microwave radiances. The framework for the evaluation is provided by the GRAPES NWP system, that forms the basis for operational NWP in China. The one-stream cloud detection method is built on collocation of AMSU-A and MHS footprints. It combines the ability of certain AMSU-A channels to detect liquid water clouds with that of certain MHS channels to detect ice clouds. The method has been previously applied in the assimilation of MHS, but not with AMSU-A.
Especially the first half of the paper is well written and easy to read. My opinion is that Section 4 needs to be improved for the sake of clarity and preciseness. I have listed my minor concerns below, and I believe the authors will have no problem to account for these while producing a revision.
Detailed comments:
Lines 60-64: It would be good to provide more information of the empirical method used in the Gridpoint Statistical Interpolation assimilation system. It does not need to go into much detail, but it should be in balance with what is said about the scattering index -based method.
Lines 137-140: Details of retrieving LWP and IWP are missing. Please either provide one or two references on how these are obtained, or describe the methodology in detail.
Lines 177-178: Please provide a reference to the ATOVS and AVHRR Pre-processing Package.
Line 225: "strongest typhoon ... since 2019" - I guess you mean some earlier year than 2019 here?
Line 233: "Figure 4 shows" - please replace by "Figure 3 shows".
Lines 235-236: I would suggest being more specific by saying "The 700 hPa air temperature is lower ...".
Figures 4 and 5: Presumably the cross sections are taken along the black lines indicated in Fig 4b, but I am not sure if this is explicitly said anywhere (it should be, in my opinion)?
Lines 249-251: It is concerning that the assimilation of extra observations causes negative differences in the air temperature analysis. The new cloud detection method apparently passes more data through as clear, but this does not necessarily mean that the data
really is clear - so I wonder if the negative temperature difference is because of the new method does not detect all clouds? Could you please include some more discussion on how the incorrect assimilation of cloud-affected AMSU-A channel 5 data would appear in the temperature analysis increment?
Lines 234-235: Do you mean to say "The geopotential height ... is decreased"? It would make more sense than saying it's increased.
Lines 266-267: Don't you mean north of the typhoon centers, rather than south? Also, it is not clear to me how this is consistent with Figure 5.
Line 286: Please replace 157-h by 123-h.
Line 298: Where do you obtain the Final analysis (FNL) data?
Lines 301-302 vs. Figure 8: The colour scheme as described in the text is inconsistent with legends shown in the figure. Please double-check which is correct and modify accordingly.
Lines 310-313: What is the verification domain in Fig. 8?
Line 355: Is there a reason to refer to NOAA-18 in particular here? I understand the experiments in the work used NOAA-19, Metop-A, and Metop-B only.
Author Response
Please see the attachment
Author Response File: 
Author Response.docx
Reviewer 2 Report
The manuscript is well written and robust and could be published with minor changes.
More specifically, authors should further investigate or justify in more detail
(a) the reason they get improved results at about 110 hours, as stated in line 315 and
(b) why after 144h the differences in ACC for ODS and CTRL simulations did not pass the significant test as stated in line 348
Author Response
Please see the attachment.
Author Response File: Author Response.docx
