Long-Term Observations of Microwave Brightness Temperatures over a Metropolitan Area: Comparison of Radiometric Data and Spectra Simulated with the Use of Radiosonde Measurements

Round 1

Reviewer 1 Report

Comments:

Compared to radiosondes, the observation of ground-based microwave radiometer (MWR) can cover the complete diurnal cycle at high temporal resolution in minute level. Nevertheless, the uncertainty of MWR observation especially in precipitation skies weakens its advantage. The manuscript remotesensing-1196675 investigate the difference between observed and simulated microwave spectra obtained from long-term observations of MWR and radiosonde, and this is good for better understanding the uncertainty of MWR and may give hints to improve the measurement accuracy of MWR. The authors indicate that different absorption models likely only change the sign of bias and a spectroscopic nature of the observed differences is responsible for it. In my opinion, the bias values in 51-54 GHz also present dependence on absorption model, and may be related to the difference in molecular oxygen absorption of model. Also, some discussion and conclusions on the difference in 51-54 GHz with elevation scanning data are not solid. Therefore, major revision should be made for this manuscript before considering its publication in Remote Sensing.

 

Detail comments:

(1) The abstract needs to be improved according to the changes in the revised version.

(2) Line 94: “radiozonde” should be “radiosonde”.

(3) Line 98: “Observations.” Should be “Observations”.

(4) Lines 115-119: In operational observation, the accuracies of temperature and humidity are 0.1 K and 1%, respectively, while in this study the sounding temperature and humidity are respectively with an accuracy of 1 K and 7-15%. As the used temperature and humidity present a relatively poor accuracy with respect to that in operational observation, does the difference make impact on or bring uncertainty in the simulated brightness temperature by using sounding data?

(5) Line 137-160: English spelling and grammar should be checked carefully.

(6) Line 196-197: “In [MPS11]…”, what does “[MPM11]” mean here?

(7) Lines 232: “…spectra simulated spectra…”, please check the sentence.

(8) Line 236: “…, One can see…”, “One” should be “one”.

(9) Line 242-243: “The |MD| in these channels are higher than corresponding SD for all models.”, what does this sentence mean?

(10) Lines 244-246: “…SD for all models is significantly less (at least 3 times) than overall standard deviation of brightness temperature…”, what is “overall standard deviation of brightness temperature”? Is that the standard deviation of observed brightness temperature with HATPRO?

(11) Line 246-247: “This allows us to treat the observed differences as small.”, how to draw this conclusion from Figure 2? In my opinion, I just see that the differences of SD for all models are small.

(12) Line 251-254: in Figure 2, the SD of observed brightness temperature is in range of 4-14 K in 22-32 GHz and 4-11 GHz in 51-58 GHz, indicating large variation with frequency in the accuracy of observed brightness temperature, which is also much higher than those of models, why?

(13) Lines 260-262: “…within the -1 +1 interval one can…”, “-1 +1” should be “[-1, +1]”, and English spelling and grammar should be checked for this sentence.

(14) Lines 266-277: as shown in Figure 3, the scanning data for zenith direction is almost the same as zenith measured data but not in 51-54 GHz, and the difference in 51-54 GHz likely depends on models as Figure 1 presents the difference in MD of different models; moreover, if the influence of observed microwave data averaging is negligible, then for the same model, the tendency should likely be the same for all elevation angles, but it is not in 51-54 GHz especially for MPM2 and MPM2a, which implies that the influence may come from both simulation and observation; furthermore, it is interesting that the difference for the lowest elevation angle is small than that in other elevation angles in 51-54 GHz, which is opposite to that in 22-32 GHz, why? Therefore, I think that the conclusions and explanations in this paragraph is not solid, and both them need be more satisfied.

(15) Lines 278-283: why the SD for the lowest elevation angle is small than that in other elevation angles in 51-52 GHz, which is opposite to the situation in other frequencies?

(16) Line 337-338: I do not think that the assumption is reasonable as Figure 1 and Figure 3 show the differences of various models in 51-54 GHz, and this is one of the concerns of this paper.

(17) Line 363-366: as mentioned in line 105, HATPRO measures brightness temperature in 51-60 GHz with a band width of 0.23-2 GHz, the authors should present the used band width in HATPRO in Figure 6.

(18) Line 369-389: it seems reasonable that possible explanation of results at 22-32 GHz is the influence of cloud contamination during the scanning, as air is usually more moisture in low levels of troposphere than high levels. However, the explanations on the results in 51-58 GHz need be more satisfied. The first 3 channels in V band are less opaque and farther from the oxygen absorption peak, with atmospheric penetration stronger than other channels, thus they are vulnerable to the influence of clouds and rainfall and the brightness temperatures in these channels are more sensitive to elevation angle than in other opaque channels. Additionally, the three models used in this study only differ in molecular oxygen absorption, and this difference also probably leads to discrepancy in simulated brightness temperature in 51-54 GHz. Maybe, the distance between HATPRO and sounding station also make impact on the results, especially HATPRO is located in the center of the metropolis of Nizhny Novgorod, in case of the sounding station is located in suburban area, the horizontal temperature gradient between HATPRO and sounding station will be obvious under the urban heat island effect, and it may be more significant for humidity, and this may partly explain the difference for elevation scanning data is larger in 22-32 GHz than in 54-58 GHz as shown in Figure 4. Therefore, the author should give more consideration to the influencing factors of the results.

(19) The conclusions need to be improved according to the changes in the revised version.

(20) Careful check on English spelling and grammar are necessary besides those mentioned in comments.

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper creates and analyzes a dataset of microwave and radiosonde observations over a metropolitan area.

General Comments:

The research is interesting and the amount of work and research done is clear from reading the paper. However, the writing needs to be corrected. For example, the authors seem to confuse the use of "a"  with "the" in many places.

Also, there are several acronyms that are missing. The authors need to remember that the paper should not be for a selected public, but to researchers in general that may be or not familiar with the area of analysis.

The same goes for some techniques. For example, when talking about the W matrix. Although I agree that going into too much detail is not of the scope of the paper, a little more explanation helps to set a clearer picture of what the authors are doing.

 

Detailed Comments:

See attached paper with comments.

 

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Please see the attachment.

Comments for author File: Comments.doc

Author Response

Dear Reviewer,

Thank you a lot for taking the time to review our manuscript and for constructive suggestions for the paper.

 

 (1) Line 156: “…limit ourselves with to…” should be “…limit ourselves to”.

The mistake was corrected; see line 156.

 

(2) Line 172: “…high (<0.01 K) accuracy…” should be “…high accuracy (<0.01 K)”.

Corrected. Line 172

 

(3) Line 177: “…widely known [30-32], widely known and used…” should be “…widely known [30-32] and used…”.

The mistake was corrected. Line 177

 

(4) Line 206: “… be treated using a…” should be “…be treated by using a…”.

Corrected in accordance with your comment. Line 206

 

(5) Line 247: “As a result of the post-processing the dataset consists of 642 cases” should be “As a result of the post-processing, the dataset consists of 642 cases”. There are many similar situations in the paper, such as lines 153, 180-181 and 471, please check them carefully.

The punctuation was changed accordingly.

 

(6) Line 261: “…see little difference…” is better than “…see no difference…”.

Changed in accordance with your suggestion. Line 261.

 

(7) Lines 264: Is the range [-0.8K, +0.6K] or [-0.6K, +0.8K]? Please check it.

The correct variant is indeed “[-0.6K, +0.8K]”. Corrected. Line 264.

 

(8) Line 438: “64-60 GHz…” should be “54-60 GHz”, right?

Indeed it should be “54-60 GHz”. Corrected. Line 438.

 

With respect,

M.V. Belikovich, M.Yu. Kulikov, D. S. Makarov, N. K. Skalyga, V.G. Ryskin, A.A. Shvetsov, A.A. Krasil’nikov, S.O. Dementyeva , E.A. Serov and A. M. Feigin