A Novel Electromagnetic Wave Rain Gauge and its Average Rainfall Estimation Method

Round 1

Reviewer 1 Report

Review comments on remotesensing-953573

Title: A Novel Electromagnetic Wave Rain Gauge and its Average Rainfall Estimation Method

Comments: This manuscript developed a new electromagnetic-based rain gauge system to measure average rainfall in a small spatial domain. The core part of this new methodology is a dual-polarization radar at K-band frequency. The average rainfall intensity was derived using mean values of multi-scan reflectivity and specific differential phase at different elevation angles. The rainfall estimation performance was compared with the results obtained from rain gauge and Parsivel disdrometer at several locations in Korea. Overall, this topic fits the scope of Remote Sensing very well. The manuscript is well written, and easy to follow. Given the continuing challenges in measuring rainfall and its space-time variability, I recommend publication of this manuscript after minor revisions.

1. In line 47, the author may want to refer to the newer version of CSU algorithm (i.e., Chen et al., 2017; JHM, DROPS2.0).
2. In Fig. 3, can you add more details about the scan elevation angles and explain different color contours in the caption?
3. Near line 131, in the calibration how far is the disdrometer from the EWRG? Are they exactly collocated as shown in Fig. 7?
4. In line 142, can you elaborate why these elevation angles were selected?
5. In lines 154-155, please explain more about the simulation, including the assumptions made in the simulation such as temperature and raindrop axis ratio, etc.
6. The results in Table 2, Figs. 10, 11, and 14are very promising. But can you provide more evaluations of rainfall accumulations at different time scales (e.g., hourly rainfall products)? 
7. Can the author comment on if the EWRG can potentially replace rain gauges? Also, is it costly effective? Can we obtain other information from this system in addition to rainfall estimates?

Author Response

Please see attached response.

Thanks for your time and kind review.

Author Response File: Author Response.pdf

Reviewer 2 Report

General Comments

This paper presents first results from the operation of a VAD scanning polarimetric micro rain radar to estimate area average rainfall. Profiling mini radars can cover areas where scanning weather radar cannot measure due to terrain effects as mentioned in section 2.2. Also, they can be a calibration reference for weather radars with about the same volume averaging. For these reasons, the paper could be useful to the readers. However, it is not clear why not use a commercially available micro rain radar (MRR), which the author does not mention but it is simpler with less cost and estimates the drop size distribution (and then rainfall rate) with the Doppler method.

Except this question, a main issue is that the symmetry axis of rain droplets is near vertical (with small up to 10 deg canting angle). Thus, Kdp (and Zdr) should decrease with elevation angle maybe by a factor of 3 from 30 to 60 deg elevation scan, while the effect on Zh is a lot smaller(what do the observations show?) This is not an issue for weather radars scanning at low elevations angles. This has not been taken into account in the proposed method. In addition, melting layer effects or generally varying Zh and Kdp with height (e.g. near cloud top or base like in Figs. 8,13) are not considered in the averaging per elevation.

Thus, the paper should be rewritten taken into account these important issues before resubmission, which probably cannot be done in the short time period for revision of Remote Sensing journal.

Specific Comments

l. 50-51: Why it is difficult to apply this method to EWRG? Actually, the difference of the method described in this paper is the averaging per elevation.

Fig. 4: EWRG Zh peaks are a lot smaller than Parsivel. Probably because attenuation correction in rain, which is quite large in K-band, was not applied?

l. 152-153: Why not use also Zdr, which a widely used polarimetric product? Against which reference rainfall data Eqs. (1) and (2) were estimated? Rain gauges also look to give quite different results than Parsivel e.g. in Figs. 10-11 (maybe a problem in Parsivel software). As it was mentioned in general comments the constants in these equations should vary with elevation angle. Finally, scatter plots for estimation of constants a,b,c would be very useful to check the accuracy of approximations in Eqs. (1) and (2).

 

Author Response

Please see attached response.

Thanks for your time and kind review.

Author Response File: Author Response.pdf

Reviewer 3 Report

This manuscript proposed a method to estimate average rainfall at multi-elevations using an electromagnetic wave rain gauge system. The author conducted the proposed method in two main rainfall events at Geoje and Yeoncheon to evaluate the average rainfall estimation. Compared with rain gauge and Parsivel disdrometer data, the promising results proved the ability of the novel EWRG to measure rainfall. The article is well-organized, but still needs further comparison to support the results. I have several comments that should be addressed before the paper could be published.

 

Line 17-19: The study only focused on heavy rainfall events, so does the method work effectively under light rainfall conditions?

 

Line 39-42: Perhaps you could cite some of the work about the EWRG.

 

Figure 1: Please label what the colors represent for in the figure clearly.

 

Line 163: It would be better if state the elevation of the two sites and what the surroundings are around the two locations.

 

Figure 10: Why the EWRG data was increasing linearly from 3 to 5 and 9 to 11’o clock? The other data were 0 mm in Fig. 10(a), and according to accumulated rainfall, the EWRG also didn’t increase (Figure 14 as well). In Fig. 10(a), why the rain gauge data (0.5 TB) was so high at 2:30? Was it a breakdown or any other reason?

 

Line 197-200: The underestimation of the EWRG also demonstrated the limitation of the system when measuring continuous rainfall in Figure 11.

 

Line 246: In section 4.3, please explain why you choose these metrics and what did the results tell us?

 

Line 274-277: Please add more conclusions drawn from section 4.1 and 4.2, including the limitation of system errors in the EWRG.

 

Author Response

Please see attached response.

Thanks for your time and kind review.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The author gave answers to the comments and made changes in the paper. However, many explanations were left in the reply and they were not included in the paper. They should be included before publication.

The author agreed with the issues mentioned in the review and justified the development of this EWRG with the volume averaging made with this method. Actually, MRR is doing an implicit volume averaging with temporal averaging. Still, MRR is not mentioned at all in the paper. It would be interesting in the future to compare EWRG with MRR.

Also, it is still not mentioned in the paper that the difference from Parsivel at Zh peaks may be due to attenuation effects, which have not been corrected. In addition, the author notes that Parsivel data are only used for qualitative evaluation due to instrument problems, but this is not included in the paper in order to explain its difference (underestimation) from rain gauges.

Add also the explanation that the significant reduction of Zh and ρhv with height-range in Figs. 10 and 15 is due to strong rain attenuation as suggested in the reply to comments.

The new text in lines 208-211 needs also some grammar corrections to be meaningful.

Author Response

Please see the attached document.

Thank you for your time and kind comments.

Author Response File: Author Response.pdf