GNSS-RO Refractivity Bias Correction Under Ducting Layer Using Surface-Reflection Signal

Round 1

Reviewer 1 Report

GNSS-RO is an unique and valuable technique for PBL characterization. This paper introduce a new approach of using reflected GNSS-RO signal for GNSS-RO N-bias removal under ducting layer, the advantage of the method is self-correcting, it doesn’t need the assisting observations or model information.

There are some comments, suggestion and questions

While the new approach do performs well in the End-to-End simulation and the selected actual RO cases, however, the usability in practice would be limited. The author also notices that the low signal strength of the reflected signal can cause large variance of the retrieved reflected bending angle. In my opinion, The surface roughness's effect on the signal amplitude of the reflected signal would be underestimated. The average global sea surface wind speed is around 6 m/s, and it means that the sea surface's roughness will affect the reflected signal's strength a lot though the grazing angle is relatively small. The new approach may be useful when the sea surface is calm, but the occurrence rate is low. The approaches proposed by Xie et al and Wang et al may be more useful when the sea surface's roughness can't be ignored.

 

Since Ducting layer height can vary from 0 m to several kilometers and there are several different Ducting types in PBL, Whether the Ducting layer height and Ducting types affect the new approach's utility? The author should state that in the discussion part or any other proper section .

 

The duct could cause the N-bias within the PBL, but it’s hard to say the N-bias is just caused by duct. I recommend the author to explain the relationship between N-bias and duct, if there also have other reason could cause the N-bias, and what extent will the duct affect, or if there are some feature for N-bias which caused by duct?

 

Specific questions to the authors

Suggest the author provide a figure to illustrate the screens utilized in this paper's Multiple Phase Screen simulation work.

 

The time order of the ray path 1,2,3,4 is not introduced clearly in Figure 3.

 

For consistency, it is better to change the sentence "Based on the simulation results, this reconstruction method is able to effectively reduce the N-bias to nearly 0% below the ducting layer." in line 363 to "Based on the simulation results, this reconstruction method is able to effectively reduce the N-bias to less than 1% below the ducting layer."

 

For some formulations, subscripts should be explained.

eq.(3) the subscripts AI, even I can guess it means Abel inversion,

eq.(5) S_T(t), what’s meaning for subscript T,

and some symbols should give the primary introduction:

eq.(13), what’s the ‘argmin’ ?

eq.(14), the a_S is a important parameter in this paper, you’d better to explain it.

 

 

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

This is an interesting paper in which the authors develop a method to address the well-known negative bias in the retrieved refractivity from radio occultation (RO) observations caused by super-refraction (or ducting) in the atmospheric boundary layer. The method uses the reflected RO signals from the Earth’s surface and a statistical least squares approach to provide a constraint on the infinite number of solutions of refractivity associated with a single bending angle profile and thereby select the refractivity profile that best matches the reflected signal. A least squares approach is used to select the solution that best matches the reflected RO signal.

My major comment on the substance of the paper concerns its practical applicability. The method requires the presence of reflected RO signals with sufficiently high SNR to obtain an accurate solution. It is not clear from the paper what fraction of RO profiles meet this condition, especially in the tropics. The authors allude to this limitation in lines 78-79 where they state that more than 45% of the RO signals poleward of 20 degrees latitude contain reflected signals. Furthermore, as indicated in lines 341-352 and lines 376-389, the reflected signal must have a large enough SNR to make the method robust. Thus the percentage of RO signals with reflections of sufficiently high SNR is likely considerably less than 45% globally. They could state explicitly that one of the advantages of RO missions with high SNR (such as COSMIC-2) is that they will have a higher percentage of occultations that produce useful reflection information. Also, it would be interesting if the authors could estimate the percentage of RO profiles from current RO missions that meet the criteria of reflections with high enough SNR to make this method useful. They may be able to make an estimation of this percentage from the ROMSAF reflection database.

The paper should eventually be published because the method is creative and the results are generally convincing. However, the method is complicated and the mathematics are extensive, and therefore, it is important that the paper be written with utmost clarity and precise English and that all the equations and the symbols therein be carefully and completely defined. Unfortunately, this draft contains many grammatical errors and awkward or misleading sentences, which detract from an otherwise good paper. There are too many for this reviewer to do a careful editing. Thus, I recommend that the authors carefully review their paper and rewrite sentences that are either grammatically incorrect, awkward, or difficult to understand. I give a few examples to illustrate this point:

Lines

11        “is” should be “are”

16        results show not shows

41        better wording: “….remote regions with few in-situ measurements.”

49        “..when the bending angle is larger than the curvature of the Earth’s surface, or ducts, the signal will….” should be rewritten something like “when the bending angle is larger than the curvature of the Earth’s surface, ducting occurs and the signal will be trapped.”

51-52   “Here we define the ducting layer as the layer where ducting occurs” is a tautology and unnecessary.

52-53 “In the case of GNSS-RO, which both transmitter and receiver locate outside atmosphere, the signal…” should be rewritten to something like “In the case of GNSS-RO, in which both the transmitter and receiver are located outside atmosphere, the signal…”

There are some equations containing symbols that are not defined, or not completely defined. An example is Equation (1), in which N_T (specifically the subscript T) is undefined. Similarly, the subscript D is not defined in Equation (2) and the AI subscripts in Equations (3) and (4) are not defined. The subscripts T, D and R are undefined in Equation (5). The authors should make sure that all variables are carefully defined. A list or table of symbols might also be useful.

The figures are generally excellent, well above average and the authors are to be commended for this. However, some figures (e.g. 4, 5, 6, 11) are missing the labels (a), (b), (c), etc.

In line 31, COSMIC-2 was launched in June 2019, not July. The authors should reference COSMIC-2 here (Anthes and Schreiner, EOS 2019 https://eos.org/science-updates/six-new-satellites-watch-the-atmosphere-over-earths-equator )

I don’t understand the first two references. They look like examples of how to format references. Delete?

 

 

 

 

 

 

 

Author Response

Please see the attachment

Author Response File: Author Response.docx