A measurement campaign to assess sources of error in microwave link rainfall estimation

van Leth, Thomas C.; Overeem, Aart; Leijnse, Hidde; Uijlenhoet, Remko

doi:https://doi.org/10.5194/amt-11-4645-2018

Articles | Volume 11, issue 8

https://doi.org/10.5194/amt-11-4645-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/amt-11-4645-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 11, issue 8

Research article

|

09 Aug 2018

Research article |

| 09 Aug 2018

A measurement campaign to assess sources of error in microwave link rainfall estimation

Thomas C. van Leth, Aart Overeem, Hidde Leijnse, and Remko Uijlenhoet

Download

Final revised paper (published on 09 Aug 2018)
Preprint (discussion started on 01 Dec 2017)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Interactive comment to 'An urban microwave link rainfall measurement campaign' by Leth et al.', Anonymous Referee #1, 01 Jan 2018
- AC1: 'author response to anonymous referee #1', Thomas C. van Leth, 16 Feb 2018
RC2: 'Review of T C Van Leth et al. “An urban microwave link rainfall measurement campaign”', Anonymous Referee #2, 10 Jan 2018
- AC2: 'author response to anonymous referee #2', Thomas C. van Leth, 16 Feb 2018
RC3: 'Referee Comments', Anonymous Referee #3, 15 Jan 2018
- AC3: 'author response to anonymous referee #3', Thomas C. van Leth, 16 Feb 2018

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Thomas C. van Leth on behalf of the Authors (16 Mar 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (26 Mar 2018) by Mark Kulie

RR by Anonymous Referee #2 (14 Apr 2018)

RR by Anonymous Referee #1 (25 Apr 2018)

Suggestions for revision or reasons for rejection

The revised manuscript mostly reflects the specific comments raised by the reviewers. The authors took effort to quantify phenomena occurring at experimental microwave link which were before described qualitatively and often in the subjective manner. Nevertheless, there are still some parts which should be clarified and formally better described. Unfortunately, the readability of the whole manuscript, and in particular the results section, which lacks clearness and reasonable conciseness, has not improved. It is often very difficult to follow the reasoning and extract from the paragraphs what it is the message the authors want to present. Some examples and suggestions are provided in the specific comments; nevertheless, especially comments on language shortcomings are not exhaustive. The authors should once more carefully read through the whole manuscript, particularly the result section, and do editing where appropriate. A stylistic and language proofread by an English native speaker might also help. The presentation quality is the main reason why I recommend major revisions and not just minor.
Finally, the authors should consider revising conclusion section. Introducing paragraph of the section is devoted to the effect of drop size distribution on power-law fit, however, I do not see any scientifically novel findings in this particular issue. The power-law fitting to DSD is even not a part of results section. In contracts, e.g. the attenuation dependence on the temperature (the whole subsection 5.4 is devoted to this), is not mentioned in the conclusions at all, despite the presented research clearly improves current understanding of baseline variability sources. Similarly, conclusions based on result subsections on solid precipitation (5.3), dew and fog (5.5), Clutter (5.7) and Compound phenomena (5.8) are not provided. I personally also miss more specific recommendations either how to cope with the effects described in the manuscript when estimating rainfall (e.g. how to use redundant information from more CMLs) or what lines of research are still needed to better cope with these effects. It should be noted, that the criticism on the conclusion section has not been raised in previous steps of the review process.

Specific comments:
P2L37: part of the sentence is missing.

P3-4 Theoretical background section: This section focuses only on the relation between rain induced attenuation and DSD, whereas the scientifically novel results of the manuscript are rather connected to the other phenomena influencing total loss such as dew and fog, wet antenna attenuation, humidity and temperature, clutter, mixed precipitation. The part related to power-law approximation might be, in my view, even replaced by a reference on a literature, as the results in the revised paper are now mostly based on comparison between observed attenuation and attenuation derived directly from DSD.

The literature review on the other phenomena is provided in the introducing paragraphs of results subsections to some extent (e.g. P11L13-21) and partly also in the introduction (wet antenna).
It would either make sense to rename the section to make it clear it is about DSD or preferably move the fragments of theoretical background connected to other phenomena from the result subsections to this section.

P4L19-20- Please, rephrase the sentence.

P7L18-21: This part is difficult to read. Do you average other quantities than DSD derived attenuation and rainfall? Being specific would improve the clearness a lot. Anyway, isn’t it at the end the same as averaging the DSD over the link path with weighted mean and then deriving either path-averaged attenuation or rainfall from this path-averaged DSD?

P8L1: Please, replace ‘somewhat’ with some quantitative measure (by XY %).

P9L4: The term ‘measure of the fitness’ sounds strange to me. Moreover, the rest of the sentence describes procedure rather than the measure itself. Maybe the term ‘reliability’ would be more appropriate here.

P9L17-21: Please rephrase. What is exactly meant with ‘it could not have been resolved from magnitude alone’?

P9L40-41: Have you experienced also longer dying times as Schleiss et al., (2013) did? (I am just curious)

P1L1: Bad English, please rephrase.

P10L14: reasonable correlation means strong correlation? Please, specify.

P10L29: What is meant with ‘but drops afterwards’ with respect to relation between visibility and rainfall intensity? Please, rephrase.

P11L13-21: Consider moving this paragraph to the Theoretical background, as this is not result.

P12L4: Do you mean more detailed analysis of your data, or really intent to research this phenomenon in general? The generic description of snow melting might be really challenging (and not so easy to study in Netherlands).

P12L23-24: Wrong syntax of the sentence.

P12-25-30:
The paragraph contains interesting findings but it is really uneasy to read (what is meant e.g. with ‘and for the most part a linear regression makes for a good fit’?). The description of residual errors is in my view rather redundant (at least without using it further for expressing uncertainty of regression coefficient). I do not see the link to the main message of the paragraph which is at least as I interpret it i) strong linear attenuation-temperature relationship during dry weather vs. milder (or none) during dry weather ii) that this relationship is hardware specific. I think that restructuring of the paragraph is really needed.
Something like this might help:
There is a strong negative correlation between temperature and attenuation (r = -0.80 to -0.92) for the periods with relative humidity below 90%. The slope of the linear fit is substantially lower for Nokia link (-0.024 dB K-1) than for the other devices (between 0.1 and 0.2 dB K-1), although Nokia link operates at the same frequency and polarization as one of the RAL devices. The linear relationship between attenuation and temperature is milder for the periods with higher relative humidity (XXX to YYY) and completely disappears in the case of Nokia device. This indicates that the attenuation-temperature relationship is clearly more influenced by link hardware than their frequency or polarization.

P12L40-42: Bad English, please rephrase.

P13L10-15: Difficult to read, please rephrase.

P13L40-42: This is unclear. How did you come to 1.5 dB and what do you mean with ‘attenuation by fog droplets themselves’? Do you mean water vapor attenuation along a link path, or something else? Please clarify.

P14L11: What it is exactly meant with ‘antennas were wetted until saturation’ and how did you recognize this saturation? Based on attenuation observations, or based on observations of antenna surface?

P14L20: ‘…decay time 3 minutes.’ 3 minutes to ‘half-life’ value or to some other quantile of max wet antenna attenuation?

P14L22: Please quantify (provide e.g. range) the initial drop of signal level.

P14L26: Minda and Nakamura (2005), do not claim/conclude that water layer on the antenna surface has an uniform layer. Their model is purely empirical. Please correct. Consider also comparing your results with Schleiss et al., (2013) who reported substantially longer drying times than Minda and Nakamura (2015).

P15L36: It is not clear how you do come to the conclusion that fog related attenuation is caused predominantly by antenna wetting. Please provide evidence.

P15L40: What it is meant with ‘net radiation flux is away from the surface’? Please rephrase.

P16L1-15: Consider removing (or moving) the paragraph about DSD. It describes known and previously well documented findings.

P16L21: ‘…additive and multiplicative bias seems quite consistent’. There is no quantitative inter event evaluation of additive and multiplicative bias provided in the results section. Please, provide this information in result section or reformulate this conclusion to be sufficiently supported by presented results. Furthermore, replace ‘quite consistent’ with a formulation indicating better (if possible quantitatively) to which extent both types of biases vary within the event.

P16L38-40: This is very general recommendation which is difficult to apply especially when additional rainfall data is not available (i.e. when one could potentially benefit from microwave links the most). I believe that more specific recommendation could be provided based on the presented results, e.g. on use of links having same/similar path, etc.

P16L40-42: Yes, however, can we recognize such link without having reference rainfall data? Could e.g. variability of baseline during dry weather period tell us something about link reliability in terms of rainfall estimation?

Rerferences:
Minda, H., Nakamura, K., 2005. High Temporal Resolution Path-Average Rain Gauge with 50-GHz Band Microwave. J. Atmospheric Ocean. Technol. 22, 165–179. https://doi.org/10.1175/JTECH-1683.1
Schleiss, M., Rieckermann, J., Berne, A., 2013. Quantification and Modeling of Wet-Antenna Attenuation for Commercial Microwave Links. IEEE Geosci. Remote Sens. Lett. 10, 1195–1199. https://doi.org/10.1109/LGRS.2012.2236074

Hide

ED: Reconsider after major revisions (27 Apr 2018) by Mark Kulie

AR by Thomas C. van Leth on behalf of the Authors (13 Jun 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (27 Jun 2018) by Mark Kulie

RR by Anonymous Referee #1 (02 Jul 2018)

ED: Publish subject to minor revisions (review by editor) (03 Jul 2018) by Mark Kulie

AR by Thomas C. van Leth on behalf of the Authors (09 Jul 2018) Author's response Manuscript

ED: Publish as is (11 Jul 2018) by Mark Kulie

AR by Thomas C. van Leth on behalf of the Authors (19 Jul 2018)

Short summary

We present a campaign to address several error sources associated with rainfall estimates from microwave links in cellular communication networks. The set-up consists of three co-located links, complemented with reference instruments. We investigate events covering different attenuating phenomena: Rainfall, solid precipitation, temperature, fog, antenna wetting due to rain or dew, and clutter.