Next Article in Journal
Genesis, Maintenance and Demise of a Simulated Tornado and the Evolution of Its Preceding Descending Reflectivity Core (DRC)
Previous Article in Journal
Combinatorial Optimization for WRF Physical Parameterization Schemes: A Case Study of Three-Day Typhoon Simulations over the Northwest Pacific Ocean
Article Menu
Issue 5 (May) cover image

Export Article

Open AccessArticle

Peer-Review Record

Observation and Simulation Studies of Three Types of Wire Icing

Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Atmosphere 2019, 10(5), 234;
Received: 26 March 2019 / Revised: 20 April 2019 / Accepted: 23 April 2019 / Published: 1 May 2019
(This article belongs to the Section Climatology and Meteorology)

Round 1

Reviewer 1 Report

The paper describes results from field observations of wire icing. Icing conditions are classified to the three different meteorological settings, freezing rain, snow and supercooled fog. Icing growth rates are presented for each of the processes and related to meteorological parameters. Moreover shedding of the ice is described and related to the ambient conditions. Finally, three existing models for the different icing processes are described and applied to the data set. The models seem to fit well to the measurements.  

The methods applied are sound and the data seem to be of very good quality. This makes it a valuable study to enhance the understanding of wire icing and also for validation of icing models. The paper is principally written well but it requires some clarification, additional information and reformulation in some sections. For example some sentence highlighting  aims, value and originality of the presented study is lacking and more information on study site, study design and methods are required. More suggestions can be found below.


L 15-17 Please provide more details on the observations

L17-18: This sentence is unclear; what is meant with “compared among...”? The “simulation” needs to be introduced – what model(s), what input, what output, what is the aim and use of the simulation...

L19 “the average measured icing growth...”

L20-22: In my opinion the finding on timing of icing rates is not very sound. I suggest to remove it (see below)

Section 1 Introduction: At the end of the sentence the research gap should be carved out. Moreover, the novelty and originality of the presented work should be highlighted. What is the step forward? Why do we need this study?

L86 A further important application of the combination between weather models and icing algorithms is spatial distributed information on icing risks as e.g. presented for Switzerland by Grünewald et al 2012; Other similar studies can be found within this paper.  

Grünewald, T., Dierer, S., Cattin, R., Steiner, P., Steinkogler, W., Fundel, F., & Lehning, M. (2012). Mapping frequencies of icing on structures in Switzerland. Journal of Wind Engineering and Industrial Aerodynamics, 107-108, 76–82.

Section 2: This section should be enlarged to present the site and the methods in more detail. I suggest to present a map, a sketch or pictures of the experimental site. Please also provide some more information on the climatic conditions of the site. Detailed information on the experimental setup is also missing.

L 114 Unclear: what are a and b, which units? A sketch might help to illustrate a,  b (and also D and phi

L116 Unclear: how was the ice measured? What was measure?

Figure 1: Legend: lightblue is missing in the legend. I would also suggest to display the “type of icing” colors in the legend instead of the caption. Freezing rain looks purple to me, not blue.

Fig 1a: What is the reason for the decrease of ice thickness at Jan 24?

Section 4: It is not clear which event or which case are described here (in relation to Fig 1 and Table 2).

L175 I suggest “reduce” instead of “avoid”

L177 Based on what criteria was the classification performed? What about transitions (e.g. between rain and snow) or mixed conditions (snow and fog)?

Figure 2: It is not clear which case(s) or period(s) are shown here. Why not using solid lines instead of points?

L196-198 I understand what you mean but the sentence does not read well and should be reformulated.

L  196-202 It is stated that icing growth increases with decreasing temperature. Then it is said that snow flakes rebound at cooler temperatures. This is contradicting the trend. Then the authors suggest that the mixture of snow fall and fog might be a reason for the increase at colder temperatures. This might well be true but the argumentation is confusing the reader and should be reformulated.

L199 how do you know that the snow flakes rebounded? Is this based on your observations? But how could you observe such rebounding if – in reality the mixture with fog resulted in sticking of the flakes?

L203 fog is the source for icing but I think one should not name it “precipitation type”

L203-211 this section is hard to read, the sentences are very long and the reader needs a while until realizing which sentence refers to which time window; I suggest to reformulate to shorter sentences and to mention the period in the beginning of each sentence: e.g.: “At 77-81h,…”  

L227f: Unclear: the growth rate was lower that which earlier observations? Values from the references? How big is the difference?

L233ff and Fig 3b: I do not see much value in analyzing diurnal variations, as presented here. Icing depends on the meteorological conditions and not on daytime. Time may not be seen as a proxy for these conditions. It is absolutely clear that temperatures drop after sunset and that this favors icing but so what? I suggest to remove this analysis.  

It would be  interesting to analyze if icing accelerates with the duration of an icing event following the increased radius of the iced wires which should alter collision efficiencies of droplets. Would the data set allow to answer this question?

L254 Please state which correlation coefficient is presented (Pearson?) Information on the number of observations should also be given.

L256 which significance test?

L257f This is contradicting to what was stated earlier (L199) that snow flakes tended to rebound at cooler temperatures and that only the combination with fog made the increased icing.  

L261ff The wind will also have an effect on the energy balance (turbulent fluxes) of the wire surface which might affect freezing

L286 I suggest to write “changed weather conditions” instead of “improved”

Figure 5b: No need for exponential notation of the y-axis labels

Figure 5d and e: Color code for temperature/wind and FLWC/precipitation are missing

Caption: “air temperature” instead of “temperature”

L306-341: This paragraph is hard to follow. It is not clear if and when the description relates to Fig 5 and when to Fig 6 or Table 3. The value of Fig 6 is not clear. It should also be stated with process (melting, gravitation…) caused shedding in the described cases.  Please rewrite this paragraph.

Section 8.1 please provide the units for the parameters used in the models (e.g. U, P, L, M…)

L347: Chang to “the ice thickness (W)….” and also give the units.

L 360: change to “the ice diameter D…”

L372 what are “dry snow processes”?

L373 the wetness of dry snow is by definition zero

L383 I don’t understand this sentence “therefore … dry snows”. I do not see the relation between Makkonen’s finding and the 0.1 used for dry snow. What is “the two dry snows”

L385 something wrong here: beta = 1/U > 1/9.1m/s = 0.1 and not 0.01; same for 0.03

L413/414 I suggest to add direct references to Fig 8 a, 8c and 8d in the text (now references with cases).

L420 should be “10:00-06:00”

Sect 8.4. it should be discussed in more detail when the model works well and when it does not and why. E.g. the first day of case 2 much icing is measured but nearly none is calculated. Why? Or case 2 31 Jan 12:00 the model calculated much icing but no icing is observed. It would also be worth to think about introducing some measure to rate model performance. E.g. a scatterplot modeled vs measured icing could help to illustrate model performance.

Fig 9: no need for this figure. All content is already visible in Fig 8;

L427: how does the model decide when it needs to simulate freezing rain/snow/fog? What about transitions?

L426: I would not call this a new model. It is just a combination of existing models with very small adaption! It is also not very complex I would say.

L430: should be “on 30 Jan”

Section 9: The conclusions section is quirt poor. It only summarizes what was described before. But a conclusions section should also provide some grading of the results, some outlook, what has  been/ can be achieved based on the results of thestudy (what are the new findings?) Where are the limitations, what additional research will be required? It may also be stated that the model seems to capture measurements well. Will that also be the case for other data sets and other study sites? What can the model now be used for? Operationally? hby whom and how?…

L455 Please add some connecting passage. First you talked about measurements. Now it’s about the model.

L457: Sounds as sticking efficiency was measured. Please state clearly that it was just calculated from wind speed. Check the numbers. And again: this is not a new model;

Author Response

Please find our point-by-point response in the uploaded Word file.

Author Response File: Author Response.docx

Reviewer 2 Report

Review comments for “observation and simulation studies of three types of wire icing” by Wang et al.


This manuscript thoroughly studied 4 observed wire icing cases in one particular location in Jiangxi, China with favorable meteorological conditions (high moisture, mountainous?) for wire icing. All four cases involved mixed precipitations (freezing rain, supercooled fog and snow). Two major conclusions that I found are concrete are: (1) icing growth rate is positively correlated with precipitation rate and wind speed, although the correlation is loose for freezing rain condition compared with other two scenarios; (2) with tuned “sticky efficiency” parameter to account for the dry snow v.s. wet snow specification in the original model, adding three icing models together to account for different scenarios would produce the best simulation results compared to observations, albeit the fact that ice shedding process still lacks a good model to mimic the observed situation. 


I found this research work has solid basis, and is conducted and described mostly in a strict sense and written mostly in a clear way. My general suggestion is that the authors can consider adding a topic sentence or two at the beginning or ending of each section (since this is a relatively long manuscript) to clarify the main purpose of each section and the major novelties of your finding. This happens in some places, but was missing in a lot of the places, so I’m lost in all numbers before realizing why you conduct certain steps of work. Secondly, I strongly suggest you re-evaluate or delete the diurnal variation part of study, since the data points are so few to generalize any concrete conclusions. There are some of other minor recommendations that I listed below. I think this manuscript deserves a final publication with minor revision following my major and minor suggestions. 


Minor suggestions:


Abstract: need 1-2 sentences at the very beginning to state the importance of your work under the big picture.


L33: add “For example,” before “Due to”.


L56-58: without specifying details about the environment in Poland, “southeast area” doesn’t help you understand your question better. I suggest to remove this paper from the literature review, and only list as a reference.


L67: what do you mean by “preparation and maintenance period”? Do you mean pre-icing and stable periods?


L106: “and it is rich…” -> “and therefore it is rich in moisture sources”. 


L108: Elevation is 1164.5 m: is this site in mountainous region? Do you consider in the topographic effect in your research? 


Table 1: Please consider add an extra column to the right showing the instrument accuracy level. 


Fig. 1: Can you consider add one more line to show the relative humidity (with respect to water since supercooled fog occurs for most of the time) so we get to know the onset threshold of RH?


L184-185: I don’t understand: why you think it’s a combined effect, rather than weak dependence or no dependence at all? This argument needs some more elaboration.


Fig. 2: I don’t understand how do you line up the calculated icing growth rate (and other variables) in time, while can separate them two three categories (1-36 hr: freezing rain, 37-66 hr: snow and 67-11 hours supercooled fog)? The x-axis is not defined as the time from the onset of icing for each case?


L219-220: Reason#1 is contradictory with L184-185, where you stated icing growth rate is poorly correlated with precipitation rate, although I do see some rough correspondence between Fig. 2d and 2a for the “freezing rain” scenario.


Fig. 3b: please add the errorbar to each datapoint, and link the data for one type together (semi-transparent shades to indicate the errobar might be useful). Right now it’s very hard to read the diurnal variation and its amplitude from Fig. 3b. 


Paragraph L233-247: since only 4 cases are studied here, it’s really hard to draw any conclusion of the diurnal cycle I think. Unless you can show the errorbar in Fig. 3b is much smaller compared with diurnal cycle amplitude, I suggest you delete the discussion around diurnal cycle.  I honestly don’t see an apparent relationship between Fig. 2b and 2a, so don’t expect to find a strong diurnal cycle under all three scenarios.


Fig. 4: Please also add the linear regression against freezing rain and the R^2 or R value in the figure. Also, please include the R^2 or R value in the right panel: it’s good to know how spread the correlation is, even the correlation is not statistically significant. 


L258: icing growth rate increased faster -> “icing growth rate increased” or “icing grew faster”.


L270-271: trend relationship -> positive relationship


Figure 5d and e: please use corresponding color on the left and right y-axes to show which variable corresponds to which y-axis.


L309: stability -> stable.


Section 8: Please start with a summarizing sentence before 8.1.1., e.g., “Icing models have been built previously for different conditions, e.g., freezing rain (ref), wet snow (ref), supercooled fog (ref). In our observed 4 cases, mixed conditions happened, so in this section, we will further explore what are the best icing model and parameters to use for our cases. “ In this way, readers will keep on track with your thoughts. I didn’t realize your purpose until reading Line 378. 


8.1.1, 8.1.2 and 8.1.3.: The variable names are inconsistent among different models, e.g., “w” seems to stand for ice accumulation rate (area or mass or diameter? It’s not explained in the text), while “w” represents the wet snow mass concentration in the air in the second model. It’s very confusing. 


L383: therefore? How do you determine the dry snow density is 1/10 of the wet snow density (I assume this is how you get beta=0.1/U)?


One question for Fig. 8: So in all these simulations, you didn’t “tune” any input parameter other than change beta to 0.1/U and 0.01/U as referred in Fig. 7, correct? 


L415: a technical question: how do your water content measurement instruments  separate super-cooled FLWC, SWC or freezing rain LWC? 


L443-446: I don’t quite like the diurnal variation part and suggest you to delete this conclusion. It’s really weak conclusion based on only four mixed-precipitation cases. 

Author Response

Please find our point-by-point response in the uploaded Word file.

Author Response File: Author Response.docx

Reviewer 3 Report

See attached review.

Comments for author File: Comments.pdf

Author Response

Please find our point-by-point response in the uploaded Word file.

Author Response File: Author Response.docx

Atmosphere EISSN 2073-4433 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert
Back to Top