Understanding the Challenges of Hydrological Analysis at Bridge Collapse Sites
Round 1
Reviewer 1 Report
Dear authors,
The paper „Understanding Challenges of Hydrological Analysis for Bridge Collapse Sites” could be interesting for many scientists who are working with risk assessment of bridge infrastructure. Thirty collapse sites with the least disturbed watersheds within the Appalachian Highland region are selected for analysis. The authors detected the inadequacy of widely used model selection criteria and fitted tests for annual peak flow data, categorized trend and seasonal variation to provide additional insights into the underlying physical processes, and compared predictor variables to inform assessments of collapse risk in relation to the impact of climate and land use change.
First of all, I would like to discuss the title of the article. There are no data about bridge collapse events in the paper. In my opinion, the paper results are related to bridge collapse risk assessment. If the authors agree, they could change the paper title.
The idea of the article is good and the conducted research is interesting. However, some inaccuracies in the scientific and technical aspects have been noted. Therefore, my evaluation is “Publication after major changes”.
I have some remarks for improvements of the paper quality:
1. Abstract. Line 11: … the Appalachian Highland region… The country title could be written too. Line 14: The results … What research results? Lines 20 – 22: this sentence is not clear.
2. Introduction. Lines 27 – 28: … with a total population of about 504,000 bridges... This phrase is unclear. Lines 32 – 34: this sentence is not clear. Line 79: …USGS stations… A full title is necessary for the first mention. Lines 81 – 82: not a clear sentence as it is not clear what database is being referred to.
3. Methods. Fig. 1 needs to be corrected: the geographical location of the basin should be shown at the national level, the main river basins should be listed, and the stations should be numbered (e.g. 1 to 30). The section “Selection of Sites” should contain a table with the characteristics of 30 USGS stations: the titles of the river basin and USGS; the flow measurement periods, the date of the collapse of the bridge, and the average and maximum discharge values. Only then the connection between the following calculations and the collapse of the bridges would be seen.
In research, the authors use hydrological data of various lengths and periods. Is it possible to compare the obtained results with each other: for example, according to the trend analysis, xx stations out of 30 stations have a positive significant trend? The direction of the trend clearly depends on the data of the selected period. You can only compare the data of the same period (for example, 1961-1990 or other periods).
The section “Analysis of Predictor Variables”. As I understood, the methodology of this section is described in the article [37]. What is the difference between the methodology and the obtained results of this article and the article [37], because in the article [37] the research object was chosen the same - the Appalachian Highland region? In the paper, the main information related to predictor variables could be described – not just a link to www.
4. Results. The numbering of stations is complicated. It would be better to do numeration from 1 to 30 and this numeration could be in Fig.1. Line 209: Only a minority of 13% of the sites … “Minority” could be removed. Lines 217 – 219: the sentence is not clear.
The Tables show a different number of observations for each station. Is it possible to compare results from time series with different lengths? Are 18 observations enough to define distributions and other parameters (Table 4)? How many minimum observations would be enough for the analysis?
Line 245: For all twelve sites considered … “All” could be removed. Lines 255 – 257: Is selecting the heavy-tail distribution depend on the length of the time series or not?
Section “Correlogram and Boxplot”. More analysis is necessary (lines 271 – 272) – not only link to www.
The y-axis legends are required in Fig. 3 and 4. Fig. 4 results should be explained in more detail. Is Fig. 5 results typical for all 30 stations? Lines 333 - 334: not clear - …significantly different in relation to two different comparisons of the collapse sites …
Lines 339 – 344. These sentences are not clear to me. Is it results from another paper [37]? The authors could rewrite these sentences more clearly and in more detail.
5. Discussion. Lines 349 – 353. If the authors used time series of the same length and period, the obtained percentages of distributions (53 and 36) might be different. Or not? Lines 358 – 360; 373 – 378; 406 - 407: the sentences are not clear.
6. The conclusions are abstract, with little relation to the research done. Findings should be more specific.
Some small corrections could be done in the English language.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
（1）for the population distributions concerned in the paper ,it's limited in GEV and Lognormal ,why do consider Person-type three or Log-Person type III distribution?
(2) Do you have some tests to know if the hydrologic time series are in the least disturbed watershed ?
(3)For a research paper in engineering field ,there is not any formula in it .Is't ok?
(4)for line 289 ,obvios should be changed to obvious
（1）for the population distributions concerned in the paper ,it's limited in GEV and Lognormal ,why do consider Person-type three or Log-Person type III distribution?
(2) Do you have some tests to know if the hydrologic time series are in the least disturbed watershed ?
(3)For a research paper in engineering field ,there is not any formula in it .Is't ok?
(4)for line 289 ,obvios should be changed to obvious
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
I appreciate the clearness of this work and the high quality of data collected.
Considering the importance of river flow-structures interactions, and the general lack in overall literature (from my point of view), of turbulence phenomena and scouring, this work can be a good starting point in the focus.
According to this, I suggested to Editors to accept after a very short minor revision. Hereafter you can find my report.
Best
Introduction
Please consider more literature works: see references below. there are some lack on analyzing statistical, hydrological and structural interactions.
row 92. Did you think about any relation between anthropical impact-if any- and hydrological aspects in Appalachian region? or critical issues are related only with "intrinsic risk conditions including high erosion at bed and bank,"?
row 116 maybe "if any" sounds too much colloquial english.
row 150 (and other rows) please change "web link" with reference avoiding intertextual connession
row 201 ". However, none of the sites exhibit heavy tailed Pareto distribution as best fitted." True and really interesting. why?
Figure 3, row 277. Please put x and y axis "name" or, at least, a legend ("x" are Jan, Feb, etc.) Again, some figure, please motivate why "8" and "9" boxplot show important outliers.
Did you try, for tab.1, 2,..., to consider also discharge value "Q"? I think it would be scientifically more correct to consider a relation between a link between hydrological conditions, discharge, flow velocity considering the aim of the paper "bridge collapse" a part from the large hydrological study
This aspect could be showed in "discussion"
References
1. Ichiro Ario https://orcid.org/0000-0002-5242-3561 Email: mario@hiroshima-u.ac.jp, Tatsuya Yamashita, Ryota Tsubaki, Shin-ichi Kawamura, Tatsuhiko Uchida (2022)
Investigation of bridge collapse phenomena due to heavy rain floods: structural, hydraulic and hydrological analysis, Journal of Bridge Engineering Volume 27, Issue 9
2. Brian Maddison,(2012), Scour failure of bridges, Proceedings of the Institution of Civil Engineers -
3. Ida Mascolo , Mariano Modano , Antimo Fiorillo , Marcello Fulgione Vittorio Pasquino and Fernando Fraternali (2018) Experimental and numerical study on lateral torsional buckling of steel c-beams with variable cross-section, Metals4. , ,
Scou
3Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
I would like to thank the authors fro corrections according to my remarks.
