Reconstructing the 2015 Salgar flash flood using radar retrievals and a conceptual modeling framework in an ungauged basin

Velásquez, Nicolás; Hoyos, Carlos D.; Vélez, Jaime I.; Zapata, Esneider

doi:https://doi.org/10.5194/hess-24-1367-2020

Articles | Volume 24, issue 3

https://doi.org/10.5194/hess-24-1367-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/hess-24-1367-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 24, issue 3

Research article

|

24 Mar 2020

Research article |

| 24 Mar 2020

Reconstructing the 2015 Salgar flash flood using radar retrievals and a conceptual modeling framework in an ungauged basin

Nicolás Velásquez, Carlos D. Hoyos, Jaime I. Vélez, and Esneider Zapata

Download

Final revised paper (published on 24 Mar 2020)
Preprint (discussion started on 25 Sep 2018)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'review of the referee 1', Anonymous Referee #1, 22 Oct 2018
- AC1: 'Respose to anonymous reviewer 1', Nicolas Velasquez, 12 Jan 2019
  - AC4: 'Zip file with text, figures and bibliography', Nicolas Velasquez, 12 Jan 2019
RC2: 'An interesting manuscript about a major FF event, but excessive conclusions drawn a theoretical modelling exercise', Eric Gaume, 10 Dec 2018
- AC2: 'Response to reviewer 2', Nicolas Velasquez, 12 Jan 2019
  - AC5: 'Zip file with text, figures and bibliography.', Nicolas Velasquez, 12 Jan 2019
AC3: 'Reviewed version of the manuscript.', Nicolas Velasquez, 12 Jan 2019

Peer-review completion

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

ED: Reconsider after major revisions (further review by editor and referees) (14 Jan 2019) by Laurent Pfister

AR by Nicolas Velasquez on behalf of the Authors (18 Jan 2019) Author's response Manuscript

ED: Referee Nomination & Report Request started (21 Jan 2019) by Laurent Pfister

RR by Anonymous Referee #1 (03 Mar 2019)

Suggestions for revision or reasons for rejection

I appreciate the revision of the manuscript that makes the paper to be more concise and better structured. Specially the rewriting of the introduction, the study site and data description and part of the methodology appears in a clearer way. I would however still suggest several rewording of the section 3.2 and 3.3. The theoretical background of the latter one must be clarified to be maintained in the manuscript, either by referring to the related literature, or by explaining the hypotheses behind the floodplain sub-model.

Finally concerning the methodology, I have one main doubt about the actual connection between the hydrological model set up and the landslide model one. The strength of the overall paper methodology is to use the soil storage dynamic simulated by the hydrological model to deduce landslide occurrence. However, it does not appear clear any more if the soil storage set up (Z) is the same for both simulations. This assumption has to be clearly clarified.

The results and the discussion appear in a clearer way as well in that second version, specially when describing the results of the hydrological model, explaining the different flow processes during the two distinct events. I think that the results could even be better inserted within the current literature, showing standardize figures or adjusting the discussion. My comments below support this point.

The discussion about the limitation of the landslide model might be more more detailed. I think, the authors should be able to settle on the reasons of the result limitations: does it come from the lack of spatial information about soil and land cover properties, or from too strong assumptions of the landslide model? Ruling on the reasons of the limitation would bring a direct outlook of the study.

Finally, I found really interesting that the soil storage capacity available before flood event impacted not only the flood magnitude but also the response time of the catchment. From my point of view, it could be appear as one of the main incomes of the paper.

Referee Report: PDF

Hide

RR by Eric Gaume (24 Mar 2019)

Suggestions for revision or reasons for rejection

The manuscript has been substantially modified after the first reviews. This new version deserves nevertheless some major improvements before a publication in HESS can be considered to my point of view. An annotated manuscript is joined to this review. Six major points need especially improvement.
1) The manuscript is based on a very limited amount of inaccurate observations: a single peak discharge, estimated at a cross-section located downstream the considered watershed, radar quantitative precipitation estimates from a radar located 90 km far away from the considered mountainous area with no ground raingauge or even amateur measurements confirming the estimated amounts and two aerial photographs for the location of landslides. This is very little to support the implementation, calibration and validation of a physically based distributed model.
2) Some additional data are mentioned in the manuscript a) direct discharge measurements in nearby watershed supporting the proposed average cross-sectional velocities between 5 and 7 m/s at line 220 and b) comparisons between radar QPE and measurements of two raingauges installed after the event at line 240. But, this important information to assess the accuracy of the estimated values is not provided in the manuscript. There is neither reference to any publication the readers could refer to. All information necessary for the analysis should be either provided or accessible.
3) The presentation of the models remains partial and confused. The model contains numerous parameters or variables that are not all defined in the text or which units are not given (see annotated manuscript). All the parameters defined in table 3 do not appear in the presented equations. The article is not clear on how the model is implemented: which parameters are calibrated, why and how, how is the value of the other parameters fixed and how are the initial conditions determined. Initial conditions are crucial for the considered rainfall events with limited rainfall amounts (less than 50 mm on average for both events).
4) The article contains some obvious major errors. The flood volumes given at line 490 do obviously not correspond to the presented results (see annotated manuscript)! By the way, the simulated runoff rates for both events exceed largely 50%. This is extremely surprising, considering the limited total rainfall amount of both events and the experience of other extreme flash floods over the world (see Marchi et al., 2010 for instance). This of course is a modelling result, but is it realistic? I have some doubts.
5) The authors are not critical enough about their own results. The landslide mapping result illustrates this particularly well. The results presented in figure 15 are considered as excellent by the authors. But, first it seems quite obvious that landslides will predominantly affect the upstream areas that have the steepest slopes and have been affected by the most intense rainfall. Moreover, a detailed look at figure 15b shows that the simulated landslides are by far less concentrated and localized on the watershed than the observed ones. Likewise, the maximum simulated number of unstable cells at the very beginning of the rainfall event (figure 15c, is a surprising and probably unrealistic results that would deserve some comments…
6) Most of the paper analyses and comments the results of the proposed poorly calibrated and not validated models. This is moderately interesting and does not support, as I did mention in my previous review, the conclusions drawn by the authors about the dominant hydrological processes during this flood event. Neither the title of the manuscript, nor the conclusions correspond to the real content of the paper. The proposed procedure, based on very little real information about the flood and moreover inaccurate information, can by no way help discriminating flood generating mechanisms.

Marchi, L., Borga, M., Preciso, E., and E. Gaume, 2010: Characterisation of selected extreme flash floods in Europe and implications for flood risk management. Journal of Hydrology, 394 (1-2), 118-133. doi:10.1016/j.jhydrol.2010.07.017.

Referee Report: PDF

Hide

ED: Publish subject to revisions (further review by editor and referees) (15 Apr 2019) by Laurent Pfister

AR by Nicolas Velasquez on behalf of the Authors (28 May 2019) Manuscript

ED: Referee Nomination & Report Request started (02 Jun 2019) by Laurent Pfister

RR by Audrey Douinot (28 Jun 2019)

ED: Reconsider after major revisions (further review by editor and referees) (14 Jul 2019) by Laurent Pfister

AR by Nicolas Velasquez on behalf of the Authors (26 Aug 2019) Manuscript

ED: Referee Nomination & Report Request started (11 Sep 2019) by Laurent Pfister

RR by Anonymous Referee #1 (23 Oct 2019)

ED: Publish subject to minor revisions (review by editor) (01 Dec 2019) by Laurent Pfister

AR by Nicolas Velasquez on behalf of the Authors (12 Dec 2019)

ED: Publish as is (24 Jan 2020) by Laurent Pfister

AR by Nicolas Velasquez on behalf of the Authors (31 Jan 2020) Manuscript

Short summary

During 18 May 2015, a storm event produced a flash flood in the municipality of Salgar (northwestern Colombian Andes), resulting in more than 100 human casualties and significant economic losses. Here we present a modeled process reconstruction of what happened during this event. For this, we only use radar rainfall estimations and a digital elevation model. Results show that with scarce data there is an opportunity to obtain acceptable tools for risk management and decision making.