Riparian Vegetation Density Mapping of an Extremely Densely Vegetated Confined Floodplain

Round 1

Reviewer 1 Report

Many thanks for the revisions to this paper. Some of the issues raised have been well clarified. Reading this version though, I am not clear my prior fears have been assuaged.

Namely, this paper still has an emphasis on removing shrubs and trees from floodplains to create more favourable flood conveyance conditions. I now fully understand that this is because flood retention at this site is not favourable. This could be expanded on in the introduction, making this point very explicit. 

Yet, a central issue still persists. First, tree removal is not a topic that should be taken lightly, especially as second, there is limited hydrological evidence in the paper as to why Amorpha and willows should be removed. Simply defining plant location and density does not translate to flood conveyance. I would recommend having the focus of the paper on the density estimates on a low slope floodplain (not talking about tree removals), and then have a more extended discussion (in the Discussion) on why these vegetation layers might need to be managed to alleviate localized flood build up. This would need good links of amorpha or similar shrubs, to roughness, i.e. how you would translate to metrics that could infer hydrology dynamics. Also, you would need extensive evidence of hydrological studies or simulations of floods over similar types of vegetation. 

 

Furthermore, the fact the other vegetation types are also invaded by Amorpha fruticosa does not negate the fact that Amorpha has been identified only in limited spatial extents. I.e. has your method explicitly dealt with understory amorpha? The limited spatial extent needs to be considered in the discussion on hydrology...i.e. how much of an effect might these amorpha thickets have on flood conveyance? 

 

 

 

Author Response

Dear Reviewer!

Thank you for re-reading and re-reviewing our manuscript. We are sorry, that still you have “fears” about it. As you indicated in your review the Introduction and the Discussion should be improved, thus we focused on these chapters. The English o the text was checked by a professional service again.

Thank you again for your useful comments!

Yours sincerely,

Timea Kiss

Author Response File: Author Response.docx

Reviewer 2 Report

Thank you for the opportunity to review the manuscript. the revised manuscript has significantly improved. The reviewer has no further comments. Recommend to accept in the current format. 

Author Response

Dear Reviewer,

Thank you very much for your positive answer!

Yours sincerely,

Timea Kiss 

Reviewer 3 Report

The manuscript has been improved. I suggest to accept it in present form.

Congratulations.

Author Response

Dear Reviewer,

Thank you very much for your positive answer!

Yours sincerely,

Timea Kiss 

Round 2

Reviewer 1 Report

I thank the authors for the clarifications, and for indulging my issues. Paper accepted. 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

This is an interesting study using Lidar to classify a floodplain in Hungary, considering winter lidar scans to best determine instances of the invasive Amorpha shrub at the location. This is interesting, and the study could be about using winter lidar to determine invasive shrubs in a forested ecosystem. Yet, hydrological discussions concerning forest clearing or dangers of Amorpha and forests in channelising and incising flow can be a dangerous think to state especially with a lack of clear and consistent hydrological evidence. For this paper to be published, there either has to be clear hydrological evidence to localized floodplain forest management, or otherwise the study could be on using lidar to classify an invasive floodplain shrub species. I think the latter ecology question could be worthy of publication, and I strongly commend the authors for doing this classification work.

 

Abstract:

• The aim is not entirely clear. You will determine the floodplain vegetation density, but will also assess the distribution? Is density foliage and branch density? Will you be defining vegetation type as well? Is type necessary to assess distribution horizontally? Or will the distribution be vertical?
• You are focusing on flood conveyance as opposed to flood mitigation. Is there a reason? Discussions of trees hindering flood conveyance can be a dangerous incentive for councils to cut down floodplain forests.

Introduction

• Similar point to above. The introduction needs an explicit section discussing not just conveyance and efficiency, but flood mitigation downstream where floodplain forests can be seen as important agents for water retention. I.e. flooding is not just a localized issue.
• You discuss erosion and deposition in the same paragraph. Vegetation can have channelization effects increasing velocity, but in which cases? Your previous paragraph discusses impeding flow.
• L 98. The denser the stand, the more difficult it is for the lidar’s NIR signal to penetrate to the understory.
• ORD and NRD seem strikingly similar, unless I am not understanding them correctly, except that NRD does not have a volume dimension.
• L 129. Depends on the lidar instrument, and elevation of flight, and the spatial resolution. This needs to be dealt with explicitly in the introduction. It is not clear what type of lidar you are discussing.
• L 153. Clearing vegetation can be a dangerous proposition in floodplains. This needs detailed deliberation, which I am not sure this paper covers.

Method:

• Section 3.1. What is the significance of winter scans? Is it to maximize understory reflectance? You need to link this to roughness, i.e. stem and branch roughness and foliage roughness importance.
• Table 1. The importance of using voxels is not clear. Explain how you can obtain accurate vertical plant structure using 9 point/m2 lidar.
• Section 3.2 is not clear, and I have not understood how vegetation types are classified. Expand the description of the decision tree.
• Figure 3. Not clear how meaningful this is.

Results:

• Figure 4. There are 5 ‘poplar plantation’ classes, 2 ‘riparian poplars’, 2 willows, and different classes arise from previous decision trees with different names. Therefore it is difficult to assess the validity of the classification method.
• Table 2 is more meaningful than Figure 4. This needs to be better linked to the results and to the method describing the decision tree.
• Figure 5. It does not seem that the Amorpha is lining the channel to justify the introduction of flood channelization and channel incision in Section 2.2 and earlier. The majority are native willows and poplars.
• Section 4.3. Not clear how meaningful the presentation of these results are. Vegetation types are different, which is why you are classifying different poplars. If you think there should be more willow categories, then classify them. Second, isn’t the ‘dense density’ category just related to an increase in understory growth?? Third, wouldn’t it be more meaningful to show NRD averages or boxplots with all species in a single graph? This would allow for direct comparisons between species. This could be done in height bins, e.g. every 2m in height.
• Section 4.4 and Figure 11 is difficult to understand and read. I recommend average or boxplot densities per height bin, showing clearly how much denser one vegetation type is to another. Furthermore, an absolute density would be better related to the area occupied by the vegetation type in the floodplain.
• Furthermore, how is density here translated to floodplain roughness.
• Section 4.5. The links to flooding here are only tenuous. It would have been more appropriate to talk about frontal area, or even roughness associated with each pixel with different height bins. Then you would be able to infer flood blockages and begin to discuss channelizations, etc.

Discussion:

• Line 655-656. Yes, the strength of your study lies in the land cover classification, using winter lidar to determine both forest and shrubland floodplain species. The second strength is about the invasive species occurrence in your area, compared to other literature of Amorpha presence on floodplains.
• Line 672-674. First, it is not known what the natural state of the floodplain was. You have 18^th century information, for this area may have been deforested, or just marshy. A marshy area is very different from a grassland floodplain ecologically and hydrologically. Therefore, advocating the ‘dangers’ of floodplain is not justified. Second, as stated before, the increase in floodwaters locally can be a good thing considering mitigation of flow downstream and loss of momentum.
• What is the effect of Amorpha ecologically? Is there ecological evidence that this invasive species needs to be limited in Hungary/Europe?

 

Conclusion:

• L 805-807. Yet you have not including Manning’s n measurements.

Reviewer 2 Report

Summary: 

The study titled “Riparian vegetation density mapping of an extremely densely vegetated confined floodplain” classify riparian vegetation and estimates species specific vegetation densities to identify their impact on the flood conveyance zone of the Tisza River. Vegetation classification and spatial density estimations are based on lidar point cloud based vegetation parameters. Normalized relative point density (NRD) has been used to estimate the vegetation densities at 1-5 m height range. The results suggest that the lidar derived structural metrics can classify vegetation with 83% accuracy. According to the study, while young poplar plantations favor the flood conveyance conditions, dense invasive Amorpha thickets and native willow are the least favorable for flood conveyance conditions. The study provides a valuable insight on flood management by mapping vegetation type and densities across the flood conveyance zone. According to authors the study can be extrapolated to other domain such as fire risk management as vegetation density and type play a critical role in fire spread and severity.

Broader comments: 

Thank you for sending the manuscript to review.

The study titled “Riparian vegetation density mapping of an extremely densely vegetated confined floodplain” classify riparian vegetation and estimate species specific vegetation densities to identify their impact on the flood conveyance zone of the Tisza River. Vegetation classification and spatial density estimations are based on lidar point cloud based vegetation parameters. normalized relative point density (NRD) has been used to estimate the vegetation densities at 1-5 m height range. The results suggest that the lidar derived structure metrics can classify vegetation with 83% accuracy. According to the study, while young poplar plantations favor the flood conveyance conditions, dense invasive Amorpha thickets and native willow provide the least favorable flood conveyance conditions. The study provide a valuable insight on flood management by mapping vegetation type and densities across the flood conveyance zone. According to authors the study can be extrapolated to other domain such as fire risk management as vegetation density and type paly a critical role in fire spread and severity.

All sections except the methodology and some sections in the results have been well written. In addition, the authors need to pay more attention to punctuation marks. Methodology required more elaboration on how and what parameters derived (add an explicit table of parameters and references if they’ve been used before) from the lidar data Also, the study mentioned several vegetation height categories in many places that might make the reader confusing. Thus, suggest rewriting or make it clear why these different height ranges (density/classification) been used. In the results, it is great to see images of each plantation and some description of them (Table 2 – first two columns). However, it might be worth to add graphs of ANOVA to show group differences of each vegetation type than just adding values (Table 2 – Selection criteria). Specific comments are below.

 

L 163 : Danube river? Would be good to be specific and not all readers are familiar with the region.

L 177 : Tisza river?

L 182 : Comma after “In the 20th century”.

L 205-210 : Any reference to these statistics?

L 234 : “by ca. 15-34 cm [8]”. What is Ca. ??

L 257 : “Gringorten formula [42]” Because this formula is a major part of the study, please add the formula and its parameters so the reader gets a better idea of flood return periods.

L 264 : Talks about 55 statistical variables per pixel. What are those 55 variables?  The table below doesn’t represent 55 variables. If they are the same variables as in the table but from different height levels, please explicitly say that.

L 271: Any reference for “Forest Web Map of Hungary”? How accurate the classification in this map to be used as the reference (at least to select field plots) as there is no indication of in-situ observations other than drone and this map?

L 309: 1 m intervals?? per class? Please clarify this.

Figure 3: Above you mentioned 1-5 m analyzed. And now says 1-2 m. This is confusing. If you used different analysis considering different height ranges of vegetation, please explicitly mentioned them in each section and in figure captions.

  Figure 4: Better to show ANOVA graphs using these specific parameters and vegetation classes, so the reader can easily capture which parameter differentiate which vegetation class/classes.  

 

 

Reviewer 3 Report

A BRIEF SUMMARY

The paper titled ““Riparian vegetation density mapping of an extremely densely vegetated confined floodplain” presents a good topic for readers of this Journal. However, some lacks emerge after reading the paper. Below is the list of serious lacks. I suggest you to resubmit this interesting paper after a moderate revision. 

 

1- Lack of an adequate definition of the bankfull discharge, bank vegetation and floristic indicators for hydraulic modelling. I strongly suggest that the authors try to add some more references especially in the "part 1 (introduction)" of the paper to make the foundation for the arguments stronger. For example you can cite some recent works (following reported), to better address this point.

• https://doi.org/10.4081/jae.2021.1140
• https://doi.org/10.1002/esp.3397
• https://doi.org/10.1002/1099-1085(200011/12)14:16/17<2959::AID-HYP129>3.0.CO;2-B
• https://doi.org/10.1016/j.ecoleng.2018.06.018

2- Why didn’t you compare obtained results, with others recent researchs on riparian vegetation? I have seen several potential links with studies above suggested. You can use these to validate results.

3- Conclusions are poor written. I suggest to replace paragraph. In particular you have to highlight novelty of your approach and to report future improvement of this interesting work.

 

SPECIFIC COMMENTS

Keywords: I suggest to add more keywords. For example “floristic indicators, bankfull discharge, bank vegetation”.