General Comments:

Review of article titled “Morphological Response of Vegetated and Urbanized Barrier Islands to Hurricane Ian” by Ilyas et al., submitted to Natural Hazards and Earth System Sciences:

This manuscript describes the morphologic responses of two barrier islands to the extreme conditions produced by Hurricane Ian. This manuscript uses a combination of multiple scenarios that incorporate different landcover classifications to determine which of the modeled scenarios most accurately replicates the observed morphological responses.

The manuscript makes a couple of important points. First, spatially variable bed roughness elements are important to include to best simulate the morphodynamic response of barrier islands during large storm events. Second, increased secondary vegetation densities can reduce erosion along dune environments.

However, this manuscript is limited in that it focuses only on dune response, instead of incorporating multiple morphological responses (i.e., changes in shoreline position, beach volume, beach slope, barrier-island interior volume, etc.) within the analyses. This study would benefit by incorporating more details about the total morphologic changes, instead of only focusing on impacts to dune crest height and dune crest cross-shore position.

Additionally, this manuscript fails to contextualize the results within the body of existing literature in the discussion. The discussion section could benefit by adding two different sections:

1. How the modeled results presented in this manuscript compare to existing interpretations of the morphologic responses of Lovers Key and Fort Myers Beach resulting from Hurricane Ian (e.g., Wang et al., 2024; McCormick et al., 2025; Hauptman et al., 2024).
2. How the modeled results presented in this manuscript compare to the existing body of literature that modeled the morphological responses of previous large storm events that have impacted developed barrier islands such as Hurricanes Michael and Sandy (e.g., Ma et al., 2024; Smallegan et al., 2016)

I think that the results presented in this manuscript are sufficiently important to publish, but the manuscript will need substantial revisions before it is suitable for publication.

Specific Comments:

*indicates more important/significant comments to be addressed

[40-45] – There is a lot of information on global mean sea level, which is not carried through the rest of the manuscript. I would suggest either limiting the amount of info about sea level, or carry it throughout the rest of the manuscript.

[70–71] – There should be more information about measured (or modeled) conditions produced by Hurricane Ian, including storm track, duration, wind speeds, water levels, etc. These conditions are published and available through Bucci et al. (2022), but should be included to provide context for the storm.

[80–99]* – There are a lot of different data sources that were used for the “observed changes” resulting from Hurricane Ian. These datasets are presented here but there are no specific details or information about the spatial uncertainties (both horizontal and vertical) of the different datasets, and the specific methodologies used to collect those data. For example, the authors discuss “LiDAR surveys” accessed by NOAA’s Digital Coast and different LiDAR surveys provided by Florida Gulf Coast University. Were these surveys collected using the same methods? Do they have the same uncertainties? Clarification and increased specificity for each of the datasets is needed here. Additionally, CEC transects extend up to 1km into the nearshore, were those collected by a total station, as noted in the manuscript, or was it done using bathymetric surveying (i.e., single beam sonar)?

[116] – The continuously updated digital elevation model that is cited through the NOAA (2018) reference is a 1/9 arc second DEM, which has a 3m pixel resolution.

[121] – How do the modeled wave and water levels compare to the measured total water levels during Hurricane Ian on Fort Myers Beach and Lovers Key? I think this should be addressed to validate the modeled water levels during the storm. These are accessible through the USGS Flood Event Viewer (https://apps.usgs.gov/fev/event/2022-ian)

[150–152] – Can the authors speculate as to why there were differences in the land cover classifications? Was it the result of differences in cell size or a different reason?

[167–170] – How did the authors differentiate beach berm crests from primary dune crests in Equation 3 at locations where beach berms (or berm ridges associated with ridge and runnel morphologies) existed following the storm?  

[175–225]* – This whole section does a nice job of comparing the modeled outputs to the observed changes but only considers dune crest height and dune position as “morphological responses”. With many other morphologic responses (including changes in shoreline position, beach volume, beach slope, barrier-island interior volumes, etc.) excluded from this section (and excluded from the entire manuscript), this section should either be renamed to be focused only on dune responses, or expand it significantly to encapsulate more of the morphologic responses resulting from Hurricane Ian.

[Figure 6] – Why are the observed Pre- and Post-storm profiles labeled as NOAA+CEC? I thought those were two different datasets, with varying spatial scales and data collection methods (i.e., one is LiDAR and one is a total station survey)?

[230–275]* – See above comment ([175–225]) about morphological responses. Additionally, there is no section here about lateral or vertical changes in dune crest. Was this section excluded due to a lack of prominent primary dunes on Fort Myers Beach, or a different reason? A short  explanation for why this was excluded from the Fort Myers Beach section should be included.

[261–263] – The observed deposition within the nearshore was a trend found in other field-based studies on Hurricane Ian (cited in the general comments section and next comment). Could you provide hypotheses as to why this is not seen in the modeled output of this study?

[278–282]* – How do your results fit into the bigger context of the growing body of literature already published on Hurricane Ian, as well as the modeled results from other previous large storm events from around the US and the World? The sensitivity analyses should still be a part of the discussion section, but the authors do not address how their work fits into the existing body of literature. For example, the authors could compare the modeled water levels from Hurricane Ian to the findings of McCann et al. (2024) and the modeled morphologic changes to the results of McCormick et al. (2025), Hauptman et al. (2024), and Wang et al. (2024). Additionally, the authors could provide more context for how their approach to incorporating variable bed roughness values for vegetation and built environments could have improved the results of previous studies that focused on other large storm events that impacted developed islands, such as Hurricane Michael or Sandy.

[294–303] – Are these examples of available datasets collected frequently enough, or at a fine enough resolution, to be useful for coastal managers?

[303–305] – This claim seems to be a bit of a stretch for the scope of this manuscript. Also, “vegetation restoration” is a little ambiguous and could use some clarification.  

[308–309] – This statement is not well supported in this context. This paper shows that the XBeach model can “reasonably” approximate the results of the inundation regime during Hurricane Ian, and does not provide citations to support that the model can accurately approximate the three other regimes. Please provide citations of other studies that have successfully replicated the three other impact regimes.

[320–334] – This section would be better suited in the results section, as it is presenting new results relating to a different land cover scenario. The results of this then could be elaborated on within the discussion and include appropriate references that corroborate the results.

Technical Corrections:

[75] – “Lovers Key State Park comprises…”

[141] – Incorrect citation; should be “Salgano (2023)”

[Figure 3] – Need scale bars and North arrows on both sets of maps with different spatial extents and different rotations; maps of Fort Myers Beach are too small to determine the differences in landcover classification

[194–195] – Elevations should include the vertical datum (NAVD88?)

[203] – What do you mean by “waning flows”? Decreased flow velocities or ebbing flows, or both?

[Figure 5] – Dashed lines and shading are difficult to distinguish in greyscale, you may want to consider making these colored lines (in addition to the different dash types) to make them more distinguishable; may want to put orientation indicator to show which direction is North/South along the x-axis to orient reader.

[212–213] – Elevations should include the vertical datum (NAVD88?)

[230] – “Hurricane Ian”

[241–244] – “A noticeable effect…”: This sentence is awkwardly worded and redundant, and should be revised.

[259] – “4.2.1 Model Versus…”

[324] – Elevations should include the vertical datum (NAVD88?)

[350–353] – “Three model scenarios… Scenario 2 is based on…”: Need to reword these two sentences as there are typos and could be combined into one sentence.

[364] – “(waves and surge)”

Overall response

This paper investigates the effects of vegetation and built environments on modelled barrier island morphodynamics in a case study in southwest Florida. The sensitivity of the model XBeach to dune vegetation cover and non-erodible surfaces was tested, using a baseline constant Manning’s roughness value, and two datasets of land cover (varying in resolution) from the National Oceanographic Partnership Program and NOAA. Using the 2022 Hurricane Ian event, the authors demonstrate firstly that including variable surface roughness in the XBeach boundary conditions significantly improves the skill of the sediment transport predictions. Secondly, using the two land cover datasets as different scenarios, they show lower resolution land cover data does not significantly impact the prediction skill. This is particularly impactful for land cover data availability when modelling coastal storm responses. What is also appreciated is the added test of the most successful model setup, which involves including vegetation patches to assess the morphodynamic response to revegetation. These results may be useful for modelling nature-based solutions to erosion in a coastal management context.

With the above being said, it would be interesting to see the study expanded beyond the dune elevation metrics used. Change in elevation maps and dune profiles are helpful for assessing before- and after-event topography, but I agree with the other reviewer that shoreline position change (and the seaward limit of vegetation) would build a richer picture of surface/environmental change being modelled by XBeach. Furthermore, the discussion does not delve into the uncertainties and potential flaws for consideration in the methodology, particularly where combining different datasets is concerned.

The manuscript makes a short but useful contribution to the field of barrier island modelling, but could be enhanced with broader analyses, to solidify its impacts suggested in the discussion.

Comments

Line 58: Please adjust the wording or grammar of this sentence, the phrasing is hard to understand.

Line 104: Is there a reason for the inconsistent numbering of the transects? Do these come from another source with breaks in the transect numbering? It might be helpful to just do 1–8 if these are contained within this study.

Line 115: Is there a difference in the capture date between these two datasets? If so, it would be worth explaining the potential impact of mismatched conditions between bathymetry models.

Line 128: Lovers Key shows a higher post-peak wave height than Fort Myers, but it sits (at least from Figure 1) behind the barrier and is mainly bay. If these timeseries are both taken from the seaward nearshore, this should probably be reflected more clearly in Figure 1. In any case, it might be helpful to briefly suggest reasons for this difference.

Line 132: It would be worth clarifying that the Manning’s roughness coefficient values used come from empirical studies in similar environments (both the uniform value and the LULC specific ones). Also Line 141 has a typo in how Salgado is cited (no brackets around author).

Line 184: It would be helpful to specify that the lack of difference between Scenario 2 and 3 are between each other, and not in relation to Scenario 1. This is clearer in Figure 4 but a slight rewording would help in the text.

Figure 7: I appreciate the choice in colour ramp for topography, but in the context of this study (which includes land cover as a metric), I feel it is misleading to include green as a colour for topography thresholds. It is partially true that higher topography tends to be because of the existence of vegetation cover in digital surface models, but not in all cases and not for these barrier islands. I would suggest steering from blue and green which are suggestive.

Line 278–282: This feels repetitive/redundant, consider shortening.

Figure 9: I would suggest colouring the added vegetation pixels in a different colour from the original primary vegetation green, to help the reader differentiate between existing and simulated veg cover.

Lines 347 and 353: Please address the spacing typos.