Bounding Surfaces in a Barchan Dune: Annual Cycles of Deposition? Seasonality or Erosion by Superimposed Bedforms?

Round 1

Reviewer 1 Report

In his paper “Bounding surfaces in a barchan dune: Annual cycles of deposition? Seasonality or erosion by superimposed bedforms?” Bristow uses satellite data and ground penetrating radar (GPR) to assess the morphology, migration, and internal structure of a barchan dune in Morocco. The methods used in this paper are robust. His results address his overarching question as to if the internal structure of the barchan dune reflects seasonal changes in wind regime or the migration of superimposed dunes down part of the slipface of the dune. The results convincingly show that both surfaces are likely present. In particular, the trough cross-stratification demonstrates that superimposed dunes do contribute to the internal stratigraphy of barchans; a novel insight that should contribute to many ongoing studies of barchan dunes and the greater role of superimposed dunes in the migration and stratigraphy of bedforms. The conclusions asserted by Bristow are supported by his results. All figures are necessary and clear.

This study presents new and interesting results in a straightforward manner that should be accessible by readers of Remote Sensing. The results should be valuable to anyone who studies bedform morphodynamics. My comments are minor and should not delay the publication of this manuscript.

Line comments:

 

Line 8: Suggest indicating the location of the barchan dune at the outset of the abstract.

Line 42: Mention here that this was a 1966 study. It will add impact to the idea that little is known about the internal structure and ward off people from thinking they could do that now(!).

Line 51: Better attributed to Brookfield (1977) with refinements from Kocurek. Brookfield, M. E. (1977). The origin of bounding surfaces in ancient aeolian sandstones. Sedimentology, 24(3), 303-332.

Line 83: Check comma usage. “The wind, blows sand…, towards…”

Line 85: Check capitalization of “Marine”.

Line 89: No units given for 0-50 scale in the upper left of the wind rose panel. Can more recent data be used to coincide with the date of the surveys?

Line 119: Also suggest reference Phillips, J. D., Ewing, R. C., Bowling, R., Weymer, B. A., Barrineau, P., Nittrouer, J. A., & Everett, M. E. (2019). Low-angle eolian deposits formed by protodune migration, and insights into slipface development at White Sands Dune Field, New Mexico. Aeolian Research, 36, 9-26.

Line 221: Section 4.1, although interesting, doesn’t clearly link to the rest of the paper. Determining the sediment flux doesn’t appear needed to support the main discussion of the paper, which is bounding surfaces within a barchan dune.

Lines 238-254: This paragraph has some redundancy with prior discussion wrt the migration rates. The numbers are also repeated needlessly in successive sentences within the paragraph. Suggest revising for simplicity and clarity.

Line 268: The locations of X-X’ and Y-Y’ should be plotted on the image of the barchan dune (Fig. 2).

Line 281: Can the position of the dune slipface during different years be plotted on the GPR profile? See the Phillips et al. (2019) paper as an example of this visualization.

Line 345: Including a figure/schematic from the 1966 paper would help the reader compare the two without having to access the paper.

 

 

Author Response

I thank R1 for the positive and constructive comments.

Line 8: Suggest indicating the location of the barchan dune at the outset of the abstract.

Reply; the words 'near Tarfaya in Morocco' have been inserted into the first line of the abstract.

Line 42: Mention here that this was a 1966 study. It will add impact to the idea that little is known about the internal structure and ward off people from thinking they could do that now(!).

Reply; yes indeed. words 'In an historic study published in 1966' inserted into the text to give context.

Line 51: Better attributed to Brookfield (1977) with refinements from Kocurek. Brookfield, M. E. (1977). The origin of bounding surfaces in ancient aeolian sandstones. Sedimentology, 24(3), 303-332.

Reply; Please note that Brookfield 1977 reference [13] is cited in the previous sentence. Brookfield used the terms first , second and third order bounding surface and the terminology of reactivation surface and superposition surface used here were defined by Gary Kocurek ref [14]. I am happy to defend the existing text.

Line 83: Check comma usage. “The wind, blows sand…, towards…”

Reply; comma moved from wind, to sand, . Hopefully that makes better sense.

Line 85: Check capitalization of “Marine”.

Reply; capital letter removed and replaced with lower case m.

Line 89: No units given for 0-50 scale in the upper left of the wind rose panel. Can more recent data be used to coincide with the date of the surveys?

Reply; The radial scale is in percentage of time. This statement has been added to the figure caption. Unfortunately I do not have any wind data from Tarfaya and have had to rely on limited published information, in this case wind roses.

 

Line 119: Also suggest reference Phillips, J. D., Ewing, R. C., Bowling, R., Weymer, B. A., Barrineau, P., Nittrouer, J. A., & Everett, M. E. (2019). Low-angle eolian deposits formed by protodune migration, and insights into slipface development at White Sands Dune Field, New Mexico. Aeolian Research, 36, 9-26.

Reply; Thank you, this is a reference to a new (2019) paper that I have not seen and has been included as reference 30, inserted in the reference list with cites in the text as well as the caption for figure 3.

Line 221: Section 4.1, although interesting, doesn’t clearly link to the rest of the paper. Determining the sediment flux doesn’t appear needed to support the main discussion of the paper, which is bounding surfaces within a barchan dune.

Reply; Yes I agree. These calculations have also prompted comments from R3 and I accept that this paragraph could be removed but having done the sums I was reluctant to leave them out. I would be happy to delete this section if the editor says so.

Lines 238-254: This paragraph has some redundancy with prior discussion wrt the migration rates. The numbers are also repeated needlessly in successive sentences within the paragraph. Suggest revising for simplicity and clarity.

Reply; I agree, one sentence, lines 244 to 246, has been deleted and the text closed up to remove the repetition.

Line 268: The locations of X-X’ and Y-Y’ should be plotted on the image of the barchan dune (Fig. 2).

Reply; I have added X-X' and Y-Y' to figure 4 rather than figure 2 as was suggested by R2. Figure 2 already has enough lines on it.

Line 281: Can the position of the dune slipface during different years be plotted on the GPR profile? See the Phillips et al. (2019) paper as an example of this visualization.

Reply; this would have been desirable but without the lidar data in Phillips et al 2019 it is not possible to match the topography, only the position of the dune toe which is indicated on figure 6.

Line 345: Including a figure/schematic from the 1966 paper would help the reader compare the two without having to access the paper.

Reply; I have considered this and gone back to the figure in question (Figure 8) in McKee 1966, and concluded that the preparation of a new figure is not necessary, if the reader is interested they should be able to access the original paper. 

 

Reviewer 2 Report

A very well researched project and well-written manuscript. I recommend acceptance with only minor editing to address the issues noted below.

Line 137-138 and following: Could you provide some estimate of uncertainty in the locations of these features. Any impact of this uncertainty on results.

Line 170 – 184 discussion. Could the variation in horn length possibly be random and unattributable?

 

Line 189: How does this migration rate compare to the entire local field? How much variation? Is this measured value close to an average for the field or is it much faster or slower? Line 220 refers to global averages if I remember correctly.

Can the two GPR profiles be plotted on an image of the dune- perhaps on Figure 4?

Line 227: different part   should be parts?

Line 285: Any speculation about the causes of these erosion surfaces and/or the preservation potential of selected surfaces?

Line 334 -335 The asymmetric troughs interpreted as blowouts by [12] might now be interpreted as superimposed dunes.  This is a great observation and conforms to what I have observed and interpreted.

Conclusions: This is one very long paragraph with several important points. I would suggest breaking it into two or three shorter paragraphs to better emphasize these points.

Author Response

Line 137-138 and following: Could you provide some estimate of uncertainty in the locations of these features. Any impact of this uncertainty on results.

Reply; Yes, I have added a sentence at the end of the paragraph stating, 'Picking independent ground control points to assess the accuracy of the measurements indicates that the image registration in Google Earth is ± 2m, which is around 2% of the length and width of the dune.

Line 170 – 184 discussion. Could the variation in horn length possibly be random and unattributable?

Reply; Good question but I do not believe in random acts of nature, there should be a reason, and in this case wind looks most likely.

 

Line 189: How does this migration rate compare to the entire local field? How much variation? Is this measured value close to an average for the field or is it much faster or slower? Line 220 refers to global averages if I remember correctly.

Reply. Local rates for dune velocity can be found in Elbelrhiti et al 2008, was ref [30] now ref [31]. New sentence added stating 'These figures for dune velocity are close to the velocities for similar sized dunes near Tarfaya [31], and slightly lower than dunes to the south near Laayoune [31]. '

Can the two GPR profiles be plotted on an image of the dune- perhaps on Figure 4?

Reply; Yes they have been added to figure 4 as suggested.

Line 227: different part   should be parts?

Reply; yest that has been changed.

Line 285: Any speculation about the causes of these erosion surfaces and/or the preservation potential of selected surfaces?

Reply; there are two issues here, first speculation on the causes of these erosion surfaces is what the paper is all about.  The emphasis on the sentence in line 285 should be on the word 'alone'. The conclusions are that you need a three dimensional view and cant just rely on a single cross-section 'alone'. Secondly the preservation potential within the rock record is limited because barchan dunes are associated with limited sand supply and it would require exceptional conditions to preserve barchan dunes. However, reactivation surfaces are not restricted to barchan dunes and they are ubiquitous within the aeolian rock record due to the natural variability of the wind.

Line 334 -335 The asymmetric troughs interpreted as blowouts by [12] might now be interpreted as superimposed dunes.  This is a great observation and conforms to what I have observed and interpreted.

Reply; thank you for that confirmation.

Conclusions: This is one very long paragraph with several important points. I would suggest breaking it into two or three shorter paragraphs to better emphasize these points.

Reply; the conclusions have been divided into four paragraphs.

Reviewer 3 Report

see the attached PDF file.

Comments for author File: Comments.pdf

Author Response

Here I understand that the seasonal cycles is recognised thanks to

truncated cross-data. The criterion to select this truncated or non-

truncated bounding interface is not clear to me and I have the feeling

that I will be able to select different surfaces if I have to reproduce

Figures 5 and 6. Is it possible to give some methods likely to help

non-expert readers?

Reply; I have tried to show this with the sketches in figure 8. The truncation should be overlain by a low-angle bounding surface. It is possible to speculate that the deflation of the slip-face by reversing winds would create a well compacted, winnowed layer that would cause a higher amplitude reflection on the GPR profile but that is just speculation.

? I am convinced that the best expression to compare sand flux data is to

express it in m2 yr-1. Especially here where the author has measured

the central slides of barchan dunes with a significant height, the best

relationship between dune morphology and dune migration rate is c =

Qdune=H, where Qdune = (1 + )Qresultant. In this expression,  is the

wind speed-up at the dune crest (usually measured between 0.5 and 2

depending on the dune aspect ratio) and Qresultant is the resultant sand

flux on a at sand bed that can be derived from the wind data (see for

example Gao et al. (2018) and its Supplementary Information). Using

 = 1:6, c = 20 myr-1 and H = 10 m, we find Qresultant _ 77 m2 yr?1,

which is on the same order of magnitude as the resultant sand ux

usually observed in Tarfaya (see Figure S4 in reference [30]).

The Eq. 1 proposed in the manuscript and in reference [9] (but not in

reference [7]) cannot be used for superimposed bedforms. In addition,

it is difficult for me to compare Qdune = 77; 414 kgm?1 yr?1 with other

data in the literature.

Reply; I am not certain what the point is here, other than there are different ways to describe sand flux in dunes. This is not an argument that I want to get involved in so perhaps I should have followed the advice of R1 and simple deleted section 4.1 if it is going to cause a problem.

 R3 appears to have solved the first part for themselves. Although they have had to assume  a wind speed-up at the crest which is not known, otherwise it would be possible to substitute the dune height H = 7.8m and dune velocity c = 21. If c= Qdune/H then that can be rearranged as c x H = Qdune, which works out at 163.8 m2 yr-1, dividing by 2.6 to get Q resultant = 63.

I accept the point that Eq.1 cannot be used for superimposed dunes but that has not been attempted here, and I can confirm that Eq. 1 is in [7], it is on page 5 of 18, paragraph [11].

? The origin of superimposed bedforms have been numerically elucidated

in Narteau et al. (2009). The morphodynamics of primary and sec-

ondary bedforms have been analysed and compared in Zhang et al.

(2010). Scours in the lee of superimposed bedforms, erosional super-

position surfaces at the crest and cross-strata of dunes under unidi-

rectional wind regimes have been numerically described in Gao et al.

(2015).

Reply; This appears to be a comment rather than a question. The model in Gao et al 2015, produces superposition surfaces that are sub-horizontal, their figure 6a. Gao et al describe them as horizontal which are difficult to relate to the observations here. If the superposition surfaces were horizontal that would be apparent in the profile X-X' along the length of the dune. But this is not the case. At the same time, the troughs visible in the cross-section Y-Y' require superimposed dunes. Therefore bounding surfaces produced by superimposed dunes must not be horizontal.

? Zhang et al. (2010) and Gao et al. (2015) have also shown that un-

der unidirectional wind regimes the dune aspect-ratio may change, an

increase in wind strength being associated with an increase in aspect-

ratio and reciprocally. Can this type of change in dune shape be seen

in sedimentary patterns?

This is an interesting question. In answer I suggest that a change in the due aspect ratio, becoming shorter and steeper, or the converse, longer and flatter, might be visible in the geometry of the sets of cross-strata. the elongation and flattening of a dune might result in an increase in the length of sets of cross-strata (by longer in this context I mean parallel to the wind direction. In contrast, a reduction in dune length and an increase in dune height by itself should result in narrower but taller sets of cross-strata. Unfortunately dune height is rarely preserved because dunes are preserved from the bottom up and the crests are almost always eroded so this would be very difficult to reconstruct from the rock-record. 

? In Figure 1, the wind data can be completed by showing at least the

distribution of wind speed, but also the distribution of sand flux ori-

entation (flux rose). Taking into account the wind speed-up effect at

the dune crest (see above), I consider that it is critical to compare this

resultant sand flux to that derived from dune migration data.

Reply; There is a sand flux rose in Elbelrhiti et al 2008, their figure A4c. I do not have any wind data, the wind roses are copied from the literature and there are no measurements to calculate the speed-up effect.

? Part of the introductory paragraph of the Discussion (Sec. 4) is a

repetition of Sec. 3.1 just above.

Reply; I have not identified the repetition and R3 has not provided line numbers to check. I did however notice that parts of 3.1 Results are actually a discussion, but nobody else has made that observation.

? Add small space in units, replacing for example myr-1 by m yr-1.

Reply; Yes thank you for that observation.

? Line 228: Qatar and not Quatar.

Reply; Yes, thank you for that observation.

Bibliography

Gao X., C. Narteau, O. Rozier, Development and steady states of transverse

dunes: A numerical analysis of dune pattern coarsening and giant dunes,

Journal of Geophysical Research Earth Surface, 120, 2200-2219 (2015).

Gao X., C. Gadal, O. Rozier, C. Narteau, Morphodynamics of barchan and

dome dunes under variable wind regimes, Geology, 46, 743746 (2018).

Narteau C., D. Zhang, O. Rozier, P. Claudin, Setting the length and time

scales of a cellular automaton dune model from the analysis of superimposed

bedforms, Journal of Geophysical Research Earth Surface, 114, F03006

(2009).

Zhang D., C. Narteau, O. Rozier, Morphodynamics of barchan and transverse

dunes using a cellular automaton model, Journal of Geophysical Research

Earth Surface, 115, F03041 (2010).