Gully Morphological Characteristics and Topographic Threshold Determined by UAV in a Small Watershed on the Loess Plateau

Round 1

Reviewer 1 Report

This study evaluated the accuracy of gully morphological characteristics extracted from UAV-images using measured data of RTK, made a statistical analysis of the morphological features of 910 gullies, and determined the topographic thresholds based a dataset of 365 gully heads. The results could provide a broader understanding of the gully morphology and gully development. However, there are still some issues need to be further considered by the authors before publication.

1.      The dataset was not described clearly enough. 1) the gullies in the study watershed can be cataloged into three types, named as bank, floor and hillslope gully (L160-161) and totally 910 gullies were identified(L192). However, it is not clear whether all gullies in the watershed were identified and what types the investigated gullies belong to. 2) Fig.1 showed the position of gully heads, but it is not clear which gullies were used to evaluated the accuracy of gully morphological characteristics and which were used to determine the topographic thresholds. Correspondingly, for the results, I wonder if the morphological parameters and topographic thresholds could be different significantly among three types of gullies.

2.      I cannot imagine the accuracy of gully measurement by a handheld RTK is better than that by UAV with RTK. So I suggest detailed and visible information be provided such as DEM or 3D models of gullies measured by a handheld RTK.

3.      Figure 2 and Figure 4 are not necessary and Figure 3 cannot show the gully edges clearly. I suggest to replace them with DOM and DEM of the three types of gullies with gully boundaries were delineated.

4.      Some quotations from literature need to be confirmed. L36-37: Gully is defined as an erosion channel with a cross-sectional area of more than 1 ft2 (929 cm^2) and a depth over 50 cm [1]. L117-119: More than 60% of the land has been eroded, and the area with an erosion modulus exceeding 8000 t /km^2 /a is 91,200 km^2 [12,14].

5.      Some statements need to be supported by evidence. L129-131：Meanwhile, great attention should be paid to gullies with a length less than 500 m, especially those less than 100 m, since they are relatively active and produce a large amount of sediment in small watershed. 

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

The abstract is well prepared but the terminology used must be taken into account. (digital elevation model or digital surface model? Is DEM automatically generated from UAV images? digital orthophoto map or orthophotoplan?).

Also in the abstract it is highlighted that: In this study, 910 gullies were visually interpreted. How were they interpreted visually? Isn't that a mistake of expression? in visual interpretation you cannot extract quantitative characteristics with a very high precision, precision absolutely necessary in these types of studies

figure 1c must be redone (a color palette must be adapted to highlight the morphometry, on a gray background nothing is understood)

Figure 3 needs to be redone (a color palette must be adapted to highlight the morphometry, on a gray background nothing is understood), specify the numeric scale of the legend for figures 3b, 3d, 3f (what does value mean?)

Insert, within the methodology, a logical scheme (flowchart) that presents the stages and sub-stages of the research approach.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

Comments to the Authors:

In the manuscript titled "Gully morphological characteristics and topographic threshold determined by UAV in a small watershed on the Loess Plateau” the authors discuss a case-study regarding the measure of gully morphological characteristics through field measurements combined with UAV monitoring. The study area falls in a typical small watershed on the Loess Plateau in China. In particular, 910 gullies were observed by unmanned aerial vehicle technology and 30 gullies were measured by field survey and compared with the parameters extracted from UAV, demonstrating the high accuracy of gully characteristic extraction from UAV-images. The addressed topic well fits the interest of the readers of Remote Sensing Journal and in my opinion the case study is interesting but the current version of the manuscript should undergo a MODERATE REVISION to overcome some limits that I found. So, I believe that the manuscript, with an additional effort by the Authors, may be considered for publication.

The main criticism that I found in this research paper consists in an absolute lacking of description about the acquisition method used to collect the images by UAV. For example, the authors should add a better careful description regarding: the scheme of the drone flight, the density map of the photos acquired by drone, how many photos were acquired by the drone? Were used Ground Control Points (markers) for georeferencing of the aerial photogrammetry? What is the mean error of geographic position in each image? etc. This information could be useful to the reader to replay the used approach. Furthermore, some specific photos of the gully forms with a geomorphological description are welcomed.

Another recommendation that I would suggest to the authors is the adding of a new table showing the main geomorphological measurements of the gullies observed during field investigations.

Finally, in the Discussion section the authors should explain the limits of the adopted approach, specifying which are its strengths and weaknesses as well as how this research differs significantly from other such ones and what makes the used approach unique and appropriate for an international journal.

Author Response

Please see the attachment

Author Response File: Author Response.docx