Long-Term Volumetric Change Estimation of Red Ash Quarry Sites in the Afro-Alpine Ecosystem of Bale Mountains National Park in Ethiopia

Round 1

Reviewer 1 Report

（1）    The objectives of this manuscript were to generate DEMs using the HAPs and the up-to-date field survey data, and to quantify the long-term volumetric changes of two red ash quarry sites in the BMNP. However, the introduction does not introduce the DEM generating methods and research status fully.

（2）    The volumetric and landscape changes of the quarry sites from 1967 to 2023 were influenced by many factors. However, the manuscript lacks a systematic analysis.

（3）    The authors used the Agisoft Metashape Professional and the most commonly used DEM interpolation methods to generate the DEMs. The method contribution to the community is very small.

（4）    It is not clear how to validate the generated DEMs in 1967 and 1984.

（5）    The grammar should be improved. For example, “The Bale Mountains National Park (BMNP) in Ethiopia comprise the largest fractions of the Afro-alpine ecosystem in Africa, which provide vital mountain ecosystem services at local, regional, and global level” in abstract should be used the third person singular. Check carefully the full text.

（6）    Check the unit carefully in full text, e.g., km2 in line 118.

（7）    Check the formula carefully in full text, e.g., equation (1) in line 239.

（8）    The map components should be included in mapping, e.g., the scale, legend, north arrow. See the figures 6-9.

In all, it is a relatively simple task to generate the DEMs using the commercial software and the existed interpolation methods. 

Can be improved.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The topic studied in the paper is relevant and important in a study area that requires conservation measures to preserve its natural values. The study involves surveying and photogrammetry but, in my opinion, these techniques were not dealt correctly, leaving many doubts about the results.

At my first reading of the abstract, what started to catch my attention was the excessive number of decimal places used to present volume and area results. All are given with 2 decimal places in square meters, or cubic meters. Unless you are making very accurate topographic measurements, result presentation with so much unrealistic precision is completely inappropriate. In fact, in the case of volumes, error estimates of around 30,000 cubic meters were given. With an error of this order, does it make any sense to write the values ​​with two decimal places? It's like calculating a person's weight with an uncertainty of 0.1 kg and writing a result as 85.3456987 kg. This is unacceptable in a scientific paper.

Along the article I came across with table 4, which confirmed the previous observation. Does it make any sense to show the position of a camera on a plane flying at 10 km altitude to the tenth of a milimeter? And altitude up to the hundredth? Reading results from a calculator, or from some computer program, and pouring them into a report without any critical spirit is not tolerable for first-year science or engineering students, much less in scientific research work of a technical nature.

These observations raised in me the doubt that the authors are in control of what they are doing. I detected a few more things that reveal a lack of rigor and knowledge of what is being done. For example, in the data ​​presented in the table we find altitude values ​​of around 13,400 meters. This altitude contradicts what had been said before (line 148), about the altitude of the flight, which was 31,000 feet above sea level (the acronym ASL stands for “above sea level”). This corresponds to about 9400 meters, and it seems that the authors added up the terrain altitude, about 4000 meters, assuming that the 31 thousand feet were above the terrain. However, a flight altitude value of 13400 meters is highly unlikely for an aerial photography flight.

In table 4 the same thing happens: values ​​of 10,500 meters appear, but it was mentioned before (line 156) that the average altitude of the flight was 7,600 meters. There is a lot of confusion here that leaves me with many doubts about any results obtained from this study.

Many other things along the paper contribute to my opinion:

Total station surveys were carried out. Was a local geodetic network used?

Was GPS used for some form of georeferencing?

What types of ground control points (GCPs) were used for the image orientation in Agisoft?

How were they distributed?

The region covered by the photos is much larger than the field survey area and there should be GCPs distributed throughout the area. In historical photos it is difficult to identify them, and even more so in a natural environment, without human constructions. Nothing is said about this.

Fiducial marks were used, but nothing is said about their photographic coordinates. Did this information exist in any camera calibration reports?

It is referred that the external orientation parameters were optimized, which the Agisoft Metashape program normally does by adjusting also the internal orientation parameters. What are the results? Did a new value for the focal length result? It would be important to see this. I don’t trust any photogrammetric result obtained…

In the topographic survey component there are also doubts. Many of the DEMs obtained from the surveyed points (figure 7) are obtained by techniques that should not be used. IDW, for example, tends to biase the DEM in the sample points, which is clearly seen in the figures. Why trying these methods? In topographic surveying, the method to generate a DEM is by creating a TIN (Triangulated Irregular Network) and converting it to a grid. Do the authors have any expertise in surveying?

Sorry but the study leaves many doubts about the quality of the results. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The article is very timely in defining how humans are altering land
in a way that is causing climate change.  The English is excellent.
These types of articles are very important in defining what we as humans
are doing to enhance the affects of climate change.  Thus the article
purpose is great.  The problem with the article is a significant amount
of quality estimates are shown with no real discussion of how they
are derived. In addition the early data sets are flown at extremely high
altitudes that produces weak elevations, and this dilemma is ignored.
If bundle adjustment error propagation is used one will see vertical
coordinate standard deviations may be larger than the data quality
that is shown.  This is a major flaw that needs to be addressed.  This
is because the authors are not photogrammetrists but are using
photogrammetric software slightly blindly as example in not mentioning
aerotriangulation/bundle adjustment.
Line 24 posing shoud be poses
Line 31 , Diff...  lower case d
Line 91-92 sentence not required as same said in previous sentence
Figure 1 - some text too small or color of text blends in with background
Line 147 - why former?
Line 151 - "and" astronomical
Line 159 - does "units" mean "pixels"?  use latter
Page 5-5 - I am not sure if Table 1-2-3-4 and Figure 2-3 provide any value to
the overall intent of the article
Line 182 - EDM is part of total station so not needed in discussion
Line 197-200 Sfm and multi stereo is now standard production not unique
Line 202 - Bundle adj. prior to SfM could resolve camera calibration
if needed so this is not new but necessary for non metric cameras
in UAV - your cameras were calibrated which is more precise than
self calibration can produce
Line 209 - and the becomes "for the"
Line 211 - how well did the estimate image quality turn out?
Line 217 - why generate self calibration not needed?
Line 219 - why no mention of bundle adjustment (it is attempted
to be discussed later in a table but not very expicit needs work)
Line 219 - discuss how many control points and their fit in bundle adj.
Line 305 - units are pixels of reprojection errors
Table 5 - you only discuss horizontal not vertical quality
Line 311-321 - how do you know the same datums exist across the
3 datasets?
Line 321 - what is stronger - you mean more change
Figure 6 - larger text not readable
Line 346 - what does chronology have to do with data quality?
Line 361 - how is uncertainty determined?
Line 364-365 - uncertainty is almost 50% of the volume - thus does
it have any real meaning?
Line 372 - again an uncertainty that is large % of total does this
have any real meaning then?
Figure 8 & 9 - text too small
Line 463-469 need multiple sentences
Line 485-486 - Due to flying height - the photogrammetry does not
produce vertical quality to the accuracies you are stating

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Dear Sir/Madam

 The paper entitled “Long-term Volumetric Change Estimation of Red Ash Quarry Sites in the Afro-alpine Ecosystem of Bale Mountains National Park in Ethiopia” by Muhammed M. A. et al. presents an effective method using HAPs to calculate long-topographical changes quantitatively. The best interpolation method was also researched to calculate height information in unobserved areas with real surveyed data by TS.

 In this regard, the paper is well-written and very interesting to read. However, I see the following major and minor issues that should be improved before publishing this paper. My recommendations for paper revision are written on a separate page.

Best regards.

Reviewer

Questions and Recommendations:
Line 160: Table 1, Summarize flight parameters It would be better to arrange the contents of Table 1 according to the generally used  format in photogrammetry. Simplifying it in this way helps readers to grasp key points easily and quickly. My recommendation is as follows:   Table 1. Flight parameters of historical aerial photographs(HAPs) for the year 1967

  This also applies to Table 3. Anyway, please make sure both tables are in the same format.   Line 162: Table 2, Summarize projection center position In the same context as above, my recommendation is as follows:   Table 2. Camera exposure position for the year 1967…

  This also applies to Table 4.   Line 167: side overlap in Figure 2. This is single strip, where side lap is not appeared.   Line 192: Scale bars in the figure 4. The value and unit in scale bars are not correct in real.   Line 195-220: 3. Methods In general, we should input prior information (initial approximate values and their estimated standard deviation) on exterior orientation for every image as geotagging in SfM-MVS. Nowadays we get this from GPS and IMU on board UAV. How could you get this information?   Line 283: Figure 5 There is no mention of the Figure 5 in the text.

Line 358-379: 4.2 Volumetric and landscape changes, and Table 7

I think the results in Table 7 are derived from the same calculation software and method. Nevertheless, why are the uncertainty volumes from 1967-1987 for both sites particularly so large? These values are considered to be abnormal statistics, reaching 43% for site 1 and 33% for site 2, respectively. I expect your reasonable answer to these results.

Line 401: Table 4 Check the correctness of Table number. It should be Table 7 instead of Table 4.     Line 666: Reference No.46 The Journal Name is missed. Fill it to complete reference.  

–The end-

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 5 Report

The article "Long-term Volumetric Change Estimation of Red Ash Quarry Sites in the Afro-alpine Ecosystem of Bale Mountains National Park in Ethiopia" is a fascinating work and combines different data sources and methods to obtain important information for the alpine environment in Ethiopia. The article can be published after some minor changes are made as follows.

Why didn’t you use the same methodology for the 2023 DEM, I think that flying a drone was not  problem for the team of authors. I don’t say that the DEM resulted from the use of a total station it is not good, but I would have used the same type of dem as previous ones.

Fig 9. – You should present the final results, the DODs as classified rasters. Now you have a continuous scale and the results of your work could be more visible.

For the Evolution of surfaces in the quarry area I think that you should create a vector file with LULC for those 3 years and obtain a land use change vector. In this way, you can evaluate the surfaces of the ecosystems affected by the ash exploitation.

Congratulation!

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The modified version is improved. congratulation. 

Author Response

We would like to thank the reviewer for his time and appreciation.

Reviewer 3 Report

The updated version is a dramatic improvement and the authors should be congratulated.  Unfortunately it is very sad in many of my critiques (see the datum issue for example) they simply answered by comment instead of adding that information to the article.  My comments are to improve the article not receive a secondary response.  The authors still do not understand the difference between true variance-covariance error propagation in the bundle adjustment vs. "adding ground control" to improve data quality.  SFM is the precursor to the bundle adjustment.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf