Monitoring of the Rehabilitation of the Historic World War II US Air Force Base in Greenland

Round 1

Reviewer 1 Report

Interesting article especially in the context of nature conservation. It is well constructed, but it has some shortcomings:

1. The title of the article does not fully reflect the research carried out (in total, monitoring of the removal of barrels).

2. There is little literature showing similar studies. The introduction lacks information on the methods used in such renovation monitoring elsewhere. In total, only 20 items were indicated (including two related to programmes).

3. 9 items in the introduction do not refer to general information, mainly related to, for example, the location of places, and those related to the possibilities of using remote sensing are missing.

4. Laser (LiDAR?) measurements are mentioned, but not included in either the objectives or the results. Were they used in any way in the analyses?

Comments for author File: Comments.pdf

English is not my native language, but in my opinion the language used is quite correct and understandable. Minor errors were noted (e.g. use of incorrect tense, see pdf).

Author Response

Dear Revewer, thank you very much for your valuable comments.

-practical corrections have been accepted and corrected in the text (in yellow), thank you very much.

“The title of the article does not fully reflect the research carried out (in total, monitoring of the removal of barrels).”

• The reviewer is correct that the only act of rehabilitation so far was the removal of the barrels. However, that does not mean that it should not be considered a rehabilitation, as no further actions had happened in the area.
• Here it is very problematic to define what is meant by rehabilitation. The Greenlandic government has neither the means nor the technology; the Danish side, after long delays, has started clearing the scrap (not just the barrels); but Covid19 has stepped in and things have slowed down. Since the author has been there and measured, we can say that there has been no major damage to nature and it is the tens of thousands of rusting barrels that are spoiling the cleanliness of the area. There is certainly no plan to remove the runway, as that would be pointless. Nature has already absorbed many of them.

“Laser (LiDAR?) measurements are mentioned, but not included in either the objectives or the results. Were they used in any way in the analyses?”

• The laser scanning is mentioned with reference to other activities carried out in related projects on the CTU FCE part, nevertheless, in presented research the laser measurements were not used.

“For completeness, it is worth supplementing the spectral curves with the values for NDVI and NDWI”

• The inclusion of the NDVI and NDWI indices in the spectral curves would certainly be interesting. However, the reflectance values of the individual bands and the values of these indices differ by two orders of magnitude in our case, and it is therefore not possible to display them together in a single plot. The values of NDVI and NDWI would appear as zero in such a plot and could be misinterpreted by the potential reader.

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear authors,

You describe a very interesting approach in studying and monitoring a remote and hardly accessible WWII site and its remediation process by remote sensing methods. The paper is doubtlessly worth being published. However, some improvements have to be made in advance:

Text & content:

·       In section 1, you could perhaps also refer to other research approaches dealing with detecting and mapping WWII remains in Greenland. Most of those I know, are dealing with detecting crashed airplanes, e.g. https://www.sensoft.ca/wp-content/uploads/2017/08/pulse_EKKO_PRO_Glacier-Girl.pdf or https://www.cambridge.org/core/journals/journal-of-glaciology/article/search-and-recovery-of-aircraft-parts-in-icesheet-crevasse-fields-using-airborne-and-in-situ-geophysical-sensors/4423F97D70891DC1BA3B7DD27C08A1F5

·         In figure caption 3, you mention that you also did a 3D documentation of some remains there. Perhaps, you could briefly show some of these results, as they also depict an interesting topic for the reader.

·         Section 3.1: You have to mention that the reasons for the failure of UAV classification will be described in detail in sections 3.2 and 4.1.

Figures:

·         Some of you figures are much too small to be understandable for the readers, e.g. figures 5, 15 & 16.

·         Same is valid for some legends, e.g. in figures 11, 12, 17, 19 & 20

·         In figure 6, some zoomed-in views would be nice to show the level of detail of your orthophoto.

·         Figure 8: Mark one sample area with barrels to show the better visibility in NIR.

·         Figure 13: The letter size of the text is too small. Furthermore, you should use the same order and colour of the channels in all sub-diagrams.

·         In figure 18, it would perhaps be helpful to add another column with the total change rate of the time span of 2020-2021 in addition to the one of the whole time 2019-2021.

 There are several minor spelling, grammar and typing errors distributed in the whole text. Please just read it once more to eliminate them.

Author Response

Dear Reviewer, thank you very much for your valuable comments.

“Figure 13: The letter size of the text is too small. Furthermore, you should use the same order and colour of the channels in all sub-diagrams.”

• We have enlarged the font and assigned a specific color to each channel. However, we believe that for clarity it is better to sort the feature importances of each channel in descending order, not so that the individual bands are always in the same place. If they were always sorted in the same order, it might not be immediately clear to the reader what the actual order of the feature importances of each channel is.

Reviewer 3 Report

This paper presents a method to estimate the number of barrels in an abandoned air bases. and to monitor the progress of rehabilitation. Although the method is based on the GRASS GIS software, the research points out a new application field of the remote sensing. There are some modifications below.

(1) The paper resampling the high resolution data to a lower spatial resolution, but the resampling method is not explained.

(2) The images from WorldView3, GeoEye1, SPOT-6 have geometric accuracy error, which means the air base in different images is not matched in the same geometric accuracy error. Dose the authors deal with this problem or it does affect the results?

Author Response

Dear Reviewer, thank you very much for your valuable comments.

“The images from WorldView3, GeoEye1, SPOT-6 have geometric accuracy error, which means the air base in different images is not matched in the same geometric accuracy error. Dose the authors deal with this problem or it does affect the results?”

• The satellite data was resampled and georeferenced, the geometric accuracy error was not further investigated. We believe that is not an issue in the task where the objective is to estimate the area instead of mapping precise location for consequent field measurements or related tasks.