On-Site VIS-NIR Spectral Reflectance and Colour Measurements—A Fast and Inexpensive Alternative for Delineating Sediment Layers Quantitatively? A Case Study from a Monumental Bronze Age Burial Mound (Seddin, Germany)
Round 1
Reviewer 1 Report
Review for Haburaj et al. : On-site measurement of RGB and VIS-NIR reflectance - a fast and inexpensive alternative for delineating sediment layers quantitatively? A case study from a monumental Bronze Age burial mound (Seddin, Germany)
MAJOR MATTERS:
The authors assume a bit too much out of a general, scientifically literate readership. I had a difficult time following their unqualified acronyms and unclear references to statistical packages, and I was reading quite closely. They need to step back from the paper a bit and offer a bit more patience in the way they explain their methods. I have offered some insights below where this could be improved, but these aren’t 100% of the copy edits that could be made. I urge the authors to reflect a bit more for readers of Heritage journal, who may not have the identical expertise in photogrammetry that they have.
The presentation of the Results (Section 3) is haphazard and confusing, often using quite subjective language. I think that Section 3.1 needs to be rewritten for clarity because I cannot see the relationships between these layers (and Figure 5 makes it even more confusing). The authors discuss “connections” and delineations between layers as if they are animated objects, but these are sedimentologic packages. What do the “clusters” mean (statistically)? Maybe this belongs more in a Discussion section as you are talking about aggregations and relationships between data points? Another instance that is emblematic of the confusing data presentation is found in Lines 367-368 where you say, “This leads to more homogeneous and accurate results, which is mostly visible in layer A and the right parts of layers B and D.” What are the “right parts” of layer A? Are there “wrong parts” of layer A? Any assessment of the results in terms of accuracy should be done in the Discussion section. Overall, the results appear compelling, but the presentation of it needs a bit of tuning up to make it easier to follow.
I think the Discussion is quite robust and clear (certainly the best section of the paper, in my opinion). One missing concept that I think would fit nicely into Section 4.2 (and/or Section 4.3) is that I think it would be a good idea to broaden the discussion to consider the applicability of this method to other types of archaeological profiles (e.g., palaeosols, house fill features, open-air sites). Burial mounds are pretty unique settings, and if this paper is going to be useful in a broad sense, it would nice to see what the authors think about its broader application. They settle on RGB images (and EDXRF) as the best tools for resolving stratigraphic boundaries, but in the contexts in which I typically work (tropical laterite soils), I am less confident this would be useful. This may not be a nice, clean answer, but it would be informative.
MINOR MATTERS:
Abstract:
Line 1: “performed in the laboratory” does not need to be set off in commas.
Line 8: Eliminate “the”
Line 11: Move “(ii)” to in front of “selectively sampled”
Line 13: You need to explain what CIE L*a*b* is before using this abbreviation like this (e.g., color spectra visible to the naked eye).
- Introduction:
Line 23: Comma needed after discipline
Line 40: northern Germany should not be in parentheses. You should just set it off with a comma.
Line 59: Same critique as in the abstract…you need to define CIE L*a*b* before just putting it into the paper with the assumption that all of your readers clearly comprehend what it means.
Line 70: What does coll. and ger. mean? (Don’t abbreviate here)
Line 73: Nordic needs to be capitalized.
Line 95: If using British English conventions, meters should be metres.
Line 97: Change datings to “ages”. Also, calibrate all 2-sigma 14C ages to the decade to eliminate false precision.
Figure 2 strata descriptions are pretty wanton. I understand that the point is to use the automated methods to refine the descriptions, but some indication of bedding features and granularity interpretations from the field context would be nice to see, especially in light of how they might change after the multispectral analyses.
- Materials and Methods
Lines 118-119: You need a hyphen between 2 and mm.
Line 125: Eliminate the hyphen after mass.
Line 128: Change “on” to “for.”
Line 137: I would call this a “meter” not a “checker.” Hanna has some instruments that are called “Checkers” but these are brands and should be capitalized.
Line 141: Add “an” in front of “anti-coagulation agent”
Line 146: Eliminate the hyphen in front of volume and don’t abbreviate.
Lines 163, 178, 303: You indicate that some mathematical operation was done in a “package,” but in the sentences you are using them, it is not entirely clear that these are R packages.
Line 198: Change “were measured” to “have been previously published”
Lines 206-207: Change hyphens to commas.
Line 209: metres not meters.
Line 217-218: No hyphen needed between georeferencer and plugin or feature and scaling. Georeferencer should also be capitalized.
Line 223: What is TPS?
Line 224: Repeats from Line 218. Eliminate the redundancy.
Line 228: Is the “raster package” an R package?
Line 241: Principle is misspelled.
Lines 249 ff: All of this should have been above after the first mention of CIE L*a*b*.
Line 286: Citation not in the correct format for publication.
Line 321: What is “1 mass-%”?
Line 323: Eliminate “very”
Line 325: Content is a non-countable noun and should always be used in singular form.
Line 327: Incorrect use of the word “remained”
Line 328: Eliminate “highly”
Line 329-330: I do not understand how a layer (A) can show “continuous trends.” What does this mean?
Figure 5: The meanings of the colours for the clusters are not apparent to me. The figure caption says that they indicate “groups,” but what are the groups? The colour variance is quite nuanced and I do not see how these fit together.
Line 331: Clustering results are presented in Fig. 5, not 4.
Line 358: Mayor should be major
Lines 362-373: This is more a Discussion paragraph than a Results paragraph.
Line 385: “Much more” is too subjective
NOTE: Due to my time constraints, I did not have time to check the Online Resources (including R code). I would need some more time to do this, but the associate editor is pressuring me to get the review completed, so this is what I can manage.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Taking the monumental Bronze Age burial mound (Seddin, Germany) as the study site, the manuscript presented and discussed the feasibility of RGB and VIS-NIR techniques for the on-site delineating sediment layers. The comparision of aforementioned techniques with classical approaches (e.g. sedimentological data) implies the potential of those remotely sensed approaches in the workflow of archaeological excavations. This manuscript is publishable in "Cultural Heritage". My only recommendation is that: an entire workflow applied in this study (in the form of a figure in the Material and Method Section) is benefical for a better understanding of the research design as well as the logical relationship of research contents.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Please see the attachment below.
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
Stratigraphic documentation and interpretation are crucial parts of archaeological research. The excavated section investigated was first analysed with usual methods and then the digital approach was applied and compared. It is a interesting work for provenance and characterization of sediments in sections.
Recently digital techniques are employed by researchers in the fieldwork of archaeological digs (1) Mediterranean Archaeology and Archaeometry, Vol.14, No 4, pp. 109-116. OPENDIG: contextualizing the past from the field to the web Vincent, M, Falko Kuester and Thomas E. Levy), also:
(2) DOI: 10.5281/zenodo.3457524 The reassessment of the roman military presence in galicia and northern portugal through digital tools: archaeological diversity and historical problems by José Manuel Costa-García, João Fonte and Manuel Gago, Mediterranean Archaeology and Archaeometry Vol. 19, No 3, (2019), pp. 17-49),
or on artifacts (Bratitsi, et al 2018) in MAA, Vol. 18, No 2, (2018), pp. 175-212. DOI: 10.5281/zenodo.1297163 “Critical assessment of chromatic index in archaeological ceramics by munsell and RGB: novel contribution to characterization and provenance studies” (it is open access www.maajournal.com).
The proposed work of using extensive colour measurements on sediments from an archaeological stratigraphy was conducted with a RGB and a multispectral camera during fieldwork, selectively sampled and visible and near infrared (VIS-NIR) hyperspectral data, both acquired in the laboratory, including the influence of colour transformation to CIE L*a*b* and the possibilities of predicting soil organic carbon (SOC) based on image data is a promising combined non destructive technique, besides other mentioned by authors and the prediction of soil properties via VIS-NIR
- Explain the peaks seen in Fig.3
- The darkroom and uniform lighting conditions should be a bit more elaborated regarding distance, reproducibility.
- Have the RGB values been clustered to explore groupings? Why not?
- CIE L*a*b*, is extremely sensitive to point differences on a surface. It is most representative to overall chromatic evaluations. And how this is related to organic content? Was this issue dealt with ? I did not see this. Yet the clustering for layers identification is a good practice and the authors did it well.
- Why phosphorus was not recorded by pXRF? It correlates with human activity.
- Fig 5 must be better discussed. Correlate and delineate the optical spectroscopy measurements with stratigraphy obtained and compare with regular techniques. The reliability of this spectral technique I expect to see a probability model of how accurate and reliable it is for adoption in future works. May be a further elucidation in the discussion with Fig 6 is worth.
- Overall the work falls within science and archaeology cultural heritage the geoarchaeology and archaeometry. A relevant recent works is worth tied with the geoarchaeological approach beyond heritage alone.
- Overall it is interesting article worth of publication.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
Thank you very much for attending my comments and suggestions.
Reviewer 4 Report
all well corrected.
