Methodology of Mosaicking and Georeferencing for Multi-Sheet Early Maps with Irregular Cuts Using the Example of the Topographic Chart of the Kingdom of Poland

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Personally, I don’t like the way authors directly referred to in the text are only indicated by the citation number.

Eg ‘… was of interest to researchers such as [1], who …’ would read better as: ‘… was of interest to researchers such as Olszewicz [1], who …’

There is a similar issue with 2 & 3, but 4 & 5 are fine as they are indirectly referred to. In many cases the referencing method is appropriate, but particularly latterly in the paper the method used is frustrating for the reader. Simply referring to a projection or datum by it reference number is not acceptable.

Many of the figures are too small or of too low resolution to be useful. For several graphs/plots it is not possible to read the figures on the axes.

Page 2, line 49-51. This is not a proper sentence.

 P.2, l.52           grammar ‘… should not only be …’

P.2, l.69/70      awkward wording – perhaps ‘It was the result of practical work by students specialising in …’

p.2 l.83            ‘consumed’ is a poor word here – perhaps ‘incorporated’

p.4, l.171         there is reference to footnote 3, but no footnotes included with text. Is this a footnote in one of the references?

 P.7, l.235         In an article like this it is more appropriate to refer to ‘sections’ than ‘chapters’

 p.9, l.300         ‘foreign authors’ is not an appropriate term here. This is an international publications, not a Polish one. You could say ‘international authors’ or just simply ‘authors’. The following sentence highlighting specifically Polish contributions is fine. In the lists of references here, there is no apparent logic to the sequence. It would be more sensible to list references in date order (earliest first) as often later works build on earlier ones. I note that 106-108 are in date order, but 109-113 are not.

p.9, l.328         ‘… programmes were used as appropriate, …’ would be better wording

p.10, point 10 on list.  Just giving the ref to [128] is not really an appropriate way to describe the process. You should state that the transformation is to ETRF2000-PL, which makes the process much clearer. Make it easy for the reader to understand what is being done without having to refer to the reference list. The references show where things come from, or where to find more information about something; they are not a substitute for telling us basic facts.

p.10, point 11  again, this is not appropriate use of referencing. State what the files are being converted to.

p.10, point 13  see comments above about use of references

p.10, l.372       here you refer to sections, rather than chapters! Be consistent if nothing else

p.10, l.375       files don’t ‘weigh’ anything. poor terminology

p.11, l.392       Not well expressed. I think what you mean is ‘… cut individual panels from the scan then re-assemble the sheet, removing the gaps at folds.’

p.11. Fig 5       it would be clearer to illustrate a zoomed in portion to show the problem of the folds, then their resolution when patched together. Currently it is not clear what the issue is from the figure.

P.11, l.413       again this paragraph makes incorrect use of citations. Tell us the basics in the text.

p.14, l.456       Figure 8. There are no obvious differences between a & b at the scale shown. If the figure is to be meaningful, it needs to be significantly bigger, or extracts used that illustrate the differences. The points used in the transformation can’t be seen, so telling the reader to note their distribution is meaningless.

P.16, l.524       ‘… allowed the reduction of the file size to 1.2 GB …’

p.16, l.532       again, tell us what system was used, not a reference number

p.20, fig16       while the proportional circles are a useful indicator of the magnitude of residuals, they don’t give the full picture. It is comon to use a plot of displacement vectors which shows not only the magnitude of residuals, but their direction. This can help to show trends in certain areas, especially where there are larger residuals. It would be hard to illustrate a vector plot for the whole map at the scale required to fit the page, but some key areas could be highlighted in additional figures.

p.21 Conclusions     again there is reference to chapters, rather than sections.

While the conclusions highlight some of the issues, especially the large residuals that have resulted for many control points, there is a lack of investigation into their cause. Looking at Figure 16, there is some evidence of systematic errors in certain areas. It would be good to see some further examination of these areas. Are there issues with particular sheets in the series? Could this be as a result of their production, scanning, or processing prior to mosaicing? While just rejecting control points based on poor fit is not appropriate, it does indicate areas that could be targeted to ensure identification errors or other processing errors are not to blame. Looking at the direction of residuals (vector arrow plots) can often highlight systematic errors in a particular area, even if it cannot point to the specific reason for it.

Comments on the Quality of English Language

Generally it is very well written. I have highlighted a few issues with language. There are a few other places where improvements could be made to sentence structure, but generally these don't impact on meaning/understanding.

Author Response

Dear Reviewer #1

We are very thankful for your deep and valuable insight in proposed manuscript. The work indeed was laborious and challenging, therefore we are very thankful for your generally positive evaluation and helpful comments. We refer to comments point-by-point and then write general conclusion.

1. Personally, I don’t like the way authors directly referred to in the text are only indicated by the citation number.

We do not like it either, but we had to follow journal guidelines. This was done manually and indeed was annoying. We now see we took it to far and now tried to find reasonable consensus. All direct references were supplemented with the names of the authors, coordinate systems, etc. All changes to the text are marked yellow (please see attached PDF file).

2. Eg ‘… was of interest to researchers such as [1], who …’ would read better as: ‘… was of interest to researchers such as Olszewicz [1], who …’

Yes, this is fixed. Thank you.

3. There is a similar issue with 2 & 3, but 4 & 5 are fine as they are indirectly referred to. In many cases the referencing method is appropriate, but particularly latterly in the paper the method used is frustrating for the reader. Simply referring to a projection or datum by it reference number is not acceptable.

Yes, this is fixed. Thank you.

4. Many of the figures are too small or of too low resolution to be useful. For several graphs/plots it is not possible to read the figures on the axes.

Figures are generally rearranged:

• fig. 2 is resized,

• fig. 5a resized and problematic canvas was marked with red, fig. 5b removed, results are easy to imagine and can be seen on final product in online application

• fig. 6a and 6b compiled as one and resized

• fig. 8a resized and redesigned to emphasize calibration points, fig. 8b removed

• fig. 9a and 9b resized, additionally hi-res plots are added in supplementary data package

• fig. 10a and 10b compiled as one and resized

• fig. 11a removed, 11b resized to present the problem clearly

• fig. 15a and 15b resized, additionally hi-res plots are added in supplementary data package

• fig. 16 is redesigned to present radius, displacement vectors and extent of source earlier maps used for compilation. We decided not to add another figure with zoomed part. Instead, hi-res original size figure 420 x 350 mm, scale: 1:1,500,000, 600dpi is added in supplementary data package along with .JGW file for QGIS georeference. All georefenced data (10-minute grid, sheet rectification points, 1-st order and spline georeferencing points, all data EPSG:2180) are added in supplementary data to enable double-checking our calculations.

5. Page 2, line 49-51. This is not a proper sentence.

Yes, this is fixed. Thank you.

6. P.2, l.52 grammar ‘… should not only be …’

Yes, this is fixed. Thank you.

7. P.2, l.69/70 awkward wording – perhaps ‘It was the result of practical work by students specialising in …’

Yes, this is fixed. Thank you.

8. p.2 l.83 ‘consumed’ is a poor word here – perhaps ‘incorporated’

Yes, this is fixed. Thank you.

9. p.4, l.171 there is reference to footnote 3, but no footnotes included with text. Is this a footnote in one of the references?

Yes, this is now referred to appropriate section in text. Thank you.

10. P.7, l.235 In an article like this it is more appropriate to refer to ‘sections’ than ‘chapters’

Yes, this is fixed in every occurence. Thank you.

11. p.9, l.300 ‘foreign authors’ is not an appropriate term here. This is an international publications, not a Polish one. You could say ‘international authors’ or just simply ‘authors’. The following sentence highlighting specifically Polish contributions is fine. In the lists of references here, there is no apparent logic to the sequence. It would be more sensible to list references in date order (earliest first) as often later works build on earlier ones. I note that 106-108 are in date order, but 109-113 are not.

Yes, this is fixed. International contributions [75-96] are listed chronologically including volumes of journals. Polish contributions [97-110] and [111-115] are listed chronologically. This was done manually and indeed was annoying. We need to introduce some automatic methods for references in future works ;-)

12. p.9, l.328 ‘… programmes were used as appropriate, …’ would be better wording

Yes, this is fixed. Thank you.

13. p.10, point 10 on list. Just giving the ref to [128] is not really an appropriate way to describe the process. You should state that the transformation is to ETRF2000-PL, which makes the process much clearer. Make it easy for the reader to understand what is being done without having to refer to the reference list. The references show where things come from, or where to find more information about something; they are not a substitute for telling us basic facts.

Yes, this is fixed. We hope its now satisfactory.

14. p.10, point 11 again, this is not appropriate use of referencing. State what the files are being converted to.

Yes, this is fixed. Thank you.

15. p.10, point 13 see comments above about use of references

Yes, this is fixed. Thank you.

16. p.10, l.372 here you refer to sections, rather than chapters! Be consistent if nothing else

Yes, this is fixed in every occurence. Thank you.

17. p.10, l.375 files don’t ‘weigh’ anything. poor terminology

Yes, this is fixed. Thank you.

18. p.11, l.392 Not well expressed. I think what you mean is ‘… cut individual panels from the scan then re-assemble the sheet, removing the gaps at folds.’

Yes, this is fixed. Thank you.

19. p.11. Fig 5 it would be clearer to illustrate a zoomed in portion to show the problem of the folds, then their resolution when patched together. Currently it is not clear what the issue is from the figure.

Yes, this is fixed. See response to point 4. Thank you.

20. P.11, l.413 again this paragraph makes incorrect use of citations. Tell us the basics in the text.

Yes, this is fixed. Thank you.

21. p.14, l.456 Figure 8. There are no obvious differences between a & b at the scale shown. If the figure is to be meaningful, it needs to be significantly bigger, or extracts used that illustrate the differences. The points used in the transformation can’t be seen, so telling the reader to note their distribution is meaningless.

Yes, this is fixed. See response to point 4. Thank you.

22. P.16, l.524 ‘… allowed the reduction of the file size to 1.2 GB …’

Yes, this is fixed. Thank you.

23. p.16, l.532 again, tell us what system was used, not a reference number

Yes, this is fixed. Thank you.

24. p.20, fig16 while the proportional circles are a useful indicator of the magnitude of residuals, they don’t give the full picture. It is comon to use a plot of displacement vectors which shows not only the magnitude of residuals, but their direction. This can help to show trends in certain areas, especially where there are larger residuals. It would be hard to illustrate a vector plot for the whole map at the scale required to fit the page, but some key areas could be highlighted in additional figures.

Yes, this is fixed. See response to point 4. Thank you.

25. p.21 Conclusions again there is reference to chapters, rather than sections.

Yes, this is fixed. Thank you.

26. While the conclusions highlight some of the issues, especially the large residuals that have resulted for many control points, there is a lack of investigation into their cause. Looking at Figure 16, there is some evidence of systematic errors in certain areas. It would be good to see some further examination of these areas. Are there issues with particular sheets in the series? Could this be as a result of their production, scanning, or processing prior to mosaicing? While just rejecting control points based on poor fit is not appropriate, it does indicate areas that could be targeted to ensure identification errors or other processing errors are not to blame. Looking at the direction of residuals (vector arrow plots) can often highlight systematic errors in a particular area, even if it cannot point to the specific reason for it.

Well, we must agree this issue was the problematic one. We cannot provide deep investigation on the cause at this stage, as this was not the goal we focused on.

• Definitely this is not the problem with production, paper ageing, scanning or mosaicking, because it would occur at the sheet rectification stage. Sheet displacements never exceed 250 m (around 2 mm on the map) in any direction. We enclose hi-res plots and full CSV table of rectification points as a proof.

• At georeferencing stage no control points were rejected, on the contrary, even GCP's with residual exceeding 4000 m (ca. 32 mm on the map) were kept. We enclose hi-res plots and full CSV table of georeferencing points as a proof.

• Vector arrows revealed some more systematic patterns that are rather region-dependent not sheet-dependent. Please take a closer look hi-res map enclosed to supplementary data. This is already georeferenced for convenience.

• We added outlines of source material used for TKKP map compilation, this may add some ideas that we described in additional paragraph (lines 601-622). To be 100% sure several other laborious analyses of source maps are necessary.

27. Generally it is very well written. I have highlighted a few issues with language. There are a few other places where improvements could be made to sentence structure, but generally these don't impact on meaning/ understanding.

As for a non-native English speakers this is always problematic. We help ourselves with several modern technology translations, spelling and grammar check apps. Yet the specific terminology and language nuances are difficult to track and fix ahead of the submission. Therefore we are very grateful for your professional and precise suggestions. We feel our manuscript was taken seriously, we appreciate that. We wish all reviews were this specific.

 

Last conclusion. As you noticed there is no separate discussion section in general. We hope that prompt evaluation of existing early map geoportals (section 1.1) combined with detailed discussion on research procedure (section 2) and following subsections would be sufficient. We believe this paper would be the first comprehensive summary of the valuable contribution of Polish researchers to maps accuracy evaluation for an international audience. We do agree that still there is a lot of work to be done with early topographical maps of Europe (particularly broadly understood Polish Territories as our main area of interest). We hope our work would be useful both for researchers and for the Public. We also hope that detailed description of methods would encourage other researchers to follow with elaborating and, more importantly, sharing their works as Creative Commons OGC services.

We hope you find this revisions satisfactory.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have undertaken a very interesting, time-consuming, but also very important work that will allow the use of the information potential of unique and valuable maps by both specialists and the public user.

In addition to the description of the implementation of the main goal, attention is drawn to the part concerning the description of the circumstances of the map's creation and technical aspects. 

There is no separate section for discussion; however, it should be noted that the authors in individual chapters refer significantly to previous research. Perhaps there is a lack of reference and discussion to the entire project being implemented, and not only its individual aspects described in subsequent parts of the article. However, in my opinion, the body of cited literature is very extensive and the cited publications refer significantly to the presented research.

 

 

 

Author Response

Dear Reviewer #2

We are very thankful for your valuable insight in proposed manuscript. The work indeed was laborious and challenging, therefore we are very thankful for your positive evaluation.

We do agree that still there is a lot of work to be done with early topographical maps of broadly understood Polish Territories. We hope our work would be useful both for researchers and for the public. We also hope that detailed description of methods would encourage other researchers in Poland to follow with elaborating and, more importantly, sharing their works as OGC services.

As you noticed there is no separate discussion section in general. We hope that prompt evaluation of existing early map geoportals (section 1.1) combined with detailed discussion on research procedure (section 2) and following subsections would be sufficient. We believe this paper would be the first comprehensive summary of the valuable contribution of Polish researchers to maps accuracy evaluation for an international audience.

The revised version of the article includes several changes to improve its clarity and quality. This especially focuses on figures to make them more representative and easier to read. Last but not the least, we are a bit concerned about the length of the article, which is 28 pages already. We hope you find this revisions satisfactory.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

I have no comments for this paper, only congrats to the authors for its quality.
The paper describes a very interesting way to approach antique maps.

Author Response

Dear Reviewer #3

We are very thankful for your valuable insight in proposed manuscript. The work indeed was laborious and challenging, therefore we are very thankful for your positive evaluation.

We do agree that still there is a lot of work to be done with early topographical maps of broadly understood Polish Territories. We hope our work would be useful both for researchers and for the public. We also hope that detailed description of methods would encourage other researchers in Poland to follow with elaborating and, more importantly, sharing their works as OGC services.

As you noticed there is no separate discussion section in general. We hope that prompt evaluation of existing early map geoportals (section 1.1) combined with detailed discussion on research procedure (section 2) and following subsections would be sufficient. We believe this paper would be the first comprehensive summary of the valuable contribution of Polish researchers to maps accuracy evaluation for an international audience.

The revised version (attached) of the article includes several changes to improve its clarity and quality (marked yellow). This especially focuses on figures to make them more representative and easier to read. Last but not the least, we are a bit concerned about the length of the article, which is 28 pages already. We hope you find this revisions satisfactory.

Author Response File: Author Response.pdf