On the Accuracy of Cadastral Marks: Statistical Analyses to Assess the Congruence among GNSS-Based Positioning and Official Maps

Round 1

Reviewer 1 Report

Thank you for the second review opportunity. I have a little different point of view on some matters, but it is not an allegation. As mentioned before, each country has its own solutions, standards and tradition in land surveying and mapping. Therefore I generally accept the authors' explanations.

In the case of Banasik P., Borowski L., “Georeferencing of the Cadastral Map in the Krakow Region” – I am aware that the article hasn’t been published yet. That is why a wrote the paragraph to illustrate the issue, not to obligate you to cite the unknown paper.

Best regards and good luck in further research.

 

Reviewer. 

Author Response

Please see file in attach.

Best regards

On behalf of the authors,
Gino Dardanelli
(corresponding author)

Reviewer 2 Report

Title is too general, does not highlight the research topic.

The whole article is structured in the form of a technical report, structure the article as a problem-based research paper - analysis and resolution.

In order to be considered a research article, the presentation of the methodology and results should be reorganized and the technical, expedient nature of the presentation should be eliminated.

Author Response

Please see file in attach.

Best regards

On behalf of the authors,
Gino Dardanelli
(corresponding author)

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors explained their point of view in concrete terms. The article submitted for publication can be accepted.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The title is too general, it should be changed to highlight what is discussed in the article proposed for publication

In the abstract (at the beginning) a sentence should be introduced to highlight the importance of the study in its global context, to introduce the issue under consideration in the general context, in order to identify the importance, topicality and necessity of the study.

The whole article is structured in the form of a technical report, the methodology shows how the maps were made and how the databases were acquired and based on the results the planimetric and altitudinal differences are identified.

In order for the article to be considered a research article, some correlations should be made with existing case studies in the world literature, etc..

As it stands, the article only highlights one system used in a particular country and the problems it raises.

 

Reviewer 2 Report

The reviewed work is a comparative study of the accuracy of determining cadastral benchmarks using various geodetic methods and information systems of various types - maps and Google Earth. The work is of a technical nature and corresponds to the subject of the journal. Written in good language, optimally structured. The abstract is clear and fully describes the content of the work. The illustrations are well made and help in the perception of the material. The history of work in this area is described fully and interestingly. The references are quite complete. It was not possible to identify any comments to the text. The work can be published in its existing form. Scientific background of Google Earth, Digital Earth concept, was not even mentioned - it's pity. 

Reviewer 3 Report

Let’s start with the least serious issues:

1.      1. What are PHs exactly? Are they boundary points, which divide the land into different estates? Or maybe they are sort of control points, the lowest class? It is written that there is no topographic connection with the cadastral control network, so what are they? Each country has its own rules, methods and traditions in land surveying and there are difficulties to translate/explaining the national solution into English. Nevertheless, the “cadastral benchmark” sounds peculiar – the word “benchmark” is rather used for vertical control marks. Maybe an explanation that these points are both (as far as I understand) boundary points and cadastral control network, but old one and lowest class/accuracy etc.

2.      2.  Line: „the PFs cartographic coordinates are available from cadastral maps.” Have coordinates been defined for the PF points? Or still, do you have to catch the coordinates from the cadastral maps? Or maybe, the coordinates were defined by reading from the cadastral map? Does Italy have database/databases with boundary points coordinates and documentation? Does it meet the assumptions of the "cadastre 2014"? [1]

3.     3.   The most important information about the Italian CORS network is that it works in ETRF2000 e2008.0. In which geodetic datum cadastral maps are defined? The Cassini-Soldner projection was used, but what was the terrestrial frame? Is there any official technical note on how to convert the cadastral datum to ETRF2000? The authors in line 239 cite their old publication [2], but it doesn’t inform about the old reference frame. This is the key aspect. Before preparing any in situ measurements there must be a conversion to the same system, frame and (if possible) epoch.  Where is it in the manuscript? The only part is very short:

“According to Timar et al. [23], given the Cassini Soldner and WGS84 geodetic coordinates of the cadastral origins, the seven parameters of the Burša-Wolf transformation [24] and the three abridging Molodensky parameters [25] between the two datums, allowed obtaining the finest horizontal accuracy [26].”

I greatly appreciate the work of prof. Timar but is his publication a professional (official geodetic) standard of such works? Does the Surveyor General of Italy (or an official responsible for Italian geodesy) recommend this approach for conversion? Maybe other ways are in use like Bursa-Wolf, Molodenskii-Badakes [6], Helmert with Hausbrandt corrections [5]  etc? The instruction (standard) has values of transformation coefficients?

If not, how the work was done by the authors? They used GNSS observations and the PHs have coordinates in old datum. What is the average systematic error for the transformation (conversion) function?

 

Let’s look at fitting the old maps into current projection coordinates [3]:

“It should be assumed that the average accuracy of georeferenced obtained in this way results from the following factors:

(a) inaccuracies of the technology of field geodetic measurements applied in the mid-nineteenth century, as well as of the technology of map drawing and their reproduction,

(b) loss of the cartometric accuracy of a map drawn up on paper (paper shrinkage) as a result of the passage of time,

(c) inappropriate selection of topographic points used to assess the accuracy of georeferencing (unnoticeable changes in the location of points resulting, for example, from the reconstruction of the façades of buildings, structures),

(d) inaccuracy of scanning and digitization of reference points of the map border on the raster, as well as topographic points on the map,

(e) inaccuracy of the georeferencing algorithm.”

The authors of such publication showed the issues which may occur as systematic errors in fitting the old map into the modern coordinates system. Now, let us look at the authors' table 1. It shows the distribution of differences between the GNSS CORS measurements and other data such as Google Earth map (GE), Numerical Technical map (CTN, in the table, is “CTC”, but I think it should be CTN), Cadastral map (MC). The analysis of table 1 gives a simple answer: the CTN is referred to as a geodetic control network and very probably realised in ETRF2000e2008. The differences are not so high, the CORS has its own accuracy and the geodetic network also. There may be a shift of control network coordinates [4]. The comparison of MC and GNSS looks different – as written above the shift between them might be caused by (a)-(e ) reasons, or maybe because of mass conversion of all the cadastral maps into the ETRF2000e2008 frame. If the author would fit the cadastral map into ETRF2000e2008 on its own, the differences would be smaller. Probably. The PFs have the map coordinates and the ETRF2000e2008 (at least some of them), so the fitting by a suitable method would give better results. The point is that there is no information about that issue. What is the a priori accuracy of that map?

Conclusion:

The authors have skipped from the introduction to observation analyses smoothly. Not much has been written about the preparation of data and its analyses. I was very surprised to see more information about the GNSS receiver (it has Wi-Fi, memory slots and Bluetooth even etc) than about the core of the geodetic analysis. One last case that is not an allegation. The analysis as performed by the authors was very popular from 2008-2015 as a “first publication” for geodesy Ph.D. students. It was the time when country by country runs its own CORS network. So, I think the authors would be better  if they focused on cadastral data and its situation in Italy, rather than writing so much about GNSS CORS, VRS, FKP …. It is just a technique. A very sophisticated tool, but still just a tool. Maybe, the right journal is the one focused on land use policy or something like that.

[1] FIG, Cadastre 2014, https://www.fig.net/resources/publications/figpub/cadastre2014/index.asp

[2] Dardanelli, G.; Pipitone, C. THE EFFECTS OF CORS NETWORK GEOMETRY AND DIFFERENTIAL NRTK CORRECTIONS 649 ON GNSS SOLUTIONS.; 2021.

[3] Banasik P., Borowski L., “Georeferencing of the Cadastral Map in the Krakow Region”, The Cartographic Journal, https://doi.org/10.1080/00087041.2021.2023963.

[4] Wolski, B.; Granek, G. Functionality and reliability of horizontal control net (Poland). Open Geosci. 2020, 12, 668–677, doi:10.1515/geo-2020-0052.

[5] Ligas, Marcin and Banasik, Piotr. "Least Squares Collocation Alternative to Helmert’s Transformation with Hausbrandt’s Post – Transformation Correction" Reports on Geodesy and Geoinformatics, vol.97, no.1, 2015, pp.23-34. https://doi.org/10.2478/rgg-2014-0009

[6] Rod Deakin, A NOTE ON THE BURSA-WOLF AND MOLODENSKY-BADEKAS TRANSFORMATIONS, 2006