Comments on the Core Cadastral Domain Model (V2) a Danish View

Comments on the Core Cadastral Domain Model (v2) – A Danish view

Erik STUBKJÆR, Denmark

Key words: Cadastre, Cadastral Systems, COST G9, Modelling, and Real Property Rights

ABSTRACT

The paper “Further Progress in the Development of a Core Cadastral Domain Model” by Christiaan Lemmen and Peter van Oosterom presents a draft cadastral core model, version 2. The purpose of the present paper is to compare the presented model with Danish institutions: To what extent does the model fit the Danish reality. The comparison gives rise to some amendments to the presented model. However, more basic doubts are expressed as regards the structuring of the classes of the legal part of the model.

A closing section briefly presents an alternative to the basic person-rights-parcel structure of version 2. This alternative draws on a view of the cadastral domain, presented by the author in 1999.

1. INTRODUCTION

The Core Cadastral Domain Model of Christiaan Lemmen and Peter van Oosterom is structured into different ‘packages’ or aspects (2003: 4):

• the real core: The Person- Right- RealEstate structure
• the legal/ administrative aspects
• the real estate object specialisations, and
• the geometric/ topological aspects

The following section 2 addresses the real estate (geographical) object classes, while section 3 discusses cross-package issues and the real core. In both cases, the proposed Model is discussed from the point of view of Danish statutory definitions and legal doctrine. Some judgements are made on the basis of research activities within the COST action: Modelling Real Property Transactions, cf. (COST G9 homepage; Stubkjær, 2002).

An introduction to the term cadastral system may be useful. In Denmark, the Cadastral Agency for many years was an instrument for implementing agricultural policies, and thus belonged under the Ministry of Agriculture. Updating and other fieldwork has always been in the hands of private practising, chartered surveyors. The Land Registry was established at the about 80 local courts. From the mid-1800s, the two property systems have used the same cadastral identifiers. A Land Registry reform (Danish: “Tinglysningsloven, 1926”) confirmed on the one hand a mutual co-operation in terms of cadastral identifier and exchange of information, while it on the other hand allowed for a certain independence, also in the definition of units of real property. The two systems taken together are in this paper termed the cadastral system, as motivated in (Silva and Stubkjær, 2002: 408f).

In the following, reference to the paper by Lemmen and von Oosterom is made only by page number, e.g. (: 5). Danish terms are provided as (“Danish term”) to allow for check of the meaning. The paper closes with a brief presentation and motivation of an alternative structure of core cadastral classes.

1. COMMENTS ON THE REAL_ESTATE_OBJECT AND ITS SPECIALIZATIONS

The conception of a unit of real property was shaped in Denmark through legislation on agricultural holdings, and on court rulings on real property transactions, respectively.

The agricultural legislation aimed for more that two hundred years at keeping holdings of a size that could sustain and occupy a family. Statutory definitions and their implementation through cadastral practise shaped holding units that comprise either one identified parcel, but more often several identified parcels. The proposed Core Model complies with this by the specialisation Parcel (Danish: “areal betegnet med matrikelnummer”) and ParcelComplex (“flere matrikelnumre, der er noteret som en samlet fast ejendom”). The text commentary (: 5) does not spell out, whether an instance of Parcel could be made up of disjunct parts and include holes, and how this relates to the class PartOfParcel.

The PartOfParcel (Danish: “delnummer”) is surely needed for change processes as the commentary indicates, but it might also be conceived as a transitory specialisation of PartOfParcel. The motivation for this is that after the digitisation of the Danish cadastral map, the National Survey and Cadastre is recording the boundary of statutorily created easements regarding forests (“fredskovgrænser”). Such boundaries and areas constitute thus a specialisation of permanent PartOfParcel. Finally, it may or may not be appropriate to establish a specialisation of PartOfParcel, the instances of which are simple (without holes) and strongly connected (cf. Worboys, 1995: 112). This issue is left open here.

In the Danish setting, Parcel is normally owned individually, but occasionally archaic joint ownership to a parcel ("fælleslod") still exists. The joint ownership constitutes (informally) a legal person, and according to modern standards, such legal persons ought to be formally established and recorded as owner of the parcel in the land registry. It might be within the competencies of the National Survey and Cadastre to establish the legal persons and the corresponding recording ex officio. Unfortunately, however, another Danish rule goes against this solution: In order to prevent scatter habitation, the tax and cadastral authorities are working together to achieve that every parcel outside urban areas is part of an agricultural holding, and thus imposed with pertinent regulations. A modernisation of the recording of joint ownership would imply that the parcel was released from the diverse holding units, which had a share in it, and made an independent property unit. This would compromise the efforts against scatter habitation, and thus the modernisation cannot be achieved without political commitment, which is difficult to achieve for such details.

The above story of joint ownership to a parcel is presented to back the position that an object class has to be available in the Core Cadastral Model for archaic constructs, which for diverse reasons are too difficult to get rid of. The class NonGeoRealEstate might serve this purpose, but the present text commentary (: 6) overspecifies the class and has to be rephrased. Also, it appears as not convincing that the Core Model treats localisation issues (e.g. incomplete geometric description) in the context of RealEstateObject specialisation, when a whole (pink and purple) ‘package’ is devoted to geometric/ topological aspects.

Summarising, the specialisations of RealEstateObject - with minor adjustment – can accommodate for the cadastral statutory definition of a unit of real property rights in Denmark, the “samlet fast ejendom”. However, it has to be taken into consideration that the cadastral unit is not the only unit of the cadastral system of practical importance. The following regards some of these related unit classes, and the ApartmentUnit and ApartmentComplex of the proposed model.

The Danish Land Registry Act (“tinglysningsloven”) uses a concept (“bestemt fast ejendom”) that has a wider meaning than the statutory cadastral unit, and has been shaped on the basis of court rulings through time. The Land Registry system thus allows for buildings being units of real property rights without the land they rest on, and allows also for condominiums. Rights in condominiums (“ejerlejligheder”) are recorded in the Land Registry system with reference to documents, which are prepared by a chartered land surveyor. Loosely speaking, the land surveyor establishes a micro-cadastre (a list of the units with identification and area, and corresponding maps) of the apartments of the property unit concerned. This is done without the involvement of the National Survey and Cadastre. The object class ApartmentUnit (: 5) would correspond to “ejerlejligheder” provided an ApartmentUnit could comprise of several parts, e.g. an ordinary apartment and a supplementary room in the basement.

The text commentary to ApartmentComplex that “there can be at most one ApartmentComplex located on one Parcel” resembles a clause of the Danish condominium act, which states that a unit of real property has to be totally subdivided into condominium units. However, sufficient correspondence with the Danish legal reality has not been achieved. Furthermore, to accommodate for the Danish building units (“bygninger på lejet grund”) you would need a class Buildings, and then relate ApartmentComplex to this class.

The Danish agricultural act and the act on assessment of real property for tax purposes have further variations of the unit of real property, which are not mentioned here. Concluding, the Core Cadastral Model provides for the cadastral notion of real property (“udstykningslovens ejendomsbegreb”), but not the other conceptualisations of units of real property in use within the cadastral system.

1. COMMENTS REGARDING THE OVERALL STRUCTURE OF THE MODEL

The overall structure of the model is discussed from several points of view: The object class Document is a component of both the geographic and the legal package. This motivates a discussion, whether there is a sufficient similarity in the way this component is treated. The other points regard the real core and the legal aspects of the Core Model.

It is characteristic of the legal/ administrative package that all object classes are related to the class LegalDocument. The motivation “that legal data [i. e. recording of rights] will not be changed without provision of a LegalDocument” is endorsed. It should be noted that rights could be changed without document, because - according to Danish legal doctrine – an agreement commits whether the agreement is evidenced on paper or not. Furthermore, positive prescription (adverse possession) changes the rights of the original owner and the user, again without provision of documents. However, the recording or ‘mapping’ of rights in the cadastral system demands an instance of LegalDocument, and it is the recording of rights, which is in focus here.

The motivation for the central position of LegalDocument was that “legal data will not be changed without provision of a LegalDocument”. This criterion could be applied for the geographical/ real estate object package as well: The change of boundary data of Parcel and other specialisations of RealEstateObject must also be based on documentation. This is in fact achieved by relating the SurveyDocument to the object classes ParcelBoundary and SurveyPoint. However, while the text commentary (: 6) fully accounts for the geometrical/ topological aspects of the relations, it does not mention the naming of Parcel and related classes, and the origin of naming in a Document. Naming is different from assigning a computer-internal identification, as Person (owner), other parties, and authorities need an easily comprehensible identification scheme. In Denmark, a document depicting a section of the cadastral map (“Ændringskortet”) is used for the naming of instances of Parcel and PartOfParcel. The concern for naming could be accommodated for by giving the class SurveyDocument a wider interpretation, and assuming that the ParcalBoundary – tp_edge structure carries naming information and can be applied to AppartmentUnit, too.

In the context of naming, mention is made of the ETeMII Reference Data White Paper, as it points to the importance of addresses and other place names (ETeMII, 2001). Place names are available in any country and culture. Furthermore, they are important in the context of cadastral systems, because they provide information on location. The location information is not precise as survey measurements can provide, but the information is relevant and understandable for the user of the cadastral system, and this is a key criterion. There is thus sufficient reason to include the object class of PlaceNames into the Cadastral Core, cf. section 4 below.

More basic problems are encountered as regards the legal aspects of the Core Model, including NonNaturalPersons like organisations, etc. It is surely relevant to account for Natural and NonNatural persons, but work so far within the G9 research activities suggest that it is the different roles of Person that are more interesting. Roles appear in the form of owner, holder of rights of different kind, surveyor, notary, cadastral agency, etc. Also, it is well taken that some sort of restrictions is only associated with the land (: 7), but to the knowledge of this author, this sort is better termed easement than encumbrance.

The hardest problem is however that it is proposed that ‘lookup tables’ shall handle legal issues of the Cadastral Core. Maintenance of such position may possibly become a kiss of death of the idea of a Core Cadastral Domain Model, because the position refrains from addressing the legal issues with sufficient detail. The appropriate structuring of the legal ‘package’ of the Core Model is surely a hard task, but if it is not addressed in a way that fits with practise, it is difficult to see how the objective of contributing to the implementation of effective cadastral systems can be accomplished.

1. AN ALTERNATIVE CORE STRUCTURE, BASED ON TRANSACTIONS

In a previous paper, the cadastral domain was described in graphical form (Stubkjær, 1999). For convenience, the diagram is presented here as Figure 4 (No figures 1..3 present here).

The diagram adopts the traditional Owner-Rights-Parcel structure, see the left column. However, TerrainObjects replace Parcels. The reason for this replacement is that Parcels are defined by human action, and as the Core Model has to be applicable across jurisdictions, such human action has to be standardised. The definition of Parcels thus has to be part of the interior of the Core Model.

TerrainObjects are the visible and discernible features of the terrain, which can be identified by software from imagery. Generally, Parcels cannot be identified in this way. To give examples: The windy West Jutland has lots of windbreaks. Some of these hedgerows mark property (and thus Parcel) boundaries, but not all. So you need supplementary evidence to establish Parcel boundaries. Furthermore, in waterlogged areas, some ditches mark Parcel boundaries, but not all. Generally, it is posited that the terrain, more specifically: the discontinuities of the terrain, does not provide sufficient evidence to establish Parcel boundaries, let alone appropriate Parcel identification. Consequently, human involvement is needed to establish Parcels (not shown in the diagram) from TerrainObjects.

The human involvement is depicted as Transactions, which update Databases, namely the database(s) of the Cadastral System. The Transactions are performed by Professionals (lawyers, notaries, etc. and civil servants), at the request of Owners. Owners have to be interpreted in the wider sense as HoldersOfRights. However as mentioned above, research within the COST G9 activity and its application in teaching suggests that roles and parties should be made more explicit; maybe even an object class of Agreements is needed. The Figure 5 below has been drafted to somehow accommodate for these considerations.

The left and centre columns of Figure 4 have largely been redrawn as the lower and centre row of Figure 5. However, the object class of RealPropertyRights has been shifted to the upper right corner of the diagram. This is due to the following considerations: The Owner-PropertyRight-Parcel structure holds in the situation, where no legal or spatial changes are made. In such situation, the cadastral system is out of focus. However, when Owner makes changes, either in the array of rights, which s/he is entitled to dispose of, and thus affecting RealPropertyRights, or in the boundaries of the holding, and thus affecting RealEsateObject, or in both together, then the cadastral system comes into focus.

Figure 5: ESt-proposal for Core Cadastral Classes, 2003

When the cadastral system is in the focus of Owner and other parties, the Owner cannot be allowed ‘direct access’ to the RealPropertyRights, because this might compromise the rights of other parties. For this reason, in every society the reorganisation of real property rights is socially mediated; it is performed as an elaborated ritual. This ritual is formalised into the institutions of cadastral systems. In Figure 5 this is expressed by introducing Transactions between Owners and RealPropertyRights. The location of the object classes of RealEstateObject and RealPropertyRights, respectively, is further substantiated by the fact that the objects of both classes are immaterial. The class diagram reflects this by layering classes of physical objects (TerrainObjects, Databases) below classes of energy consuming objects (Persons, Transactions), and further under the mentioned classes of communicative, immaterial objects.

The design criteria that data could be maintained by different organisations (: 3) is not neglected.

5. CONCLUSION

The Core Cadastral Domain Model by Oosterom and Lemmen has been discussed on the basis of Danish rules and practises, and COST G9 research activities. It was concluded that the proposed Core Model (version 2) does accommodate for the Danish statutory cadastral unit, but not for other Danish conceptions of units of real property. Amendments to the Core Model have been proposed, but also a more basic weakness pointed out, namely that the legal package was incomplete. An alternative Cadastral Core was presented. Design criteria for this included that the core classes should be robust across jurisdictions, and that the configuration of classes should reflect the fact that Owners are alienated from their RealPropertyRights during the socially mediated reorganisation of property rights and property boundaries.

REFERENCES

COST G9 homepage: with reference to COST homepage:

ETeMII (2001) Reference Data White Paper, 31. July 2001 cf.

Lemmen, C., van Oosterom, P (2003) Further progress in the Development of a Core Cadastral Domain Model. Paper to be presented at FIG Working Week, April 13.-17., 2003.

Silva, M. A., Stubkjær, E. (2002) A Review of Methodologies used in Research on Cadastral Development Computers, Environment and Urban Systems 26 (5) 403 - 423

Stubkjær, E (2002) Modelling Real Property Transactions. Paper presented at the XXII FIG Congress, Washington, DC, USA, April 19-26 2002.

Stubkjær, E (1999) Cadastral Research – Issues and Approaches Kart og Plan 59 (3) 267 – 278

Worboys, M (1995) GIS – A computing perspective. Taylor & Francis, London. 376 p.

BIOGRAPHICAL NOTES

Erik Stubkjær is professor of cadastral science at the Department of Development and Planning, Aalborg University, Dennmark, since 1977.

He has written numerous research articles and conference contributions with a view to articulate theoretical foundations of the European cadastres.

Interest in educational development made Erik Stubkjær the co-ordinator of the EU's Phare/ TEMPUS project: Improved Education on Environment and Infrastructure (1996-99). The project regarded the restructuring of the study programme of the Department of Geodesy, Faculty of Civil and Geodetic Engineering, University of Ljubljana, Slovenia, and resulted in a formally agreed renewal of the study programmes.