Incremental update and upgrade of Spatial Data

Incremental update and upgrade of Spatial Data Martin Scheu, Wolfgang Effenberg and Ian Williamson Summary This paper examines issues relating to the update and upgrade of the digital cadastral map. Of central interest is the transfer of incremental update and upgrade data from the custodian to the user and possible solutions for the associativity problem. The paper clearly differentiates between spatial update and upgrade and categorises three different methods for the transfer of incremental upgrade information. The impact on both the users and the custodian of the cadastral map are discussed in view of providing the most efficient and timely incremental upgrade information. A strategy that provides the maximum information to the users about the shift in the digital cadastral map and additional serves as input information for any adjustment procedures the user utilise to minimise the associativity problem. While this paper draws on the authors´ experience of maintenance of cadastral maps in the Australian states of Victoria and New South Wales and the German state of Berlin, the efficient delivery of updates and especially upgrades of the spatial cadastral data to users is of interest to users of spatial data. Zusammenfassung Die Geo-Basisdaten des Liegenschaftskatasters bilden die Grundlage für eine Vielzahl von GIS-Anwendungen. Dieser Artikel beschreibt den organisatorischen Kontext und die technischen Randbedingungen einer differentiellen Fortführung von digitalen Katasterkarten an Beispielen aus Australien und der Bundesrepublik Deutschland. Dabei wird zwischen Einfüge- und Löschoperationen (Update) und geometrischen Verschiebungen für existierende Objekte der Katasterkarte (Upgrade) unterschieden. Bei jeder Fortführungs-Homogenisierung tritt bei einem Großteil der Nutzer von digitalen Katasterkarten das Problem der nachbarschafttreuen Anpassung der Geo-Fachdaten auf. Diese Problematik wird aufgegriffen und anhand eines Beispiels aus dem Bereich der Versorgungswirtschaft erläutert. Darauf aufbauend wird eine Übertragungsstrategie für Upgrade-Informationen vorgeschlagen, die eine Minimierung des Aufwands für die Anpassung der Geo-Fachdaten beim Nutzer ermöglicht. Keywords Digital cadastral maps, Spatial data, Incremental Update and Upgrade, Associativity, Utility Industries 1

Introduction

Australian and German cadastres are rapidly moving to multipurpose systems. The spatial component of these systems is the digital cadastral map. These maps are now widely used by utilities and local government for planning, asset mapping, monitoring and documentation purposes. An important role of the custodian of the multipurpose cadastral map is to ensure its accuracy and currency and to efficiently transfer any changes to the spatial data to the users. This paper deals with the maintenance of the digital cadastral map and discusses some technical issues, especially the transfer of incremental update data from the custodian to the user and possible solutions for the associativity problem. The term digital cadastral map (DCM) is used in this paper as a synonym for the tem digital cadastral databases (DCDB) in 1

Australia (EFFENBERG and WILLIAMSON 1997) and for Automated Real Estate Map (ALK) in Germany (VOGEL 1993; EBERHARDT 1993). 2

Context of incremental update and upgrade

In common with many countries in the world, the individual states within Australia and Germany generated computerised cadastral maps using digitising techniques. In urban areas, survey control, field notes and field surveys were used to enlarge the number of control points and geometrical constraints. The derivation of digital cadastral maps from paper maps means that most of the features are represented at a graphical accuracy. This means that the DCM is subject to two different types of maintenance activities by the custodian, updates and upgrades. While from the customers’ viewpoint the ongoing process of maintaining the currency of the DCM is handled in the same incremental process it is still necessary to differentiate these two processes. This paper simply defines the processes of update and upgrade as follows: •

Spatial UPDATE of DCM refers to those processes that ensures that all geospatial objects appearing or disappearing in the real world are recorded in the DCM within a specified time period, i.e. the DCM is up to date;

•

Spatial UPGRADE of DCM refers to activities associated with increasing the level of accuracy for existing, geospatial objects.

The term maintenance is used to encompass both the processes of update and upgrade. In the paper Framework for Discussion of Digital Spatial Data Flow within Cadastral Systems Effenberg and Williamson (1999) show the similarities amongst a number of diverse western cadastral systems particularly in term of the flow of spatial data to update the jurisdiction wide DCM. This paper further contends that the problems associated with these digital cadastral systems are also similar world wide. The network of organisations involved in the flow of spatial data to maintain the jurisdiction wide DCM for both Germany and Australia are also surprisingly similar. This is evident from the following Figures 1 and Figure 2 where the major institutional organisational entities and their interaction are depicted for the cadastral systems in Australia and Germany. Figure 1:

Organisational network and spatial data interaction in Australia

Figure 2:

Organisational network and spatial data interaction in Germany

Clearly there are obvious similarities in these two figures but importantly there are some common spatial data transfers. The shaded arrow highlights the common need to transfer changes to the digital data from the custodians of the DCMs to the users of DCMs. Many of the custodians of the DCMs in both Australia and Germany are addressing the need to efficiently and effectively transfer these spatial data changes to the users. The thesis of this paper is that there are many shared problems and common possible solutions for Australia and Germany and indeed other cadastral systems world wide. Specifically this paper will focus on the transfer of changes from the custodian’s DCM to utility industries with their specific need to map and associate their underground assets to the boundaries depicted in the digital cadastral map. In the past changes have been transferred by supplying the users with a refresh copy of their complete map base. Alternatively the supply of refresh copies for lesser areas based on tiles or other predefined areas were also used. These refresh methods effectively hid the changes from the users who required a more transparent update method to allow an assessment of the impact 2

of the changes on associated spatial data. The incremental update, highlighted in the figures 1 and 2, and its benefits are clearly argued by DOMINGUEZ et al. (1994). There is of course a large number of users and suppliers to the DCM. In focusing on the transfer of incremental update information to the utilities only ignores other possible solutions that involving the whole cadastral system network. To fully explore the possibilities of including all these users and suppliers an analysis methodologies in the context of the Zachman Framework are useful (EFFENBERG and WILLIAMSON 1999). This approach is useful both analyse and accelerate processes of spatial data exchange and DCM maintenance. 3

Requirements of the Utility Industry

The Utility Industry (UI) represents a group of highly experienced users of GIS-Technology with specific requirements for their digital mapping based on spatial cadastral data. Most utility companies rely on the custodians for the supply of an up to date DCM as reference layers in their GIS. Broadly the cadastral layer is used as spatial reference for utilities digitising asset or infrastructure data without reference to its exact location on the ground. This digitising of UIfeatures associated with the cadastral layer may be visual process or by using the offset command from computer aided drafting systems (CAD). The latter technique is widely used in UI, especially water and natural gas where there is a need geospatial data representing the shape, true size and location of their facilities for the purposes of construction maintenance and administration. The description of the requirements of UI in relation to the DCM is provided by WAN and WILLIAMSON (1994a) in a paper "Problems in Maintaining Associativity in LIS with Particular Reference to the Needs of the Utility Industry" for the states Victoria and New South Wales; similarly for Germany by BERNHARDT (1994) in a book titled “GeoInformationssysteme in Energieversorgungsunternehmen“. These can catagorised as four main needs: (1) Easy to use data exchange standard for DCM first take-on and incremental update. (2) Just in time transfer of maintenance information for the DCM. (3) Adequate accuracy of the DCM to avoid problems with accurately surveyed points locating facilities of Utility Industry. (4) Minimised effort and costs for processing of update and upgrade data transferred from the custodian. This paper concentrates on point (4) by examining the content of the incremental update and upgrade information transferred from the custodian to users of cadastral information. 4

The Associativity Problem

Maintenance Associativity is defined as the association between features in different layers in a layer-based LIS/GIS (WAN and WILLIAMSON 1994b). The reference to layers has its origins in computer aided design (CAD) systems. This paper seeks to define associativity as the association between objects assigned to two different data sources, for example the DCM and UI asset data. When objects shift in the DCM in the maintenance process, every user, utilising the DCM as spatial reference for the asset information, may need to change their related (associated) information.

3

A shift in a DCM object (building footprint or fence) resulting in the subsequent incorrect position of the asset objects is diagrammed in Figure 3. Clearly the asset data must be repositioned to maintain the integrity of the spatial asset information. Figure 3:

Associativity problem due to shift in cadastral object

WAN (1993) described three methods for tackling the associativity problem: (a) Database method: The relationship between a feature in a cadastral layer and a feature UI asset layer is represented in the database (b) Object oriented method: The features in each layer are broken into objects. In the cadastral layer, a cadastral feature such as a building footprint will know what it is associated with to calculate its location. (c) Transformation method: Viable in situations where systematic factors dominate the association between the cadastral and UI asset layers In general there are a number of software packages available in Australia and Germany utilising one or more of these methods. Most operate as an additional software package in combination with particular GISs targeted specifically for the UI. Example descriptions of these adjustment programs are described in ELDRIDGE et al. (1998) and ROSE (1990). Another approach is a standalone package where the data transfer between the GIS and the adjustment software is managed using ASCII-files (BAKER and PAXTON 1994) (ASCHOFF et al. 1999). The more information available to these adjustment packages will maximise the successful automated adjustment of the asset data resulting in the most efficient and economical treatment of the associativity problem 5

Associativity data

Custodians and users of DCM can each contribute to minimising the effect of the associativity problem. Firstly the users should store maximal information concerning the associativity of their asset data with the DCM. Associativity information may for example include geometrical constraints and offsets used at the time of asset data capture (see underlined descriptions in Figure 4) Secondly the custodian should transfer as much information as possible concerning spatial changes in the DCM. This obviously means supplying all the necessary data to reflect any changes in the DCM, whether this in the form of new spatial data or the shifting of existing spatial data associated with an accuracy upgrade. The contention of this paper is the information about the movement of the cadastral object is necessary to fully describe the upgrade. This will require a differentiation in the incremental update information between update and upgrade information and supplying information about the precis amount cadastral objects are moved. Figure 3 depicts a shifting of the DCM due to an upgrade of the building footprint and the need to recalculate the position the asset data to its true relative or associated location. Preferably this recalculation is done automatically by an adjustment software package which is supplied the information representing the shift of the DCM and geometrical constraints concerning objects in the cadastral layer and the UI asset data. Figure 4:

Shifting cadastre and maintaining associativity with UI asset data

The movement of the cadastral object is described as vector information. Each of the five vectors representing the movement of cadastral objects shown in Figure 4 is a result of establishing a higher level of accuracy for the existing DCM and also important input data for adjustment software packages. Figure 5 shows one of theses vectors as a connection between an old and a new position of an existing point of the DCM. 4

Figure 5: 6

Vector representing the shift of DCM object.

Technical process of incremental update and upgrade

This paper focuses of incremental maintenance as a transfer of update and upgrade data from the custodian to the user of DCM. Figure 6 shows the technical process of this update process. Figure 6:

Transfer of incremental updates and upgrades from custodian to user

An incremental update relates to logical units consisting of points, lines, polygons and text, this grouped data is identified by Unique Feature Identifier (UFI) in Australia and with an unique Object Number in Germany. In this paper the term Object Identifier (OID) is used as a synonym for both of them. Every user participating in this process has to adopt the custodian´s data model, must have software in place to transfer the incremental updates into their GIS (parser) and establish advanced version management procedures for their GIS objects to correctly process for subsequent incremental updates (Hesse and Jacoby 1995). Without doubt every update set of data deletes objects with their old OIDs and inserts new objects with new OIDs, this process is well defined and established in the process of incremental update. The transfer of upgrade information rather than update information generally causes the associativity problem. Three different approaches to the transfer upgrade information are discussed in this paper and are shown in Table 1. Table1:

Methods of data transfer for upgrades of the DCM

Method (A) (B) (C)

Upgrade Shift of one object of DCM Delete – old OID Insert – new OID Delete – old OID Insert – old OID Delete – old OID Insert – new OID with explicit reference to old OID Transfer Vectors (Fig. 4)

Of course every of these three methods has advantages and disadvantages, these are described as follows Method (A) The upgrade information deletes the object and inserts the identical object with a new OID at the more accurate location. The process is similar to an update process for objects but is not caused by any changes in the real world. Advantages: •

Requires no additional functionality of the custodian’s and users’ GIS;

•

Fulfils the demands of object oriented software development (OOSD) strategies.

Disadvantages: •

High-level version management necessary for calculating shift vector information when comparing old and new objects;

5

•

Textual data sets attributed to old OID in users‘ GIS must be reconnected to the new OID.

Method (B) Any upgrade information is transferred as deleting an object and inserting the identical object without a change of OID. Advantages: •

Requires little additional functionality of the custodian’s and users’ GIS;

•

Supports the calculation of shift vector information, the users’ GIS only has to compare deleted with added OID to separate upgrade from update information;

•

Attributed textual data sets still refers to constant OID.

Disadvantages: •

Advanced version management necessary to calculating the shift vector information for input into adjustment software packages;

•

Deleting and adding objects with identical OID necessitates a significant rearrangement of data in the upgrade file. The delete information must be preprocessed, otherwise the users GIS will refuse to create of a "new" object with existing OID.

Method (C) Incremental upgrade information generates new OID for objects but there is reference to the old OID. Additional shift information is given in the form of a vector, which includes the coordinates of the point, the orientation and magnitude (see Figure 4). Advantages: •

No advanced version management necessary;

•

Shift vectors as a result of custodian´s calculation process are transferred directly as an input information for adjustment software packages;

•

Fulfils the demands of OOSD strategies;

•

Shift vector information is transferred with a georeference, no new OID on the level of single points is necessary;

•

In case of upgrade changing of attributed textual data to the new OID is supported.

Disadvantages:

7

•

Higher volume of data in the incremental update process;

•

Additional functionality of custodian’s GIS and for the parser in users’ GIS.

Conclusions

It is incumbent on both partners in the process of incremental maintenance to establish a transfer strategy for upgrade information, which minimises the effort to solve the associativity problem. All three upgrade strategies discussed in this paper have their advantages and disadvantages. At this stage the method (C) seems to provide the most economical way to transfer upgrade information in an incremental update environment. This strategy provides the maximum information for users about the shift in the DCM and additional serves as input information for any adjustment software packages. 6

The information highway enables the users to explore data from various sources without regard to their actual location. The DCM along with other infrastructure data will be stored in various databases owned and maintained by different companies or departments of government. The future development of Internet/Intranet-technology will provide mechanisms to distribute and access geospatial data such as DCMs. In this environment the necessity of storing the DCM in a physically different computer will disappear and hence the necessity of an incremental update process. The transfer of update information will be superfluous. On the other hand, in this new environment, there will be a continuing and increased need for the transfer of incremental upgrade information. The associativity problem will exist and become more serious in an environment using the spatial data infrastructure as the unique link between data stored in various databases along the information highway. At a minimum, method (C) provides a separation of update and upgrade information and enables custodians to transfer well defined information for the incremental, spatial data, maintenance process. 8

Acknowledgments

The authors wish to gratefully acknowledge the support of the Deutsche Forschungsgemeinschaft (DFG), Land Victoria, Government of Victoria, Australia, the Land Information Center (LIC) Government of New South Wales, Australia, the German Association of Surveying (DVW) and The Australasian Urban and Regional Information System Association Inc. (AURISA). This paper is a direct result of the staff exchange program between the Technical University Berlin and The University of Melbourne, Australia. The views expressed in this paper are those of the authors and do not necessarily reflect the views of the agencies mentioned. 9

References

ASCHOFF, B., GIELSDORF, F. and GRÜNDIG, L.: Data Acquisition for GIS – Integration of heterogeneous geometrical data. Paper presented at the Third Turkish German Joint Geodetic Days, Istanbul, June 1999. BAKER, A.J. and PAXTON, I.:The Associativity Problems in the Utility Environment; A Progress Report on the Problem Solving Process. Proceedings of AURISA 94, the Annual Conference of Australasian Urban and Regional Information Systems Association Inc. Sydney, Australia 1994. BERNHARDT, U.: Geo-Informationssysteme in Energieversorgungsunternehmen, VWEWVerlag. Frankfurt am Main, Germany, 1994. DOMINGUEZ, G.; LAWFORD, G.; WILLIAMS, D. and CARROLL, D.: The Generation of Incremental Updates and their Role in Helping Your System’s Bottom Line. Proceedings of AURISA 94, the Annual Conference of Australasian Urban and Regional Information Systems Association Inc. Sydney, Australia. November 1994. DVW, Deutscher Verein für Vermessungswesen: Organisation of Surveying and Mapping in the Federal Republic of Germany, Schriftenreihe des DVW, Band 10, Konrad Wittwer Verlag, Stuttgart 1993. EBERHARDT, H.: German Cadastre-System – Realisation, Technics. In DVW. 1993 Page 67-77, 1993. EFFENBERG, W. and WILLIAMSON, I.: Digital Cadastral Databases : The Australian Experience. Proceedings of AGI 97 Conference, Birmingham, UK 1997. 7

EFFENBERG, W. and WILLIAMSON, I: Structured Methodologies for Cadastral Systems. (submitted to Computer, Environment and Urban Systems), 1999 ELDRIDGE, J.; HOMBURG, R.; PATCHETT, T. and HESSE, W.: Problems Maintaining Your Spatial Data Reference System? Smallworld International Conference, Hotel Fira Palace, Barcelona, Spain, May 1998. HESSE, W. and JACOBY, S.: Incremental Updates, Unique Feature Identification and Version Management in Modern GIS. Proceedings of AURISA 95, the Annual Conference of Australasian Urban and Regional Information Systems Association Inc. Melbourne, Australia, 1995. ROSE, A.: Homogenisierung in SICAD/SICAD-DIGSY, SICAD-Kurier, Muenchen: Siemens AG, Nr. 52. - S. 67-75, 1990. VOGEL, F. W.: German Cadastre System – Realisation at State Level. In DVW 1993, Page 55-59, 1993. WAN, W.Y. and WILLIAMSON, I:. Problems in Maintaining Associativity in LIS with Particular Reference to the Needs of the Utility Industry. The Australian Surveyor, Vol. 39, No. 3, 187-193, 1994. WAN, W. and WILLIAMSON, I.: Solutions to Maintaining Associativity in LIS with Particular Reference to the Needs of the Utility Industry, The Australian Surveyor, Vol. 39, No. 4, page 290-297, 1994.

Dr.-Ing. Martin Scheu Department of Geomatics Technical University Berlin SEK H20 Straße des 17. Juni 135 10623 Berlin, Germany email: [email protected]

Wolfgang Effenberg BSc, DipEd, GradDipCompStud (Canberra), MAppSc (Melb). Department of Geomatics The University of Melbourne Victoria, Australia, 3010. email: [email protected] Ian P. Williamson PhD (NSW), DrHC (Olsztyn), FTSE, FISAust, FIEAust, HonFMSIAust, LS, CPEng Department of Geomatics The University of Melbourne Victoria, Australia, 3010. email: [email protected] 8

10 Biographical Notes 10.1.1 Martin Scheu Dr.-Ing., Vermessungsassessor Martin Scheu ist wissenschaftlicher Assistent am Institut für Geodäsie und Ausgleichungsrechnung der TU Berlin. 1995 promovierte er im Bereich der GeoInformationssysteme und legte die II. Stattsprüfung ab. Seitdem lehrt er Fach GIS I an der TU Berlin im Studiengang Vermessungswesen und wird in diesem Jahr aller Voraussicht nach sein Habilitationsverfahren eröffnen. 10.1.2 Wolfgang Effenberg BSc, DipEd, GradDipCompStud (Canberra), MAppSc (Melb). Wolfgang Effenberg is a lecturer in Information Technology at Latrobe University, Bendigo, Victoria, Australia. He obtained his Masters at Melbourne University in 1994 with research into temporal considerations of digital road networks. He is part time PhD candidate in the Department of Geomatics at The University of Melbourne, Victoria, Australia. His current major research focus is the investigation and utilisation of current computing technologies and communication in managing the update and upgrade of spatial cadastral data. 10.1.3 Ian Williamson PhD (NSW), DrHC (Olsztyn), FTSE, FISAust, FIEAust, HonFMSIAust, LS, CPEng Ian Williamson holds the Chair of Surveying and Land Information in the Department of Geomatics at The University of Melbourne, Victoria, Australia. He holds bachelors, masters and doctorate degrees in Surveying, is a Registered Professional Land Surveyor and a Chartered Professional Engineer. Professor Williamson is both a Fellow of the Institution of Surveyors, Australia and the Institution of Engineers, Australia. He has consulted widely to state and federal governments in Australia and overseas, United Nations agencies and the World Bank.

9

10

11