A Survey-accurate Cadastral Model Aids in Railway Infrastructure Development

In Australia, the New South Wales (NSW) government owns and operates the electrified metropolitan railway network throughout Sydney and surrounding commuter areas. As the population of Sydney has grown outward into new areas of urban development, the government has recognized the need for major upgrading of parts of that system. To facilitate those upgrades, the NSW Transport Infrastructure Development Corporation (TIDC) was enacted to focus on the construction of major projects.

One such project involved the upgrade of 10 kilometers of an existing single line that serviced a historical rural area identified as a major growth center for Sydney’s future. To commence the planning and design stage, the project managers needed a survey-accurate cadastral model of the route, and for political reasons with a minimum of survey fieldwork.

TIDC engaged Australian ESRI business partner Geodata Information Systems to prepare a survey-accurate cadastral fabric of the railway corridor. Geodata worked in conjunction with the Sydney office of the engineering project managers, surveyors, and the engineering and management firm Connell Wagner.

As a “big picture” government infrastructure developer, TIDC envisioned that when it completed the construction stage of the project, the firm would provide the railway operations department with a valuable database of the highest spatial quality for ongoing asset management. The ArcGIS Survey Analyst technology produces a survey database of field measurement, geodetic survey control, and the spatial outcome of surveyors’ boundary determinations to create an integrated geodatabase dataset designed to store a continuous parcel fabric to cover a certain jurisdiction. The dataset is built on the survey measurements from registered subdivision plans and geodetic survey control without loss of any information from the original survey records. The dataset includes parcels, boundary lines, parcel corners, and control points.

The underlying technology for the cadastral fabric dataset and the new workflow Cadastral Editor is Geodata’s GeoCadastre, a survey software that has been in development for the last 20 years for a user group of Australian companies and utilities.

In Cadastral Editor, the user enters individual TIDC parcels and subdivisions using coordinate geometry (COGO)-based plan entry. The integrated cadastral fabric is developed from all relevant survey and plan data using least-squares adjustment for the best-fit representation of the parcel layer.

To form the desktop model for the railroad corridor, Geodata manually entered survey information including: 
    • the parcel dimensions on the subdivision plans of the current titles, 
    • survey information connecting to other cadastral marks confirming boundary definition, 
    • survey measurements to coordinated survey control marks, and 
    •  survey information from historical plans where survey measurements have not been made on recent plans.

As each new parcel is joined to the survey data model, the accuracy of the fit is also measured, checking the spatial quality of the new data and providing another level of survey rigor.

The data is then adjusted using least squares. The adjustment report compares all the computed lines with the original measurements. If the residuals are large, it usually indicates a problem with the plan, and additional field survey definition on the ground may be needed to properly isolate and address these apparent inaccuracies. The plans concerned have been given a weighting (linked to the age or accuracy of the survey) that either removes them from the least-squares adjustment or substantially reduces their impact such that their overall effecton the model is negligible.

The electronic compilation is based on the intuitive method a surveyor would otherwise manually undertake to createa cadastral database. Model accuracy is directly related to the quality of the survey information available (e.g., in a new sub-division, the model accuracy would be 20–30 millimeters). Prior to data entry, it is difficult to gauge the quality of the survey data.

In the TIDC model, a reasonable length of the route corridor is adjacent to urban areas,
which led the project team to believe the survey information would be reasonable and would not require a significant amount of extra field survey control.

More than 4,000 parcels were entered to provide a rigorous model for the adjustment. In the model area, more than 130 coordinated survey control marks with survey connections tothe cadastre were found in the NSW Lands Department Survey Control database. The coordinates of approximately 90 ofthose marks were of the highest accuracy and considered suitable to be used as accurate control for the adjustment.

Following completion of the desktop model, the surveyors undertook two days of fieldwork in several areas where properties were defined by older survey plans and required more definition. The survey consisted of finding cadastral survey marks, such as pegs and reference marks, that were identifiable on the survey plans manually entered into the model. These marks were accurately coordinated by RTK GPS. The same points inthe model can then be used as fixed control points in the adjustment to improvethe spatial precision of the cadastre inthe immediate area, or the accurate coordinates can be compared with the coordinates computed by the model.

On one field day, the model was taken into the field with GPS; the desktop model coordinates
allowed surveyors to navigate accurately (0.05–0.5 meter) to where marks on the ground or buried marks were located. In many cases, the model quickly allowed recognition that survey marks had been destroyed by cables or other work.

The survey connectivity and rigor ofthe desktop data allowed the model to be created and adjusted with only one extra survey field control mark being necessary.  In the final model, 16 control points were held fixed with the remainder assigned a “not active” status to provide
independent verification of the model’s spatial quality.

The spatial accuracy of the final TIDC model is in the order of 30–100 millimeters in urban areas where reasonable survey information is available and 100– 200 millimeters in rural or other areas of older survey plans where there is limited survey information.

The project now has a survey-accurate cadastral databasewith property attributes to underpin the project GIS databasein the ArcGIS environment. It is important to correctly maintain this valuable dataset for the duration of the project and to pass it on as a legacy for future operational efficiencies and asset management. You cannot quantify the true value of such an asset for the future. Unfortunately, people who have had to deal with spatially poor GIS databases in the past are the only ones who truly recognize the value.

The NSW Lands Department is promoting permanent global navigation satellite system base stations, such as the Continually Operating Reference Station (CORS), across the state. These stations will provide surveyors with centimeter geodetic accuracy for all their measurements in the near future. The cadastral layer in the geodatabase is the only way we can meet the spatial challenges of both survey and GIS professionals today and in the future.


Ian Harper is a surveyor with Geodata Information Systems.
Karen Richardson is a writer for ESRI in Redlands, California.

All trains are manufactured by Commonwealth Engineering (Comeng) and A Goninan & Co. All carriages are made of stainless steel, are doubledeck construction, and aren’t air-conditioned.

» Back to our Online-only 2009 Issue

Website design and hosting provided by 270net Technologies in Frederick, Maryland.