A 2.290Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE
Surveyors are working to improve the spatial accuracy of GIS layers in Montana.
Because Montana based most of its cadastre upon the Bureau of Land Management’s Geographic Coordinate Database (GCDB, a digital representation of the locations of the Public Lands Survey System corners), the process for improving the spatial accuracy of the GIS cadastre first requires upgrades to the accuracy of the GCDB. Montana’s surveyors play a significant role in the GCDB enhancement projects where the ultimate goal is to improve the spatial alignment of GIS layers. This article discusses the surveyor’s role in making GIS better.
Background
Although Montana is the fourth largest state in land area at more than 145,552 square miles, and has a population density of about 6.2 people per square mile, it is one of the few states in the nation to have a statewide cadastral GIS. The number of parcels in the Montana Cadastral dataset is about the same as the number of people in the state a little more than 900,000. The annual economic benefits of the Montana Cadastral dataset average in the milliondollar range an obvious demonstration of the value of this system. The benefits accrue from the variety of applications that one can do against a statewide cadastral dataset and the savings in time and money realized by the availability of the data via the Web (Figure 1). Nevertheless, the methodologies used to create this statewide cadastre in a few short years in the 1990s resulted in less than optimum spatial accuracy in some areas of the state. "Less than optimum" means that certain applications fail in some areas, such as GPS-derived address points not falling in the correct parcel, or calculations for impermeable area are difficult to apply uniformly due to the spatial misalignment between GIS layers. Fortunately, spatial accuracy of GIS layers can be improved and surveyors are part of that process. Montana has a spatial accuracy improvement program dedicated to enhancing the spatial accuracy of the cadastral layer in problem areas.
Sources of Errors
Errors arise from a variety of causes. Because Montana based its cadastral framework layer on the Bureau of Land Management’s GCDB in most areas, the accuracy of the parcels is dependent on the accuracy of the GCDB, thus the parcels can be no more accurate than the GCDB. The BLM constructed the GCDB by digitizing coordinates off the U.S. Geological Survey’s topographical maps, supplemented in some areas by GPS-derived coordinates and/or data entry of recent recorded survey bearings and distances. The majority of coordinate pairs of the GCDB are approximations at best. Additionally, other sources of error lie in the reliability of the source data (assessor maps that were hand drafted over the course of decades), and the processes used to digitize the parcels. The digitizing process may have inadvertently introduced errors.
In practice, best results come from minimizing the error in areas where the density of parcels (and other GIS layers) is highest, that is, the higher the parcel’s density is, the smaller the spatial error should be. Nevertheless, in Montana, the spatial error varies somewhat randomly so we have some urban areas (high density) with very large errors (Figure 2), and some rural areas (such as where one whole section might be a single parcel more or less one square mile) with relatively high accuracy (Figure 3).
Regardless of the source of error, one can improve the spatial accuracy in a couple of ways. For rural areas, which make up the majority of Montana, the process to improve the parcels starts by improving the GCDB. To improve the GCDB, one must develop coordinates that are more accurate usually by field observations (such as GPS surveys) or by entering, recent (and presumably more accurate) survey measurements. After entering new coordinates into the GCDB, the GCDB is readjusted to fit the new control, then the parcels are readjusted to fit the new GCDB. Surveyors and survey measurement data are the key to this process. Surveyors perform the field survey to generate new coordinates, or may even already have coordinates on some Public Lands Survey System corners (tied to the National Spatial Reference System). Surveyors also perform the GCDB re-adjustments, under the guidance of the BLM.
Montana received a federal appropriation through the U.S. Department of the Interior/Bureau of Land Management to maintain and enhance the accuracy of GCDB based cadastral framework data. The federal appropriation enabled Montana to enhance the spatial accuracy of those areas of the state where there are gross errors. DJ&A, PC of Missoula, Montana is under contract with the state of Montana Department of Administration’s GIS Bureau to do some of this work using the following process.
Research and Planning
The first step is to research corner records, certificates of survey and plats, and BLM records to identify PLSS corners that are suitably monumented and that are likely candidates for coordinating with GPS. Common practice is to completely surround a project area with survey control in order to constrain the adjustment within the project area, and to be able to control the adjacent townships. All corner records retrieved are scanned into TIFF format, given the appropriate GCDB ID, and the relevant information is added to a geodatabase and referenced to the corner location so that searches are readily performed and maps easily produced showing the location of monumented corners that have corner records associated with them. In order to discover whether or not other coordinate data are available on the PLSS, stakeholder meetings are held with local engineers and surveyors in private and public organizations (such as engineering firms, surveying firms, utility companies, the BLM, the U.S. Forest Services, the Bureau of Reclamation, etc.). If they have existing survey control on the PLSS that is tied to the National Spatial Reference Systems (NSRS) that they are willing to contribute to the project, then their coordinate data can also be used to control the GCDB re-adjustment. In some instances, such as when no reliable monumentation exists or access is cost-prohibitive, then data entry of existing recorded survey, plat, and deed information may be more a cost-effective means to improve GCDB coordinate values.
As a planning tool, a map is produced showing where possible PLSS corner candidates are, based on the GCDB reliability estimates (need), the location of existing control coordinate data, the suitability for field ties of PLSS corners (existing corner record indicating a suitable monument in locations that will strengthen the GCDB township, and minimal access and safety issues). Based on that information a survey plan is developed (Figure4).
Survey Steps
The next steps are to perform the survey fieldwork (new corner ties, verify others’ coordinates, if any), then reduce, calculate and analyze survey data. Field surveys are performed primarily using GPS survey methods to obtain 5 cm or better accuracy (typical results) on control points. All coordinate information obtained is referenced to the National Spatial Reference System (NSRS) and entered into the GCDB control files as geog raphic coordinates (NAD 27).
All new and existing control coordinate data are entered into a GCDB control file and each GCDB township is re-adjusted using the BLM’s GMM software and approved BLM methods and procedures. The adjustment is analyzed for errors and blunders and re-adjusted if necessary, then submitted to the BLM for review. The BLM then incorporates the re-adjusted GCDB into the agency’s framework data
base.
It is important to note that GCDB error estimates are somewhat subjective, so some estimates may be overly optimistic. Because of this it is possible for the error to appear to increase after the adjustment, even though the coordinates may be more reliable. In these instances the accuracy is better, but merely appears to have worsened because the original estimate overstated the accuracy. However, it may also be true that the accuracy got worse, which does happen in some areas due to bad survey data or blunders in the GCDB. These errors can usually be isolated and measures taken to improve them. Generally, however, one will see an improvement in the GCDB accuracy after new data are entered and the GCDB is adjusted (Figure 5). That is, after all, the point.
Parcel Adjustment
After the GCDB is readjusted to the new coordinates, the parcel layer is adjusted by the Montana Department of Administration GIS Bureau, using a least squares adjustment. The least squares adjustment program, developed by Tim Hodson of ESRI, uses the changes in positions of the new GCDB points compared to the old GCDB points for the same corner. This is based on the point identifier (GCDB ID) for points that are found within a user-specified search radius. The application automatically finds the matches for each GCDB point (old to new). However, in areas where the new coordinates change significantly (greater than the search tolerance), which sometimes happens, linkages can be made manually. The program then calculates the changes in position and D applies those changes to the parcel layer in what is essentially a rubber-sheeting process. Other GIS layers that are linked to the GCDB, such as fire districts, or resource management area boundaries, may also be adjusted in this way.
Conclusion
Construction of the Bureau of Land Management’s Geographic Coordinate Database was a huge undertaking that is still ongoing in some parts of the country and completed in other parts. For those areas that are complete, maintenance in the form of accuracy enhancements can occur to help any number of GIS layers fit together better. Surveyors are key players in this process because of their skills and knowledge for research, field reconnaissance, field observations, data reduction and analysis. Through the surveyor’s effort, the GIS data used by many is improved for all.
Rj Zimmer is Director of GIS and Mapping for DJ&A, PC of Montana, an engineering, surveying, and mapping consulting company.
A 2.290Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE