Put to the Test

Professional Surveyor Magazine - September 2008

Figure 1: WA State ref net was one of three DGPS correction methods used

For years, public and private entities have traditionally carried out resource inventory and position tracking projects using hardcopy maps and other manual techniques. As GPS technology has become more widely available and cost effective, agencies have adopted mapping-grade technology for resource inventory, management, and other activities.

Today, most entities use mapping-grade GPS for data collection and storing this information in geographic information systems (GIS), which can range from single analyst to enterprise-level multi-user environments. Many mapping-grade GPS users employ differential correction through post-processing to remove errors from the GPS data and produce more accurate results. However, the adoption of real-time differential correction has yet to gain widespread use by mapping-grade GPS users.

Why this historic overview? It sets the stage for our side-by-side test. Marshall and Associates (Marshall) and partner Bush, Roed, & Hitchings (BRH), a civil engineering and land surveying firm in Seattle, Washington, are conducting a small, ongoing, unscientific study near Green Lake Park in Seattle to compare GPS receiver grades and DGPS correction methods typically used by surveyors. The various GPS receivers include survey-, mapping-, and recreational-grade units. The evaluated DGPS methods focus on real-time correction and include the WAAS (Wide Area Augmentation System) and VRS (Virtual Reference Station) correction from the WSRN (Washington State Reference Network). Marshall has used mapping- and recreational-grade equipment and BRH survey-grade GPS receivers to collect location data for approximately 20 deciduous trees.

The impetus behind the GPS receiver comparison study was to investigate and evaluate the use of real-time differential correction with mapping-grade units. Marshall is continuing to collect data this summer to further compare differential correction methods with these receivers. The study has also become an exercise to identify and understand the differences between the data requirements for surveying versus GIS applications. As most readers acknowledge, a historic disconnect exists between surveyors and GIS practitioners. The following preliminary study results and subsequent discussion seeks to explain the application and standards for GIS data collection.

Sources of Real-Time Differential Correction

Together, the Federal Aviation Administration and Department of Transportation developed WAAS for use in precision flight approaches. WAAS consists of approximately 25 ground reference stations spread across the United States, including two master stations on either coast. The master stations collect data from the reference stations and create a correction message, and the reference stations then transmit the GPS correction message to one of two geostationary satellites. GPS users with WAAS-enabled GPS receivers can read the correction signal broadcast from these satellites (http://www.gps.faa.gov/index.htm).

The WAAS correction signal is only available in the United States but does not require a subscription or extra equipment. The WAAS signal reception is ideal for open land or marine applications but can be obscured by trees and topographical features. Also, the GPS receiver must have an unobstructed view of the southern hemisphere to obtain the WAAS signal.

Located throughout Washington State, the WSRN is a regional cooperative of GPS reference stations. The network of Continuously Operating Reference Stations (CORS) provides real-time services along with free downloadable differential correction data files for use with post-processing software. Real-time differential correction services are available through partnerships, memberships, and subscriptions (www.prsn.org). The network consists of numerous CORS operated by different owners around the state; these GPS receivers send data to a central processing center at Seattle Public Utilities (Figure 1).

This CPC is the key communication component of the WSRN; users access both traditional real-time kinematic (RTK) GPS corrections from individual CORS or the improved Virtual Reference Station (VRS), a correction computed from multiple stations. Users initiate a connection by contacting the CPC via a cellular (phone) data connection previously configured to communicate with the GPS receiver and choosing the type of correction (RTK or VRS). Most industry-standard GPS receivers, both survey- and mapping-grade, can recognize and apply real-time corrections.

Intermittent cell phone coverage limits connectivity to the internet in Eastern Washington and can be a problem. The system is reliable more than 80 percent of the time with a hard phone line or radio connection.

Marshall found it easy to configure and enable the WAAS correction on a mapping-grade receiver. The initial configuration of the mapping-grade receivers with the WSRN proved time-consuming because of additional hardware and Bluetooth communication requirements. BRH had previously configured the survey-grade receivers and routinely used them with the WSRN.

BRH used the survey-grade GPS receiver with the WSRNVRS correction to establish baseline tree locations. Marshall used a mapping-grade GPS receiver with the WAAS correction enabled and a recreational-grade unit to collect data. Marshall has configured and tested the WSRN connection with the mapping-grade receiver in the field but has not yet successfully collected data with this method. Plans for future data collection include using the following GPS and DGPS configurations: mapping grade with VRS/WSRN, mapping grade with VRS/ WSRN and post-processing, mapping grade with WAAS and post-processing, and mapping grade with post-processing.

Preliminary Results

Figure 2 depicts the difference in measured distance (in five-foot increments) from the survey-grade receiver location to the mapping- and recreational-grade units' positions. The recreational-grade receiver is missing data where Marshall moved the unit, but the data point collected did not register at the new location. We deleted these erroneous location points from the dataset. In general, the mapping-grade receiver with the WAAS correction provided better GPS data than the recreational unit. The largest error was associated with tree number 11, where the mapping-grade receiver recorded a location approximately 35 feet from the survey-grade point. We anticipate that further comparisons with future data collection efforts would provide more insight into accuracy levels of mapping-grade GPS units with the range of differential correction techniques.

From experiences collecting data and configuring real-time correction with the mapping-grade GPS receiver, Marshall has compiled the following recommendations:

  • Allow plenty of configuration time when using real-time correction. Trying to determine how to use a particular cell phone as a data modem can be a challenging and lengthy process.
  • The cell phone must maintain a strong signal while using the WSRN. This is more important than connection speed.
  • For the cell phone to communicate with the GPS unit or other data collection device, maintain good Bluetooth connectivity between the two components by keeping the devices within approximately 10 meters (32 feet). The initial "pairing" of the phone and GPS device can be tedious.
  • Follow GPS best-practice methods, including acquiring a current almanac and establishing a satellite connection before collecting data. This is critical.
  • Determine the datum of the real-time correction information. Data shifts may occur if the real-time correction source and GPS unit are not set to the same datum.
  • Using both real-time correction and post-processing techniques will achieve the best possible data accuracy.

Further Considerations

The accuracy of the data collected during the tree survey is outside surveying standards; however, in most cases, this data would probably end up in a GIS. Why would GIS practitioners be willing to collect and use such inaccurate data? The answer lies with the intended application.

Engineers and surveyors often collect and download survey-quality GPS data for integration into CAD-based programs. They have traditionally used CAD as a powerful design and construction tool with survey-grade data. Since CAD is a graphics program, lines themselves are important, and the overall drawing is the information. Thus, accurate survey-grade information is represented through the spatial depiction of the elements making up the drawing.

Data collected using mapping-grade GPS units is usually incorporated into a GIS. Planners and resource managers traditionally use GIS for map production, spatial analysis, and as a decision support tool. The GIS is often a collection of data from many different documented sources and accuracies, including mapping-grade GPS units. GIS stores and accesses data in a database-oriented manner, which provides capacity for creating and representing geographic features and related attributes over wide geographic extents. Spatial elements used to create maps from GIS data are simply a representation and link to the information stored in the database. GIS can store all accuracy levels of data; however, since the data is often used to conduct further analyses and modeling procedures, getting a baseline mapping-grade accurate spatial dataset is the first crucial step to building a robust GIS in an efficient and cost-effective way.

Mapping-grade receivers will continue to be used by GIS practitioners and others for a variety of applications. Using best practices while acknowledging receiver limitations and applying differential correction methods long used by surveyors and engineers will help ensure the best possible accuracy for spatial data collected with mapping-grade GPS units. Granted, this data will never have the accuracy of survey-grade information or be appropriate for design and construction projects. But conversely, survey-grade data collected for a design or construction project could provide further utility when incorporated into a GIS.

A GIS can add spatial reference, spatial context, and environmental perspective that would enable the data to integrate with other information for maintenance and management applications and provide the basis for powerful decision tools. Spatial data is a key link between GIS practitioners and surveyors. Recognizing and understanding the broad applications for data collected through different methods and accuracies provides synergistic benefits to both fields.

About the Author

  • Dana Trethewy

» Back to our September 2008 Issue

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