New York is experimenting with differential GPS, which offers greater accuracy and translates to better GIS data for user agencies.
LATHAM, N.Y. - The latest step in the evolution of global positioning system technology is being tested by a New York state department which maps and analyzes endangered species habitat. The Fish & Wildlife Division of New York's Environmental Conservation Department is conducting a pilot program to determine the feasibility of using a differential global positioning system (DGPS) to support a broad range of geographic information-based applications in ecosystem management.
One of the main potential applications for this technology is mapping New York's 2.5 million acres of wetlands. Other applications include mapping forest areas, hiking trails, rare plant and animal communities and monitoring toxic-substance areas.
The pilot program - directed by Senior Wildlife Biologist John Ozard and Senior Fish and Wildlife Ecologist Scott Crocoll - was prompted by the need to ensure accuracy and currency in the division's geographic information system (GIS) and other biodiversity databases. The division was also looking for a faster, more accurate and economical method of mapping and data collection.
PROBLEMS WITH EXISTING METHODS
According to the project's leaders, existing methods for mapping wetlands and locating species habitats are slow, error prone, and costly in manpower. "Where GPS is going to come into wetlands mapping," explained GIS Project Coordinator Wayne Richter, "is in the data gathering and data updating stage."
"For instance, freshwater wetlands have traditionally been mapped by air photography or ground surveys," Richter said. "Boundaries were then transferred to a quad map by estimating locations. Surveys conducted with GPS will be used as a basis to amend maps already in the GIS, [but] with much greater accuracy."
The differential in DGPS is a referencing technique used to overcome natural and artificial errors in a GPS. Natural errors result from atmospheric conditions, timing differences, and minor perturbations in satellite orbits. Another source of error, called selective availability (SA), is randomly activated by the Defense Department to degrade the accuracy of GPS for civilian use. The Pentagon's SA is ostensibly motivated by national security concerns.
Raw GPS with SA activated produces accuracies of 100 meters; without SA, predictable accuracies for civilian use are between 20 meters and 30 meters. DGPS, however, can produce accuracies in the millimeter range, depending on the type of measurement used by the receiver.
Differentially-corrected static files obtained with pseudorange measurement produce accuracies of between 2 and 5 meters. The higher-price carrier-phase measurement, an option with the Magellan ProMark V, produces differentially-corrected accuracies in the sub-meter range.
The differential principle has been used in electronic navigation systems for years. In DGPS, a computer measures the differences between the known geographic location of a GPS base station and its satellite-reported positions, and generates time-stamped corrections, which can be transmitted as real-time corrections to remote receivers within a 300-mile radius, or downloaded to a hard disk for post processing (applying differential corrections after remote files have been collected).
POST PROCESSING - FROM GPS TO GIS
The base station, located at the agency's Wildlife Resources Center in Albany, takes a fix each second, 10 hours a day, seven days a week, and stores the data on an internal file. Every hour, a computer automatically downloads the file to its hard disk, then clears the receiver's memory. The base station's recording rate and hours of operation are user selectable and can be changed locally, or remotely by PC, modem and telephone.
GPS files collected in the field can be post-processed at the Resources Center or from remote locations. Files from the remote receiver and the base station are first converted to RINEX, a receiver-independent exchange format that enables post-processing software to work with files from different brand GPS receivers. The software then time-matches data from the two receivers and applies corrections to the remote files. Corrected files imported to the GIS are first converted by a utility program in the software to a language understood by AutoCAD, Arc/Info, MapInfo and other databases.
Currently in the developmental research phase of the pilot program, DGPS is being evaluated with a broad range of applications for the bureaus of Wildlife, Fisheries, and Environmental Protection. The project team, for example, mapped Tern and Piping Plover communities on Long Island. Both populations are declining along the Atlantic Coast and the species have been put on the state's endangered list. Other endangered or threatened species habitats being mapped by the team include the Massasauga Rattlesnake, found in wetlands, and the Eastern Timber Rattlesnake, which is found mainly in mountainous areas.
Crocoll explained that wetlands are plotted on New York state planimetric maps using 1:24,000 scale, with thick lines indicating the approximate outer edge of the wetland. "Based on the scale of the map, that line could be 50-feet wide," he said. But "GPS is going to give us an accuracy of 2 [meters] to 5 meters."
Differentially-corrected static files taken in open terrain are producing this kind of accuracy with occupation times of nearly five minutes. Forested areas with heavy canopy require longer occupation times.
Data collected can be checked by a crew still in the field using a portable PC, Ozard said. "The flexibility is especially useful because it lets us check the accuracy of our product before we return to the home office. We can also differentially correct our data from a remote office and produce an output map before returning home."
BASE STATION RANGES
Part of the DGPS test includes verifying the accuracy of known horizontal control stations. When a National Geodetic Survey marker is in the area, the team will take one or two static files. If those correct to better than 5-meter accuracy, they are fairly confident that the data will be in the same accuracy range. Most of the tests thus far have shown accuracies within 5 meters, Ozard said.
Another test is to see how far from a base station DGPS measurements can be taken while maintaining accuracy within 5 meters. Points taken from as far away as 311 miles from a station have produced errors of less than 3 meters. "The accuracies are better than those required for wildlife and land management applications," Ozard said. "It looks like one base station will adequately cover the entire state."
Although it will take another year to complete the evaluation, Ozard believes DGPS will be a very useful tool in determining locations of rare and endangered species, particularly in areas without landmarks. "It should also assist us in determining the movements of animals, their home range, and how far they travel in various time periods. Another practical benefit will be using mobile files to map locations of smaller communities and habitats that cannot be determined from aerial photographs."
Following training and product evaluation, the New York project leaders chose a Ranger, 12-channel, all-in-view Base Station from Ashtech, and a package of three Magellan NAV 5000 PRO GPS Receivers, Hewlett-Packard HP-95LX Dataloggers, and Magellan's mission-planning, data-collection and post-processing software. The NAV 5000 units were subsequently upgraded to the more powerful ProMark V's. Their expanded data-storage capacity eliminates the need for external data recorders. Important factors in the selection of remote receivers were portability, weight and compactness. "We were looking for receivers that could easily be carried through rugged terrain and brush," Crocoll explained.