Testing airborne vs. conventional GPS

Many in the GIS industry are touting airborne GPS as a feasible and cost-effective alternative to conventional GPS ground control for photogrammetry. Some argue that airborne GPS means few, if any, control points must be targeted prior to photography.

Others say airborne GPS is cost-effective because only one GPS crew member and at least two receivers (one in the plane and one on the ground) are needed, while conventional GPS typically requires four crew members and four GPS receivers.

Airborne GPS requires a GPS receiver in the aircraft, one or more receivers on the ground at known geodetic control station points and at least five GPS satellites in a good geometric constellation.

During airborne GPS, the x,y,z coordinates of the perspective center of each aerial photograph are obtained at the instant of exposure.

Coordinates collected are then processed and used in analytic triangulation, the mathematical means for densifying and extending surveyed control.

Results from analytic triangulation are used to georeference stereo models to ground coordinates during the photogrammetric mapping process.

Consequently, when the St. Johns River Water Management District in Florida needed a GPS survey, engineers used the opportunity to kill two birds with one stone. The team wanted to study the accuracy, reliability and cost-effectiveness of airborne-GPS control for photogrammetry versus the conventional GPS surveying and mapping method.

The engineering team had many questions to answer. Would the results from airborne GPS meet National Map Accuracy Standards (NMAS) for 1 inch=100 feet scale mapping with two-foot contours?

Would results be good enough for 1 inch=50 feet scale planimetric/orthophoto mapping?

Would it be possible to produce 1 inch=50 feet scale mapping with one-foot contours and still meet NMAS?

In scenario number one, 12 known GPS points were held fixed to benchmark results from airborne GPS. Points computed met NMAS for 1 inch=100 feet scale mapping with two-foot contours but not for 1 inch=50 feet scale mapping with one-foot contours.

In a real-world scenario, using airborne GPS would have saved 50 percent of the cost of targeting and performing a conventional GPS ground survey for 1 inch= 100 foot scale mapping with two-foot contours. On large-size projects with rugged, wooded and inaccessible terrain, this approach could save municipalities tens of thousands of dollars.

In scenario two, 20 perimeter control points were held fixed to benchmark airborne GPS results. While points computed met NMAS for both 1 inch=50 feet scale mapping, computed points did not meet NMAS for producing one-foot contours. The degree of improvement compared to scenario one was slight.

Therefore, in a real-world scenario, the benefit of obtaining these eight points via conventional GPS prior to airborne GPS would not be worth the additional cost.

In scenario three, 35 perimeter control points were held fixed to benchmark airborne GPS results.

While the points computed met NMAS for 1 inch=50 feet scale planimetric/orthophoto mapping, they still did not meet NMAS for producing one-foot contours, even with as many as 35 perimeter control points known and targeted.

In scenario four, only the four corner points were held fixed to benchmark airborne GPS results in a square project area. The results met NMAS for 1 inch=100 feet scale mapping with two-foot contours.

Therefore, this approach would be effective for mapping “perfectly shaped” project areas — squares and rectangles — at the same scale and contour interval.

If this had been a real-world scenario, airborne GPS would have saved 75 percent of the cost of targeting and performing a conventional GPS ground survey.

The study ended with mixed results. While airborne GPS is feasible for some projects, it cannot be considered the universal solution for all photogrammetric mapping projects requiring GPS control.

To ensure accuracy, crews still need to conduct some type of ground control survey, such as targeting a few existing ground control points and/or surveying additional points.

Targeting existing controls and performing a conventional GPS ground survey not only provide valuable checkpoints for the airborne GPS data collected, but also greatly reduce the potential for systematic, hidden errors, which may not become apparent until the digital mapping is used for engineering design, construction projects or some other application.

If no existing control is available, additional control at selected locations would be required to avoid the risk of undetectable, systematic and hidden errors in the airborne GPS data.

In addition to a good constellation of satellites, an airborne GPS mission requires good flying weather for aerial photography and a good sun angle. The potential for GPS signal loss should be minimal.

Many other factors, such as ground crew coordination, climate conditions in the project area, geographic location, site accessibility, site geometry/shape and size and accuracies, must be considered.