At the outset, I will admit that I am not an expert in GPS. That said, I have a few questions for surveyors now engaged in the practice of real-time kinematic (RTK) surveys achieving "survey-quality" results—in contrast to "GIS-quality" locations, which, I understand, can mean anything from sub-meter accuracy to sub-football-field accuracy, depending on the equipment. The questions I pose are not intended to be rhetorical (a shift in gears for me, I know), but arise from what I understand to be constraints on the present use of GPS, and about which I hear nothing in the national dialogue. Since apparently a number of surveyors across the country conduct these surveys, some have possibly identified and overcome the factors prompting my questions. However, those surveyors I have spoken with on the subject were unable to offer answers.
My questions are these:
•Is it usually possible to get a "shot" on all four sides of a building in the same general time period (say, within 15 minutes)?
•How does one get shots exactly at building corners given that the antenna has at least a four-inch diameter, and many times a six-inch diameter, on top of the rod?
Satellite Geometry
Signals received from the satellites are "line-of-sight" signals, that is, reception requires an unobstructed sight line between satellite and antenna. Just as tree canopy obstructs signals from satellites, so also will a building obstruct signals. Depending upon the location of the satellites at any given time, some corners of a building might be "GPS-able" but, unless a period of time passes, other corners will not. The addition of the Russian GLONASS constellation to the equation might help this situation—since it would add several more satellites to the sky at any time. But would that cure the problem? Is it possible now, or will it be in the near future, to walk around any building and get a shot at each corner?
Assuming it is, or will be, possible, let’s examine those shots. How does one get a measurement exactly at the corner of the building, when the size of the measurement device itself prevents getting closer than, say, within three inches of it? Am I missing something?
Digital Topography
Most of us have clients who demand digital topography as the final product. GPS constraints force one of three options:
•A method is found to obtain the desired results without having to switch equipment midstream.
•An alternative measurement system, such as a total station, is employed in non-GPS-able areas.
•Centimeter-level accuracy is no longer expected.
I would suggest that, if it has not already been done, someone devote sufficient energy to achieving option number one. Option number two will prove to be very cumbersome to field personnel, since the missed locations would be sporadic shots across the project and not "that 30 percent over there." In practice, the crew chief would have to remember specific shots missed, and come back later and get them individually. On a project of any complexity, that task would be nearly impossible.
The third option is the default. In other words, unless specific solutions are found, the field work will have "approximate" building corner shots, stemming from the constraints on the method used to obtain them. This will occur when crews jam the rod as close to the corner as it will go, get a reading, and move on. With this method, buildings can grow by up to six-inches on each side. If this is the case, we are less than honest in remaining silent about the diminished accuracy of our product.
Digital Design
When stadia topography was in vogue, crews routinely measured the lengths of the buildings in addition to getting shots at the corners. Since the final product of the survey was analog (a drawing of some sort), the draftsman had the ability to publish only the desired measurement (usually by labeling the dimension of the building). With the advent of total stations and data collectors, the practice of directly measuring the buildings largely fell out of favor, since the locations from the total station were potentially as precise as any measurement using a chain along the face of the building. Total stations and data collectors ushered in digital topography—normally a preface to digital design. While "prism-pole" shots within a tenth of a foot or so at the corner would have posed no problem to manually drafted plans, digital plans suffer because designers "snap" to the existing features in the design file to begin their design. Now with RTK locations, potentially greater errors could be introduced because of the physical size of the measurement device. Rightly or wrongly, few designers question the accuracy of field-run digital topography. They merely assume it as the basis for the design and proceed. Moreover, memories often get short where digital files are concerned, and new uses for old files get devised—even when the use is inappropriate because of the original method of data collection.
Errors Discovered During Stakeout
Who discovers the inaccuracy? Ironically, the same people who introduced it—the survey crew. This time, however, it will be at stake-out time, and they will notice that the building addition has been designed a few inches longer than the existing building. In the interim, the entire design/approval process will have been completed.
I know that some will suggest that rooftop surveying will avoid most, if not all, of these complications. Generally, however, that is more inconvenient than using alternative methods to obtain the measurements, especially if there are many buildings within the area of survey. Offsets are another potential solution, but they again introduce complications. For instance, does one assume that the ground elevation at the unattainable building corner is the same as that at the offset point?
Surely someone has identified and, if using this method for building locations, overcome these obstacles. Perhaps that someone would be kind enough to share the solution with the rest of us.
Copyright © 1997 By Joel M. Leininger, LS