GIAA Mailbag: On Multipath and Trigonometric Leveling
Q: I was told that fences (chain link) could be the cause of multipath. Is this true?A: Multipath refers to GPS signals received by the antenna that have reflected off some surface. The path taken by these signals is longer than the most direct path from the satellites. They can cause the position determination to be incorrect or to have a high uncertainty because of measurement “noise.” A chain link fence, “holes” notwithstanding, is excellent for reflecting GPS signals. It is more than likely where a chain link fence surrounds a base station that the measurements are affected. But note that most giant radio telescopes have dishes constructed of a metallic mesh or grid. Only properly designed experimentation (survey the same points using a base station located in an extremely clear area) will confirm whether you are affected by error-producing multipath. Surveyors should be especially watchful for this cause of errors with a rover, static measurement or base station. It is rare for GPS observations to not have some multipath. The objective is to reduce the level of multipath signals. It is likely that your GPS system has been designed to mitigate the effects of multipath. However, no system is completely immune.
Q: Why don’t my total station-derived elevations check those I get with my automatic level?A: Some general trigonometric leveling mistakes include: incorrect elevation of the point you are set up over or backsighting, incorrect instrument and prism heights, and sighting the incorrect point on the target. Trigonometric height is calculated using the vertical angle and slope distance. Small variations in the vertical angle cause dramatic variations in elevation. If you have eliminated the general causes, and the horizontal distance to the point you are surveying is short (under 500 ft), then the error may be due to one of two primary causes of vertical angle error.
Most total stations allow the user to turn the tilt compensator on or off. If the compensator is off, the vertical angle will be incorrect in the magnitude in which the instrument is out of level in the observed direction. A typical total station’s level vial specification is 30"/2mm. A vertical angle error of 30" (one level vial graduation) would cause an elevation error of 0.073 ft in a 500 ft distance with the compensator off.
Another possibility is an instrument calibration called collimation—relating to the proper adjustment of the horizontal cross hair. Check an instrument’s collimation by reading face 1 (direct) and face 2 (reverse) vertical angles. Add the direct and reverse zenith angles; the sum should be 360?00'00". If the sum is 30" more, then the error in each reading is 15". Consult your user manual or your supplier for more information on checking and calibrating your instrument.
For longer distances (over 500 ft), check that your total station or data collector (not both!) has earth curvature and refraction (C + R) correction turned on. Error due to (C + R) increases proportionally with the square of the horizontal distance to the point. For example, the (C + R) error in a 500 ft baseline is only 0.005 ft. The (C + R) error in a 1,000 ft horizontal distance is 0.021 ft (four times the error at 500 ft). The (C + R) error in a 2,000 ft horizontal distance is 0.083 ft, which is 16 times the error at 500 ft.
One final note: you wouldn’t observe a level rod at 2,000 ft with your automatic level and expect the results you would get at 150 ft. The lesson is to understand why and make appropriate allowances in your design of observations and expectations for trigonometric leveling.
Thanks for the assistance of Matt Delano of Nikon in preparing this column.
If you have a question for the GIAA Mailbag, E-mail your questions to Diana Brown at firstname.lastname@example.org. Although we can’t print all questions, we will attempt to cover as many as possible.
In a further effort to reach users of survey instrumentation, GIAA will present a half-day workshop at the ACSM convention in Las Vegas this month. Mark your calendars for an informative presentation on “How Things Work” on Sun., March 18, 2001. You will hear from industry experts on the design concepts and other technical aspects of various components of optical and satellite surveying systems.