Pole Height and Leveling
A rover pole that is even slightly out of plumb at the moment of collection displaces the computed ground point sideways, silently, in a direction and amount that depends entirely on which way the pole was leaning.
Tilting a rover pole moves the antenna horizontally away from the true point directly below it. That horizontal displacement grows with both the pole's length and the size of the tilt, and it affects the horizontal position even though the mistake is a leveling error, not a horizontal measurement error.
The Geometry Behind the Error
A pole tilted by a small angle displaces its top, where the antenna sits, sideways by roughly the pole's length multiplied by the sine of that tilt angle. A two-meter pole tilted by just one degree displaces the antenna, and therefore the computed position, by roughly three and a half centimeters. Double the tilt, and the displacement roughly doubles as well. This is a small, easily overlooked amount of lean producing an error well outside RTK's own centimeter-level precision.
A tilt error is not a defect in the RTK solution at all. The carrier-phase math can be perfect, and FIX can be entirely genuine, while the antenna itself simply was not sitting directly above the point the crew intended to measure.
Bubble Levels and Their Limits
A traditional bubble level has a finite, published accuracy, often on the order of a small fraction of a degree at best under ideal conditions, worse if the bubble itself is dirty, worn, or improperly calibrated. Combined with an operator's own visual judgment of when the bubble is centered, some residual tilt is nearly always present, and that residual scales directly into a horizontal error the moment it is combined with pole length.
A tilt-compensated pole using an internal motion sensor does not make leveling irrelevant. It compensates for tilt within a calibrated range and typically requires periodic recalibration. A pole used outside its calibrated tilt range, or one whose calibration has drifted, can still introduce the same displacement a traditional bubble level would.
A Simple Habit With an Outsized Payoff
Because the displacement scales with pole length, a taller pole magnifies the same small tilt into a larger error than a shorter one would. Extending a pole only as long as the job actually requires, checking bubble condition periodically, and treating tilt-compensation calibration as a maintenance item rather than a one-time setup step are all simple habits that directly address one of the most physically straightforward, and most commonly overlooked, sources of horizontal error in RTK fieldwork.
A point collected with a fully extended pole on a windy day sits a few centimeters off from an independent check, with excellent PDOP, a short baseline, and calm atmospheric conditions otherwise. What is worth checking first? Whether the pole was actually plumb at the moment of collection, since wind and a long pole are a classic combination for exactly this kind of error.
A tilted pole displaces the antenna, and therefore the computed position, horizontally by an amount that grows with both pole length and tilt angle. Bubble levels have real, finite accuracy limits, and tilt compensation manages the problem within a calibrated range rather than eliminating the need to care about it entirely.
Physical setup errors sit close to the ground and close to the crew's own control. The next chapter returns to something farther upstream, and much less visible: the reference frame and datum a survey was actually tied to.