Baseline Length and RTK Performance
RTK's entire precision advantage, established earlier in this series, comes from canceling error that is shared between a base and a rover. That sharing depends on the two receivers experiencing nearly the same atmospheric and orbital conditions, and the farther apart they sit, the less that assumption holds.
RTK performance degrades gradually as the distance between rover and reference grows, because atmospheric and orbital errors become less correlated over longer distances, leaving a larger uncorrected residual behind.
Why Distance Erodes the Cancellation
A patch of atmosphere directly overhead affects a base and a rover sitting a few hundred meters apart in nearly the same way. Stretch that same baseline out to twenty or thirty kilometers, and the rover may be looking through a meaningfully different slice of atmosphere than the base, especially for ionospheric conditions, which can vary significantly over tens of kilometers. The shared error that RTK depends on canceling becomes less shared, and the residual left behind after correction grows accordingly.
A Practical Rule of Thumb
A commonly cited rule of thumb treats RTK positional error as growing by roughly one part per million of baseline length, on top of a fixed baseline error that exists even at very short distances. A ten-kilometer baseline, under that rule, might add on the order of a centimeter of additional uncertainty beyond what a very short baseline would produce. This is a rule of thumb, not a guarantee, since actual performance depends heavily on atmospheric conditions and the correction service in use at the time.
Modern equipment has not eliminated the effect of baseline length. Network RTK corrections model atmospheric behavior across a region and can extend usable range considerably, but they do not remove the underlying relationship between distance and degraded cancellation. They manage it, they do not erase it.
What This Means for Trusting a Position
A rover working a long distance from its correction source, whether a single base or a network's modeled reference, is operating with a wider margin for uncorrected atmospheric residual than the same rover working close in. That does not automatically make the position untrustworthy, but it changes how much precision is reasonable to expect, and it is worth knowing the approximate baseline in play before treating every FIX solution as equally tight.
Two points are collected using the same rover on the same day, one five kilometers from the correction source, one forty kilometers away. Should both be trusted to the same level of precision? Not automatically. The longer baseline carries a larger expected residual error, even under an otherwise identical FIX solution.
RTK precision degrades gradually with baseline length, because atmospheric and orbital errors become less correlated between rover and reference over distance. Network RTK extends usable range by modeling that variation, but the underlying relationship between distance and residual error never disappears entirely.
Baseline length matters because of the atmosphere sitting between rover and reference. The next chapter looks directly at that atmosphere, and at the conditions that make it a bigger factor on some days than others.