Tier Four: Multiple Independent Re-Occupations
A single session, however well corrected, shares one unavoidable limitation: it only ever sees one day's satellite geometry and one day's atmospheric conditions. Occupying the same point again, on a different day, catches a category of error no amount of correction within a single session could ever reveal.
Reoccupying a base point on separate days and averaging the independent results catches a bias tied to one specific day's satellite geometry or atmospheric conditions, the same logic behind forcing a fresh ambiguity resolution, discussed in the third companion of this series, just applied across days instead of minutes.
What a Single Corrected Session Still Cannot See
Even an excellent, well-corrected session is still only one sample of a specific, if not literally random, set of conditions, one satellite constellation arrangement, one day's atmospheric state. A subtle bias tied specifically to that day's geometry or conditions would look completely normal from inside that single session, with nothing about the data itself suggesting anything was off.
A forced reinitialization within one session tests whether that session's ambiguity resolution was a fluke. Reoccupying on a separate day tests something a single session cannot: whether the day itself, its geometry or its atmosphere, was quietly responsible for the result.
How Many Occupations Is Enough
There is no single universal answer, but the underlying principle is straightforward: each additional independent occupation, ideally spread across genuinely different days rather than back-to-back sessions, adds real evidence against a day-specific bias. Two or three well-separated occupations, each established with whatever method from the earlier chapters fits the connectivity available, already meaningfully outperforms a single session of any kind.
Reoccupying the same point three times in one afternoon is not the same evidence as reoccupying it on three separate days. Same-day reoccupations share the same satellite geometry and the same atmospheric conditions, and cannot reveal a bias tied to either one.
A Middle Tier, Not a Final Answer
This method sits deliberately between the quick, single-session establishments already covered and the full rigor of the chapters still ahead. It costs real calendar time, spread across days rather than compressed into one sitting, in exchange for catching an entire category of error the faster methods simply cannot see. It is not yet the most rigorous option this companion covers, and the next chapters explain what that fuller rigor actually adds beyond this.
A base coordinate was established using a single, well-corrected network RTK session. Does that session's precision rule out a day-specific bias in the result? No. A single session, however well corrected, cannot distinguish its own result from a bias tied to that one day's conditions. Only an independent session on a different day can test that.
Reoccupying a base point across genuinely separate days and averaging the independent results catches a bias tied to one day's satellite geometry or atmosphere, something no single session, however well corrected, has any way of revealing on its own. It is a real, meaningful step up in rigor from any method covered so far.
The remaining methods in this companion stop relying on a real-time correction entirely, and instead lean on logging raw data for processing after the fact, starting with a method built specifically to work well near existing infrastructure.