Tier One: Plain Survey-In
The simplest way to give a new base a coordinate requires no correction service at all: let the receiver compute its own autonomous position and average that position over several minutes. It is also the least accurate method this companion covers, and understanding exactly why matters more than the technique itself.
A survey-in averages an uncorrected, autonomous position over time. Averaging reduces random noise effectively. It does nothing at all for systematic error, orbital, clock, and atmospheric bias that does not average toward zero no matter how long the session runs.
What Averaging Can and Cannot Fix
Random, moment-to-moment noise in an autonomous position genuinely shrinks as more samples accumulate, which is exactly why a longer survey-in produces a tighter, more repeatable result than a short one. Systematic error behaves completely differently. An uncorrected broadcast ephemeris error, or an uncorrected atmospheric delay, biases every single sample in roughly the same direction, and averaging a hundred biased samples together produces a very precise average of a biased answer, not a correct one.
A longer survey-in is more precise than a shorter one. It is not necessarily more accurate. Precision and accuracy are answered by entirely different evidence, a distinction this series has returned to since its third companion.
Where This Method Actually Belongs
A plain survey-in is a reasonable choice when a project's accuracy needs are genuinely modest, when the resulting base will only support relative, same-session work where the absolute coordinate matters less than internal consistency, or when no correction source of any kind is reachable. It is a poor choice whenever the base's absolute position needs to hold up to outside comparison, since that is exactly the kind of accuracy this method structurally cannot guarantee.
A survey-in that reports a small, tight standard deviation is not evidence that the resulting coordinate is accurate. That figure describes how consistently the receiver agreed with itself, not how close the average sits to the truth.
A Method Worth Knowing, Not Avoiding
None of this makes survey-in a poor tool. It makes it a specific tool, honestly understood, for situations where its particular weakness, uncorrected systematic bias, does not actually matter to the job at hand. The next several chapters cover methods built specifically to remove that weakness, each trading some amount of speed or connectivity for a meaningfully more accurate result.
A base is survey-in with no correction source, over a full hour, producing an extremely tight, repeatable position. Has that hour bought meaningfully better absolute accuracy than a five-minute survey-in would have? Not necessarily. The extra time improves precision, not the systematic bias that a longer average cannot see or remove.
Plain survey-in trades speed and simplicity for an accuracy ceiling set by uncorrected systematic error, not by how long the session runs. It is the right tool exactly when that ceiling does not matter to the work at hand, and the wrong one whenever a base's absolute coordinate needs to withstand outside comparison.
The very next method removes plain survey-in's core weakness entirely, using a correction source that needs no cellular connection or nearby infrastructure at all.