Measurement
Data gap filling
Data gap filling is estimating readings that are missing from a record, because a sensor failed, a battery ran out or a site visit was missed. It is done by a method stated in advance, and every filled value is flagged so it can never be mistaken for a measured one.
The method matters less than its being declared before the gap appears. Choosing a method after seeing the gap invites choosing the one that flatters the result.
Where a methodology asks for it, the filled value is the conservative one, so a gap cannot add credits that a measurement would not have.