River History Needs a Stable Point Identity

A river name can stay the same while the measurement point changes. Long-term comparison requires knowing which point, source, method, and units belong to each part of the history.

Suppose the system later begins using a different model grid point. Joining old and new values into one continuous line can look like a real environmental shift even when part of the difference came from the selection change. If metadata does not record that transition, the explanation is lost.

A useful archive preserves stable identifiers, location, units, and source version. When corrected data arrives, the relationship between the old and new values matters too. If two versions exist for the same timestamp, we should know why they differ.

For me, a data archive is not only a collection of old numbers. It also preserves the conditions under which those numbers are allowed to be compared. Much of historical trust lives in the metadata.

Source: official reference.