Closed BaptisteVandecrux closed 1 month ago
Of course there are complications with the use of persistence QC.
At several sites, lat/lon/elevation have degraded precision, which, depending on the magnitude and direction of their movement, lead to long periods of constant longitude or latitude that are not a malfunction, but many good measurement rounded and saved with identical digits.
If we use persistence QC in those situation, we will throw out true measurements.
See an example below with NUK_L:
Issue spotted at numerous sites. Example here at TAS_A:
https://github.com/GEUS-Glaciology-and-Climate/PROMICE-AWS-data-issues/issues/114