search

DOV-Vlaanderen / groundwater-logger-validation

Analysis on validation methods for groundwater logger data
MIT License
2 stars 2 forks source link

DYLP222A LS when TC expected #41

Closed fredericpiesschaert closed 5 years ago

fredericpiesschaert commented 5 years ago

The second levelshift seems to be a temporal change to me. Source file: https://github.com/DOV-Vlaanderen/groundwater-logger-validation/commit/aa8dc75889a4ca47b7e3d81a29fab4fa2eb6238b image

DavorJ commented 5 years ago

Good example. Seems related to (cf. presentation): image

In think in your particular case, the reason why it is a LS and not TC, is because it stays low for several points (LS), and then gradually starts decaying (TC). Exponential decal is strongest at the start, so TC doesn't fit well for the first couple of points where the decay effect should be strongest.

fredericpiesschaert commented 5 years ago

aha, that makes sense indeed. Perhaps a nice example to add to the documentation? It is an eye-opener to me.