Closed casadoj closed 1 year ago
I have asked Juliana for the observed discharge data.
I have used the KGE that Cinzia has provided from the calibration. However, several reporting points don't have KGE values, so I can't filter them.
Using a KGE limit of 0.5 has filtered out the reporting points with a large number of "observed" events.
I have identified some stations with very poor performance. So far, I have removed them from the analysis by removing stations in which there were 5 of more observed flood events in 2021-2022.
It could be interesting to do this selection of good/bad-performing stations using the KGE between the observed time series (actually measured) and the reanalysis time series. For that, I will have to ask Juliana for the discharge series of all the fixed reporting points for the 2 years of analysis.