Lilli Sun,
This is for H1 reference model uncertainty test (see 43260 for L1).
The 20190118 model is a bit confusing. The model parameters are rescaled according to 46605, which are not consistent with MCMC results. Using the model file and the MCMC results leads to systematic errors. Need to further investigate if the systematic error seen in the attached uncertainty plot is caused by the inconsistency.
There's still time-dependent, frequency-dependent systematic errors in the DARM loop that the model does not capture. See 46806
At the moment it is not unexpected to see significant systematic errors in uncertainty plots. The sensing GPR plot does show a large systematic error.
Attached are the resulting MCMC corner plots, model vs measurements plots, GPR plots. The TDCFs are all set to 1.0 since it's for the reference model.
