Lesley De Cruz

@sidekock and I will start working on a "control member" which does not include the noise cascade and does not become smooth over time. First idea: create a simple non-decomposed...

Thanks @aitaten! Yes, very interesting! Both are interesting but I think they contain different information. We will try to create a control member according to the description I posted above...

Update: we actually already created a cascade level-2 blended nowcast for the probability matching (see comment "8.7.1 first blend the extrapolated rainfall field (the field that is only used for...

The naieve approach to just use the combination blended with the weights of the second cascade level doesn't seem to work too well (apologies for the crappy plots). The weights...

OK, so I didn't manage to solve it in time ;-) The control member now "dies out" according to the AR(2) process which is not ideal, but closer to the...

Hi all, thanks a lot for the insightful remarks above, there's indeed a fundamental inconsistency in trying to create an unperturbed control member which would also lie within the STEPS...

I will reopen this soon, will just blend with the first NWP member as a sane default.

Additionally, dependencies are defined all over the place: - environment.yml (and the corresponding _dev file) - requirements.txt (and the corresponding _dev file) - setup.py - pyproject.toml Which is the source...