jservonnat
commented
5 years ago You are right that CliMAF should be improved on this area. We consider adding an argument to .explore() : allow_shorter_period = True/False If False, this would meet your initial need. If True, it would return an the largest possible ensemble with covers the periods intersection. In both cases, a warning should be issued to explain what occurred.
We will let you know when the feature will be available.
Hello,
I am trying to define an ensemble of files that all cover the requested period. However, in some CMIP6 experiments, there are multiple ensemble members and sometimes they do not cover the same period. That's the case for the abrupt-4xCO2 run for example. My idea is thus to search all the ensemble members available for an experiment that covers the requested period.
I tried this :
define dictionnary and climaf project
`d = dict(project='CMIP6',mip='CMIP', model='CNRM-CM6-1', experiment='abrupt-4xCO2', realization='*', variable='tas', table='Amon', period='1850-2349') tmp = ds(**d)
Get the members available
members = tmp.explore('choices')['realization']
if one member, then create a list
if type(members) is not list: members = [members]
update the dataset
d.update({'realization' : members}) var = eds(**d)
The issue is that climaf will return all ensemble members available for this experiment. However, some of them (in that particular case) does not respect the requested period.
Is there a way to fix this ? Thank you,