Another thing we might want is a very small set of core variables (at the moment one) that are initialised from the configuration before models initalise.
The motivating case is layer_heights, which is initialised by the Plants model at the moment. That makes some sense, since most of it is canopy height, but the actual initial state is all from config and the core components, but it is used by all sorts of models as a central reference for vertical layer heights.
So, we could introduce those core variables - possible as part of the initialisation process of CoreComponents - and maybe flag them as core somehow. I don't think this is part of the variables PR, just something that I'm thinking about right now and wanted to log.
Another thing we might want is a very small set of core variables (at the moment one) that are initialised from the configuration before models initalise.
The motivating case is
layer_heights
, which is initialised by the Plants model at the moment. That makes some sense, since most of it is canopy height, but the actual initial state is all from config and the core components, but it is used by all sorts of models as a central reference for vertical layer heights.So, we could introduce those core variables - possible as part of the initialisation process of
CoreComponents
- and maybe flag them as core somehow. I don't think this is part of the variables PR, just something that I'm thinking about right now and wanted to log._Originally posted by @davidorme in https://github.com/ImperialCollegeLondon/virtual_ecosystem/issues/388#issuecomment-2167599389_