This is a follow-up to #81, which dealt with extracting the calculation of global radiation and sunshine duration from larger models. Unfortunately, we did so before introducing the submodel concept. So, our HydPy-Meteo submodels currently only work as stand-alone models, not as submodels in the sense of #90. This is the reason for the unfavourable situation that, for example, HydPy-L uses HydPy-MORSIM as a "real" submodel, but HydPy-MORSIM needs to be connected to HydPy-Meteo_V1 via the old node-based mechanism.

I hope we will not require many discussions because the main functionalities of all Meteo models are already available. The only problematic point I see coming is that one might wish to use the same Meteo submodel instance twice, for example, for calculating global radiation for the snow routines of HydPy-L and the evaporation submodel used by HydPy-L, which is currently only possible when using the old node-based mechanism.