mone27
commented
3 years ago I was thinking about doing it in this way:
- Every submodel needs to have a function that takes as argument the current input and the previous state (as a dataframe row) and returns the output and the new state.
- for input we read the hainich fluxnet data (at least this is what I understood)
- Then there is a main loop which steps over the time and calls the submodels with the correct input while keep track of their state and their outputs.
- for the output we can just collect everything in a big dataframe and save it as csv (or I am missing something?)
In this way the model should be able to run, but probably we are going to have to spend some time to make sure that the submodels interacts well together (eg. emitted longwave radiation is calculated at least twice). Also If I remember correctly Alexander was saying that we are not going to use the leaf temperature model because it is already embedded in the photosynthesis one.
Finally we could centralize also the parameters management and think to have a consistent structure in the submodels.
Can this be a viable working plan? @femoyano what is your proposal?
femoyano
Some necessary initial planning
Specific questions
How will the different code parts be organized/located? What input will be used? How will the output be written and saved?