Need flexible index for bare soil

[dsolyga]

Bare soil should not be set necessarely to the first element of PFT-array. It is very risky to have a fixed index.

Proposal :

Lot of work in perspective. The risk has diminished since the externalization (the model will crash if are soil is not in first position).
Need to create a label to distinguish bare soil from the other PFTs. Very complicated because most of the model should be recoded.

[jpolcher]

Replying to dsolyga:

Bare soil should not be set necessarely to the first element of PFT-array. It is very risky to have a fixed index.

Proposal :

Lot of work in perspective. The risk has diminished since the externalization (the model will crash if are soil is not in first position).
Need to create a label to distinguish bare soil from the other PFTs. Very complicated because most of the model should be recoded.

Why is that ?

If the index of the bare soil PDF is a parameter of the model than the bare soil PFT can be shifted at will.
It will take some time until in all parts of the code the assumptions ibaresoil=1 are used. But with some time it can be done and it is easy to test.

[jpolcher]

We should have this discussion on the surface fraction definition/slowproc document ... for which I still have not had any return.

[jgipsl]

This will be part of a bigger discussion on how to better handle the PFTs and move towards SFTs(surface functional types) including lakes, ice, cities,...