Adding moss and lichen as two new plant functional types (PFT)s in CLM-FATES

Purpose with this issue

Discuss how to implement moss and lichen as two new PFTs in FATES, which can be run with the host model CLM5. Both biophysical and biochemical processes of moss and lichen will be considered. In the end, the dynamics and distribution of moss and lichen can be predicted by CLM-FATES. There is also a corresponding issue on FATES repository (issue #707)

Brief science information

Moss (in more general term, bryophyte) and lichen are unique in their biophysical and biochemical feedback to climate compared to vascular plants, and are important components in the cold and dry environment. Moss and lichen have been implemented several land surface models, such as JULES and ORCHIDEE and JASBACH, with varying complexity. With the implementation of moss and lichen, we hope to further improve CLM-FATES in representing vegetation and its climatic feedbacks in the cold environment. More details can also be found in this presentation

Plan for the implementation

The following figure presents a simplified flow-chart for the suggested changes:

Current plans

[x] Step1: Biophysical impacts of moss and lichen (i.e., photosynthesis, thermal insulation, albedo, and water holding capacity)
- Key features have been and will be implemented :
  - Relatively large water holding capacity (CLM)
  - Strong insulation for heat and moisture transfer from ground to air, which is also a function of moisture content (CLM)
  - No root, no btran, no transpiration (FATES)
  - Vcmax limited by relative moisture content (FATES)
  - Internal leaf CO2 partial pressure (Ci) is not influenced by stomatal conductance but by relative moisture content (FATES)
  - Optical parameters influenced by water content (FATES)
  - High snow interception (CLM) and quick snow burial for moss (FATES)
- Following files are changed:
```
Environmental variables: 
$HOME_CLM   = https://github.com/ESCOMP/CTSM
$HOME_FATES = https://github.com/NGEET/fates
```
  - In FATES:
```
$HOME_FATES/main/EDParamsMod.F90
$HOME_FATES/main/EDPftvarcon.F90
$HOME_FATES/main/FatesInterfaceMod.F90
$HOME_FATES/main/FatesInterfaceTypesMod.F90
$HOME_FATES/main/FatesRestartInterfaceMod.F90
$HOME_FATES/biogeophys/FatesPlantRespPhotosynthMod.F90
$HOME_FATES/biogeophys/EDSurfaceAlbedoMod.F90
$HOME_FATES/biogeochem/EDCanopyStructureMod.F90
$HOME_FATES/biogeochem/EDMortalityFunctionsMod.F90
$HOME_FATES/parameter_files/fates_params_default.cdl 
```
  - In CLM:
```
$HOME_CLM/src/biogeophys/CanopyFluxesMod.F90
$HOME_CLM/src/biogeophys/CanopyHydrologyMod.F90
$HOME_CLM/src/utils/clmfates_interfaceMod.F90
```
[ ] Step2: Add biogeochemical and demographic processes (e.g., carbon allocation, nutrient cycle, reproduction)
[ ] Step3: Add moss and lichen as a understorey layer or grow on other plants (Epiphytic)

Long term goal

Build comprehensive database or testbed for modeling moss and lichen.
Include Moss and lichen in NorESM or CESM for the next phase of CMIP.

Major issues

Water interception is treated in CLM, which can not describe water interception by upper- and under-storey canopy in FATES, respectively.
- Should we copy water interception code into FATES? Or Creat addition variables to keep upper and understorey information from FATES, and send them to CLM?
Need to keep LAI, SAI, roughness length of moss and lichen (from FATES) separated from the other PFTs in CLM, so that the resistance of moss or lichen layer to moisture and heat transfer can be calculated in a way different from the other PFTs in CLM?
Need to further improve thermal impact of moss and lichen. Plan to develop a new resistence model for moss and lichen based on newly available BHS code (@swensosc). Can this be done together with solving issue #1274 (@ekluzek,@rgknox)?
Need to add moss and lichen into the surface data file of CLM. How should this be done? We need global/regional cover fraction of moss and lichen.
Any suggestions on the key features of moss and lichen to be implemented in the model are welcomed.

Plan to move things forward

Current implementation of Step1 is available on github (ctsm and fates) (Further improvement is still needed). To divide up the review work we think it is reasonable to file a PR when each step is finished. In this manner there is room for a restructuring of the code before too much new functionality is added.

Would you (@billsacks,@ekluzek on CLM side, @rgknox,@glemieux on FATES side) be interested in having a PR at each step?

People involved:

@huitang-earth, @kjetilaas, @rosiealice, @ckoven, @jkshuman, @adrifoster, @ingealthuizen, @eaasmof, @srg101, @sunnivin

ESCOMP / CTSM