moving the time stepping of canopy water content out of the canopy (energy balance) routine

[penguian]

keyword_keepgit owner:[email protected] type_JAC readiness | by inh599

---

In preparation for JAC we would be advised to modify the structure of CABLE so that appropriate parts of the model can be called cleanly. At the moment the canopy water content state variable (canopy%cansto) is time stepped within the cable_canopy routine - this is the only state variable that is updated this way, all the rest are updated in soil_snow, SLI, CASA-CNP etc.

In JAC we anticipate that cable_canopy will be called multiple times per time step and will only need to evaluate the surface energy balance (and associated fast flux of carbon). Correspondingly we need to reorganise the code so that the time stepping of the canopy water content can be completed separately from cable_canopy - while maintaining conservation and current performance.

A suggested method to do so is (in some detail based off the trunk at 9037)

1. remove the update of %cansto at line 49 of cbl_SurfaceWetness
2. update the calculation of %fwet at line 52 of cbl_SurfaceWetness to use %cansto + %wcint
3. modify lines 854-879 of cable_canopy to remove the updates to %cansto while retaining the correct evaluation of the flux terms %spill, %through and %precis. Note that %wcint needs to be added to %cansto at various points when evaluating the fluxes to preserve the intent (because of change 1. above)
4. remove the evaluation of %delwc (line 879) from cable_canopy
5. create a new SUBROUTINE to time step %cansto and evaluate %delwc. This would be called immediately before soil_snow in cbl_model_driver_offline, before the call to soil_snow in cable_implicit_driver (in ESM1.5 and CM2) and before the call to soil_snow in the WRF and CCAM driver routines.

Connections/interdependence with the ground water model and SLI may need to be followed through.

In CM2 %cansto would still need to be placed within the prognostic bank structure.

There are other 'working state variables' that are updated in a similar manner within cable_canopy - canopy%rtsoil, canopy%fes_cor and canopy%ga (if the Penman_Monteith option for latent heat flux from the ground is used). These are variables whose values (currently) need to be updated correctly from time step to time (e.g. in order to conserve) but this is a result of the algorithm design not because they represent a biophysical state of the land. JAC will need to be mindful of (i.e. address) these working state variables at some point in the development cycle.

---

Issue migrated from trac:352 at 2023-11-27 11:43:56 +1100

[penguian]

@[email protected] set keywords to maygit

[penguian]

@[email protected] changed keywords from maygit to keepgit