ESM15 reconciliation: casa_delsoil phosphorus

[rml599gh]

Lines 1446-1466 in casa_cnp.F90, subroutine casa_delsoil includes alternate ifdef ESM15 versions of the calculation of casapool%dPsoillabdt and casapool%dPsoiloccdt.

             casapool%dPsoillabdt(nland)= casaflux%Psnet(nland) + fluxptase(nland)         &
                  + casaflux%Pdep(nland) + casaflux%Pwea(nland)      &
                  - casaflux%Pleach(nland)-casaflux%pupland(nland)   &
!jhan: ESM15 is effectively using xkpsorb**2 - inadvertently?!?!?!?!
#           ifdef ESM15
                  - casabiome%xkpsorb(casamet%isorder(nland))*casaflux%kpsorb(nland)*casapool%Psoilsorb(nland) &
#           else
                  - casaflux%kpsorb(nland)*casapool%Psoilsorb(nland) &
#           endif
                  + casaflux%kpocc(nland) * casapool%Psoilocc(nland)
             ! here the dPsoillabdt =(dPsoillabdt+dPsoilsorbdt)
             ! dPsoilsorbdt  = xdplabsorb
             casapool%dPsoillabdt(nland)  = casapool%dPsoillabdt(nland)/xdplabsorb(nland)
             casapool%dPsoilsorbdt(nland) = 0.0

#           ifdef ESM15
             casapool%dPsoiloccdt(nland) = casabiome%xkpsorb(casamet%isorder(nland))*casaflux%kpsorb(nland)* casapool%Psoilsorb(nland) &
#           else
             casapool%dPsoiloccdt(nland) = casaflux%kpsorb(nland)* casapool%Psoilsorb(nland) &
#           endif
                                         - casaflux%kpocc(nland) * casapool%Psoilocc(nland)

In CABLE2.4 xkpsorb is defined in cableparm as 0.01*deltcasa where deltcasa=1/365. This is used to calculate kpsorb in casa_coeffsoil. In delsoil kpsorb is then multiplied by xkpsorb but now xkpsorb is dependent on soil order with values specified in delsoil in a data statement (ranging from 0.5-0.99). Including the 0.01*deltcasa factor gives a range of 1.37e-5 to 2.71e-5.

In CABLE3 xkpsorb is read in from the pftlookup file and are dependent on soil order AND are consistent with the 0.01*deltcasa factor being included, i.e. xkpsorb ranges from 1.37e-5 to 2.71e-5. In casa_coeffsoil, kpsorb uses the soil order dependent xkpsorb values.

This means that the non-ESM15 version of the code gives the same behaviour as the CABLE2.4 code given appropriate values of xkpsorb in the pftlookup file. The ESM15 versions should be removed as they are inconsistent with how xkpsorb is now input.