xlsn not defined in coupled mode for LIM3

[ufla]

Amongst some other constants, xlsn (volumetric latent heat fusion of snow) is only declared in OPA_SRC/DOM/phycst.F90 if key_lim3 is NOT defined. However, in coupled mode (key_coupled) the variable is referenced in OPA_SRC/SBC/sbccpl.F90, thus preventing the code from compiling.

[gm]

In fact this problem concerns both the xlsn (used for melting solid precip by multiplying by xlsn/rhosn) and xlic (use for melting iceberg by multiplying by xlic/rhoic).

One has to realize that xlsn/rhosn = xlic/rhoic = lfus = latent heat of ice melting = 334.e⁺⁶ -- therefore this is the variable that should be used.

the phycst.F90 module has been modified to introduce lfus in LIM2 with a value of xlsn/rhosn =333.7e⁺⁶ in order to keep the implicit value already used in both xlsn and xlic. In next version (v3.3) a same value of 334.e⁺⁶ will set for both LIM-2 and LIM-3.

in sbccpl.F90 module, we now use only lfus value to compute both solid precip. and iceberg melting heat.

NB: for CICE guys, first be sure of defining lfus, and second, please check the value of latent heat of ice you are using. We should share a same value for the 3 ice models.

[charris]

Gurvan - we currently use 0.334e+6 for lfus in CICE. We also use 2.501e+6 for cevap and 2.835e+6 for lsub so these values are very slightly different to the LIM ones. Should all three be defined independently or one calculated from the other two?

Chris