Bug with so_capa_ice

[maignan]

Presently, so_capa_ice is defined in thermosoil as:

so_capa_ice(:) = so_capa_dry + mcs(:)*capa_ice*rho_ice 

and is later used here as:

            pcapa(ji, jg) = so_capa_dry_ns(jst) * (1-mcs(ji)) + so_capa_ice(ji) * tmc_layt(ji,jg) / mille / dlt(jg)
...
	           IF (ok_freeze_thaw_latent_heat) THEN
	              pcapa(ji, jg) = so_capa_dry_ns(jst) * (1-mcs(ji)) + &
	                water_capa * tmc_layt(ji,jg)/mille / dlt(jg) * xx + &
	                so_capa_ice(ji) * tmc_layt(ji,jg) / mille/dlt(jg) * (1.-xx) + &
	                shum_ngrnd_perma(ji,jg)*mcs(ji)*lhf*rho_water/fr_dT
	           ELSE
	              pcapa(ji, jg) = so_capa_dry_ns(jst) * (1-mcs(ji)) + &
	                water_capa * tmc_layt(ji,jg)/mille / dlt(jg) * xx + &
	                so_capa_ice(ji) * tmc_layt(ji,jg) / mille/dlt(jg) * (1.-xx)
	           ENDIF

so that the mineral soil capacity is counted twice, and the coefficient mcs should not be present.

We propose to modify in:

so_capa_ice(:) = capa_ice*rho_ice 

Ideally, and to be consistent with water_capa expressed in J/m3/K and defined in constant_soils_var.90, we should have ice_capa defined directly as capa_ice*rho_ice in the same file.

The change is to be made in versions 2.2/3.0/4.0.

Fabienne & Agnès. To be validated by Catherine.

[cottle]

It should be correct if ice_capa is really defined as the ice specific heat capacity in J/kg/K, then the multiplication with the density will transform it in a volumetric heat capacity in J/m3/K. I have still a doubt since water_capa is already a volumetric heat capacity in constantes_soil_var.f90 (in J/m3/K)and I cant find where ice_capa is defined and what are its units.

Besides, I see that some of the comments and units of the capacity are not coherent in the codes, when the units are written in J/m3/K, it is written only heat capacity and not volumetric heat capacity (it is the case for all the soils values).

I see also a variable called mx_eau_nobio in constantes_soil_var, that is used apparently to calculate a volumetric water capacity for nobio which has not the correct units,

!$OMP THREADPRIVATE(qsintcst)
122 REAL(r_std), SAVE :: mx_eau_nobio = 150. !! Volumetric available soil water capacity in nobio fractions
123 !! @tex $(kg.m^{{-3} of soil)$ @endtex}

It is also defined in constantes_soil.f90, still with an error in the units!

!Config Key = NOBIO_WATER_CAPAC_VOLUMETRI

114 !Config Desc =
115 !Config If =
116 !Config Def = 150.
117 !Config Help =
118 !Config Units = [s/m^{2]
119 CALL getin_p('NOBIO_WATER_CAPAC_VOLUMETRI',mx_eau_nobio)}

So all these lines if they are not used should be deleted, and all the definitions of the heat capacities variables should be also cleaned with a title coherent with the units , in order to not mislead the reader.

[jgipsl]

According to the comment in src_parameters/constantes_var.f90, de unit of capa_ice is J/kg/K:

REAL(r_std), PARAMETER :: capa_ice = 2.228*1.E3       !! Heat capacity of ice (J/kg/K)

[jgipsl]

-------- Message transféré --------
Sujet : 	Re: so_capa_ice
Date : 	Tue, 1 Mar 2022 13:59:45 +0100
De : 	Catherine Ottle <catherine.ottle@lsce.ipsl.fr>
Pour : 	josefine.ghattas@ipsl.fr <josefine.ghattas@ipsl.fr>


Bonjour Josefine,

La valeur est bien celle d'une capacité spécifique et pas volumique, et l'unité est aussi 
cohérente, donc c'est bon pour moi de faire la correction du ticket. Je ne peux pas faire les 
corrections dans les commentaires car je n'ai pas le droit de committer, mais à l'occasion, je 
veux bien prendre le temps de te lister ce qui doit être corrigé.

Bonne semaine et à bientot,

Catherine

[jgipsl]

Modification on so_capa_ice as decribed in the description of the ticket done in ORCHIDEE_2_2 here : [7506]
And in the trunk here: [7507]

[jgipsl]

Still to be done:

• integration in ORCHIDEE_3 to be decided
• clairification on comments (Catherine). Is it enough to do it in the trunk?

[jgipsl]

Comments are corrected in trunk [7520] and ORCHIDEE_2_2 [7519].

[nvuilsce]

Done in ORCHIDEE_3 Commit [7633]