getin in hydrol_main at each time step, why ?

[jgipsl]

Following is seen in hydrol.f90 in the trunk rev 7392, I've not checked when it was introduced. Why making a call to getin at each time step ?

   !! KFACT_ROOT_TYPE = cons is used to impose that kfact_root = 1 in every soil layer,
    !! so that ks does not increase in the rootzone; else, kfact_root defined as usual
    !Config Key   = KFACT_ROOT_TYPE
    !Config Desc  = keyword added for spmip exp1 and exp4 to get a constant ks over soil depth and rootzone
    !Config If    = spmip exp1 or exp4
    !Config Def   = var
    !Config Help  = can have two values: 'cons' or 'var'. If var then no changes, if cons then kfact_root=un
    !Config Units = [mm/d]
    kfact_root_type='var'
    CALL getin_p("KFACT_ROOT_TYPE",kfact_root_type)
    IF (kfact_root_type=='cons') THEN
    kfact_root(:,:,:) = un
    ENDIF

[jgipsl]

Done in branch 2_2 [7476]. Still to be done in the trunk. Note that the calculation of kfact_root is kept in hydrol_main because it needs a variable (vegetsoil_max) calculated in hydrol_vegup.

[jgipsl]

Done in the trunk: [7499]
The kfact_root calculation was previously done in the loop over the soiltiles in hydrol_soil. The calculation has now been moved to the begninning of hydrol_soil.

The calculation was previously moved into the soiltile loop because the calculation used the variable root_depth which is calculated in the same loop (see #773). This is not done anymore but it might be put back. In that case, root_depth could be added into the restart file and could then be used in the beginning of hydrol_soil.