The comments in the 2 following lines in OPA namelists:

ln_dynvor_ene = .false. ! enstrophy conserving scheme ln_dynvor_ens = .false. ! energy conserving scheme

should be just inverted to give:

ln_dynvor_ene = .false. ! energy conserving scheme ln_dynvor_ens = .false. ! enstrophy conserving scheme

Default parameters in dynadv_ubs lines 27-28 are not coherent with the documentation, a fourth order centered scheme with a quickest diffusive part is activated by default instead of second order with 1/3 order diffusive part.

There is a bug about the dynamic allocation in sbcblk_clio.F90 in subroutine blk_oce_clio, blk_ice_clio, blk_clio_qsr_oce and blk_clio_qsr_ice.

• In subroutine blk_oce_clio, checks on wrk_in_use and on wrk_not_release are wrong. You have to test 2d variables and not 3d varaibles.
• In subroutine blk_ice_clio, some 3d workspace arrays are unavailable. I suggest to change wrk_3d_1 and wrk_3d_2 by wrk_3d_4 and wrk_3d_5.
• In subroutine blk_clio_qsr_oce, some 2d workspace arrays are unavailable. I suggest to change wrk_2d_1 to wrk_2d_6 by wrk_2d_4 to wrk_2d_9.
• In subroutine blk_clio_qsr_ice, some 2d workspace arrays are unavailable. I suggest to change wrk_2d_1 to wrk_2d_6 by wrk_2d_5 to wrk_2d_10.

In attachment the corrected file with my suggestions.

Best regards,

Pierre Mathiot