Context

Output of ice-ocean sensible heat flux has the wrong sign

Fix

replace this line:

IF( iom_use('hfxsensib'  ) )   CALL iom_put( 'hfxsensib'  ,     -qsb_ice_bot * at_i_b         )

By this one:

IF( iom_use('hfxsensib'  ) )   CALL iom_put( 'hfxsensib'  ,      qsb_ice_bot * at_i_b         )

Hi all,

Just a minor bug. In limadv.F90 one has several calls to ensure the proper exchange of data between the different CPUs, the cyclic boundary condition if needed, and the north fold of the domain if also required:

      !-- Lateral boundary conditions
      CALL lbc_lnk( psm , 'T', 1. )
      CALL lbc_lnk( ps0 , 'T', 1. )
      CALL lbc_lnk( psx , 'T', 1. )
      CALL lbc_lnk( psxx, 'T', 1. )
      CALL lbc_lnk( psy , 'T', 1. )
      CALL lbc_lnk( psyy, 'T', 1. )
      CALL lbc_lnk( psxy, 'T', 1. )

Well, because psx and psy are gradients of ps0, they should be treated as velocity, i.e., with a sign change through the north fold. So the correct calls should be for psx and psy:

      CALL lbc_lnk( psx , 'T',-1. )
      CALL lbc_lnk( psy , 'T',-1. )

Fred.

In routine floblk.F90, the velocities zwinfl and zwoutfl (lines 139-142 of revision 1152 on the trunk) involve the vertical velocity with its direction reversed. However, the z direction has already been made positive upwards, which is consistent with the coordinate system of the unadulterated w velocity. So I believe this sign change is wrong (but so could I be!). Steven