SI3 heat budget for melting/growing sea ice

I think there is an issue in the way we deal with the heat budget in SI3 for forming/melting sea ice. It is probably not dramatic but it is not as accurate as it should be. The problem comes from the fact that we need to somehow guess the energy budget in the leads without having information on all the fluxes. And this guess is (I think) not so accurate. Whether it is linked to occurence of supercool waters in some specific configurations, I do not know.

As far as I understand, I think we need to replace these lines in icesbc.F90:

   IF( zqld >= 0._wp .AND. at_i(ji,jj) > 0._wp ) THEN
     fhld (ji,jj) = zswitch * MAX( 0._wp, ( zqld - zqfr_pos ) * r1_Dt_ice / MAX( at_i(ji,jj), epsi10 ) - qsb_ice_bot(ji,jj) )
     qlead(ji,jj) = 0._wp
   ELSE
     fhld (ji,jj) = 0._wp
     qlead(ji,jj) = MIN( 0._wp , zqld - ( qsb_ice_bot(ji,jj) * at_i(ji,jj) * rDt_ice ) - zqfr )
   ENDIF

by those ones:

   IF( ( zqld - zqfr ) < 0._wp .OR. at_i(ji,jj) < epsi10 ) THEN
     fhld (ji,jj) = 0._wp
     qlead(ji,jj) = MIN( 0._wp , zqld - zqfr )
   ELSE
     fhld (ji,jj) = zswitch * MAX( 0._wp, ( zqld - zqfr_pos ) * r1_Dt_ice / MAX( at_i(ji,jj), epsi10 ) )
     qlead(ji,jj) = 0._wp
   ENDIF