JacTechnicalReviewJhan

JULES (formerly MOSES) was developed within the UM before it had a standalone option. Perhaps this is a contributing to factor as to why there are multiple calls to the LSM per timestep, even neglecting the further complicating role of ENDGAME dynamics (which calls the LSM twice per timestep [down the implicit pathway]).

Earlier discussions with UKMO-JULES team have prompted some simplification of the calling points to the "LSM". Online, JULES points of contact with the UM atmospheric model comprise the surf_couple* routines;

surf_couple_radiation()
surf_couple_explicit()
surf_couple_implicit()
surf_couple_extra()

which we also refer to as the radiation, explicit, implicit and hydrology (extra) pathways.

These are also the same calling points in the standalone model (following the IO section which we intend to co-opt) in a timestepiing routine (control.F90). A pseudo code description of this folllows:

subroutine control()

do i=1,end_step

   call surf_couple_radiation()
   call surf_couple_explicit()
   call surf_couple_implicit()
   call surf_couple_extra()

end do

Our approach is to tackle these in the same order in which they are executed:

surf_couple_radiation

surf_couple_explicit

surf_couple_implicit

surf_couple_extra

Previous Home