Skip to content

Neutral grid mapping in geographic coordinates onto periodic GEMINI #104

Open

Description

@paulinchin

The neutral=4 (geographic coordinates) cannot be properly mapped onto periodic GEMINI grid. This can result from the way it is replicating the glat/glon variables there and then using them to interpolate the neutral data in space. It works fine in neutral=3 though.

The block that should potentially be the issue is:

!> assign a reference meridian alt,lat,lon across periodic grid
  subroutine set_periodic(self,flagperiodic,refalt,refglon,refglat)
    class(curvmesh), intent(inout) :: self
    integer, intent(in) :: flagperiodic
    real(wp), dimension(:,:), intent(in) :: refalt,refglon,refglat
    integer :: ix3

    ! flag appropriately
    if (flagperiodic/=0) then
      self%flagper=.true.
    else
      self%flagper=.false.
    end if

    ! In the special case where the user wants flagperiodic==1, we take the additional step of forcing
    !   glat/glon to be the same across the x3 dimension so that the neutral atmosphere and photoionization
    !   will be constant vs. x3.  This is most typically used in instability simulations when one wants to
    !   explicitly remove any dependence of background parameters on the x3 direction.  One would not want to use
    !   this when simply trying to model an angular coordinate (e.g. longitude) across the full globe as the
    !   atmospheric and SZA changes are needed to realistically capture the system.
    ! force periodicity in geographic locations using reference meridian data
    !! FIXME: specify ghost cells for glat/glon arrays?
    if (flagperiodic==1) then
      do ix3=1,self%lx3
        self%glat(1:self%lx1,1:self%lx2,ix3)=refglat(:,:)
        self%glon(1:self%lx1,1:self%lx2,ix3)=refglon(:,:)
        self%alt(1:self%lx1,1:self%lx2,ix3)=refalt(:,:)
      end do
    end if
  end subroutine set_periodic

Activity

  1. mattzett commented on Jul 25, 2023

    @mattzett
    Member

    I think when I was doing my initial tests with neutral forcing of bubble I was using interptype=3. I think that is probably the only way to use neutral perturbations with periodic grids because geographic input neutral perturbations will not along with the longitude coordinate in the model. So they need to be aperiodic, but I'm not sure that would work for plasma bubble simulations because there is a tendency for things to grow on the edges of the domain.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions