Even Newer Dynamics for General atmospheric modelling of the environment (ENDGame),

Institution:  U.K. Met Office, Exeter, U.K.

Model Metadata

• Metadata Entry Form
• View Metadata

Description:

ENDGame is new dynamical core designed for the UK Met Office's unified model for weather and climate prediction. It uses a two-time-level iterated semi-Implicit semi-Lagrangian discretisation of the full incompressible Euler equations and is designed to work across a range of horizontal scales from O(100Km) to O(1Km). Physics schemes are coupled via a mixed parallel/sequential split for slow and fast processes respectively.

Typical horizontal resolutions, physics and dynamics time steps, and dissipation coefficients:

ENDGame uses a latitude-longitude grid, operationaly the horizontal resolution is chosen to give isentropic resolution at mid-latitudes this results in a 4:3 ration of grid points EW-NS. The the grid resolution is denoted by NXXX where XXX is the number of 2 grid length waves that can be supported in the EW direction. The operational resolution is N512 which yields 1024x768 grid points EWxNS. For DCMIP the resution will be a fixed number of degrees EW and NS.

ENDGame has no resolution dependent dissapation. The only additional form of damping is a rayleigh damping layer on the vertical velocity.

The semi-Lagrangian semi-implicit method used in ENDGame also has inherent damping mechanisms from the semi-Lagrangian interpolation (which is cubic by default) and if used the temporal off-centriing in the semi-implicit scheme.

Information on the computational grid:

ENDGame use a latitude-longitude grid with a C-grid stagering such that the meridional velocity is held at the poles. In the vertical it uses a Charney-Philips staggering with vertical velocity and potential temperature held at the surface and model lid.

Table of completed test cases

Results

Test Set 1

1-1 Deformational flow

        error norms and correlation plot (with SLICE)

        error norms and correlation plot (without SLICE)

1-2 Hadley cell flow

       error norms and convergence rates

1-3 Flow over orography

       error norms

Test Set 2

2-0-0 Resting atmosphere over orography

2-0-1 Resting temperature inversion over orography (Not currently implemented)

2-1 Flow over Schar mountain 

        Nonhydrostatic results: T'. U', W 

        Hydrostatic results: T', U', W

2-2 Sheared flow over Schar mountain

        Nonhydrostatic results: T', U' (perturbation from initial field),  W

        Hydrostatic results: T', U' (perturbation from initial field), W

Test Set 3

3-1  Nonhydrostatic gravity waves

Test Set 4

4-10 Dry Baroclinic wave X = 1

      -Hydrostatic results:  T850  VORT850   EPV850 EPV 315K PS W 1840m Kinetic Energy

      (411, 412, 413 hydrostatic are all visually same as 410 when W magnitude scaled by X)

      nonhydrostatic kinetic energy at day 20

      -For comparison nonhydrostatic results with increased damping

4-11 Dry Baroclinic wave X = 10

4-12 Dry Baroclinic wave X = 100

4-13 Dry Baroclinic wave X = 1000

       nonhydrostatic kinetic energy at day 20

4-2 Moist baroclinic wave with large scale condensation

          1 degree resolution:

                  -Nonhydrostatic: T850, PRECIP, PS

                  -hydrostatic: T850, PRECIP, PS

          0.5 degree resolution:

                 (note runs crash betwen 14.5 and 15 days)

                 -Nonhydrostatic: T850, PRECIP, PS

                 -Hydrostatic:  T850, PRECIP, PS

4-3 Moist baroclinic wave with simple physics

          1 degree resolution:

                  (note runs crash shortly after 12 days)

                  -Nonhydrostatic: T850, PRECIP, PS

                  -Hydrostatic: T850, PRECIP, PS

          1/2 degree resolution

                  -Nonhydrostatic (model fails after 10.5 days)

4-1A Dry deep atmosphere baroclinic wave

Test Set 5

5-1 Idealised cyclone with simple physics

5-2 Idealised cyclone with full model physics  (Not currently implemented)

References:

• An inherently mass-conserving semi-implicit semi-Lagrangian discretisation of the shallow-water equations on the sphere. Zerroukat et. al.
• Coupling a mass-conserving semi-Lagrangian scheme (SLICE) to a semi-implicit discretization of the shallow-water equations: Minimizing the dependence on a reference atmosphere. Zerroukat et. al.
• An inherently mass-conserving iterative semi-implicit semi-Lagrangian discretization of the non-hydrostatic vertical-slice equations. Melvin et. al.

Members of this modeling group during DCMIP-2012 and room location:

Room: to be determined