The ICON Project: Motivation, Current Status and Perspectives

1
The ICON Project Motivation, Current Status and
Perspectives

• A joint project of the Max Planck Institute for
  Meteorology (MPI-M)
• and the German Meteorological Service (DWD)
• MPI-M M.A. Giorgetta, E. Roeckner, L.
  Bonaventura(1), M. Esch, A. Gassmann, P. Korn, L.
  Kornblueh, H. Wan
• DWD D. Majewski, Th. Heinze, P. Ripodas, B.
  Ritter, H. Frank, D. Liermann, U. Schättler, J.
  Steppeler
• (1) Now at the Politecnico di Milano

2
Contents

• Motivation
• Development Plan
• Perspectives
• Summary

3
Motivation Dynamical cores are at the heart of
ESMs

• Schematic view of the Earth System as we want to
  represent in numerical models
• Dynamical cores are at the base of the general
  circulation models and their properties determine
  the quality of the solutions

4
Motivation at MPI Consistent cont. and transport
eq.

• Original reason
• Inherent problems with tracer mass conservation
  in the existing ECHAM5 atmospheric general
  circulation.
• ECHAM5
• Spectral transform dynamics with semi-implicit
  time stepping, Asselin time filter and hybrid
  vertical coordinate, global mass fixer
• Gridpoint transport scheme (LinRood), which is
  mass conserving provided winds are derived from a
  mass conserving dynamics
• Discretized continuity and transport equations
  are not consistent
• Sinks/sources of tracer masses, unless
  corrected winds are used
• No theoretical basis for correction of winds
• Need for new dynamical core with consistent
  solution of continuity and transport equations.
• Plan for a new global and regional model system

5
Motivation at DWD Extend the model system

• Include more processes in operational models as
  necessary for the increased use of satellite
  radiances.GME is not mass conserving
• Example ECMWF has extended the model to the
  stratosphere and included a (linearized) ozone
  scheme
• Extended forecasts / seasonal forecasts requires
  coupled atmosphere ocean model.GME is not
  coupled to an ocean model
• Need of a new model system
• Joint ICON project of DWD and MPI-M

6
Goals of the ICON project of MPI-M and DWD

• Develop a mass conserving dynamical core with
  consistent discretization of the continuity and
  transport equations
• Use icosahedron for grid construction to allow a
  quasi uniform horizontal resolution (Sadourny,
  1968 GME, CSU, NICAM)
• Hydrostatic and non-hydrostatic cores
• Global and regional domain atmospheric models?
  unify codes for global and regional dynamics
• Local mesh refinement in selected regions with
  conservative 2 way interaction (2 way nesting).
• Use same grid structure discrete operators for
  ocean model

7
Global and regional Icosahedral grids
Icosahedron 12 vertices 20 equilateral triangles

• Example for local grid refinement
• Quasi uniform base gridicosahedron edge ? 6
  cell edges
• 2 step refinement in a lat-lon regionover Europe
  by bi-section of edges1 triangle ? 4 ? 16

8
Development plan Done/Published In work To be
done

• Shallow water model prototype
• Grid generator
• triangular grids based on icosahedron
• Globally uniform refinement
• Shallow water model (ICOSWP)
• C-grid discretization (? Heinze)
• Vector valued interpolation by radial basis
  function
• Evaluation (? Bonaventura)
• Williamson test suite
• Benchmarking against STSWM and GMESWM (? Ripodas)
• New data structure
• Parallelization and refinement
• Global and regional domains, boundaries
• Great circle and small circle grids
  optimizations (? Heinze)
• Shallow water model (ICOSWM)
• Hydrostatic atmospheric model (ICOHAM)
• Dynamical core

9
Perspectives

• MPI-M
• Single numerical modeling framework for
  atmosphere and ocean
• ECHAM/MPIOM REMO ? ICONAM/ICONOM
• DWD
• GME LM ? ICONAM with global domain and regional
  refinement
• ICONAM/ICONOM for seasonal forecasting

10
Summary

• Shallow water models have been developed and
  evaluated for globally uniform grids
• Next steps
• Local grid refinement in shallow water model
• Hydrostatic atmospheric dynamical core
• To be discussed
• Equations for non-hydrostatic atmospheric model

11
Documentation and Publications

• http//www.icon.enes.org/
• Publications
• Bonaventura, L., L. Kornblueh, T. Heinze, and P.
  Ripodas, A semi-implicit method conserving mass
  and potential vorticity for the shallow water
  equations on the sphere, Int. J. Numer. Meth.
  Fluids, 47, 863-869, 2005.
• Bonaventura, L., and T. Ringler, Analysis of
  discrete shallow-water models on geodesic
  Delaunay grids with C-type staggering, Mon. Wea.
  Rev., 133, 2351-2373, 2005.
• Baudisch, J. , L. Bonaventura, A. Iske, E.
  Miglio, Matrix valued radial basis functions for
  local vector field reconstruction applications
  to computational fluid dynamic models, MOX Report
  75, 2006.
• Wan, H., M. A. Giorgetta, and L. Bonaventura,
  Held-Suarez test with ECHAM5, Berichte zur
  Erdsystemforschung, 20, Max Planck Institute for
  Meteorology, Hamburg, Germany, pp. 41, 2006.
  (submitted to Mon. Wea. Rev.)
• Manuscripts
• Bonaventura et al., Wave dispersion and stability
  analysis for C-grid discretizations on triangular
  meshes

12

• The End

13
C-grid discretization on icosahedral grids

• Icosahedral grids and C-type discretization

Delaunay (triangular) and Voronoi
(hexagonal/pentagonal) grid at the first
refinement level.
Primal and dual cells in C grid staggering
on the triangular icosahedral grid. (i)
mass (l) normal velocity (v)
vorticity points