Operational AladinBelgium

1
Operational Aladin-Belgium

• SGI Altix 3700 Bx2, 56 Itanium2 1.5 Ghz 6 Mbyte
  CPUs
• Model version AL29t2
• Domain size 240 x240 points
• Horizontal resolution 7km
• Number of layers 46
• Time step 300 s 2-time level SISL advection
  scheme
• Forecast range 60h and 4 runs/day (0,6,12,18 UTC)
• Lateral boundary conditions Aladin-France and
  Arpege
• LBC update every 3
  hours
• Initialization digital
  filter initialization
• Hourly post processing

2
Progress with the 3MT scheme in Alaro-0

• Alaro-0 was developed for operational forecasts
  at high resolution, with meshes between 10 and 2
  km.
• 3MT implements in Alaro-0 the package described
  by Gerard (2007) and ideas of Piriou
• Tests have been performed at 9 and 4km resolution
• The forecast at 4km, in the socalled grey zone
  of the convection, is consistent with the other
  scales.

3
Monitoring the Coupling-Update Frequency

• When a fast moving storm system crosses the
  border of a limited area model, the coupling
  frequency may be too low to capture this event
  adequately. An index has been developed to
  monitor the data loss due to the coupling.

New approach to lateral boundary conditions

• P. Termonia and F. Voitus propose an approach
  where the LBCs are computed with a numerical
  finite-difference scheme that is different than
  the semi-implicit semi-lagrangian scheme of the
  dynamical core

4
Including atmospheric layers in vegetation and
urban offline surface schemes

• A formulation to include prognostic atmospheric
  layers in offline surface schemes is derived from
  atmospheric equations.
• Classic multi-layer surface schemes need complex
  coupling between atmospheric models levels and
  surface scheme levels
• The proposed coupling remains simple because the
  atmospheric layers interacting with the surface
  scheme are independent of the atmospheric model
  that could be coupled above (Masson, 2007).
• The Surface Boundary Layer (SBL) (both inside and
  just above the canopy) is resolved by a
  prognostic way, taking into account large-scale
  forcing, turbulence and, if any, drag and canopy
  forces and surfaces fluxes.

5
Downscaling of the ERA-40 reanalysis

• As a first approach, we opted for a double
  nesting strategy, with domain resolutions at 40
  km and 10 km.
• A third nesting at 2 km was also introduced, but
  this consisted only of a dynamical adaptation run
  (30' and limited physics) for a more detailed
  wind field (Zagar Rakovec, 1999). This work is
  part of the ECMWF special project SPFRCOUP.