Status of HIRLAM activities

1
Status of HIRLAM activities

• HAC, 20061026

2
Mesoscale forecast model

• Meso physics development of EDMF scheme ahead of
  schedule 3D-turbulence scheme awaiting stable BL
  developments.
• Meso dynamics Mercator map factor description
  ready for coding delay in VFE, transparent LBC.
• Syn model parametrizations / surface developments
  relevant for mesoscale largely on schedule.
  Surface/surfex workshop in Toulouse, December
  2006. Lack of qualified manpower/expertise in
  soil surface data assimilation/modelling gt shift
  manpower from upper air physics to this area,
  training of staff needed.
• Verification routine monitoring tools set up
  visualization of experimental models on Hexnet
  validation cases selected, being worked out.
• Monitoring experience AROME realistic in
  precipitation, but systematic errors in
  near-surface behaviour (no SURFEX yet). Root
  causes to be studied. Computationally expensive
  gt need to study how model can be brought within
  operational time constraints

3
Mesoscale data assimilation

• 2006 preparations for mesoscale, development of
  synoptic scale 4D-VAR. 2007 creation and testing
  of (3D-VAR) meso DA system.
• Preparatory activities training in ALADIN
  3D-VAR/ODB obs operator description, to achieve
  convergence implementation, testing of ALADIN
  3D-VAR. Detection of things to be improved in
  ALADIN 3D-VAR.
• Use of obs radar, satellite radiances, GPS
  developments on target. Radar simulation model to
  be brought under Reference System, and to be used
  in mesoscale validation.
• Status obs operator convergence somewhat
  delayed, rest on schedule.
• Planning meeting on mesoscale DA in Zürich,
  October 13 detailed common plans for 2007,
  division of tasks.

4
Synoptic scale data assimilation

• Emphasis in 2006 mainly on finishing development
  and operational implementation of 4D-VAR.
• Results of 4D-VAR testing systematic and big
  improvement over 3D-VAR, scores similar to
  3D-VARECMWF blending. Impact highest in
  situations with large forecast errors.
• Because of 4D-VAR work less manpower available
  in 2006 for use of observations, but progress in
  most areas as planned.
• Surface DA algorithms little progress, due to
  lack of qualified manpower.
• For 2007 emphasis on coherent impact studies of
  high-resolution remote sensing data, and on setup
  of mesoscale 3D-VAR.

5
4D-VAR - improvements consistent for all
periods studied- Introduction of DFI control
variable spinup reduced to a point where
explicit initialization seems no longer
necessary- however reproducability problems on
ECMWF/ IBM
o3d 3D-Var o4d 4D-Var
Wind speed
Temperature
6
Synoptic scale forecast model

• Moist turbulence scheme ready for implementation,
  MSO/SSO and sloping surface radiation adaptations
  nearly so.
• Stable boundary layer adaptations of Sukoriansky
  et al. tested, no great improvements seen wrt
  version 7.0.
• Alternative treatment of detrainment/entrainment
  gives promising results, implemented for testing
  in STRACO, KF, ECMWF. Preparations of comparison
  STRACO KF nearly completed
• Surface Snow-forest adaptations work in right
  direction, but some serious problems remaining.
  Good progress with lake modelling. No manpower
  yet for inclusion of urban effects.
• Little effort yet in coupling with air quality
  modelling. Needs more attention in 2007.

7
Predictability

• Request made for SP resources for laboratory
  environment at ECMWF. Outcome not yet known.
• Workshop June 2006 global division of tasks made
  between HIRLAM / ALADIN partners inventory of
  tools/scripts to build GLAMEPS. Start was made
  with building GLAMEPS script environment.
• COSMO GM, EWGLAM/SRNWP, Thorpex Eur. Planning
  meeting GLAMEPS concept presented, invitation
  to other consortia to join. SRNWP common
  European LAM EPS adopted as priority project.
• Vienna meeting Nov. 2006 finalize plans for
  GLAMEPS system setup.
• Development of new methods concentrate in 2006
  on SVs, in 2007 on lower boundary conditions,
  physics perturbations
• Development of assessment/ postprocessing tools
  on schedule.
• Discussion with ECMWF GLAMEPS as PAF?

8
Impact of Bayesian Model Averaging
calibration/weighting on PEPS ensemble German
stations, Dec.2005-Jan.2006
9
(No Transcript)
10
(No Transcript)
11
System and applications

• Ref.system version 7.1 in preparation, release
  expected Jan./Feb. 2007
• System cleanup of code and scripts good
  progress, first results to be included in 7.1.
  Simplified script system for experimental use
  under construction.
• Portability benchmarking facility used to screen
  forecast code. Screening of analysis code just
  started.
• Hexnet data portal gradually expanding.
• System management training/workshop in Madrid,
  October
• Documentation point of attention for system
  group sci.secr. in 2007.

12
Monitoring and verification

• RCR routine monitoring gradually extended more
  data usage/ obs statistics, BL monitoring
• Setup to be generalized to include national
  operational models as well
• Methodology dispute
• Present RCR verification system deemed
  inadequate.
• Integrate elements of existing national
  verification systems
• Working week to define this.
• Standard EWGLAM verification of national
  operational models underway, to be put on Hexnet,
  automated.

13
Reduction of a long-standing problem
14

• Operational cooperation
• Hexnet data portal possibility created to
  include national operational/experimental models.
  Gradually increasing number of models
• Development of initial common postprocessing
  package, to be included in version 7.1. To be
  extended later (e.g. with radar simulation model)
• Harmonization/verification issues
  promoted/accepted as projects in SRNWP
• Harmonization of obs preprocessing under
  consideration in 2007
• Training of (new) local system managers (Madrid,
  October)
• Contacts with users
• Experiences of users explicit topic of discussion
  during member state visits to INM, met.no, FMI
• On Council request regular (yearly) inventory of
  national user contacts
• Main problems stable BL/fog under/overestimate
  of high/low winds description of snow and snow
  melt conditions
• Request/interest for high resolution models, more
  model postprocessing users ambivalent about
  probabilistic forecasting
• Suggestion HIRLAM-wide users forum?

15
International cooperation

• COSMO information exchange at COSMO General
  Meeting. Common interests e.g. in mesoscale
  turbulence, EPS, verification, education and
  training.
• ECMWF cooperation in physics developments
  informal contacts on IFS/phasing discussions on
  GLAMEPS proposal for PAF?
• SNRWP initiative for formulation of more
  ambitious new SRNWP programme. All major items
  suggested approved at EWGLAM/SRNWP meeting.
• Climate modelling community cooperation in
  physics, surface developments. No specific
  initiatives.
• EURRA no activities undertaken.