COSMO Status Report 2003

1
COSMO Status Report 2003 2004Tiziana
Paccagnella, ARPA-SMR

• Organizational Structure of COSMO
• Lokal-Modell (LM) Overview
• Operational Applications of the LM
• Activities of the COSMO WGs
• Future developments

2
Organizational Structure of COSMO (2)
3
Organizational Structure of COSMO (3)
4
Internal Exchanges

• WG1 Meeting about satellite assimilation by
  1-Dvar
• July 2004, DWD
• WG2 Workshop 21 September 2004 Milano
• WG3 Workshop 21 September 2004 Milano
• WG4 Workshop May 2004 Geneve
• 21-22 September 2004 Milano
• WG5 Workshop May 2004 Geneve
• WG6 Meeting on Code Optimization and
  Portability
• MeteoSwiss, CSCS and DWD
• And many internal visits to work on specific
  topics.

Two meetings of Working Groups Coordinators Two
meeting of the Steering Committee
5
COSMO Status Report 2003 2004Tiziana
Paccagnella, ARPA-SMR

• Organizational Structure of COSMO
• Lokal-Modell (LM) Overview
• Operational Applications of the LM
• Activities of the COSMO WGs
• Future developments

6
Lokal-Modell (LM) Dynamics and Numerics
7
Lokal-Modell (LM) Dynamics and Numerics
8
Lokal-Modell (LM) Physics
9
Lokal-Modell (LM) Physics
10
Lokal-Modell (LM)
11
Lokal-Modell (LM) data assimilation
12
COSMO Status Report 2003 2004Tiziana
Paccagnella, ARPA-SMR

• Organizational Structure of COSMO
• Lokal-Modell (LM) Overview
• Operational Applications of the LM
• Activities of the COSMO WGs
• Future developments

13
Operational Application of LM
14
Lokal Modell at DWD
15
Lokal Modell aLMo atMeteoSwiss
16
LAMIItaly
17
Lokal Modell at HNMS
18
Lokal Modell at IMGW
19
COSMO-LEPSat ECMWF
20
COSMO Status Report 2003 2004Tiziana
Paccagnella, ARPA-SMR

• Organizational Structure of COSMO
• Lokal-Modell (LM) Overview
• Operational Applications of the LM
• Major changes during past year
• Activities of the COSMO WGs
• Future developments

21
Major changes to the COSMO systems during last
yearGME

• GME Operational GCM at DWD
• Icosahedral-hexagonal grid
• Operators of second order accuracy
• 60 km mesh size, 164000 (369000) gridpoints/layer
• 31 layers (hybrid, sigma/pressure)
• Prognostic variables ps, u, v, T, qv, qc, qi, o3
  
• Programming Fortran90, MPI for message passing
• Intermittent data assimilation (OI, 3-hourly)
• From 00 and 12 UTC Forecasts up to 174 hours
• From 18 UTC Forecasts up to 48 hours

22
Major changes to the COSMO systems during last
yearGME
23
Major changes to the COSMO systems during last
yearGME

• 27 September 2004 40-km, 40 layers
• Grid spacing reduced from 60 km to 40 km
• Number of layers increased from 31 to 40
• Lowest model layer now 10 m above ground
  (before 33 m)

• New multi-layer soil model with solution of heat
  conduction equation
• inclusion of the effects of freezing/melting of
  soil water
• improved snow model

24
Major changes to the COSMO systems during last
yearLM

• LM Version 3.6 11.12.03
• Changes to the new multí-layer soil model
• New option to assimilate single SATOB winds
• LM Version 3.7 18.02.04
• Prognostic precipitation for Leapfrog scheme
• Additional options for Runge-Kutta scheme
• Synthetic satellite images for Meteosat, MSG
• Adaptations in the TKE scheme

25
Major changes to the COSMO systems during last
yearLM

• LM Version 3.8 23.03.04
• Technical changes to Runge-Kutta scheme (also
  prognostic precipitation with SL-advection)
• LM Versions 3.9 22.04.04
• Bug corrections
• LM Versions 3.10 23.06.04
• Bug corrections
• LM Version 3.11 28.07.04
• First ifdef in lmorg (because of floating point
  trapping on the IBM)

26
Major changes to the COSMO systems during last
yearLM

• LM Version 3.12 15.09.04
• Preparations for assimilation of 1dVar satellite
  retrievals
• New option for revised quality control for
  humidity
• Initialization of fields for prognostic
  precipitation

27
COSMO Status Report 2003 2004Tiziana
Paccagnella, ARPA-SMR

• Organizational Structure of COSMO
• Lokal-Modell (LM) Overview
• Operational Applications of the LM
• Activities of the COSMO WGs
• Future developments

28
Working group 1 Data Assimilation Coordinator
Christoph Schraff from DWD

• Further developments have been carried on about
  assimilation of radar data.
• A huge amount of experimentation has been carried
  on as regards Latent Heat Nudging (LHN) but
  results are not very promising.
• The introduction of the prognostic precipitation
  scheme in LM has brought another serious problem
  of consistency with the basic assumption of the
  present LHN scheme.
• More work will be devoted to this task next year
  but it is under planning the organization of a
  workshop, open to scientist outside COSMO, to
  better address future programs in radar data
  assimilation.
• As regards GPS data, some more work has been
  doing about the application of TOMOGRAPHY
  technique to estimate the 4-dimensional
  distribution of atmospheric water vapor
  (cooperation with Marc Troller from ETH).

29
Working group 1 Data Assimilation Coordinator
Christoph Schraff from DWD

• Wind-profiler data are now assimilate on an
  operational basis positive impact especially on
  upper air fields. Work on assimilation of SATOB
  winds is ongoing.
• The nudging assimilation code has been adapted to
  allow preliminary tests, on real case, of the
  LM-Z version of the model.

30
Working group 2 Numerical Aspects Coordinator
Juergen Steppeler from DWD

• The development of the Z-coordinate version of LM
  has been continued during the year
• The eulerian version, by Steppeler and Bitzer, is
  almost ready for testing on real cases by using
  the available interface to the terrain-following
  physics of LM.
• A new Z-coordinates version of LM with
  semi-implicit two time-level semi-implicit
  discretizations and semi-lagrangian advection has
  been developed, by Cesari and Bonaventura, and
  tested on ideal 2-D cases.
• Formulation of a new dynamical lower Boundary
  Condition
• Investigation and testing of RK schemes and
  high-order spatial discretization

31
COSMO Status Report 2003 2004Tiziana
Paccagnella, ARPA-SMR

• Organizational Structure of COSMO
• Lokal-Modell (LM) Overview
• Operational Applications of the LM
• Activities of the COSMO WGs
• Future developments

32
LMK A Very Short Range Forecasting
SystemSlides from a presentation by

• Günther Doms, Michael Baldauf

33
LMK - (LM-Kürzestfrist)

• Goals
• Development of a model-based NWP system for very
  short range forecasts (2-18 h) of severe weather
  events on the meso-? scale, especially those
  related to
• deep moist convection (super- and multi-cell
  thunderstorms, squall-lines, MCCs, rainbands,...)
  
• interactions with fine-scale topography(severe
  downslope winds, Föhn-storms, flash floodings,
  fog, ...)

34
LMK Assimilation and forecast cycle
(LAF-Ensemble)
18h
15h
12h
9h
6h
3h
00
03
(UTC)
00
21
18
15
12
09
06
LMK new forecast every 3 hours
35
LMK will require not only adjustments of existing
schemes but also the development of new components

• Further developments in the method to assimilate
  radar data(Latent heat nudging, LHN)
• 5-min radar composits (DX-data) including quality
  control
• Development of an efficient time-integration
  scheme allowing for high-order spatial
  discretization- modified metrics, new
  symmetric thermodynamics, BCs, cov. metrics
• Full 3-d budget equations for precipitating
  hydrometeors
• Development of new physics schemes for
  kilometer-scale simulations (graupel and hail
  microphysics, 3-d turbulence, shallow convection)
• Development of new tools for model validation and
  verification using remote-sensing
  data(pseudo-satellite and pseudo-radar products,
  Lidar, cloud-radars)
• Development of products and methods to interpret
  model results from deterministic forecasts (on
  stochastic scales) and to estimate the
  predictability (probabilistic products)

SKIPPO
36
Current Activities

• Radar Data
• Generation of a DX composite for the DWD network
• Quality control
• Clutter corrections
• Latent Heat Nudging (LHN)
• Diagnostic studies -gt Reasons for the short
  memory time of LHN?
• Use of DX data
• Code transfer to LM
• LHN lt--gt prognostic precipitation ? problem
  different basic assumptions

37
Current Activities (2)

• Dynamics, Numerics, Boundary Data - New
  symmetric thermodynamics within the RK3 solver
  - Lateral boundary conditions pressure drop
  (solved) - LBCs for prognostic precipitation
  - Update of LM2LM ( SLEVE coordinate, ex. Param
  for data assimilation) - Idealized cases
• Physical Parameterization - Implementation of
  3-d turbulence for the RK3 solver - Test and
  evaluation of the new Graupel-Scheme -
  Development of a shallow convection scheme -
  Sensitivity tests for radiation update frequency
• Verification and Validation - Synop
  Verification of LMK testsuites -
  Pseudo-satellite and pseudo-radar products

38
Project LMK Milestones

• Autumn 2004Prototype version of the LMK-System
  with data assimilation but without LHN running.
• Summer 2005Prototype version of the LMK-System
  with LHN is running in a quasi-operational
  mode.Further testing and evaluation of new
  numerical schemes and physical parameterizations.
  
• Early 2006Start of a pre-operational
  test-phase.Fine-Tuning and final evaluation of
  all components of the system.
• End 2006Start of the operational application .

39
Working group 3 Physical Aspects Coordinator
Marco Arpagaus from MeteoSwiss

• Planetary Boundary Layer formulation in LM still
  under revision.
• Some tuning of the scheme has been performed and
  some work has been also done to allow higher
  vertical resolution of the lowest part of the
  boundary layer
• 1-D single column model developed
• Implementation of 3D turbulence formulation the
  prognostic equations have been extended to
  include three-dimensional flux-terms. The scheme
  is under testing to improve LM use at very high
  horizontal resolution (as the planned 2.8 km
  version of LM). .

40
Working group 3 Physical Aspects Coordinator
Marco Arpagaus from MeteoSwiss

• The Cloud Ice scheme (Doms, 2002), has been
  implemented in the operational applications of
  LM.
• The prognostic precipitation scheme has been
  completed (Gassman, 2002) operational
  implementation is ongoing.

41
Working group 3 Physical Aspects Coordinator
Marco Arpagaus from MeteoSwiss

• Development of a Shallow convection scheme on the
  meso-? scale started and ongoing
• A first theoretical formulation of a new
  convection scheme has been done by Guenther Doms
  to better represent convective processes in the
  meso-gamma version of LM.

42
Working group 3 Physical Aspects Coordinator
Marco Arpagaus from MeteoSwiss

• The new multi-layer soil model (Schroding and
  Heise, 2002) is under operational implementation
  at DWD.
• A lake model based on a two-layer self-similar
  representation of the temperature profile in the
  mixed layer and in the thermocline has been
  developed by Dmitrii Mironov and is now being
  tested.

43
Working group 4 Interpretation and Applications
Coordinator Pierre Eckert from MeteoSwiss

• The COSMO-LEPS (see COSMO status report SRNWP
  meeting 2003) suite is running regularly and
  products are disseminated to COSMO partners.
• Test of experimental suites to assess possible
  modifications of the operational suite are
  ongoing. Another Special Project by ARPA-SIM and
  MeteoSwiss has been approved by ECMWF, which aims
  at further development of the system.
• Quite a lot of progresses have been done in the
  development of COSMO-LEPS verification. The 24-h
  precipitation forecast by COSMO-LEPS has been
  verified against observations over 3 COSMO
  countries (Italy, Germany and Switzerland) for
  the period November 2002 - January 2003.
  Different methodologies.
• Present results, also coming from statistical
  analyses of EPS, give two major indications
• Clustering on 10 members shows better results
  than using 5 clusters
• A super-ensemble of 1002 members (the two most
  recent EPS) is better than using 1503 members
  (the present configuration)
• The COSMO-LEPS suite was modified in accordance
  with these results since July 2004.

44
Working group 4 Interpretation and Applications
Coordinator Pierre Eckert from MeteoSwiss

• As regards SLEPS project, simulations for winter
  storms Lothar and Martin have been performed
  using ECMWF EPS with moist and operational
  singular vectors, respectively, and using two
  different horizontal resolutions (EPS vs. LEPS
  resolution). Implementation of Brasseur wind gust
  formulation in SLEPS LM version.
• Guidelines for the use of bench forecasters and
  Collection of forecasters feed back are under
  preparation. Workshop has been organised in
  Geneva 5-6 May 2004.

45
Working group 5 Verification and Case Studies
Coordinator Frances Schubiger from MeteoSwiss

• Operational verification of national applications
  is ongoing regularly in each COSMO country.
  Verification will be extended also to the
  assimilation cycle.
• Many activities addressed to high resolution
  verification using regional radar data,satellite
  and non-GTS data are carried on in Germany,
  Switzerland and Italy.

46
Working group 5 Verification and Case Studies
coordinated by Francis Schubiger from MeteoSwiss

• The procedure to collect not GTS data from
  Germany, Switzerland, Italy and Poland (more than
  4500 stations for precipitation) is now
  operational. LM national applications running on
  overlapping areas can be compared.
• Operational model output from the different COSMO
  countries are exchanged in real time at the COSMO
  web-site grib files are plotted on the member
  area with the same graphycs package to make
  easier forecasts intercomparison.
• The development of a COSMO common verification
  package at ECMWF is well ongoing.

See!!!
47
COSMO non-GTS precipitation data for verification
48
Working group 6 Reference Version and
Implementation Coordinator Ulrich Schaettler
from DWD

• Installation, updating and running of the LM is a
  permanent task of the COSMO member sites.
• The last released version of LM is the 3.12 (7
  versions released during the year). The present
  Reference Version of LM is still the 3.1 the
  next reference version will be released very
  soon.
• The SW updating procedure is still under
  completion but big progresses have been done
  thanks also to the ongoing implementation of the
  Common Verification Package at ECMWF.
• The unified version of INT2LM (boundary and
  initial conditions from LM, GME and IFS) is now
  available and tested.

49
Working group 6 Reference Version and
Implementation Coordinator Ulrich Schaettler
from DWD

• The sub-working group on code optimization and
  portability is working. Software engineering
  guidelines will be released.
• Due to problem in Greece connected to the olimpic
  games, the COSMO web site was moved to Manno
  (usually the mirror site). The principal web site
  in Greece will be restored in few months from
  now.
• Documentation is still under completion.
• SW tools are available for internal COSMO
  exchanges through the web.
• Towards ISO 9000. DWD and MeteoSwiss are about to
  be ISO certified (October / November)

50
LME LM Europe

• Model Configuration
• Grid spacing 0.0625 ( 7 km)
• 665 x 657 grid points per layer
• 40 vertical layers
• Timestep 40 sec
• Daily runs at 00, 12, 18 UTC, 78h
• Boundary Conditions
• Interpolated GME forecasts with
• ds 40 km and 40 layers (hourly)
• Hydrostatic pressure at lateral
• boundaries
• Data Assimilation
• Nudging analysis scheme
• Variational soil moisture analysis
• SST analysis at 00 UTC
• Snow depth analysis every 6 hrs

51
Time schedule for LME

• March 2004
• Implementation of LME at ECMWF
• May 2004
• Agreement on model domain between DWD
  departments
• October 2004
• Pre-operational in parallel suite with data
  assimilation at DWD including current model
  changes
• 1. Prognostic precipitation
• 2. New soil model
• First quarter in 2005
• Operational at DWD

52
COSMO Status Report 2003 2004Tiziana
Paccagnella, ARPA-SMR

• Organizational Structure of COSMO
• Lokal-Modell (LM) Overview
• Operational Applications of the LM
• Activities of the COSMO WGs
• Future developments

53
The COSMO perspectives driving scientific
developments for the next years are

• Improve of LM system at the current resolution
• Doing high resolution forecast at the meso-gamma
  scale
• Enhance Use of Remote-sensing data in data
  assimilation
• Development of Ensemble Systems Based on LM
• Improve internal exchange and communications
• Improve exchange with other consortia

54
Future Activities and Developments

• The present COSMO-LM system will be improved to
  produce better quantitative precipitation
  forecast at scales at which deep convection must
  still be parametrized.
• Even going to higher horizontal resolution with
  Cloud Resolving Models (CRM), coarser models will
  be an important component of operational
  meteorological suites for a long time.

55
Future Activities and Developments

• Improving assimilation system for the meso-gamma
  scale implies more extensive and better use of
  remote sensing data
• variational technique to assimilate satellite
  radiances will be developed and implemented.
• different methodology to fully exploit the use of
  radar data will be evaluated LHN will be anyway
  further tested and developed.
• Much effort must be spent to produce a good
  analysis of the 3-D moisture field
• to this purpose GPS data will be used also with
  tomography technique.
• The nudging assimilation scheme will be tuned and
  further optimized also giving more emphasis to
  the use of surface level data.
• The further development of the COSMO nudging
  scheme will incude also modifications required by
  the use of the LM-Z version of the model.

56
Future Activities and Developments

• LM_Z will be tested on real cases also comparing
  the two dynamical core options developed
  eulerian and Semi-implicit semi-lagrangian.
• Further developments of higher order RK schemes
  to reach a suitable level of accuracy.
  Computational efficiency and numerical accuracy
  must be anyway pursued together in the coming up
  generation of meso-gamma CRM.

57
Future Activities and Developments

• Most of the ongoing projects related to physical
  aspects are already oriented toward meso-gamma
  applications of LM. Particular emphasis must be
  given to
• Implementation of 3D turbulence
• Development of a new radiation scheme
• Full implementation of the new soil model
• Advanced microphysical scheme
• Improvement of PBL
• Shallow convection

58
Future Activities and DevelopmentsWG4
Applications

• LAM Ensemble System(s)
• COSMO-LEPS performance must be completely
  evaluated also comparing the system with other
  lower cost system.
• Alternative Ensemble prediction methodologies
  should also be explored like multi-model
  ensemble.
• At longer time scale, a Short Range Ensemble with
  own perturbations could be developed.
• Post-processing methodologies and LM products
  applications will be fully reported to allow
  exchanges among COSMO partners.

59
Future Activities and Developments

• Verification methodologies will be improved
  and/or developed as regards
• Verification at the meso-gamma scale
• Boundary layer processes
• Conditional verification
• The list of COSMO case studies for verification
  will include cases selected by model developers
  to test the impact of new model developments.

60
Future Activities and Developments

• The improvement of scientific and information
  exchanges, both within COSMO and between COSMo
  and other consortia or scientists, is at high
  level of priority in the short period.
• the COSMO web-site will be progressively enriched
  and the establishment of a solid web-management
  will assure a constant and timely updating of the
  site content.
• Cooperations with other European consortia will
  be stimulated also opening the working groups
  workshops/meetings.
• Visiting scientist programmes will be defined.

61

• Thanks for your attention !!!

62
(No Transcript)