EWGLAM/SRNWP Annual Meetings 2005 SRNWP Business Meeting
Post on 27-Jan-2016
25 Views
Preview:
DESCRIPTION
Transcript
EWGLAM/SRNWP Annual Meetings 2005
SRNWP Business Meeting
5 October 2005
Ljubljana (Slovenia)
Short-Range Numerical Weather Prediction Programme
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 2
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 3
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 4
The Network of ConsortiaHIRLAMDenmark*, Finland*, Iceland*, Ireland*,Netherlands*, Norway*, Spain*, Sweden*
COSMOGermany*, Greece*, Italy*, Poland (Warsaw Branch), Romania, Switzerland*
ALADINAustria*, Belgium*, France*, Portugal*,Bulgaria, Croatia, Czech Republic, Hungary*,Moldova, Poland (Cracow Branch), Romania,Slovakia, SloveniaALADIN – LACEAustria*, Croatia, Czech Republic, Hungary*,Slovakia, Slovenia
UKMOUnited Kingdom*
26 NWS are Members of the SRNWP Programme * EUMETNET Member
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 5
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 6
OPERA II2004-2006Responsible Member: FMI ; Project Manager: Asko Huuskonen
Data hub at the Met Office: rain. Also VAD and VVP (wind profiles).
From the OPERA Programme Decision, Point E:- To compile a set of European reference product and quality algorithms.
Planned from the OPERA Programme was a European composite that theMet Office would produce every day (00-24 UTC accumulated rain).
From the “List of Decisions” of the 23rd EUMETNET Council, 14th Dec. 2004under OPERA:“20’000 euro will be released to start the developments of the hub system,excluding any work on compositing”
Reason (Minutes Final Draft of C23, page 3):“rain accumulation products had a commercial value and, therefore,[...] would place the OPERA programme in competition with the private sector”
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 7
Recommendation
At the 12th Meeting of the SRNWP Programme held the 5th of October 2005 in Ljubljana, the NWS delegates passed the following recommendation:The European Short-Range NWP community asks the EUMETNET Council to recall its decision taken at its 23th Meeting (14th of December 2004 in Reading) to exclude from the OPERA Programme “any work on compositing” (Point 1 of the List of Decisions).
Radar compositing is needed by the Short-range NWP community for- validation and verification of the precipitations of the meso-scale LAMs- definition of better initial conditions by assimilation of precipitations
It would be a waste of time and resources if the same work (holding the radar data from the hub and making a composite) should be done in several NWS.
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 8
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 9
Purpose of EUCOS
Agreed by EUMETNET Council in November 2000:“EUCOS is the ground-based observing system designed to serve
the needs of General Numerical Weather Prediction (GNWP) over Europe”
Aim of the EUCOS preparatory phase (2000-2001):“To define an optimised design for a composite observing system
which would serve regional scale numerical weather prediction requirements”
Aim of present phase of EUCOS (2002-2006):Deployment of this composite observing system.
EUCOS terminates the 31st of December 2006.A second phase (2007-2011) is planed
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 10
Red: EUCOS Upper Air Design Blue: Other radiosondes
This figure illustrates the current complete Upper Air Network in Europe from which thecurrent EUCOS designated radiostations were selected
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 11
The revised observing system
EUCOS considered that the experienced development and implementation these last years of new observing systems such as
- wind profilers
- radar wind profiles (VAD and VVD)
and
- the massive growth of AMDAR
would allow to significantly reduce the number of upper-air soundings stations.
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 12
E-AMDAR Network The E-AMDAR network has developed significantly since the time of preparing the original EUCOS upper-air design, as illustrated below:
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 1331 selected radiosondes and 2 ocean platforms
The EUCOS Radiosonde Network Redesign
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 143 and 6 hourly AMDAR profile 2008 potential
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 15Selected daily E-AMDAR profile coverage
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 16The proposed revised EUCOS design
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 17
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 18
Recommendation
Considering that the Proposed Revised Design presented in the “Review of the EUCOS Upper-Air Network Design” of 18 May 2005 does not consider the observational requirements needed for the high-resolution short-range NWP models, the European Short-Range NWP community asks the EUCOS Programme Board and the EUCOS Advisory Group to significantly increased the density of observations and to make sure that this density remain sufficient at night.
More generally, the delegates of the 12th Meeting of the SRNWP Programme held the 5th of October 2005 in Ljubljana ask that in the second phase of the EUCOS Programme (2007-2011) the same attention, priority and resources be given for the observation of the meso-scale as it will be done for the observation of the synoptic scale.
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 19
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 20
Migration to Binary Code
BUFR code for dissemination of radiosonde data would permit- exact time of measurement
- transmission of the actual sonde coordinates during ascent
- higher vertical resolution data
- simpler code (no A-, B-, C-, D-group)
Vaisala has indicated that his sonde and software are already able to provide the information in BUFR, but some implementation work is needed at the NWS.
The WWW Department of the WMO is making efforts to encourage migration to binary based code. Europe could play in this effort a leading role.
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 21
Recommendation
At the 12th Meeting of the SRNWP Programme held the 5th of October 2005 in Ljubljana, the NWS delegates ask the EUMETNET Coordinating Officer to take the necessary administrative and technical steps in order to promote between the European NWS the dissemination on the GTS of the radiosonde data in two formats: in TEMP code and in BUFR code.
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 22
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 23
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 24
Model comparison
• Three times – in Council or ICWED Meetings – our Directors have expressed the wish of a model comparison over Europe
• It has to be cheap. Thus only our basic models will be compared:
- LM (version run by the DWD)
- Aladin-France (Toulouse)
- HIRLAM (reference version at FMI)
- UM (North-Atlantic / Europe version)
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 25
Facts
• Today the Met Office compares daily the precipitations of these 4 models over Great-Britain (24-hourly accumulated)
• The Met Office has been chosen for this comparison as the task will consist in an extension of the above mentioned work
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 26
Plan
• Verified will be from T+0 to T+48 at 6h intervals:- precipitations
- mean sea level pressure- 2m temperature- 10m wind
• Scores for precipitations:
- frequency bias
- equitable threat score
- log-odds ratio
- Hansen-Kuipers skill score
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 27
Results and Costs
• Results
Monthly mean statistics will be calculated and published on the Met Office external internet site under password protection
• Costs
The costs that will be submitted to the EUMETNET Council will amount to the salary of a scientist for 9 months per year
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 28
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 29
Data hub for high resolution, non GTS,24-hourly accumulated daily precipitations
• The spatial density of the SYNOP precipitations is absolutely inadequate for the validation and verification of the very high resolution models
• Creation of a “data hub for precipitation observations” as already done by EUMETNET at the Met Office for the radar, wind-profiler and GPS ZTD data (“one stop chop” strategy)
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 30
Situation today in Europe
Who is collecting today in Europe 24-hourly accumulated high-resolution precipitationobservations?
DWD (for Consortium COSMO)ECMWF (data from the Members)EU Project ENSEMBLESEUROGRID show-case (non yet accepted)
Report in the SRNWP web sitehttp://srnwp.cscs.ch/Documents/HighResolutionPrecipitations.htm
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 31
Candidates
Two very good candidates:
= DWD- does already this work for the COSMO
Members- collects 24-hourly accumulated precipitationsfrom some 4’500 stations
= ECMWF- receives the data from the Members and
associated Members- collects 24-hourly accumulated precipitationsfrom some 17’000 stations- makes gridded analyses
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 32
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 33
Dissemination of hourlySYNOPs
Action now endorsedby EUCOS
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 34
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 35
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 36
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 37
GPS Zenital Total Delays
• At its 24th Meeting (April 2005), the EUMETNET Council adapted the E-GVAP Programme (EUCOS-GPS Water Vapour).
• DMI as Responsible Member (Henrik Vedel)• The Met Office will run the data hub.
• The Met Office already collects radar data (rain and winds) and wind profiler data.
• >>> the data hub strategy (one stop chop) begins to become effective
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 38
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 39
SRNWP-PEPSa regional multi-model ensemble in Europe
Internet: www.dwd.de/PEPS
Short-Range
Numerical Weather Prediction
Programme
EWGLAM/SRNWPAnnual Meeting 2005
3-5 October 2005Ljubljana (Slovenia)
Presentation ofMichael Denhard
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 40
20 weather services / 23 forecast products
Denmark HIRLAM 16 ECMWF +60h 0, 6, 12, 18 Finland HIRLAM 22 ECMWF +54h 0, 6, 12, 18 Ireland HIRLAM 16 ECMWF +48h 0, 6, 12, 18 Netherlands HIRLAM 22 ECMWF +48h 0, 6, 12, 18 Spain HIRLAM 22 ECMWF +24h 0, 12 Norway I HIRLAM 11 ECMWF +30h 0, 12 Norway II HIRLAM 22 ECMWF +30h 0, 12Sweden I HIRLAM 11 ECMWF +48h 0, 6, 12, 18 Sweden II HIRLAM 22 ECMWF +48h 0, 6, 12, 18
Belgium ALADIN 15 ARPEGE +60h 0, 12 Austria ALADIN-AUSTRIA 9.6 ARPEGE +48h 0, 12 France ALADIN-FRANCE 11 ARPEGE +48h 0, 12 Croatia ALADIN-LACE 8.9 ARPEGE +48h 0, 12 Czech Rep. ALADIN-LACE 11 ARPEGE +48h 0, 12 Hungary ALADIN-LACE 11 ARPEGE +48h 0, 12 Slovakia ALADIN-LACE 11 ARPEGE +48h 0, 12 Slovenia ALADIN-LACE 9.5 ARPEGE +48h 0, 12
UK I UKMO-LAM 12 UM global +48h 0, 6, 12, 18 UK II UKMO-EU 12 UM global +48h 0, 6, 12, 18
Germany LM 7 GME +48h 0, 12, 18 Switzerland aLMo 7 ECMWF +72h 0, 12 Italy Euro LM 7 EuroHRM +60h 0 Poland LM 14 GME +72h 0, 12
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 41
PEPS-Grid with a grid spacing of 0.0625° (~7 km) covering Europe
The ensemble size depends on location and every PEPS grid point has its
own probability distribution
thresholdaisTandipointgridPEPSatforecastsofnumbertotaltheisNwhere
N
iatTexceedingforecastsofNumberP
i
ii
Probabilities(Nearest Neighbour)
SRNWP-PEPS
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 42
Ensemble Products
Ensemble mean. Forecast periods +06...+30h (24 hours), +06...+18h and +18...+30h (12 hours)
• Total precipitation (accumulation), sum of convective and large scale precipitation• Total snow (accumulation) ), sum of convective and large scale snow• Maximum 10 m wind speed• Maximum 10 m wind gust speed• 2 m minimum/maximum temperature
Probabilistic products. Forecast period +06...+30h (24 hours)
• Probabilities of total precipitation Thresholds: > 25, > 40, > 70 mm• Probabilities of total snow Thresholds: > 1, > 5, > 10, > 20 cm• Probabilities of maximum wind speed Thresholds: > 10, > 15, > 20, > 25 m/s• Probabilities of maximum wind gust speed Thresholds: > 10, > 15, > 20, > 25, > 33 m/s
Probabilistic products. Forecast periods +06...+18h and +18...+30h (12 hours)
• Probabilities of total precipitation Thresholds: > 20, > 50, > 100 mm • Probabilities of total snow Thresholds: > 1, > 5, > 10, > 20 cm • Probabilities of maximum wind speed Thresholds: > 10, > 15, > 20, > 25 m/s• Probabilities of maximum wind gust speed Thresholds: > 10, > 15, > 20, > 25, > 33 m/s
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 43
Maximum Ensemble Size
depends on main run and meteorological parameter
MaximumEnsemble Size
Totalprecip.
Totalsnow
Windspeed
Wind gustspeed
Temperature
00 UTC06 UTC12 UTC18 UTC
20 19 20 8 207 6 7 - 720 19 20 8 208 7 8 1 8
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 44
Ensemble Mean
21/01/2005 00 UTC +06...30
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 45
21/01/2005 00 UTC +06...30
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 46
Cut-off times
Model run cut-off time
00 UTC 05.30 UTC
06 UTC 11.30 UTC
12 UTC 17.30 UTC
18 UTC 23.30 UTC
SRNWP-PEPS runs operationally since December 2004
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 47
Mask of areas without sufficient models
Wind gusts
provided by COSMO and some ALADIN countries
statistical estimation of wind gusts within PEPS?
Statistics of availability of models
Additional productsmore synoptic oriented parametersindices of convectivity
Precipitationmedian instead of meanlower thresholds
PEPS-Meteograms (provided by Meteoswiss)
running
under way
planned
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 48
ValidationComparison with COSMO-LEPSScoring probabilistic forecasts
Ensemble Calibration Bayesian Model Averaging (BMA) Raftery et. al., 2005
BMA for precipitation
running
under way
planned
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 49
Thank you to the
contributing Weather Services !
EWGLAM/SRNWP Annual Meeting 2005, 3-5 October 2005, Ljubljana (Slovenia) 50
Agenda
• The Network of Consortia• Programme OPERA: Compositing necessary• EUCOS II (2007 – 2011): Action required• Soundings: BUFR besides TEMP• STORMNET (Dominique)• Model comparison• Hub for high resolution precipitation observations• Dissemination of hourly SYNOPs• GPS Zenital Total Delays• SRNWP PEPS• Other points
top related