The Beijing 2008 FDPRDP project

Document Sample
The Beijing 2008 FDPRDP project Powered By Docstoc
					The Beijing 2008 FDP/RDP project

                            Yong WANG
                              ZAMG, Austria

Presentation thanks to: Y-H. Duan, J-D. Gong, Y-L. Li and L. Li (NMC/CMA),
H. Seko and K. Saito (JMA) and B. Kuo (NCAR), ISSC of B08FDP/RDP
 What is B08FDP/RDP?
A five years WWRP research project (2005-2009)
Forecast and Research Demonstration:
August 2008, the Beijing Olympic Games.
Two sub-projects:
- B08FDP: Forecast Demonstration for 0-6 hour
  forecast based on Nowcasting.
- B08RDP: Research and Development for 6-36 hour
  forecast based on mesoscale ensemble prediction
Beijing 2008 Olympics Meseoscale Ensemble Prediction
Research and Development Project
       Beijing Olympic Games 2008, 8-24 August

Climate at Beijing in August
Observation   Observation
Network       Network
(2004)        (2008)
B08FDP/RDP Management Organization Structure
International Science Steering
       Committee (ISSC)
  Geoff DiMego (NCEP)
  Bill Kuo (NCAR)
  Martin Charron (MSC)
  Kazuo Saito (JMA)
  Yihong Duan (CMA)
  Lawrence Wilson (MSC)
  Yong Wang (ZAMG/Meteo-France)

  Jiandong Gong (CMA), secretary of ISSC
              Time Schedule of B08RDP
Year\Mont   January - May     June       July     August   Septembe        October       Nov.-Dec.
    h                                                          r
                                       WWRP                                 WWRP          Beginnin
  2004                                  team                               approve            g
                                       visit to                           B08FDP/RD      implemen
                                       Beijing                            P proposal       -tation
                                            Technique & System
            1st   B08RDP workshop              development                Pre-test on data transfer
                                             Relevant research
                                         Preliminary running & data          Products archive;
  2006       Initial setup of meso-                                             verification;
                                          transfer test; 2nd B08RDP         Continued work on
            scale ensemble system
                                           workshop;basic training          setup of Meso-EPS

            Modification of              Quasi real-time running &         Products      Report to
  2007                        -ring
             the systems               data transfer test; 3rd B08RDP      archive;       WWRP
                                       workshop; forecaster training      Verification

  2008                        -ring      Quasi operational running            Full verification

  2009                                4th B08RDP international workshop
Products in GRIB 2:
Horizontal   Lat / lon
Vertical     pressure
Ensemble     > 10 Mems
Output       3 hours
                CMA Data base                                         Decompress obs.        NCEP MEP
                                                      NCEP ftp
                                                      server         Format & Compress       MEP data
   Start             Obs

            Format & Compress                                         Decompress obs.        JMA MEP
                                                      JMA ftp
                                                      server         Format & Compress       MEP data
                CMA ftp
                                                                      Decompress obs.        MSC MEP
                                                      MSC ftp
     NMC MEP               CAMS MEP                   server         Format & Compress       MEP data

     MEP data              MEP data
                                                                      Decompress obs.        ZAMG MEP
 Format & Compress         Format & Compress          FR/ZAMG
                                                      ftp server     Format & Compress       MEP data

             CMA Data base

                              MEP Products                      BMB Data base      Display
   MEP decompress                                                                              forecaster
                              Multi-center super EP
                                                                        Training forecast
                              MEP verification

The data flow chart of B08RDP between CMA and
participant countries are setup completely
Experiment: 24 July 2007 to 31 August 2007
       The preliminary result of 2007 Tier 1 Experiment
                  Bias scores                             Bias scores
 3.5                                               CMA
   3                                               CAMS
 1.5                                               ZAMG
        0.1   1     3    5    10   20   50   100

                        ETS                                     TS
 0.12                                              CMA
  0.1                                              CAMS
 0.06                                              NCEP
 0.04                                              ZAMG
        0.1    1    3    5    10   20   50   100

              6h precipitation                            2m temperature

                         Ensemble mean, 26 July --- 5 Aug. 07
                                                                           After K. Saito (2007)
Thunderstorm on 31 July 2007
After K. Saito (2007)
                    Participants System:
                    • JMA/MRI: JMA-NH 3 km
Tier-2 experiment   • CMA/MMC: WRF 3km
                    • NCAR-BMB: MM5 3 km, WRF 3 km
Thunderstorm case   • NCEP: WRF 3km

                                       After H. Seko (2007)
Tier-2 experiment
Bias corection and combination of the MEP systems

   Bias Correction                             Combination

  • Goal                       EPS systems • Goal
   Improve reliability while                 Super EPS system
   maintaining resolution       - CMAS/CMA
                                -   NCEP
  • Method                      -   JMA       • Method
    NCEP, moment-based,         -   MSC        equal weights
    Kalman Filter type                         weights by CC or error
                                -   ZAMG/MF
                                -   NMC/CMA
  • Parameters                                • Parameters
   H500, T850, T2m, RH2m, RR                   H500, T850, T2m, RH2m, RR
                                      Bias correction and combination: T2m
                                            RMS error (2007.8.14-2007.8.24)
                      left – before BC                                            right –after BC

            5                                                             5
                                                                        4.5                                 NMC/CMA
            4                                             NMC/CMA         4                                 CAMS/CMA
RMS error

                                                                        3.5                                 NCEP
            3                                             NCEP            3                                 MRI/JMA
            2                                             MRI/JMA       2.5
                                                                          2                                 MSC
            1                                                           1.5                                 ZAMG
                                                                          1                                 combination
            0                                             combination   0.5                                 spread
                  6      12     18     24      30   36    spread          0
                              forecast hours                                  6    12   18   24   30   36

                •RMS errors of all EPS after calibration have been decreased obviously.
                •Diurnal changes have been lowered after calibration.
                •Combined-Ensemble RMS error after calibration is the smallest of all, before
                calibration, it is the smallest in most cases.
                •After calibration spread is more closer to RMS errors than before.
     Training activities: EPS training workshop for
                          forecaster, 1 week, 2007

                        Dr. H. Gmoser, ZAMG

Dr. R. Grumm, NCEP
B08RDP and the other international projects
“ It was recommended that B08RDP work be considered as a TIGGE LAM experiment,
 and coordination should be established between the TIGGE LAM subgroup of the
 TIGGE/GIFS WG and B08RDP activities.” by GIFS-TIGGE working group

MAP D-Phase & B08FDP/RDP
 It was recommended by WWRP that B08FDP/RDP
and MAP D-Phase should cooperate and exchange
information, in particular, on short range EPS,
nowcasting, verification, data format etc.
Wang for MAP D-Phase in 2nd B08RDP workshop 06.
Dr. Gong for B08RDP in 1st MAP D-Phase workshop 06.
Dr. Dorninger for MAP D-Phase in 3rd B08RDP workshop 07.
Wang for B08RDP in 2nd MAP D-Phase workshop 08.

B08FDP/RDP list in GEO work plan weather Task WE-07-02.
Implementation plan 2008

 • Real time demonstration period: 24 July --- 24 Aug. 2008
    • starting in June and July as training period for calibration
 • Bias correction and multi-system combination in real time
 • Objective and subjective verification by forecaster
 • Preparation of joint papers on research results

 ZAMG contribution for 2008
 • Use of multi-physics, NCBB and clustering
 • Introduction of blending (SV+Breeding)
 • AROME 2.5km case studies
Thanks to all the partners of the B08FDP/RDP,
in particular, the members of the ISSC and the
technical support team.