Preliminary%20Results%20from%20CliPAS/APCC%20Multi-Model%20Ensemble%20Hindcast%20Experiments

1
Preliminary Results from CliPAS/APCC Multi-Model
Ensemble Hindcast Experiments
Bin Wang and June-Yi Lee IPRC/ICCS, University of
Hawaii, USA In-Sik Kang Seoul National
University, Seoul, Korea Chung-Kyu Park APCC,
Busan, Korea
Acknowledge contributions from all CliPAS
investigators
2
About APCC
APEC (Asia-Pacific Economic Cooperation)

APCN (APEC Climate Network)
APCC
(APEC Climate Center)
3
APEC Participating Member Economies
4
Background From APCN to APCC

• APCN (1999-2004)

APCN, The APEC Climate Network, is a regional
climate program aimed at realizing the APEC
vision of regional prosperity through mitigation
of economic losses induced by abnormal
climate. APCN produces real-time operational
climate prediction information based on a
well-validated multi-model ensemble system
(MMES).

• APCC (2005-

In order to enhance the activities of APCN,
Korea proposed and the APEC Science and
Technology Ministry endorsed establishment of
APEC Climate Center (APCC) in Korea with a core
staff of scientists and computing facilities.
The APCC Opening Ceremony will be held on
18-20th November 2005 during the APEC Summit
Meeting in Bussan, Korea,.
5
APCC is an international institute and serves as
a hub for APEC regional climate research and
prediction
To make an effort toward accomplishing the
WCRP/COPES vision
To provide core facilities and man powers to
accomplish the vision.
6
CliPAS
Climate Prediction and Its Application to
Society A Joint US-Korea Research Project in
Support of APCC
Objectives Investigate a set of core scientific
problems on multi-model ensemble (MME) climate
prediction Establish well-validated MME
prediction systems for intraseasonal and seasonal
prediction Develop economic and societal
application models.
7
Participating Institutions in CliPAS
APCC
IPRC- ICCS / UH
KMA
CES/SNU
NCEP
COLA
GFDL
NASA
FRCGC
FSU
8
CliPAS Investigators (Oct. 2005)
PI Bin Wang (UH/IPRC/ICCS)
Co-PIs J. Shukla (GMU/COLA), I.-S. Kang (CES/SNU), L. Magaard (ICCS/UH)
Co-Is J.-Y. Lee (UH/ICCS) B. Kirtman, J. Kinter (GMU/COLA) T. Krishnamurti, Steven Cocke (FSU), N.C. Lau, T. Rosati, W. Stern (NOAA/GFDL), M. Suarez, S. Schubert, W. Lau (NASA/GSFC), A. Kumar , J. Schemm (NOAA/NCEP), J.-S. Kug (CES/SNU), W.-T. Yun (KMA) C.-K. Park (APCC), S, Kar (APCC), J.-J. Luo (FRCGC/JAMSTEC), T. Yamagata (UT) J. Marsh (UH/ICCS), W.-D. Grossmann (GKSS/ICCS)
9
APCC/CliPAS Project
RESEARCH THRUST AREAS

• Establish a pilot operational APCC-MME SPS

• New methodology for integrating MME predictions

• Strategy for Intraseasonal prediction

• Interactive multi-model ensemble prediction
  experiment

• Coupled model initialization and data
  assimilation
• Perturbed physics experiments

• Climate information system model and
  socio-economic value assessment models

10
Current CliPAS/APCC MME Hindcast Experiments
Two-Tier systems
One-Tier systems
SNU SST prediction system
CGCM
AGCM
NASA
CFS/NCEP
FSU
GFDL
SNU
CAM2 (UH)
SINTEX-F
SNU/KMA
ECHAM(UH)
Hybrid CGCM (UH)

• 1981 2004 summer and winter season for 24
  years
• Summer from May 1 to September 30
• Winter from November 1 to March 31

Experiment Period
11
MME Hindcast Skill Temporal Correlation/
1981-2001
2m Air Temperature
DEMETER MMEP
APCC MMEP
Summer Mean Prediction
Winter Mean Prediction
12
(No Transcript)
13
MME Hindcast Skill Taylor Diagram/ 1981-2001
2m Air Temperature
DEMETER MMEP
APCC MMEP
JJA
DJF
14
MME Hindcast Skill Temporal Correlation/
1981-2001
Precipitation
APCC MMEP
DEMETER MMEP
JJA
DJF
15
MME Hindcast Skill Taylor Diagram /1981-2001
Precipitation
DEMETER MMEP
APCC MMEP
JJA
DJF
16
MME Effective Index/ Precipitation
DJF
JJA
17
Wang et al. (2004)
18
(No Transcript)
19
Correlation Coefficients between the observed and
5 AGCM MME hindcasted June-August precipitations
(1979-1999)
Wang et al. (2005)
Fig. 1
20
Area averaged correlation coefficients (skills)
21
MME Hindcast Skill in AAM region 1981-2001
Precipitation
Southeast Asian and WNP region 80-150E, 5-30N
DEMETER MMEP
APCC MMEP
JJA
DJF
22
MME Effective Index/ Precipitation
DJF
JJA
Southeast Asian and WNP region 80-150E, 5-30N
23
One-Tier 1 vs Two-Tier Anomaly PCC over AAM (JJA)
ENSO vs Precipitation SST vs.
Precipitation
24
Probabilistic MMEP Range of Area of ROC Curve/
Above Normal Precipitation
APCC
DEMETER
JJA
DJF
Probabilistic forecast for above normal
precipitation greater than 0.5 standard deviation
25
Deterministic and Probabilistic MMEP Potential
Economic Value/ Above Normal 2m Air Temp
APCC
DEMETER
Probabilistic forecast for above normal 2m air
temperature greater than 0.5 standard deviation
over ENSO Region
26

• Summary of the Preliminary Results
• The CliPAS blended one- and two-tier MME
  hindcasts
• have skills comparable to DEMETER in
  precipitation and
• surface temperature prediction, although their
  individual
• modles performance is lower that those of
  DEMETERs.
• b. The CliPAS MME is more effective due to their
  larger
• mutual independence as evidenced from their
  larger
• range of their skills .
• c. The MME is more effective when and where
  individual
• models have moderate performances while
  potential
• predictability is large. MME is more applicable
  to the
• summer monsoon regions.
• d. In A-A summer monsoon heavy precipitation
  regions,
• one-tier is superior to two-tier system due to
  increased
• feedback from the local surface SST and improved
  ENSO
• teleconnections.