APECAPEC ClimateClimate CenterCenter (APCC)(APCC) ClimateClimate InformationInformation ServiceService

Australia AAsiasia Brunei Darussalam Canada Chile People’s Republic of China Hong Kong, China PPacificacific Indonesia Japan EconomicEconomic CooperationCooperation Korea Malaysia Mexico New Zealand CClimatelimate Papua New Guinea Peru Philippines Russia Singapore CCenterenter Chinese Taipei First AMY/MAHASRI Workshop Thailand 8-10th, January, 2007, Tokyo, Japan United States 8-10th, January, 2007, Tokyo, Japan Viet Nam

APEC Climate Center VisionVision ofof APCCAPCC

ƒƒ RealizingRealizing thethe APECAPEC visionvision ofof regionalregional prosperityprosperity ¾¾ thethe enhancementenhancement ofof economiceconomic opportunitiesopportunities ¾¾ thethe reductionreduction ofof economiceconomic lossloss ¾¾ thethe protectionprotection ofof lifelife andand propertyproperty APEC Climate Center (APCC)

ECMWF/EU APCC/APEC

Iceland

Norway Finland Sweden Denmark Ireland The Netherlands Germany United Czech Republic Belgium Kingdom Austria Luxembourg Hungary France Slovenia Switzerland Croatia Italy Portugal Spain Greece Turkey

1973

2005 GoalsGoals ofof APCCAPCC ƒ Facilitating the share of high-cost climate data and information

ƒ Capacity building in prediction and sustainable social and economic applications of climate information

ƒ Accelerating and extending socio-economic innovation FunctionsFunctions ofof APCCAPCC

ƒ Developing a value-added reliable climate prediction system ƒ Acting as a center for climate data and related information ƒ Coordinating research toward the development of an APEC integrated climate- environment-socio- economic system model APCCAPCC StructureStructure

Member Science Working Director Advisory Group Committee

Administration Science System division division division

Climate Climate Information R&D Prediction Service ClimateClimate Prediction:Prediction: APCCAPCC MultiMulti--ModelModel EnsembleEnsemble

APCC MultiMulti--InstitutionalInstitutional CooperationCooperation

APEC Climate Center MMEMME PredictionPrediction MethodMethod

Training Phase Forecast Phase

Observed Analysis

t=0 MME N Real-time S = O + ∑ai (Fi − Fi ) Prediction i=1

train 2 The weights are computed at each grid point by minimizing the function: G = ∑()St − Ot t=0

APEC Climate Center APCCAPCC DeterministicDeterministic MMEMME ForecastForecast

APEC Climate Center Probabilistic Forecast Method

ƒ Defining terciles using Normal Fitting Method

- For the middle/upper tercile boundary : mean plus 0.43 times the standard deviation Æ μ + 0.43σ - For the lower/middle tercile boundary : mean minus 0.43 times the standard deviation Æ μ -0.43σ

ƒ Forecast probability

- Above normal case (For example) Xc A μ PP =1− N( ,σ) ∫−∞

A Probability of Above-normal

N Probability of Near-normal

-Xc Xc B Probability of Below-normal

APEC Climate Center Probabilistic Forecast Method

ƒ Combine different models

Na : num. of above-normal Nn : num. of near-normal Combine : according to each model’s square root of ensemble size Nb : num. of below-normal

ƒ Merged 3-category distribution

χ2 (Chi-square) TEST Above-normal k 2 3-Categorical distribution ∑(Oi −Ei ) 2 i=1 χ = Near-normal Ei O : Forecast frequencies E : Hindcast frequencies

Under 0.05% significance level Below-normal

APEC Climate Center APCCAPCC ProbabilisticProbabilistic MMEMME ForecastForecast

APEC Climate Center MME-basedMME-based statisticalstatistical downscaling:downscaling: precipitationprecipitation inin ThailandThailand

Bangkok

Correlation coefficient before and after dow nscaling in each station

0.8

0.6

0.4

0.2 Before 0

Cor. coeff After -0.2

-0.4

-0.6 Sattahip Lop buri Nakhon Bangkok Average Donmuag ratchsima Chata buri Prachin buriPrachin Suphan buriSuphan

APEC Climate Center ForecastForecast Verification:Verification: DeterministicDeterministic MME MME

Taylor Diagram : 1983~2003, JJA Hindcast

APEC Climate Center ForecastForecast Verification:Verification: DeterministicDeterministic MME MME

Anomaly Correlation Coefficient for hindcast in each region 0.6(precipitation, JJA, 1983-2003) 0.5 0.4 0.3 0.2

ACC 0.1 0 -0.1 -0.2 AUSTRALIA EAST ASIA NORTH RUSSIA SOUTH SOUTH AMERICA AMERICA ASIA

individual model multi-model

ensemble schemes APEC Climate Center ForecastForecast Verification:Verification: ProbabilisticProbabilistic MMEMME

Globe,Globe, AboveAbove--Normal,Normal, PrecipitationPrecipitation

Reliability Diagram ROC Curve

1.0 CWB 1.0 0.9 GCPS 0.9 GDAPS_F 0.8 0.8 GDAPS_O 0.7 0.7 HMC 0.6 0.6 IRI IRIF 0.5 0.5 JMA 0.4 0.4 METRI rate Hit

Frequency 0.3 MGO 0.3 0.2 NCC 0.2 NCEP Observed Relative Relative Observed 0.1 PMME 0.1 0.0 계열14 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Forecast Probability False alarm rate

APEC Climate Center ClimateClimate InformationInformation ServiceService

APEC Climate Center RealReal--timetime ClimateClimate MonitoringMonitoring

El Nino Monitoring

APEC Climate Center Data Service System

MAHASRI Transformation Visualization

MODEL Data Data format Quality Data MME Acquisition Unification Check Dissemination openDAP

OBS (NCEP, CAMS_OPI, Reduce Data ftp the data size CMAP, JRA-25) Backup

SATELLITE

APEC Climate Center 기상청 Data Service System 장기예보지원 ¾¾ 2D2D 그래픽그래픽 표출 표출 시스템 시스템 구축 구축 및 및 시험 시험 운영 운영

GDS 기반의 원격 기후 자료 표출 시스템 시범 운영 http://gds.apcc21.net:9090/dods (외부접속) 웹에서 제공되는 http://testbed.apcc21.net:9090/dods (내부접속) 엔트리 정보를 이용한 기후자료 표출

APEC Climate Center ResearchResearch && DevelopmentDevelopment

Multi-CumulusMulti-Cumulus convectionconvection schemescheme EnsemblesEnsembles Climatology and zonal mean of precipitation during the boreal summer (JJA)

z Convection scheme of the four different individual models and unified model SAS Kuo MCA ZM UNI

Heating rate

Change rate of (SAS + Kuo + moisture MCA + ZM ) / N Rain fall , (N=4)

Snow fall

Cloud water in cumulus (SAS + ZM ) / N Areal rate of , (N=2) cumulus Cloud mass flux

APEC Climate Center Multi-CumulusMulti-Cumulus convectionconvection schemescheme EnsemblesEnsembles

EOF of CISO for Precipitation

Unified Scheme best captures the CISO evolution

Unified scheme

Observation

APEC Climate Center TC activity/monsoon in high-resolution GCM (2007~)

1997 2004

•Examine high-resoultion GCM performance •Compare with SNU, NASA results

1999 2005

APEC Climate Center FutureFuture PlanPlan Improving Quality of Service

1st phase : Innovative technology development (~2010) 2nd phase : Socio-economic application technology development (2011~2015) 3rd phase : Providing climate and application information to the end users as growth engine (2016~2020) CollaborationCollaboration withwith MAHASRIMAHASRI Development of RCM schemes and application models: ƒAPCC will provide hindcast datasets from participating models and from the CLiPAS project, as well as MME seasonal prediction data for dynamical downscaling efforts of MAHASRI, in order to better understand the predictability of the Monsoon systems and for the development of application models. Data service and management: ƒAPCC will serve as one of the data centers for the intensive observational data during the AMY-2008 ƒDatasets will be made available to model providers for verification and ultimately model improvement CollaborationCollaboration withwith AMY/MAHASRIAMY/MAHASRI

Expects high resolution (spatial and temporal) station data, LS data, for ƒhigh-res model verification ƒStatistical/RCM downscaling development ƒapplication model development ƒClimate monitoring ((intraseasonal,intraseasonal, drought, flood, TCTC,, other extreme events) FocalFocal points/memberspoints/members inin MAHASRIMAHASRI WG:WG:

ƒPrediction & application: C.K. Park/Nohara ƒRCM downscaling: Tam/Song ƒData Management: Tam/Lee Thank you