SLIDE 1 The Operational Subseasonal to Seasonal Climate Forecast System and Development at CWB
Mong-Ming Lu Jhy-Wen Hwu Chih-Hui Hsiao Yea-Chin Tung & CWB CFS Team
Research and Development Center Central Weather Bureau, Taiwan
盧孟明 盧孟明 盧孟明 盧孟明 胡志文 胡志文 胡志文 胡志文 蕭志惠 蕭志惠 蕭志惠 蕭志惠 童雅卿 童雅卿 童雅卿 童雅卿
5th Conference on East Asia and Western Pacific Meteorology and Climate cum Hong Kong Meteorological Society 25th Anniversary 11:00 Sunday, Nov 03, 2013
SLIDE 2
Long-range Weather Forecast at CWB History
Period I (1978 – 1994)
1978 - Issue the Monthly Weather Outlook 1994 - Issue the Seasonal Weather Outlook Prediction methods: pure statistical Focus area: Taiwan
Period II (1995-2001)
Conduct systematic tropical and global climate monitoring Access the dynamical model prediction products of IRI, ECPC, NCEP, JMA, ECMWF on web Continuously improve the pure statistical prediction methods Focus Area: Taiwan
SLIDE 3 Long-range Weather Forecast Monthly-to-Seasonal Climate Forecast
(short-tem climate forecast)
Period III (2002-present) A Modernization Breakthrough
Goal: Objective Probabilistic Prediction
- Prediction Strategy: Multi-type (statistical, dynamical,
dynamical-statistical), Multi-model Multi-member Prediction Lead Time: extend to 4-6 months Focus Area: Taiwan and Southeast Asia
- Improve the understanding of regional climate variations
- Science-based representation of prediction uncertainties
SLIDE 4
The CWB Climate Information System Framework The CWB Climate Information System Framework The CWB Climate Information System Framework The CWB Climate Information System Framework Users
Climate Information Dissemination System Climate Monitoring System Climate Analysis System Climate Forecast and Monitoring Decision Supporting System Climate Data Base Climate Data Process and Display System (in CWB Virtual Data Center) Prediction Dynamical-Statistical Climate Prediction System Statistical Prediction System Climate Forum
2002 ~ 2009
SLIDE 5
The Backbone of Climate Service
SLIDE 6
Prediction Procedure
SLIDE 7
Prediction Schedule
SLIDE 8
Global SST forecast for NDJ 2013
the first forecast month: Nov 2013
ensemble mean anomaly probability forecast
No skill forecasts are masked.
Product
SLIDE 9
- 850 wind monthly forecast
ensemble mean anomaly
Nov 2013 Dec 2013 Jan 2013
Above normal winter monsoon in Taiwan area
Product
SLIDE 10
- Precipitation Forecast for
NDJ 2013 ensemble mean anomaly probability forecast
No skill forecasts are masked.
Product
SLIDE 11
Precipitation Temperature
Probabilistic forecasts of precip. and temp. for NDJ 2013 derived from the statistical downscaling methods
Product
SLIDE 12 850 wind T2m Precip.
Product Example
IRI/ECHAM4 + (CWB-RSM,NCEP-RSM) Prediction for NDJ 2013 ensemble mean anomaly
Product
SLIDE 13 Probabilistic forecasts
NDJ 2013 derived from the dynamical downscaling methods
Product
SLIDE 14
On Going Developments
SLIDE 15
Correlation of T2m Monthly Forecast and Reanalysis
Retrospective forecast: 1983-2012 Reanalysis: NCEP CFSR Forecast: CWB 2-tiCWB AGCM/CWB OPGSSTv2) Forecast initial time: Dec. Forecast lead: 3 months
CWB 2-T CFSv2 – CWB AGCM/CWB OPGSSTv2 CWB 2-T CFSv1 – CWB AGCM/CWB OPGSSTv1
SLIDE 16
Correlation of T2m Monthly Forecast and Reanalysis
Retrospective forecast: 1983-2012 Reanalysis: NCEP CFSR Forecast: CWB 2-tiCWB AGCM/CWB OPGSSTv2) Forecast initial time: Dec. Forecast lead: 5 months
CWB 2-T CFSv2 – CWB AGCM/CWB OPGSSTv2 CWB 2-T CFSv1 – CWB AGCM/CWB OPGSSTv1
SLIDE 17 Summary
- CWB is capable of producing weather and
climate forecast information up to 2 seasons.
- Since the main source of predictability of the
long-term predictability of the climate around Taiwan is ENSO, we expect better forecast skill of the CWB 2-tier CFSv2 than v1.
- The retrospective forecast with new MME
strategy, make 4-member runs of 7 months everyday, generates a rich data set that can be used to study predictions from subseasonal to seasonal scales.
SLIDE 18
SLIDE 19
SLIDE 20 THE ELEMENTS THE ELEMENTS THE ELEMENTS THE ELEMENTS
Dynamical-Statistical Prediction System Statistical Prediction System Supplemental Conceptual Models Interannual Climate Variations
(ENSO, AA Monsoon, Ind Ocean Dipole, TBO, Typh Tracks, Pacific Sub-High)
Extreme Weather a Climate Events
(Heavy Rainfall, Drought, Typhoons, Temperature Extremes, Weather Hazar
Decadal-scale Clim Variations
(AO, NAO, PDO, ENSO)
G G G Gl l l lo
b b ba a a al l l l C C C Cl l l li i i im m m ma a a at t t te e e e C C C Ch h h ha a a an n n n AA Monsoon Monitoring AA Monsoon Monitoring AA Monsoon Monitoring AA Monsoon Monitoring Tropical Convective Tropical Convective Tropical Convective Tropical Convective Systems Monitoring Systems Monitoring Systems Monitoring Systems Monitoring
(MJO, Easterly Waves, Typhoons )
Local Climate Monitoring Local Climate Monitoring Local Climate Monitoring Local Climate Monitoring Midlatitude Purterbation Midlatitude Purterbation Midlatitude Purterbation Midlatitude Purterbation Systems Monitoring Systems Monitoring Systems Monitoring Systems Monitoring
(Blockings, Waves, Upper-level Jet Streams, Storm Tracks)
PREDICTION
SST/Subsurface Ocean SST/Subsurface Ocean SST/Subsurface Ocean SST/Subsurface Ocean Temperature & ENSO Temperature & ENSO Temperature & ENSO Temperature & ENSO Monitoring Monitoring Monitoring Monitoring
MONITORING ANALYSIS
ASSESSMENT ASSESSMENT ASSESSMENT ASSESSMENT
SLIDE 21
Prediction – probabilistic nature of the forecast beyond 7 days Analysis – climate extremes attribution, understand climate variations on the interannual to interdecadal and longer time scales
Monthly and seasonal outlooks are prepared for mean temperature and accumulated precipitation in three categories: Below-, Near- and Above-normal (median). The likelihoods of three categories are assesed.
Monitoring – climate modes on the subseasonal to seasonal time scales
Watch the status of major climate modes (ENSO, NAO, AO, Blockings, MJO, BSISO, Tropical cyclones, … etc.) Project possible development and the influence on local weather Enable CWB to deliver science based climate services
