![Page 1: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/1.jpg)
Page 1 JH JEONG
KMA-APCC-CNU updates for dynamical seasonal climate prediction
Jee-Hoon JeongChonnam National University
Nov 2016, WGSIP18
![Page 2: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/2.jpg)
Page 2 JH JEONG
HindcastForecast
ICEOCEAN
ICEOCEAN
KMANWP(N512L70)
Met OfficeNWP
GloSea5 GloSea5
Results Results
Joint System
InitialField
EnsemblePrediction
KMA-Met Office Joint Seasonal Forecasting SystemKMA: the same configuration with Met Offices. Only, initial condition of weather model is different
I. Soil moisture initialization for KMA operational seasonal forecast (CNU-UNIST team)
ERA-interim soil moisture for hindcast,
But climatology for forecast
![Page 3: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/3.jpg)
Page 3 JH JEONG
2016 summer heatwave in Korea/East Asia
Nearly 1-month long heatwave in
Korea/East Asia this summerDry condition from spring (JULES-JRA55 soil moisture)
Dry condition – atmospheric high pressure system led such a mega
heatwave
![Page 4: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/4.jpg)
Page 4 JH JEONG
Soil moisture initialization from offline JULES(GloSea5’s LSM) run
• Atmospheric forcing's from JRA-55 reanalysis
• Precipitation is adjusted to meet observed monthly precipitation
• Scaling of soil moisture in considering bias between JULES and
GloSea5
• Hindcast was performed to examine the impact on skill
JULES-JRA55 soil moisture is quite consistent with JULES-observation soil moisture
![Page 5: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/5.jpg)
Page 5 JH JEONG UNIST Climate Environment Modeling Laboratory
Skill (R) comparison: GloSea5 hindcast
1~30 days mean Tsfc forecast skill (R) for 1996~2009
before
after
![Page 6: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/6.jpg)
Page 6 JH JEONG
II. KMA/NIMS-SNU: stratospheric influences on dynamical climate prediction for East Asia
Anomalously warm Dec 2015Was associated with strong polar
vortex
Polar Vortex
There is a strong correlation between polar vortex and Eurasian
temperature
Work led by SNU: prof S.-W. Son, KMA/NIMS: Dr. H.-S. Kang
![Page 7: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/7.jpg)
Page 7 JH JEONG
II. KMA/NIMS-SNU: stratospheric influences on dynamical climate prediction for East Asia
MJO-extratropic teleconnection is strengthened during easterly phase of MJO
MJO signal strengthened during easterly phase of QBO
Higher MJO prediction skill during EQBO
QBO (Quasi-Biennial Oscillation)
![Page 8: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/8.jpg)
Page 8 JH JEONG
II. Skill of polar vortex and QBO in GloSea5 forecast
Polar vortex QBO
Polar vortex: 1-2 month predictable
Polar vortex: predictable more than 6 months
Two factors will be utilized for winter seasonal climate prediction by KMA (dynamical-statistical)
![Page 9: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/9.jpg)
Page 9
III. updates for APCC
![Page 10: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/10.jpg)
Page 10
DMME: SCM Forecast (2008JFM-2015DJF)
The skill of forecasts has continuously increased. - SCM MME outperforms individual models
APCC MME
Improvements of prediction skill
Real time forecast
Hindcast skill of each year’s MME
SCM: Simple Composite MME (Multi-Model Ensemble)
![Page 11: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/11.jpg)
Page 11
Experimental: Prob. ENSO Category & Strength Forecast
Definition: based on 3-month mean Nino 3.4 index (+1.5oC /+1.0oC /+0.5/-0.5oC/-1.0oC /-1.5oC : Strong/Moderate/Weak El Nino/La Nina)
Parametric estimator using a Gaussian fitting method for categorical probabilities
Simple composite of the individual model’s probabilities with equal weightings
3M Mean Nino 3.4 Index: 2013-15
2013 2014 2015
OBS
Forecast
To become fully operational in 2017
![Page 12: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/12.jpg)
Page 12
OASIS3-MCT
ECHAM v5.3T159, L31
POP v2.0.10.3 – 1.0 deg, L40
CICE v4.1
• SST
• Surface current
• Tau over water
• LH, SH
• Radiation fluxes
• Total water fluxes
• Height, Wind
• Humidity, Temperature
• Radiation fluxes
• Rainfall, Snowfall
• Ice fraction
• Ice surface temperature
• Albedo
• SST, SSS
• Surface current
• SSH gradient
• Ice freeze/melt heat fluxes
• Ice fraction
• Stress between ocean and ice
• Melting fresh water flux
• Salt flux
• Net ocean-ice heat flux
• Shortwave ice to ocean
Once a day (36 variables)
12
Development of in-house seasonal forecast model: Seamless Coupled Prediction System (SCoPS)
![Page 13: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/13.jpg)
Page 13
CFSR atmosphere variable
Integrated Atmosphere-Ocean Initialization Scheme
Development of in-house seasonal forecast model: Seamless Coupled Prediction System (SCoPS)
3-D Atmospheric Initialization
Ensemble Adjustment Kalman Filter (EAKF)
XBT & ARGO ocean variable
• Total Ensemble members : 10
• Time lagged forecasting with perturbation ensemble spread by EAKF
• ATM/OCN/ICE ICs: the 1st, 5th of every month (5 perturbation initial data)
• Hindcast period: 1982-2014 (33 years) / Forecast period: 2015-present
• 7month integration
Hindcast Simulation
![Page 14: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/14.jpg)
Page 14
Multi Model Subseasonal forecast of BSISO(Boreal Summer Intraseasonal Oscillation)
The canonical northward propagating component The AMS pre-monsoon and onset component
Lee, J.-Y., B. Wang, M. C. Wheeler, X. Fu, D.E. Waliser, and I.-S. Kang, 2013: Real-time multivariate indices for the boreal summer intraseasonal oscillation over the Asian summer monsoon region. Clim. Dyn., 40, 493-509.
![Page 15: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/15.jpg)
Page 15
Multi Model Subseasonal forecast (from July 2013): BSISO (Boreal Summer Intraseasonal Oscillation, Lee et al. 2013)
NCEPCFS
GFS
BoM POAMA
ECMWF EPS
UKMO MOGREPS-15
CWB CWB EPS T119
• Participating models
![Page 16: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/16.jpg)
Page 16
Multi Model Subseasonal forecast (from July 2013): User friendly information from BSISO forecast
Heavy Rainfall Probability Forecast (left)and its Skill Assessment over Mekong river basin (right)
ECMWF
ECMWF
![Page 17: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/17.jpg)
Page 17
Thanks!
![Page 18: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/18.jpg)
Page 18
Experimental: Forecast skill for ENSO Category
ENSO Category Forecast: 1983-2005, All seasons
• E: El Nino• N: ENSO-neutral• L: La Nina
ENSO Intensity Forecast: 1983-2005, All seasons
• SE: Strong El Nino• ME: Moderate El Nino• WE: Weak El Nino
• SL : Strong La Nina• ML: Moderate La Nina• WL: Weak La Nina
To become operational in 2017
![Page 19: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/19.jpg)
Page 19
2016-17 seasonal TCs track density forecast
Contour: Seasonal TC track density in each gridShaded: Anomaly compared to the climatology
Seasonal Tropical Cyclone Forecast: South Pacific Ocean
![Page 20: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/20.jpg)
Page 20
Selection of EAWM Index
Predictability of EAWM in climate models
Predictability of East Asian Winter Monsoon (EAWM)in APCC MME Forecast System
Prediction skill in hindcasts
Model hindcast period : 1983-2006Model forecast period : 2012-2015
Fig 1. The temporal correlation coefficients between the observed and simulated normalized DJF-mean EAWM indices for the period 1983–2007. Dashed lines denote the threshold values for the 95% and 99% significance levels.
Representability of the T2m ano. related to EAWM
Fig 2. Scatter plots of the normalized EAWM index vs DJF-mean surface temperature anomalies averaged over East Asia (100–145°E, 20–50°N) for the period 1983–2007 based on the EAWM index defined by Wang and Chen (2014). The strong (weak) EAWM year with cold (warm) winter defined in Table 3 is shown at the lower right (left) corner with blue (red) characters of each plot.
Prediction skill of EAWM Index
Prediction skill in real-time forecasts
Fig 3. Correlation coefficients between observed and simulated indices from POAMA for different lead months. Red solid line and blue dotted line indicate original and modified index, respectively.
![Page 21: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/21.jpg)
Page 2121
East Asian Winter Monsoon
COR: 0.51/0.74
Lee et al. (2014)
East Asian Summer Monsoon
COR: 0.23/0.58
Li and Yang (2010)
COR: 0.24/0.46
COR: 0.44/0.69
Wang and Fan (1999)
Development of in-house seasonal forecast model: Seamless Coupled Prediction System (SCoPS)
SCoPS predicts EASM and EAWM related circulation better than APCC CCSM3
![Page 22: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/22.jpg)
Page 22
Skill assessment for BSISO real-time forecast
- ECMWF produces skillful BSISO1 forecasts up to a lead time of close to 30 days, whilethe same level of skill in other models extends only to a lead time of about 10 days.
- Predictable period of BSISO2 is about 8 ~ 20 days.
![Page 23: KMA-APCC-CNU updates for dynamical seasonal climate prediction€¦ · Nov 2016, WGSIP18. Page 2 JH JEONG Hindcast Forecast ICE OCEAN KMA NWP(N512L70) Met Office NWP GloSea5 Results](https://reader034.vdocuments.net/reader034/viewer/2022050111/5f49045dce6e2d5e5f3c433d/html5/thumbnails/23.jpg)
Page 23
Multi Model Subseasonal forecast (from July 2013): Possible application of BSISO forecast
BSISO forecasts service
f (Amplitude, Phase)BSISO index =
Probability forecast of heavy rainfallestimated by BSISO forecasts
Week-2 forecastWeek-1 forecast
..
Flood Inundation Mapping