PREDICTABILITY of WPSH and EA SUMMER MONSSON and WNP TS

PREDICTIONBin Wang, Baoqiang Xiang, June-Yi LeeDepartment of Meteorology and IPRC,

University of Hawaii

PNAS 2013International Conference on S2S Prediction,

College Park, Maryland, USA 10-13 February, 2014

The CliPAS-ENSEMBLES MME System (9 models)

InstituteModelName

AGCM OGCMEnsemblemember

Reference

ABOM POAMA1.5BAM 3.0dT47 L17

ACOM30.5-1.5o lat x 2.0o lon

L3110 Zhong et al (2005)

GFDL CM2.1AM2.1

2olat x 2.5olon L24MOM4

1/3olat x 1olon L5010 Delworth et al (2006)

NCEP CFSGFS

T62 L64MOM3

1/3olat x 5/8olon L2715 Saha et al (2006)

SNU SNUSNU

T42L21MOM2.2

1/3olat x 1olon L406 Kug et al (2008)

CMCC-INGV CMCCECHAM5T63 L19

OPA 8.22.0o x2.0o L31

9Alessandri et al (2011)

Pietro and Masina (2009)

ECMWF ECMWFIFS CY31R1

T159 L62HOPE-E

1.4o x 0.3o-1.4o L299

Stockdale et al. (2011)Balmaseda et al. (2008)

IFM-GEOMAR IFMECHAM4T42 L19

OPA 8.22.0o lat x 2.0o lon L31

9Keenlyside et al. (2005)Jungclaus et al. (2006)

MF MFIFS

T95 L40OPA 8.0

182GPx152GP L319

Daget et al. (2009)Salas Melia (2002)

UKMO UKMOECHAM5T42 L19

MPI-OM12.5o lat x 0.5o-2.5o lon

L239 Collins et al. (2008)

Current skills for Prediction of SM rainfall is far from satisfactory

Given the limited skills of the current models, Is there anything

else we can do beside MME?A new thrill:

Use the high predictability of the controlling circulation system to

improve rainfall prediction and TC activity prediction that the models

cannot do

Asian summer monsoon Circulation systemAsian summer monsoon circulation system

Variations of WPSH is of utmost importance for seasonal forecast of EASM rainfall and WNP TS activity, (Tao and Chen 1987, Nita 1987, Wu et al. 2002, Wang et al. 2001, Lau and Weng 2002),

Yet the causes of the WPSH variability remains debated and the predictability of the WPSH has not been established.

I. How to measure the year-to-year variation of the WPSH?

II. What determines WPSH interannual variability?

III. How predictable is the WPSH?IV. Can WPSH predictability

benefit EASM rainfall and WNP TS predictions?

I. WPSH Index and its representation of the EASM

and WNP tropical storms (TSs)

WNPSH has largest variability in the global subtropics

WNPSH index:

Normalized JJA mean

H850 anomaly

(15°N-25°N, 115°E-150°E)

80,83,87,88,93,95,96,98,03

81, 82, 84,85,86,90,94,01,02,04

Sui et al. 2007Wu and Zhou 2008: 500 hPa (120-140E, 10-30N)

Wang et al. 2009

MV-EOF1: Precipitation, winds at 850 hPa and 200 hPa, and SLP.

WPSH Index and the EA-WPSM PC1: R=0.92

WPSHI represents the leading mode of EA-WPSM

WPSHI reflects WNP TS Days (TSD)4 extremely strong (left) vs. 4 extremely weak (right) years

WPSHI and Subtropical WNP TSD r=-0.81

WPSHI reflects the TS numbers that affect EA coasts

WPSHI and reversed TS numbers in six coastal regions (R=0.76 ).

JJA TSDs: Climatology (contours) and the difference (shading) between 9 strong and 10 weak WPSH cases.

II. What control interannual variability of

the WPSH?

Approach to answer: What controls major modes of

H850 variability?

r=0.60 r=0.73Cor with WPSHI

First two leading EOF modes of JJA 850 hPa GPH anomaly

They account for 75% of WNP H850 variance

The normalized WNPSH index (black) and the reconstructed index based on

PC1 and PC2 (1.226×EOF-1+1.245×EOF-2) from the EOF analysis of H850 anomaly (red). They have a correlation of about 0.94.

WPSHI can be reproduced by PC1 and PC2

Origin of EOF2?

Anomalies associated with EOF 2

H850 & precipitation

SST & 850 winds

Anomalies forced by ECP SST coolingECHAM ensemble experiments

Numerical Exp: EOF 2: Forced response to CP cooling

Origin of EOF 1?

Anomalies associated with EOF 1

H 850 &rainfall

Winds 850 & SST

A

EOF 1: WPSH and Indo-Pacific SST coupled Mode

Positive A-O thermodynamic feedback between the WPSH and the Indo-Pacific warm pool ST dipole

warming WPSH

Wang et al. 2000, Wang et al. 2003

Coupled Model (POEM) Experiments

Orgin of WPSH variability

1. remote cooling/warming in the equatorial central Pacific (EOF2)

2. local positive thermodynamic feedback between the WNPSH and the

Indo-Pacific warm pool SST dipole. (EOF1)

III. How predictable is the interannual variation of the

WPSH ?Estimate the practical predictability

byPhysically based Empirical (P-E) model

Multi-dynamical models ensemble

Predictors selected based on the major modes dynamics

IO-WNP = SSTA in IO (10°S-10°N, 50°E-110°E) minus SSTA in WNP (0°-15°N, 120°E-160°E) during April-May. r= 0.76

CPSST = May minus March (SSTA (5°S-5°N, 170°W-130°W)r = - 0.50

NAOI during April-May is a precursor for EOF2 r = - 0.41

WPSH intensity reproduced by a Physically based empirical model

WPSH index =1.756×(IO-WNP)–0.435×CPSST–0.282×NAOI

Hindcast the WPSH intensity by coupled climate models (1982-2005)

The WPSH intensity is highly predictable

IV. Using Subtropical Predictability to forecast

EASM rainfall and WNP TS prediction

EASMIntensity index

Tropical storm days (WNP)

Number of TS impacting EA coasts

Prediction of EASM strength, Tropical storm days (TSDs) and the TS numbers influencing EA

The P-E model prediction is constructed by the product of the predicted WPSH index and the regressed precipitation onto the observed WPSH index.

3 coupled models’ MME P-E model Prediction

Temporal correlation skill of JJA rainfall

Observed WPSH index (blue) and simulated and predicted WPSH (red)

The empirical model is constructed based on the time period between 1979-2009 and the last three years (2010-2012) are predicted.

1980 1985 1990 1995 2000 2005 2010

-2

-1

0

1

2 r=0.80r=0.80

Real forecast test (2010-2012)

Conclusions1. Importance: The IAV of the WPSH faithfully represents

the strength of EASM (r=–0.92), the TS days over the subtropical WNP (r=–0.81), and the total number of TSs impacting East Asian coasts (r=–0.76).

2. Dynamics of WPSH: The IAV of WNPSH is primarily controlled by equatorial central Pacific warming/cooling and a positive atmosphere-ocean feedback between the WNPSH and Indo-Pacific warm pool ocean (SST dipole).

3. Predictability: The WNPSH is highly predictable. 4. Prediction of precipitation and TS: Predictability of the

WNPSH paves a promising pathway to predict EASM, ASM and WNP tropical storm activity.

ImplicationsPositive monsoon-ocean interaction can

provide a new source of climate predictability.

Subtropical dynamics is important for understanding monsoon/TS predictability.

Making use of the global models’ strength in prediction of large scale circulation may

substantially improve direct rainfall and TC prediction

Thank You

Comments?

How can anomalous western North Pacific Subtropical High intensify in late summer?

Xiang et al. 2013, GRL

1) The majority of strong WNPSH cases exhibit anomalous intensification in late summer (August), which is dominantly determined by the seasonal march of the mean state. 2) The local convection-wind-evaporation-SST (CWES) feedback relying on both mean flows and mean precipitation. This mechanism is more effective in August.

Coupled mode of WPSH intensify from June to August

14 extreme WPSH composite evolution from June to August

Xiang et al. 2013 GRL

Mean precipitation (shading) and 850 hPa wind

Model simulated response in June and August

June August

Xiang et al. 2013 GRL

The local convection-wind-evaporation-SST (CWES) feedback relying on both mean flows and mean precipitation is more effective in August.

Observed intensification of WNPSH in August

EOF-2 (forced mode)

Modeling Evidence for EOF-2

June August

Precipitation (shading), H850 (black contours) and SST forcing (blue contours)

Why Intensified?

Simply answer:---- Mean state change

Remarks

Two leading EOF modes of H850 in JJA