1. CLIVAR/ISVHE Intraseasonal Variability Hindcast Experiment Supporters Design and Preliminary results ISVHE was initiated by an ad hoc group: B. Wang, D. Waliser, H. Hendon, K. Sperber, I-S. Kang Research Coordinator: June-Yi Lee Preliminary results were prepared by June-Yi Lee and Bin Wang

2. ECMWF JMA CWB ABOM EC NCEP CLIVAR/ISVHE Intraseasonal Variability Hindcast Experiment The ISVHE is a coordinated multi-institutional ISV hindcast experiment supported by APCC, NOAA CTB, CLIVAR/AAMP & MJO WG, and AMY. UH IPRC Supporters SNU PNU GFDL NASA CMCC http://iprc.soest.hawaii.edu/users/jylee/clipas.htm

3. ISVHE Participations Institution Participants ABOM, Australia Harry Hendon, Oscar Alves CMCC, Italy Antonio Navarra, Annalisa Cherichi, Andrea Alessandri CWB, Taiwan Mong-Ming Lu ECMWF, EU Franco Molteni, Frederic Vitart GFDL, USA Bill Stern JMA, Japan Kiyotoshi Takahashi MRD/EC, Canada Gilbert Brunet, Hai Lin NASA/GMAO, USA S. Schubert NCEP/CPC Arun Kumar, Jae-Kyung E. Schemm PNU, Korea Kyung-Hwan Seo SNU, Korea In-Sik Kang UH/IPRC, USA Bin Wang, Xiouhua Fu, June-Yi Lee Current Participating Groups

4. Motivation of the ISVHE The Madden-Julian Oscillation (MJO, Madden-Julian 1971, 1994) interacts with, and influences, a wide range of weather and climate phenomena (e.g., monsoons, ENSO, tropical storms, mid-latitude weather), and represents an important, and as yet unexploited, source of predictability at the subseasonal time scale (Lau and Waliser, 2005). is one of the dominant short-term climate variability in global monsoon system (Webster et al. 1998, Wang 2006). The wet and dry spells of the MISO strongly influence extreme hydro-meteorological events, which composed of about 80% of natural disaster, thus the socio-economic activities in the World's most populous monsoon region. The Boreal Summer Monsoon Intraseasonal Oscillation (MISO) Need for a Coordinated Multi-Model ISO Hindcast Experiment The development of an MME is the intrinsic need for lead-dependent model climatologies (i.e. multi-decade hindcast datasets) to properly quantify and combine the independent skill of each model as a function of lead-time and season. There are still great uncertainties regarding the level of predictability that can be ascribed to the MJO, other subseasonal phenomena and the weather/climate components that they interact with and influence. The development and analysis of a multi-model hindcast experiment is needed to address the above questions and challenges.

5. Numerical Designs and Objectives Free coupled runs with AOGCMs or AGCM simulation with specified boundary forcing for at least 20 years Daily or 6-hourly output Control Run ISV hindcast initiated every 10 days on 1 st, 11 th, and 21 st of each calendar month for at least 45 days with more than 6 ensemble members from 1989 to 2008 Daily or 6-hourly output ISV Hindcast EXP Additional ISO hindcast EXP from May 2008 to Sep 2009 6-hourly output YOTC EXP Three experimental Designs for aiming to  Better understand the physical basis for ISV prediction and determine potential and practical predictability of ISV in a multi-model frame work.  Develop optimal strategies for multi-model ensemble ISV prediction system  Identify model deficiencies in predicting ISV and find ways to improve models’ convective and other physical parameterization  Determine ISV’s modulation of extreme hydrological events and its contribution to seasonal and interannual climate variation.  Better understand the physical basis for ISV prediction and determine potential and practical predictability of ISV in a multi-model frame work.  Develop optimal strategies for multi-model ensemble ISV prediction system  Identify model deficiencies in predicting ISV and find ways to improve models’ convective and other physical parameterization  Determine ISV’s modulation of extreme hydrological events and its contribution to seasonal and interannual climate variation.

6. Output Request temperature, salinity, ocean currents (U and V), and vertical motion from surface to 300m III. Upper Ocean 3D Field total precipitation rate, OLR, surface (2m) air temperature, SST, mean sea level pressure, surface heat fluxes (latent, sensible, solar and longwave radiation), surface wind stress, and geopotential, horizontal wind fields (u and v) at three specific levels: 850, 500, and 200 mb I. Atmospheric 2D Field 17 standard pressure levels: humidity, temperature, horizontal wind and vertical pressure velocity (Pa/s), and each of the components of the diabatic heating rates (e.g., shortwave, longwave, stratiform cloud, deep and shallow convection). II. Atmospheric 3D Field The requested outputs are as follows. 1.Output requested from the control simulations: 6-hour values of items I, II and III. 2.Output from the hindcasts initiated from January 1989 up to May 2008: Daily mean values of items I and III. 3.Output from the hindcasts initiated during the YOTC Period (May 2008 – October 2009): 6-hour values of items I, II and III.

7. Description of Models and Experiments Model Control Run ISO Hindcast PeriodEns NoInitial Condition ABOM POAMA 1.5 & 2.4 (ACOM2+BAM3) CMIP (100yrs)1980-200610The first day of every month CMCC (ECHAM5+OPA8.2) CMIP (20yrs)1989-20085Every 10 days ECMWF ECMWF (IFS+HOPE)CMIP(11yrs)1989-200815Every 15 days GFDL CM2 (AM2/LM2+MOM4)CMIP (50yrs)1982-200810The first day of every month JMA JMA CGCMCMIP (20yrs)1989-20086Every 15 days NCEP/CPC CFS v1 (GFS+MOM3) & v2 CMIP 100yrs1981-20085Every 10 days PNU CFS with RAS schemeCMIP (13yrs)1981-20083The first day of each month SNU SNU CM (SNUAGCM+MOM3) CMIP (20yrs)1989-20081Every 10 days UH/IPRC UH HCMCMIP (20yrs)1994-20086Every 10 days One-Tier System Model Control Run ISO Hindcast PeriodEns NoInitial Condition CWB CWB AGCMAMIP (25yrs)1981-200510Every 10 days MRD/EC GEMAMIP (21yrs)1985-200810Every 10 days Two-Tier System

8. Evaluation on Control Runs Variance of 20-100-day Bandpass Filtered Precipitation in Observation and Control Simulations (NDJFMA)

9. Evaluation on Control Runs Pattern Correlation Coefficient and Normalized Root Mean Square Error for Mean Precipitation and 20-100-day Variance The PCC and NRMSE between the observed and simulated seasonal mean precipitation (mean) and variance of 20-100-day bandpass filtered precipitation (variance) for individual control simulations.

10. Evaluation on Control Runs 20-100-Day U850, U200 and OLR along the equator (15 o S-15 o N) The first two multivariate EOF modes of 20-100-day 850- and 200-hPa zonal wind and OLR along the equator (15 o S-15 o N) obtained from observation and control simulations. The percentage variance explained by each mode is shown in the lower left of each panel.

11. ISV Forecast Skills/ ONDJFM 1 Pentad Lead 2 Pentad Lead 3 Pentad Lead 4 Pentad Lead Temporal Correlation Coefficient Skill for U850 ABOMJMANCEP ISOISO+IAV ISOISO+IAVISOISO+IAV

12. Can the MME approach improve MJO forecast? The MME and Individual Model Skills for MJO Common Period: 1989-2008 Initial Condition: 1 st day of each month from Oct to March MME1: Simple composite with all models MMEB2: Simple composite using the best two models MMEB3: Simple composite using the best three models

13. The MME and Individual Model Skills for MJO Normalized RMSE

14. ENSO Dependency Total anomaly vs ISO component & ENSO Dependency Taking into account IAV anomaly, the practical TCC skill for the RMM1 and 2 extends about 5 to 10 days depends on model. In particular, the skill improvement is remarkable for the RMM1 verifying the interference of the ENSO phase onto the MJO phase. It is noted that the skill in the La Nina years is better than El Nino years in most models.

15. ENSO Dependency The dependence of the forecast skill on the initial phase of the MJO NCEP model has less skill for the MJO with phase 3 and 4 initial condition than other phases. CMCC model has better skill when the MJO initially locates in the eastern Indian Ocean and west of the Maritime continent. Other models have less sensitive to the initial phase of MJO.

16. Summary The ISVHE has been coordinated to better understand the physical basis for prediction and determine predictability of ISO. 12 climate models’ hindcast for ISO have been collected from research institutions in North America, Europe, Asia, and Australia. Using the best three models’ MME, ACC skill for RMM1 and RMM2 reaches 0.5 at 27-day forecast lead. An enhanced nudging of divergence field is shown to significantly improve the initial conditions, resulting in an extension of the skillful rainfall prediction by 2-4 days and U850 prediction by 5-10 days. This suggests that improvement of initial conditions are a very important aspect of the ISO prediction (Fu et al. 2011). It is noted that the skill in the La Nina years is better than El Nino years in most models. This may be related to the previous findings indicating the presence of weaker MJO activity in El Nino conditions and stronger activity in La Nino conditions (Seo 2009). Each model’s forecast skill has different sensitivity to the initial phase of MJO. NCEP model has less skill for the MJO with phase 3 and 4 initial condition than other phases. CMCC model has better skill when the MJO initially locates in the eastern Indian Ocean and west of the Maritime continent. Other models have less sensitive to the initial phase of MJO.

17. Thank You!

18. Individual Model Skills for MJO Evaluation of the temporal correlation coefficient (TCC) skill for the RMM1 and RMM2 using available hindcast data Validation dataset: NOAA OLR, U850 and U200 from NCEP Reanalysis II (NCEPII) Each model has different initial condition and forecast period.

19. MV EOF Modes for BS MISO Major Northward Propagating MISO modeGrand Onset mode (LinHo and Wang 2002)

20. Hindcast Skill for the MISO index ISVHE from ABOM, CMCC, ECMWF, and JMA Period: MJJAS from 1989 to 2008

21. Jun 15, 2006 Jun 30, 2006 Jun 15, 2006 Jun 30, 2006 Example for the MISO forecast