Climate Diagnostics and Prediction Workshop State College, PA

1. Recent Evolution of the MJO and Description of an Experimental Global Tropics Hazard Assessment Jon Gottschalck1, Vern Kousky1, Wayne Higgins1, Marco Carrera1, Klaus Weickmann2, Ed Berry3, Eric Blake4, Chet Schmitt1, Brad Pugh1, Wanqiu Wang1, Kyong-Hwan Seo1, Michelle L’Heureux1, Qin Zhang1, Wassila Thiaw1,Wei Shi1, Huug van den Dool1 1 Climate Prediction Center 2 Climate Diagnostics Center 3 NWS, Dodge City, Kansas 4 National Hurricane Center Climate Diagnostics and PredictionWorkshop State College, PA October 24-28, 2005

2. Outline • Review of MJO Activity October 2004-September 2005 • 1. Overview of the MJO • 2. Details of two MJO events (spring, mid-winter) • Experimental Global Tropics Hazard Assessment • 1. Description and Prototype Product • 2. Methods and Forecast Tools • 3. Verification • 4. Upcoming Plans

3. MJO Activity October 2004 – September 2005

4. Overview of MJO Activity -- OLR

5. Overview of MJO Activity – Velocity Potential

6. Overview of MJO Activity – 850 mb Wind

7. Overview of MJO Activity – Oceanic Heat Content Warming triggered at onset by a period of persistent westerly wind anomalies

8. Spring Event – (March – May 2005)

9. Spring MJO Event – Characteristics • MJO phase and amplitude • Multivariate EOF analysis • 850 hPa / 200 hPa zonal wind, OLR • (Wheeler and Hendon, MWR, 2004) Period: ~ 45 days (1st), 30 days(2nd) Phase Speed: ~ 4.5 m/s (OLR) Start Phase: Western Indian Ocean Ending Phase: Africa

10. Spring MJO Event – Historical Context Spring 2005 MJO activity Greatest seasonal MJO activity during the last several years dating back to 1997

11. Spring MJO Event – Velocity Potential Animation

12. Spring MJO Event – Velocity Potential Animation

13. Spring MJO Event – Impacts Mar 27-Apr 15, May 6-15 May 11-20 Apr 26-30 Beneficial rains across Indonesia Heavy rains in Central America and northern South America Enhancement in the rainy season in northeast Brazil Above average rainfall

14. Spring MJO Event – Impacts May 26-30 Apr 6-15 Exacerbation of existing dry conditions during several time periods Beneficial break in rains across regions in Central America Break in the rainy season in northeast Brazil Below average rainfall

15. Spring MJO Event – Impacts Mid-late April Apr 16-25 Early-mid April 1. Tropical cyclones Isang and Adeline-Juliet 2. Typhoon Sonca Enhanced SPCZ Tropical Cyclone Development and South Pacific Convergence Zone

16. Winter Event – (December 2004 – January 2005)

17. Winter MJO Event – OLR Enhanced convection associated with the MJO shifts eastward from Indonesia to near the date line. Moisture plume across Hawaii and into southern California.

18. Winter MJO Event – 200 mb wind Upper-level wind anomalies stretching from the tropics into the southwest United States

19. Winter MJO Event – California Rains

20. Experimental Global Tropics Hazard Assessment

21. Background and Status • MJO weekly update since summer of 2004 1. Assess and predict the MJO 2. Discuss evolution  Increasing interest / emphasis on impacts • Global Tropics Benefits/Hazards Assessment • 1. Prototype product  3 page PDF (outlook, discussion, cumulative verification) • 2. Physical basis: ENSO, MJO, other coherent and/or persistent anomalies • 3. Above/below average precipitation, tropical cyclones, etc.  weeks 1 and 2 • 4. Weekly conference call with our collaborators: • Klaus Weickmann – CDC, Ed Berry – NWS Dodge City, Kansas • Eric Blake, Richard Pasch – NHC, Bill Boos – MIT • 5. Outlooks are consistent with US / Africa Hazards and NHC / JTWC forecast tracks

22. Prototype Product – Outlook with Key – Page 1 Outlook: September 26, 2005 Week 1 1. Hurricane Kenneth and Tropical Storm Norma will impact the east Pacific Ocean 2. Increased chance of above average rainfall across west central Africa 3. Increased chance of below average rainfall across India and the Bay of Bengal 4. Increased chance of above average rainfall in the equatorial Indian Ocean 5. Increased chance of tropical cyclone development, Typhoon Longwang will impact the western Pacific 6. Increased chance of above average rainfall in the equatorial western Pacific Week 2 1. Increased chance of above average tropical cyclone activity in the western Atlantic and Caribbean Sea region 2. Increased chance of above average rainfall in the equatorial Indian Ocean 3. Increased chance of below average rainfall in the vicinity of the Philippines

23. Prototype Product – Discussion – Page 2 • Forecast reasoning and basis for outlook areas • Information related to regions and/or impacts not placed on outlook maps (lower confidence) Discussion: September 26, 2005 With ENSO neutral conditions continuing, the amplitude/phase of the MJO remains the dominant forcing across the global tropics. During the last week, the MJO remained weak but there are signs that the MJO may strengthen during the next few weeks. Currently, the enhanced phase of the MJO is located in the western hemisphere with weak upper-level divergence mainly situated across the Atlantic and Africa regions. The suppressed phase stretches from the Indian Ocean to the eastern Pacific Ocean. Above average SSTs, however, have aided the redevelopment of convection in the eastern Indian Ocean and the far western Pacific despite the large scale upper-level convergence in these areas. Hurricane Kenneth and tropical storm Norma continue to impact the eastern Pacific Ocean while Typhoon Longwang is slowly moving to the west towards Asia. During the next 1-2 weeks we expect an increase and expansion of convection from the eastern Indian Ocean into the far western Pacific as a result of a few factors. Above average SSTs will continue to support enhanced convection in these areas while the residual enhanced phase of the MJO propagating eastward from the western Hemisphere will interact with other modes of intraseasonal variability and result in a consolidation of convection in the region. There is a high level of uncertainty, however, of how quickly this area of enhanced convection will propagate eastward. Statistical model forecasts are mixed in how quickly this area propagates east while the GFS dynamical model depicts are more rapid shift. Based on the scenario described above we expect above average rainfall during week 1 across western Africa as a result of the remaining enhanced phase of the MJO and in smaller areas in the eastern Indian Ocean and far western Pacific primarily due by local SSTs. We anticipate a larger region of above average rainfall by week 2 in the eastern Hemisphere. Drier than average areas are expected across India and the Bay of Bengal during week 1and later in week 2 in the vicinity of the Philippines as the weakening suppressed phase of the previous MJO event continues to evolve. In the near term (week 1), areas in the eastern Pacific will continue to be impacted by tropical cyclones. Hurricane Kenneth will move east of Hawaii and tropical storm Norma off the west coast of Mexico. Also, Typhoon Longwang will move east towards China and Taiwan and is expected to produce substantial rainfall, wind, and wave action concerns in these areas late during week 1. Due to continued positive SST anomalies and low vertical wind shear we expect the threat of tropical cyclone activity to remain higher than average across the western Pacific throughout the period.

24. Forecast Resources • Frequent observational monitoring (satellite review, SST, etc.) • ENSO composites • MJO composites • Statistical MJO forecasts • Dynamical forecasts • Frequent monitoring of other coherent tropical modes of variability • Synoptic–Dynamic Model of subseasonal variability (Weickmann and Berry, 2005)

25. Forecast Resources: Statistical Models Jones Forecast October 3, 2005 Empirical Wave Propagation November 5, 2004 Wheeler Forecast April 3, 2005 A change in sign of convection is forecast during the period Convection forecast to shift eastward Remote signals also evident

26. Forecast Resources: Dynamical Models -- GFS Velocity Potential Eastward propagating large scale velocity potential field

27. Forecast Resources: Monitoring Other Coherent Modes Consolidation of convection across eastern Indonesia and the far western Pacific

28. Forecast Resources: Monitoring Other Coherent Modes A fast moving atmospheric Kelvin wave aids in the development of several tropical cyclones in the eastern Pacific during September

29. Initial Verification • Qualitative verification on a case by case basis to date April 2005 Weekly average OLR 1ºx1º Tropical cyclone tracks

30. Planned Quantitative Verification • Evaluate rainfall regions and tropical cyclone development outlooks • Utilize CMORPH, CMAP, OLR global gridded datasets to verify precipitation 1. Short-term: CMORPH weekly data with CMAP climatology 2. Long-term: Apply a scaling methodology between datasets in (1) • Create spatial maps for following: • 1. Hit Rate • 2. Threat Score • 3. Probability of Detection • 4. False Alarm Rate • 5. Bias • 6. Heike Skill Score

31. Upcoming Plans • Fully incorporate the use of GIS software • Release experimental combined product via WWW • Finalize verification strategy and verify forecasts from Mar 2005 - Feb 2006 • Continue to expand collaboration / users  monitoring and research areas • Accelerate research component • Develop additional objective input  Clear link with NOAA CTB

32. Questions, Suggestions, or Comments?

33. Spring MJO Event – Impacts Easterly anomalies east of enhanced convection Westerly wind anomalies located in the western area of upper level divergence associated with enhanced convection

34. Winter MJO Event – Characteristics Duration: ~ 30 days Phase Speed: ~ 4.5 m/s (OLR) Start Phase: Indian Ocean Ending Phase: Western Pacific Ocean Brown and blue lines indicate MJO event

35. Resources: Synoptic-Dynamic Model • Model of sub-seasonal variability • Core is based on four stages based on the MJO recurrence time • Focus on circulation changes and extreme events • Global and zonal atmospheric angular momentum important • Weickmann and Berry, J. Climate, 2005