1. MMF Tasks Global Visualization of 4-month Simulations 30-day MJO Simulations Preliminary Analysis on Simulated Cloud Liquid Water Path (removed from this file) Preliminary Analysis on MMF simulations for two golden years (2006, 2007)

2. This animation shows simulated precipitating rain with the Goddard Multiscale Modeling Framework (MMF), which consists of a global model (fvGCM) and 13,104 cloud models (GCE). Numerical simulations are initialized on April 1 2006, and the animation starts on August 31 2006, and ends on January 3, 2007 (4-month simulations). Several remarkable weather system are captured with some degree of realism (see followings). 1) Tropical waves move off the coast of Africa and propagate westward. [ It is known that tropical cyclogenesis can be initialized (or triggered) by these tropical waves. Therefore, accurate simulations of their interactions with small- scale convection are important for improving the simulations of TC genesis. ] 2) The eastward-traveling system in the southern hemisphere (SH) are the so-called the polar vortex, which is most powerful in the hemisphere's winter (JJAS, in the SH). 3) The equatorial Amazon has abundant rain between November and May. During the Brazilian spring season (October/November/December), most of the countries get wetter, except for the Brazilian northeast. 4) In comparison, during this period (winter in the northern hemisphere), mid-latitude periodic frontal systems move eastward across the USA. 5) Near the end of simulations, heavy precipitations appear near the ITCZ. Visualization was made by Bryan Green of CSC/NASA Ames. The global model simulation was made by Bo-Wen Shen and Jiundar Chern of NASA GSFC. Tao, W.-K., J. Chern, R. Atlas, D. Randall, X. Lin, M. Khairoutdinov, J._L. Li, D. E. Waliser, A. Hou, C. Peters-Lidard, W. Lau, and J. Simpson, 2008: Multi-scale modeling system: Development, applications and critical issues, Bull. Amer. Meteor. Soc. (accepted). Goddard MMF Simulated Hourly Rainfall 08/31/2006 - 01/03/2007

4. Scale the MMF up to 364 or more CPUs with a 2D (X-Y) domain decomposition

5. The MJO (Madden and Julian 1972, 1994) is one of the most prominent large-scale features of the tropical general circulation; it is typically characterized by deep convection originating over the Indian Ocean and subsequent eastward propagation into the Pacific Ocean. Current understanding (including theory and hypotheses) indicates that (1) moisture convergence (e.g., Lau and Peng 1987; Wang 1988), (2) surface heat and moisture fluxes (e.g., Emanuel 1987; Neelin et al. 1987), (3) cloud-radiation feedback (e.g., Hu and Randall 1994, 1995), (4) convection-water vapor feedback (e.g., Woolnough et al. 2000; Tompkins 2001), and (5) “ discharge- recharge ” associated with moist static energy build-up and release (e.g., Blade and Hartmann 1993) are important for the MJO ’ s initiation, intensification, and propagation (see a review by Zhang 2005).

6. It is known that accurately predicting tropical activities at a sub-seasonal scale (~30days) is crucial for extending the predictability of numerical weather prediction (NWP) beyond two weeks. Among challenges in predicting tropical activities is the accurate forecasting of an MJO, which is a large- scale convective organized system with a 45-60 day scale. It is believed that improving moist process associated with cloud-scale motion is crucial for improving simulations of MJOs in numerical models. By taking advantage of existing global and cloud models, the so-called MMF provides an innovative approach for understanding these multiple processes and multiscale interactions. While MMF approach has shown promising long-term simulations, its performance on short- term and/or extended range simulations is less understood. As compared to climate simulation which is viewed as a boundary value problem, short-term weather simulation/ forecasting is an initial-value problem. Therefore, it is argued that accurate sub-seasonal forecast may depend on accurate representation of both initial and boundary conditions, suggesting the importance of model initializations. In this case study, we will address if and how the Goddard MMF is suitable for short-term and extended-range weather forecasts, aimed at improving model’s capabilities in simulating sub-seasonal weather systems. In a 30-day simulation of the MJO in late 2006 and early 2007, it is found that the location and propagation speed of the MJO are simulated reasonably well, including the formation of large-scale convective system in the Indian Ocean and its eastward propagation. Additional numerical experiments for this case suggest that the following additional factors for improving MJO simulations may be required: (1) accurate initial conditions (e.g., active or passive phases of an MJO), (2)accurate representations of the mechanical and thermal effects of the Maritime continent. NASA Goddard Multiscale Modeling System simulated MJO

7. Figure Caption A 30-day simulation of an MJO initialized at 0000 UTC December 13, 2006. Left panels (from top to bottom) are 200 hpa velocity potential at 0300 UTC December 13, 16, and 21, respectively. Right panels (from top to bottom) are the 200 hpa velocity potential at 0300 UTC December 26, 31, 2006 and January 5, 2007. Compared to the NCEP/GFS reanalysis, this MMF simulation captures several major features usually associated with an MJO: (1) initiation of large-scale organized convection in the Indian Ocean in panel (b), (2) intensification as shown in panel (c), (3) slow propagation (prior to reaching the Maritime continent), (4) followed by fast propagation, and (5) weakening. However, this simulated MJO also produces stronger vertical motion than does the NCEP/GSF reanalysis.