Ensemble Forecasts for Weather and Seasonal Climate III Posters
Presiding: T Krishnamurti, Florida State University; B Rajagopalan, University of Colorado
A41D-01 0830h
Improved Seasonal Climate Forecasts of the South Asian Summer Monsoon Using a Suite of 13 Coupled Ocean-Atmosphere Models
Thirteen state-of-the-art coupled atmosphere-ocean models were used to construct a consensus forecast called the synthetic superensemble. The merit of this technique lies in assigning differential weights to the member models based on their past performance. Seasonal forecasts obtained from the synthetic superensemble were compared with the individual models and their ensemble mean over the South Asian monsoon region during 13 years of the Northern summer. The skill of the synthetic superensemble, in terms of RMS errors, were best among all the models and their ensemble mean. The RMS errors of the synthetic superensemble were less than that of the ensemble mean at a significance level of 95% or more (based on a student's t-test). Anomaly correlation from the synthetic superensemble were particularly high when the anomaly was high over the South Asian monsoon domain. This ensures the reliability of the present approach in predicting extreme events. This was illustrated with case studies during the switch on and off of the Indian Ocean Dipole which seems to modulate the Indian monsoon rainfall. The results of this study suggest that superensemble provides some what consistent forecasts on the seasonal time scale. This methodology needs to be tested for real time seasonal climate forecasting over the South Asian region.
A41D-02 0830h
On the Utility of a Modest Physics-Based High-Resolution WRF Ensemble for Hurricane Prediction: Hurricane Ivan as an Example
The remarkable 2004 Atlantic hurricane season featured six "major" hurricanes (according to the Saffir-Simpson (SS) intensity scale), four of which directly impacted the state of Florida (Charley, Frances, Ivan, Jeanne) within a six-week period. Of these, Hurricane Ivan distinguished itself with impressive statistics in terms of lifespan (22 days), maximum intensity (SS category 5), damage (est. 13 billion dollars) and U.S. deaths (26). A variety of tools are currently available to forecasters at the National Oceanographic and Atmospheric Administration's (NOAA) Tropical Prediction Center (TPC), including several deterministic and statistical models as well as the Florida State University superensemble (e.g., Shin and Krishnamurti, 2003a,b). Often a blend of solutions from the various packages is utilized, though in other cases the TPC forecasters will follow the solution from a preferred model based on recent performance for the tropical cyclone of interest. Given that the NOAA National Centers for Environmental Prediction (NCEP) continue to move towards an operational environment where a relatively modest ensemble (roughly 6 members), constructed from within the Weather Research and Forecasting (WRF) framework, will figure prominently in the near future (DiMego 2004), a timely question to ask is whether the performance of such an ensemble for tropical systems will add value to the tool box now available to TPC forecasters. While a fully robust answer to this question demands a period of extensive testing under operational conditions, individual case studies can provide significant insights into some aspects of the expected performance of such a modeling system. Therefore, in this presentation we will present early results and limited performance metrics for such a case study, focusing on the 30-hour period beginning with Hurricane Ivan's entrance into the Gulf of Mexico. We note that while our 7-member ensemble consists entirely of WRF model members, in line with NCEP's planned approach, for this case study we limit the ensemble diversity to WRF physics schemes while utilizing higher resolution (2.5 km) for each member. As such, we hope to gain insight into the utility of such ensembles, at such resolution, for prediction of the meso-beta to meso-gamma structure and evolution of tropical cyclones, particularly strong systems such as Ivan.
A41D-03 0830h
Impact of soil moisture on precipitation in regional climate simulations over Rocky Mountains and High Plains.
Regional climate simulations over eastern slopes of Rocky Mountains and adjacent High Plaines are performed in summer 2004 to examine impact of soil moisture on precipitation and to study performance of convective closures and ensemble of closures. Results are compared with cloud resolving simulations and observations to assess feedback between convective closures and soil moisure and identify shortcomings of the closures.