Recent implementation of the Emanuel cumulus arameterization scheme (Emanuel 1991, Emanuel and Zivkovic-Rothman 1999) in the U. S. Navy Operational Global Atmospheric Prediction System (NOGAPS) in replacement of the Relaxed Arakawa-Schubert Scheme (Moorthi and Suarez 1992) has lead to an overall improvement in the performance of NOGAPS, especially for tropical cyclone predictions. The Emanuel scheme, however, shows some weaknesses. Among them, the most significant characteristics are a warm bias at upper levels, a weak wind bias at all levels, and under-prediction of heavy-precipitation events. In this paper, we will discuss the results from our recent modification to the scheme.

Unique features of the Emanuel Scheme include a prognostic determination of the cloud-base mass flux to maintain a subcloud-layer quasi-equilibrium, vertical distribution of mixing of undilute air rising through cloud base according to the buoyancy gradient, and the breakdown of the cloud mass flux at each level into subcloud drafts based on the buoyancy sorting concept. Guided by the comparison of the vertical heating profile with tropical observations, we modify the scheme so that the vertical partitioning of mixing is dependent on the buoyancy instead of buoyancy gradient. This modification removed the warm bias at upper levels, improved the weak wind bias and improved tropical cyclone forecasts. However, the scheme suffers from insufficient precipitation locally for heavy events, and the onset of precipitation has a delay phase error. We have identified the determination of cloud base mass flux as an important element for further improvement of the scheme and provided suggestions in this regard.