Journal of the Atmospheric Sciences, 71, 3327-3349.

Gross moist stability and MJO simulation skill in three full-physics GCMs.

James J. Benedict and Eric D. Maloney
Department of Atmospheric Sciences, Colorado State University, Fort Collins, CO.

Adam H. Sobel
Department of Applied Physics and Applied Mathematics, Department of Earth and Environmental Sciences, and Lamont-Doherty Earth Observatory, Columbia University, New York, NY.

Dargan M. W. Frierson
Department of Atmospheric Sciences, University of Washington, Seattle, WA.


Previous studies have demonstrated a link between gross moist stability (GMS) and intraseasonal variability in theoretical and reduced-complexity models. In such simplified models, moisture modes --- convectively coupled tropical disturbances akin to the MJO whose formation and dynamics are linked to moisture perturbations --- develop only when GMS is either negative or "effectively" negative when considering additional sources of moist entropy. These simplified models typically use a prescribed, time-independent GMS value. Limited work has been done to assess GMS and its connection to intraseasonal variability in full-physics general circulation models (GCMs).

The time-mean and intraseasonal behavior of normalized GMS (NGMS) are examined in three pairs of GCMs to elucidate the possible importance of NGMS for the MJO. In each GCM pair, one member produces weak intraseasonal variability while the other produces robust MJOs due to a change in the treatment of deep convection. A highly correlated linear relationship between time-mean NGMS and MJO simulation skill is observed, such that GCMs with less positive NGMS produce improved MJO eastward propagation. The reduction in time-mean NGMS is primarily due to a sharp drop to negative values in the NGMS component related to vertical advection, while the horizontal advection component has a less clear relationship with MJO simulation. Intraseasonal fluctuations of anomalous NGMS modulate the magnitude of background NGMS but generally do not change the sign of background NGMS. NGMS declines ahead of peak MJO rainfall and increases during and after heaviest precipitation. Total NGMS fluctuates during MJO passage but remains positive, suggesting that other sources of moist entropy are required to generate an effectively negative NGMS.