Sensitivity of SCM Convective Clouds During AMIE-Gan to Cumulus Parameterization Assumptions

 

Authors

Anthony D. Del Genio — National Aeronautics and Space Administration
Audrey B. Wolf — NASA - Goddard Institute for Space Studies
Jingbo Wu — Columbia University

Category

MJO: Madden-Julian Oscillation

Description

Moist convection during an MJO event is characterized by time variations in the relative populations of shallow cumulus, congestus, deep convective cells, and mesoscale organized clusters. These variations appear to be regulated by large-scale dynamical processes that determine atmospheric humidity. The AMIE-Gan AMF deployment during the DYNAMO field experiment provides an excellent opportunity to test cumulus parameterization assumptions that determine the sensitivity of convection to humidity. We used the AMIE-Gan constrained variational analysis forcing to drive different versions of the GISS SCM. The GISS Model E2 cumulus parameterization partitions the mass flux into two parts with different entrainment rates based on the Gregory scheme. The CMIP5 model version, which does not produce a propagating MJO in YOTC hindcasts, simulates a transition from shallow to deep convection at a smaller column water amount than that observed in KAZR reflectivities and produces excessive updraft speeds rarely observed in oceanic convection. A modified model version that increases the entrainment rate of the less entraining plume and does simulate an MJO delays the transition somewhat, but still creates too much deep convection at intermediate column water amounts. This version produces a bimodal vertical velocity distribution, with peaks at < 1 m/s and at 6 m/s. A further modification that only allows the less entraining plume to exist when a downdraft cold pool is present produces a fairly realistic transition from shallow to deep convection and updraft speed peaks at < 1 m/s and 2-3 m/s because cold pools only form on a couple of occasions during the AMIE-Gan forcing period. However, the GCM with this physics produces more frequent cold pools and a weaker MJO in YOTC hindcasts.

Lead PI

Anthony D. Del Genio — National Aeronautics and Space Administration