PNNL

Abstract

Convective momentum transport (CMT) is the process of vertical redistribution of horizontal momentum by small-scale turbulent flows from moist convection. Traditional general circulation models (GCMs) and their multiscale modeling framework (MMF) versions poorly represent CMT due to insufficient information of subgrid-scale flows at each GCM grid. Here the explicit scalar momentum transport (ESMT) scheme for representing CMT is implemented in the Energy Exascale Earth System Model-Multiscale Modeling Framework (E3SM-MMF) with embedded 2-D cloud-resolving models (CRMs), and verified against E3SM-MMF simulations with 3-D CRMs and observations. The results show that representing CMT by ESMT helps reduce climatological mean precipitation model bias over the western Pacific and the ITCZ regions, which is attributed to the weakened mean easterlies over the Pacific. Also, CMT from simulations with 2-D and 3-D CRMs impose a similar impact on Kelvin waves by reducing their variability and slowing down their phase speed, but opposite impacts on the MJO variability. The ESMT scheme readily captures the climatological mean spatial patterns of the zonal and meridional components of CMT and their variability across multiple time scales, but underestimates their magnitude, particularly in the middle troposphere. CMT mainly affects the MJO by decelerating its winds in the free troposphere, but accelerates its near-surface winds. This study serves as a prototype for implementing CMT scheme in the MMF simulations, highlighting its crucial role in reducing model bias in mean state and spatiotemporal variability.