On The Representation of Mesoscale Convective Systems Across West Africa in Kilometer-Scale Global Simulations

Across West Africa, better understanding of mesoscale convective systems (MCSs), which is responsible for most of the rainfall over the region, is crucial for prediction of precipitation from short (seasonal) to long (decadal) timescales. Simulation of MCSs in numerical models is a challenge due to the importance of multi-scale interactions between convective-scale dynamics, microphysics, and the large-scale environment. Most global simulations are currently performed at relatively coarse resolutions (~50-200 km) with parameterized convection; however, recently there has been an effort to run global kilometer-scale simulations, such as those run as part of the DYAMOND (DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains) project, which allow convection to occur without parameterization. In this study, we examine the ability of three global GEOS simulations, with resolutions of 3 km, 12 km, and 50 km, to produce MCSs across West Africa. We find that although the 3km GEOS simulation better represents the timing of the MCS characteristics, it generally produces fewer MCSs with shorter durations than the MCSs detected from observations. The representation of the African Easterly Jet (AEJ), wind shear, and the diurnal cycle of radiative and turbulent fluxes in these simulations is explored to better understand biases in MCS initiation times and locations. This analysis can be extended to other DYAMOND simulations, which allow us to investigate how other kilometer-scale global simulations represent MCS initiation and growth across West Africa.