Tracking Mesoscale Convection Systems in the US in E3SM Multiscale Modeling Framework

Organized mesoscale convective systems (MCSs) contribute a large amount of precipitation in the US during spring and summer, which impacts the availability of freshwater and flooding events. However, many current global climate models cannot capture MCSs well and misrepresent the timing and intensity of precipitation in the Central US. In this study, we investigate how the representation of MCSs can be improved in the Energy Exascale Earth System Model (E3SM) by either using a finer global grid (~25 km) to better represent intermediate-scale dynamics (HR) or adding an explicit representation of convection via the multiscale modeling framework (MMF) configuration. Both experiments are configured to use an equivalent amount of computational resources, and MCSs are identified and tracked using TempestExtremes. The E3SM-HR reasonably represents precipitation in some aspects but simulates too few MCSs over the continental US. This failure to capture MCSs may be related to underestimating deep convection and high-level cloud ice with the conventional convective parameterization. On the other hand, E3SM-MMF represents the seasonal number of MCSs and the associated annual cycle reasonably well compared to the E3SM-HR. Nevertheless, the location and the initial time of the MCSs in E3SM-MMF show some differences from observations.