Consequences of cloud superparameterization for climate simulations of land-atmosphere feedback and the microphysical sensitivity of US mesoscale convective systems.

Description

Preliminary findings are shown from a new DOE Early Career superparameterized (SP) climate modeling project focusing on twin issues of land-atmosphere coupling and microphysical sensitivities of US summer MCS. The latter are an exotic form of rainfall that can be captured in SP climate models but are in need of tuning against DOE station and field campaign data. For land-atmosphere feedback, a robust effect of SP is discovered to be enhanced Bowen ratio and Bowen ratio sensitivity to climate change. This is evident across many regions and model versions with different resolution, cloud microphysics, and land-surface processes. It is associated with a broadening of the temperature distribution to include more extreme heat events in SP models. For microphysical sensitivities of MCS, SP pilot tests show responses that are subtle compared to classic CRM benchmark experiments due to strong regional internal and event variability. Composite microphysics sensitivities (such as would be needed for tuning) are indetectable even in samples of 20 storms per season implying hundreds of storms are necessary to effectively tune SP MCS signals against ARM data. This argues for a long-term hindcast experiment design validated against long radar and radiometer records at the DOE SGP site rather than a targeted seasonal field campaign hindcast experiment targeting MC3E aircraft data constraints, as initially envisioned for this project.