Process-Level Diagnostics of Marine Stratocumulus in TaiESM1: Insights into Parameterization Successes and Deficiencies

Abstract. Marine stratocumulus clouds strongly influence Earth’s radiation budget by reflecting incoming solar radiation, yet many general circulation models (GCMs) struggle to simulate them, primarily due to parameterization deficiencies. In contrast, the Taiwan Earth System Model version 1 (TaiESM1) can realistically simulate marine stratocumulus clouds and associated cloud radiative effects, although the reasons for this performance remain unclear. To investigate this, we conduct short-term hindcast simulations for July 2001, coinciding with the DYCOMS-II field campaign over the northeastern Pacific, together with complementary single-column model (SCM) simulations of its RF01 and RF02 flight observations. Detailed tendency analysis of cloud liquid water and related variables is used to quantify contributions from individual parameterizations and their interactions. Results show that stratocumulus maintenance depends on (1) sufficient moisture transport by the turbulence scheme; (2) conversion of moisture to cloud liquid by the cloud macrophysics scheme; and (3) cloud liquid diffusion by the turbulence scheme. In the cloud layer, the turbulence and macrophysics schemes exhibit a delicate balance, and their residual determines cloud evolution. In the subcloud layer, the turbulence scheme unrealistically diffuses cloud liquid to the surface layer, which is then removed by the macrophysics scheme, causing spurious cooling and moistening tendencies. This unrealistic behavior requires correction. This study demonstrates that combining field observations, short-term hindcast simulations, SCM experiments, and tendency analysis provides a powerful framework for diagnosing parameterization behaviors, successes, and deficiencies. This framework is broadly applicable to other cloud-related biases in GCMs.