Bretherton, Christopher SCaldwell, Peter MSantos, Sean Patrick2021-03-192021-03-192021-03-192020Santos_washington_0250E_22352.pdfhttp://hdl.handle.net/1773/46726Thesis (Ph.D.)--University of Washington, 2020Time integration in climate models is complex, typically involving several levels of nested sets of substepped processes, potentially using a different time integration scheme at each level. Many climate models are also subject to a high degree of time step sensitivity, producing a significantly different climate when the model's time step is reduced. In order to reduce the effect of time integration error on model output, it is therefore important to attribute a model's time step sensitivity to particular processes or sets of processes in need of improvement. This study focuses on attributing time step sensitivity to sets of processes in the Energy Exascale Earth System Model version 1 (E3SMv1), and in particular in E3SMv1's atmosphere model (EAMv1). The first stage of this study focuses on EAMv1's stratiform microphysics scheme, the Morrison-Gettelman microphysics version 2 (MG2). Numerically relevant timescales in MG2 are derived by computing the eigenspectrum of its Jacobian. These timescales are found to often be smaller than the default 5 min time step used for MG2. The fast timescales are then heuristically connected to individual microphysics processes. By substepping a few particular rain processes within MG2, the time discretization error for those processes was considerably reduced with minimal additional expense to the overall microphysics. While this improvement has a substantial effect on the target processes and on the vertical distribution of stratiform‐derived rain within EAMv1, the overall model climate is found to not be sensitive to the MG2 time step alone. The second stage of this study examines E3SM as a whole. We find significant time step sensitivity in EAMv1, leading to large decreases in the magnitude of cloud forcing when the time step is reduced to 10 seconds. When MG2 is substepped together with other cloud physics processes, the effect of MG2 substepping is amplified, causing increased precipitation, a drying of the atmosphere, and an increase in surface evaporation. Coupling the model’s dynamics and physics more frequently reduces cloud fraction at lower altitudes, while producing more cloud liquid at higher altitudes. Reducing the deep convection time step also reduces low cloud mass and cloud fraction. We conclude that the cloud physics in a global circulation model can depend strongly on time step, and particularly on the frequency with which cloud-related processes are coupled with each other and with the model dynamics. Substepping individual processes often seems to be an attractive method for improving time integration without excessive computational cost. However, substepping can easily fail to reproduce the effect of changing the entire time step for a model. Effective use of substepping therefore requires careful analysis to group processes that jointly contribute to a model's time step sensitivity.application/pdfen-USCC BYclimate modelingmicrophysicstime integrationAtmospheric sciencesComputational physicsApplied mathematicsApplied mathematicsThesis