Exploring a Lower-Resolution Physics Grid in CAM-SE-CSLAM
A lower-resolution grid is implemented into the Community Atmosphere Model with spectral element dynamics and conservative semi-Lagrangian tracer transport (CAM-SE-CSLAM) for evaluating the physical parameterizations. Building on previous work, the physics grid is effective at reducing spurious numerical noise from the spectral element dynamics and contains 5/9th fewer grid columns than the dynamical core. Algorithms are presented that map fields between the dynamic core and the lower-resolution physics grid while maintaining numerical properties ideal for atmospheric simulations such as mass conservation and mixing ratio shape and linear-correlation preservation.
Despite the use of a lower-resolution physics grid, the range of scales resolved by the model, which impact precipitation processes, is indistinguishable from the default method of evaluating the physics at the same resolution as the dynamical core. And since the physics is a primary contributor to the large computational burden of current generation atmosphere models, the lower-resolution physics grid provides significant cost savings with little to no downside.
The lower-resolution physics grid is implemented into a high-order finite-element dynamical core. It is therefore compatible with the atmosphere model of the Energy Exascale Earth System Model (E3SM) and would provide a more accurate solution (i.e., precipitation) while providing a significant reduction in computational costs.