Seamless Stratocumulus Simulation across Model Resolutions

Due to the inherent nonlinearity of microphysical process rates, biases up to a factor of 4 can be produced by not correctly representing the unresolved variability within a model grid-box. Often model tuning is used to correct this for a certain configuration, but a better approach would be to design the parametrization to implicitly adapt to the model resolution, and therefore represent the required amount of sub-grid variability and bias correction. A scale-aware microphysical parametrization is therefore developed and tested on a case study forecast of stratocumulus evolution. The parametrization uses observational data on spatial variability of cloud and precipitation, gained from research aircraft, ground based radar and lidar, and CloudSat, to upscale microphysical process rates to the appropriate grid-scale. Simulations at a range of model grid-lengths between 1km and 100m are compared to aircraft observations. The improved microphysical representation removes the correlation between precipitation rate and model grid-length, allowing all resolution simulations to produce similar precipitation rates.