Modeling the Effects of Turbulent Supersaturation Fluctuations in Large Eddy Simulations

Clouds play a significant role in Earth’s climatological and hydrological systems. However, many aspects of collision-coalescence and warm rain formation remain poorly understood. One hypothesis to explain these processes is the impact of turbulence. Turbulence may have significant implications on cloud microphysics due to supersaturation fluctuations which can result in enhanced evaporation, activation, or growth of cloud droplets beyond what is found in a quiescent environment. However, accurate representation of these processes in large eddy simulations (LES) remains a significant challenge due to a lack of existing models for small- scale turbulent fluctuations of the supersaturation in the subgrid scales, which are not resolved in an LES. Several studies have been conducted to better understand these small-scale processes including those using the Pi Chamber at Michigan Technological University. The Pi Chamber is a laboratory chamber which uses a warm plate on the floor and a cool plate on the ceiling to produce and sustain turbulent Rayleigh-Bénard convection. Further, aerosol particles are injected into the chamber to support cloud particle activation. Following this approach, this study conducts a suite of direct numerical simulations (DNS) of the Pi Chamber with varying droplet injection rates. Using these simulations, a priori testing is conducted and used to evaluate new subgrid scale models for the subgrid scale supersaturation variance in LES. This study focuses on supersaturation variance of one simulation case. Supersaturation data is low-pass filtered and used calculate two subgrid scale models: a scale similarity model and a gradient model. Through statistical comparisons of these models to the true subgrid scale supersaturation variance, it is hypothesized that one model may perform better suggesting it may be a candidate to model these subgrid scale processes in the future.